WO2009074708A1

WO2009074708A1 - Use of a smart camera for controlling an industrial ac drive

Info

Publication number: WO2009074708A1
Application number: PCT/FI2007/050552
Authority: WO
Inventors: Veijo Pulkkinen
Original assignee: Euroelektro International Oy
Priority date: 2007-10-11
Filing date: 2007-10-11
Publication date: 2009-06-18
Also published as: EP2203794A1; EP2203794A4

Abstract

The invention is based on the idea of connecting a smart camera directly to a frequency converter to control an industrial production process. The smart camera comprises process-specific analyses software. The smart camera takes pictures of an object in the production process and analyzes each picture immediately. Results of the analyses the smart camera sends directly to the control unit of the frequency converter that in turn controls the speed and / or torque of the AC-motor as needed. Smart cameras are capable of taking consecutive pictures at very short intervals and analyzing each picture between shots. Frequency converter's control units are also capable of fast communication and motor control. Thus this invention makes it possible to reach significantly faster response time than what is possible in a system that makes use a programmable logic controller (PLC).

Description

Use of a Smart Camera for Controlling an Industrial AC drive

Field of the invention

[001] This invention relates to a process automation system generally and more particularly in setting quality standard to a product in an industrial process operated by an electrical drive system where a frequency converter converts constant line supply voltage and frequency connected to the input terminals of the frequency converter to variable voltage and frequency at the output terminals of the frequency converter to feed an alternative current (AC) motor at controlled rotational speed. Background of the invention

[002] Advantages of an AC-motor over the DC-motor ( Direct Current motor) are ruggedness, minor need for maintenance and lower price. Disadvantages are low starting torque, nominal torque is only available at rated speed and the rotational speed of the motor shaft is constant and dependent on the fixed frequency of the supply network. Most applications require higher or lower speed than constant rotational speed of the motor shaft and also high starting torque therefore mechanical gear or belt transmission must be used. Advantages of a DC-motor are good controllability over a wide speed range and high starting torque. Therefore they are used widely in industrial processes that call for high starting torque in addition to wide speed control range.

[003] Frequency converter has made it possible to replace a DC-motor by an electrical drive system, which consists of an AC-motor, and a frequency converter that converts constant voltage and frequency of an AC-network to controllable voltage and frequency to feed an AC-motor. Changing output frequency of the frequency converter changes rotational speed of an AC- motor and thus the speed range of the AC-motor can be controlled from the zero speed to over nominal speed of the motor. The frequency converter controls not only the rotational speed of the connected AC-motor but also the torque of the motor. This kind of electrical AC-drive systems have widely superseded DC-motors in process industry and consumer goods industry and also in energy, oil, gas, and petrochemical industries. Application examples are e.g. roller tables, conveyors, pumps, fans, extruders and cranes.

[004] Figure 1 shows the basic construction of the frequency converter. The frequency converter is a housed unit that contains all of the necessary components and has all of the necessary connectors on the housing wall. The supply network's constant three-phase voltage and frequency is rectified in the rectifier 11 and resulting DC-voltage is filtered in the filter circuit integrated in the intermediate circuit 12. After that, the constant DC-voltage is led to the inverter 13 which produces the desired, controlled three-phase voltage and frequency. The inverter is formed out of Insulated Gate Bipolar Transistors (IGBT). The supply network's voltage is, as mentioned before, mainly three-phased but in small frequency converters that regulate low power three-phase motors, the supply network's voltage can be single- phased. This way, the frequency converter changes the single-phase incoming voltage to a three-phase outgoing voltage.

[005] The frequency converter's control unit 14 makes up the IGBTs gate control so that not only the three-phase voltage but also its frequency can be adjusted at the converter's output. Because the AC motor's 15 rotational speed is determined by the feed voltage's frequency, the rotational speed can be regulated with the help of a control unit 14 when a frequency converter is connected to the motor. The user is able to install all of the frequency converter's necessary settings or in other words, is able to program the frequency converter using the control panel 111. Furthermore, the frequency converter can have multiple interfaces with external devices/systems. The control unit is a programmable logic consisting of programmable functional blocks that execute logical and mathematical functions.

[006] Frequency converter can be connected to fieldbuses used in industry using fieldbus modules 16. Analog and digital I/O modules 17 and 18 make it possible to send control information to the frequency converter and receive monitoring and measurement information from the frequency converter. The frequency converter can have a network module 19 such as an Ethernet module, so it can be connected to a computer network. This way, the frequency converter can be monitored and even programmed remotely. Fieldbuses can even be replaced by radio communication. Alternatively, frequency converters can have optical cable connectors 110. In the case where one frequency converter is subordinately connected to another frequency converter (master/follower arrangement) the follower converter's commands are usually transmitted using an optical cable.

[007] Typically in electric drives the frequency converter is programmed with initial values such as frequency, motor speed, torque, current and start-up acceleration values, as well as stopping deceleration values. Start-up values are programmed according to the application that the AC motor operates so that the values are different, for example in pumps and conveyors. Should there be need for changing e.g. motor speed or torque or other major value while the frequency converter is in operation this can be done using the control panel or monitoring connection (I/O, fieldbus or Ethernet connection). Thus, the frequency converter functions independently and in many applications, the AC motor's on/off command is the only external control signal to the frequency converter. This method of operation is used e.g. in servo-control, whose frequency converter-controlled AC motor has replaced a traditional DC-motor.

[008] Figure 2 shows a figure of how electrical drive is typically used in process industry. In this process example there are three (3) stages that use an AC motor. For simplicity's sake, we can assume that the AC motors operate conveyors: motors 1 and 2 operate the first process stage's conveyor 1, motor 3 operates the second process stage's conveyor 2 and motor 4 operates the third process stage's conveyor 4. The conveyors' speeds can be independent of each other even though speeds are dependent of each other in continuous production. Motor 1 is controlled by frequency converter 21 and correspondingly, frequency converter 22 controls motor 2. In this case, frequency converter 1 is connected through a fieldbus to programmable logic 29 which gives the frequency converter new speed reference command when necessary. Instead of being connected to the fieldbus, frequency converter 22 is connected directly via optical cable to frequency converter 21 , which gives frequency converter 22 its speed reference command. Frequency converter 21 is then the master of frequency converter 22, which is then the follower. Frequency converters 23 and 24 operate independently and are both connected via the fieldbus to process control logic 29. At start-up, each frequency converter is programmed with speed reference commands, which produce the desired initial voltage and frequency and the AC motor's rotational speed. Various control and alarm limit values are programmed in as well. [009] Process control logic takes care of all process stages. The conveyors in the figure can only be a part of the industrial process. In this example, the process control logic 29 receives information about each conveyor's speed from sensors (sensor 1... sensor 3). During the logic's program cycle, sensor information is read and based on that, decisions are made for each frequency converter on whether or not its conveyor's speed should be increased or decreased and if so, by how much. If a conveyor's speed must be changed, the process control sends a speed reference command to the conveyor's frequency converter, which then changes the output voltage and frequency accordingly. Subsequently the AC motor that receives its feed voltage from the frequency converter changes, as does the conveyor speed. In a master-follower situation, the master frequency converter 21 receives its speed reference command through the fieldbus and its control unit calculates not only its own inverter's control values but also the follower's inverter control values, which it then sends along optic cables to the follower frequency converter 22.

[010] A typical characteristic of state-of-the-art electric drives is that frequency converters are pre-programmed to produce an output voltage and frequency so that the AC motor that is connected to it has the desired rotational speed. If the AC motor has to function in a certain way according to a predetermined speed curve in a given application, its commands can also be programmed into the frequency converter.

[011] The challenge in this kind of an independent application is that although it is possible to receive information from various sensors in the process, the information from those sensors is handled in different controllers, such as PCs for example. When a motor's rotational speed needs to be changed according to one sensor's measurement, it has to be done separately through the frequency converter's user interface, either locally or remotely. In the simplest case there are no sensors, only processes or their results are inspected visually or by taking samples, which are then analyzed. The AC motor's speed is then changed if necessary. These types of adjustment methods are slow and it is difficult to maintain the desired quality of the finished product. More specifically, the problem is that the frequency converter receives its speed reference commands from somewhere else and it is not an active part of the product's quality control and quality level maintenance.

[012] Frequency converters are usually used in environments where the frequency converter's controlled AC-motor is one of many motors. The process stage that the motor operates is dependent on other process stages, in which case the wide range of motor speed controllability, made possible by the frequency converter, is exploited. The configuration introduced in figure 2 is an example of this kind of environment. A state-of-the-art solution in a highly automated industrial environment is that a dedicated unit is responsible for the process control, which is most often a Programmable Logic Controller (PLC). The information from a large number of different sensors and detectors in the process is fed into the inputs of the PLC. The PLC uses this information in its calculations and then, if necessary, sends control commands to actuators in the process like actuators and frequency converters. Connections between the actuators and the PLC can be hardwired even though in large automated processes actuators and the PLC are connected to a common fieldbus.

[013] The challenge with these kinds of environments is that for the external controller, like PLC, the frequency converter is just one of many actuator controller. Information about the part of the process that the frequency converter's AC motor operates first comes from a sensor to the process controller, which then executes its calculations and then sends control commands to the frequency converter. This takes time and there are applications where this kind of delay is detrimental. Furthermore, frequency converter functions in this type of environment inflexible: AC motors rotate at the exact speed that the rigid logic program has determined. It is laborious to program "what if" actions in to a logic program, or in other words, how much can a single frequency converter-fed motor's speed be raised or lowered so that the process is not disturbed but so that the whole process' efficiency grows. More specifically, the problem in these environments is that the frequency converter receives its speed reference commands from external controllers and it is not an active part of the product's quality control and quality level maintenance.

[014] The objective of this invention is to create a control method and arrangement for a frequency converter fed AC-motor which resolve the aforementioned problems. The objective is to create an independent arrangement, made up of one or many frequency converters, which actively participates in the quality maintenance of the final or intermediate product even without being connected to an external controller.

Summary of the invention [015] This invention is based on the idea of connecting a smart camera, with a process-specific analysis program, directly to a frequency converter. With this method, the smart camera is used to take a picture of the product, which can be anything like an external piece, liquid, granule, sheet and filmstrip, etc. The picture is then analyzed in the smart camera by handling the pixel values according to a predetermined task. The smart camera's analysis results contain a message which is sent to the frequency converter's control unit. It handles the message and adjusts the AC motor's rotational speed as necessary.

[016] The invention's electrical drive system firstly includes a smart camera that is connected to the frequency converter's control circuit by a data transmission link. The smart camera has handling equipment for processing the picture's pixel values according to a predetermined task and sending the result of the task to the frequency converter's control circuit. Secondly, it includes control circuit's connection equipment, which has been configured to handle the results of the smart camera's analysis and which can modify the

AC motor's operating voltage and frequency and rotational speed according to those results.

[017] One advantageous way that the invention can accomplish this is a servo drive that consists of a frequency converter and servo motor that functions as an AC motor, whose operating voltage has been determined as the output voltage of the frequency converter and that positions industrial process devices via mechanical transmission. In this invention, the smart camera that is connected to the servo drive takes a picture of the desired stage of the industrial production process. The camera has handling equipment installed for analyzing the picture's pixel values according to a predetermined task and for sending the results of the analysis via data transfer connections directly to the control circuit of the frequency converter. The control circuit of the frequency converter has equipment for handling the analysis results and for adjusting the servomotor's voltage, frequency and timing based on those results so that the device is positioned according to the picture taken by the smart camera.

[018] Smart cameras are very fast; they can take and analyze hundreds of pictures in a second. Logic circuits in frequency converters are also very fast, so the arrangement according to this invention allows for a faster reaction time than is possible in an environment where PLC is used. The frequency converter or group of frequency converters does not need to connect to a PLC and it can still achieve the benefits of a process control system. Brief description of the drawings

[019] The invention is described more closely with reference to the examples hown in the figures of the accompanying drawings, in which

Figure 1 presents the frequency converter's functional blocks, Figure 2 presents examples of electrical drives in industrial processes, Figure 3 describes the invention's basic principles, Figure 4 is a simplified flowchart of how a smart camera functions, Figure 5 is a simplified flowchart of how a frequency converter functions, Figure 6 describes a practical application, Figure 7 presents how the smart camera functions in the practical application showed in figure 6, Figure 8 describes another practical application, Figure 9 is a simplified flowchart of how the smart camera functions in the practical application showed in figure 8, Figure 10 is a simplified flowchart of how the follower frequency converter functions in the practical application showed in figure 8, Figure 11 describes the functions of the master frequency converter in the application showed in figure 8,

Figure 12 describes the functions of the follower frequency converter in the application showed in figure 8, Figure 13 presents the invention's execution method using a servo drive, Figure 14 is a schematic figure of how the smart camera functions in the execution method described in figure 13, Figure 15 presents the functions of a frequency converter, and Figure 16 shows the function of an AC motor serving as a servo motor.

Detailed description of the invention

[020] In figure 3, one can see the invention's basic principle. In this example, the AC motors' 31 and 32 process conveyor is shown. This way, motors use a conveyor that proceeds at a set speed. At each end of the conveyor is an AC motor that moves the conveyor and the master frequency converter 310 provides the operating voltage to motor 31 and correspondingly the follower frequency converter 320 provides the operating voltage to motor 32. Frequency converters are connected to the three-phase feed grid 323 from the inlet side. In large automated systems, the master frequency converter 310 can be optionally connected to the controller like the (PLC) 33, where it gets at least its start and stop commands. In this type of use, the draw-in roller's motors regulate not only the conveyor's speed but also the conveyor's sag D. Sag is regulated by adjusting the motor's torque using the frequency converter. [021] In addition to its own properties, a conveyor's sag is affected by the weight of the product/products on the conveyor and possible conveyor stretching. In state-of-the-art solutions, it is common to use a detector on the lower edge of the conveyor's upper belt, whose depression is directly comparable to the sag. The frequency converter's torque is then adjusted through the user interface according to the detector's readings.

[022] With the invention, a smart camera 32 is used, which takes pictures periodically from the side of the conveyor. Because of the limits of the drawing technology used, the camera in the figure is pictured above the conveyor but in actual use, it is on the side of the conveyor at the height of the upper belt so that the belt is seen in the picture as a line. The command for taking a picture is given by a detector (not shown), by the camera's own timer or by the master frequency converter 310. After each picture is taken, the smart camera's analysis application analyzes the picture using its pixel values . according to the predetermined calculation tasks. The analysis application is, of course, task specific. For example, in calculating sag D, the distance from the centre point of the line seen in the picture and the reference level is measured.

[023] Smart cameras have selection of software, from which the most appropriate program is selected for this type of application, ldealistically the smart cameras have a wide software library, from which to select the most appropriate program for each industrial application. The software is initialized at start-up: the pixels of the picture that is taken by the smart camera are scaled to millimetres using the camera's reference picture, in which the exact reference points' measurements are known. The distance between the reference picture and the smart camera is exactly the same as the product's distance from the camera in the production process. So when scaling the picture, appropriate parameter values are used. In practice, whenever the camera has taken a picture, the application software analyzes it and forms an analysis message. In this example, the result of the analysis is the value for sag D in millimetres that is included in the analysis message. Technically it is entirely possible to upload a nominal sag value into the camera's initialization, so that as a result of the analysis, the camera can ask the frequency converter to either increase or decrease the AC motor's torque as necessary.

[024] However, in practice it is economical to keep the master frequency converter as the master and let it make decisions on changes to speed or torque. In that way, the analysis message only contains the results of the analysis and no commands. This means that the frequency converter's control unit uses the analysis results to calculate if the line speed or torque needs to be changed. The decision making control is with the frequency converter, which has more information about the whole process than the smart camera does.

[025] The analysis results are sent in a message via optic cable 321 to the control unit of master frequency converter 310. When the message is received, the control unit processes the results and makes decisions about possible changes to the line speed or motor torque. If the master frequency converter 310 has decided to decrease sag D, it changes the frequency of the output voltage to the level indicated by the frequency converter and then correspondingly changes the speed and torque of the AC motor 31. The master frequency converter also gives a new speed/torque command to the follower frequency converter 320, which then changes the speed or torque of the AC motor 32 connected to it.

[026] The smart camera 32, continuously takes and analyzes pictures of the specified object and sends the results to the master frequency converter. If it is known that change of the sag is slow, it is enough that the master frequency converter gives the smart camera an execution command when it wants to measure the amount of sag in the line.

[027] Figure 4 shows a simplified block diagram of how the smart camera functions in figure 3's example. Once the smart camera is installed and connected by optic cable or fieldbus to the frequency converter, then initialization is carried out. The initialization can be done from the frequency converter or from an external PC that is temporarily connected to the smart camera. During initialization, firstly an appropriate analysis application program is chosen, if the smart camera comes with many different analysis applications ready installed in its software library. Then, the analysis area's size and position is determined because it is possible that only one specific area of the picture is of interest. It is important to calibrate the pixel area of the picture that the smart camera takes, so that one knows how many millimetres the picture corresponds to in reality. This is done by taking a picture of the reference picture, which is taken from the same distance as the product is during production and whose exact dimensions are known. Once the application has been selected, parameters are programmed into the application, which determine what is being calculated from the pixel values in the analysis application and what results are included in the analysis message. Correspondingly, the frequency converter has to receive the appropriate initialization, after which the system is ready to be taken into use.

[028] Depending on the application environment, stage 42, the smart camera either receives the picture execution command from an external detector, its own clock pulse or from a picture execution command sent by the frequency converter. The external detector could be a detector connected to the conveyor belt that gives a trigger pulse when the object to be pictured is in position in front of the smart camera. When the picture is taken, stage 43, the camera's software analyzes it, or in other words, carries out the predetermined task, stage 44, whose results it includes in the analysis message and sends to the frequency converter, stage 45. Depending on the object to be pictured, the camera can take even as many as 100 pictures a second, so that the time taken to complete the stages mentioned above would be 10 milliseconds. [029] The analysis software can also be one that requires multiple consecutive pictures in order to execute the task. The same things are calculated in each picture and once the picture sequence is complete it is easy to calculate the average, dispersion, etc.

[030] Figure 5 shows a simplified block diagram of how the frequency converter functions in the example shown in figure 3. The frequency converter's logic has been initialized to execute predetermined tasks. When the results of the analysis are received, stage 46, the results or more specifically, the group of numeric values is brought to the inputs of the logic circuit, stage 47. Logic circuits calculate the need to change the motor's speed or torque, stage 48, and if it is necessary to form a new speed/torque command for the inverter, stage 49, which is then sent to the gate of the inverter's transistor, which in this case is an Insulated Gate Bipolar Transistor (IGBT), stage 410. Now, the AC motor uses the changed frequency from the inverter's output as its operating voltage, which is brought to the motor's connector, stage 411. In stage 48, the logic compares the suggested speed from the analysis results with the comparison values. The frequency converter that acts as the master in the smart camera's view, makes decisions on changes to speed and it uses the comparison values that were programmed during initialization to do so. Furthermore, the master frequency converter gives the acting follower frequency converter its speed reference command.

[031] The logic circuits are very fast, so changes in the inputs of the logic are seen almost immediately in the outputs. Taking into account that modern smart cameras are capable of very fast picture speeds and are still able to do the necessary analyses, the speed at which the combined smart camera and frequency converter solution can make adjustments is very fast.

An application example of the invention

[032] Figure 6 presents an example of using the invention in cement production. As it is well known, finished cement is produced by milling cement clinkers in a mill. The desired coarseness of fine aggregate is then separated from the milled material by a rotating separator. The rest of the aggregate is then returned to the mill. There are many types of milling/separating mechanisms. In some cases, the milling/grinding is done in separate milling/grinding units, from which the milled/ground aggregate is then blown into a separating unit. In other mechanisms, these units are integrated. From the invention's point of view, there is no significant difference as to how the material is milled and separated.

[033] According to the figure, the cement clinker is brought to a milling vat that has a hammer mill or ball mill at the bottom (not pictured). The clinkers are milled in the mill and the milled material rises up in the vat. Air is blown into the vat so that a fluidized bed is formed. The upper part of the vat has a separator 64, which is rotated by an AC motor 63, which receives its operating voltage from the frequency converter 61. The separator consists of a cylinder that has rotating, vertically attached blades inside of it. Due to the air flow's effect, the milled material rises up in the cylinder. The blades then hit larger particles of the milled clinker and knock them back into the lower part of the vat for re-milling. The finer milled material goes past the blades and is then blown along with the airflow to an exit pipe 65. The separator's blades' speed, or in other words, the AC motor's 63 rotational speed determines the separated material's particle size.

[034] In a state-of-the-art solution, a sample of the material is taken from the exit pipe, which is then analyzed by either the naked eye or by using known particle size analysis methods such as sifting or separating. The purpose is to make sure that the largest particle observed stays below the set limits. If larger particles are found, the rotational speed of the separator is increased by manually entering a new speed reference command into the frequency converter. This method is slow.

[035] In this invention, the smart camera is connected to the frequency converter. Additionally a powerful flash, such as a laser flash, is also connected to the smart camera. The camera's initialization parameters are programmed via Ethernet for example from the frequency converter. The purpose of the initialization is to scale the dimensions of the picture area in millimetres, which the camera's image cell recognizes as a picture element (pixel). What fits into one pixel's area depends on the distance of the object from the camera and this is why the initialization uses a picture, whose grain size is known. The camera's software is made to calculate the object's grain size in the picture. The camera's scaling parameters are changed from the frequency converter until the camera indicates that the grain size is correct. The camera is now initialized.

[036] After initialization, pictures are taken with the smart camera 62 of the particle current flowing from the exit pipe at appropriate intervals. The picture rate is for example 20 pictures per second. In practice, a side current is created from the particle current for picturing purposes. It is guided to the picturing point and the distance of the side current to the smart camera is the same as the distance used in the initialization of the camera. Because the speed of the particles in this side current is high, the camera has to use very short shutter times in order to get freeze-frame pictures, which means that very powerful lights have to be used at the moment the picture is taken. It is practical to use a laser flash as the light source so that a very strong and focussed light pulse can be flashed at the exact moment the picture is taken, resulting in a successful freeze-frame picture.

[037] After each picture, the smart camera's analysis program calculates the size of the largest particles in the picture and their average and sends the results to the frequency converter 61. It is also possible for the smart camera to calculate the average size of particles from many consecutive pictures and then send the results to the frequency converter.

[038] The frequency converter then sends the results to its logic program, which has been initialized to carry out the necessary logic operations. In these operations, the logic decides whether or not to increase or decrease the rotational speed of the blades in the separator. If the largest particle size surpasses the reference limits, the logic forms a new speed reference command for the inverter, which slightly increases its output voltage's frequency so that the rotational speed of the blades in the separator increase. This way the blades knock down the largest particles back into the mill. The logic can be initialized so that it increases or decreases the speed reference command in very small increments. Then, the analysis results from smart camera's picture can control the process so that the optimal particle size is reached in a slow enough way. [039] Figure 7 shows an even further simplified flow chart of the smart camera's function in the milling example. The camera is initialized in stage 71 so that it is known specifically how many millimetres in the picture correspond to one picture element (pixel). This is done by placing the camera at a known distance from a reference object, whose particle size is known. A picture is taken of the object, which is made up of n x m picture elements, where n is the number of picture elements on the light cell on the x axis and m is the number of picture elements on the y axis. The software calculates the reference point particle's dimensions in both directions and gives the results in millimetres. The camera's parameter values are adjusted until the parameters that give the right millimetre values in both directions are found. The camera is now ready for use.

[040] When the picture execution command comes from the frequency converter for example, stage 72, the camera takes a picture of the particle, stage 73, and analyzes it by calculating the sizes of the largest particles in the picture. If the particle size surpasses the limit values given, stage 75, the camera sends that information to the frequency converter, stage 76, which increases the speed of the separator's blades. Stage 75 can be left out as well, in which case the camera reports the largest particle sizes to the frequency converter.

Another application example of the invention

[041] Figure 8 shows the invention in a situation where there are many consecutive process steps, 1-5, and each is connected to an electrical drive. For simplicity's sake we can assume that the steps are conveyors that carry products between handling units (not shown). One of the frequency converters, frequency converter 83, is the master and the other frequency converters are followers.

[042] At step 4 of the process, there is a device 86, that could be for example, the jam filling station in a bakery's production line. It dispenses jam from its dispenser tube 890 onto dough disks. The success in dispensing the jam depends not only on the dispenser's features but also the line speed at step 4 of the process. The more the frequency converter's AC motors increase the whole line's speed, the more mistakes occur at the jam filling station. The quality of the jam filling function is a compromise between many factors and it is attempted to maintain quality at a good level, for example only a certain amount of mistakes at the jam filling station are allowed. Mistakes can be measured by, for example, how much the jam filled area is off the dough disk's centre point. The desired place for the jam is the centre of the disk however some of the disks may have jam dispensed more to the side. In a state-of-the-art solution, the master frequency converter 83, controls process step 84's follower frequency converter by standard settings so that each step's speed corresponds with each other so that the speed is not so high that the quality of the jam dispensing suffers. The quality is inspected using the naked eye.

[043] According to the invention, a smart camera 87 is used to take pictures of the jam topped dough disks on line 4 from above. The smart camera is connected to line 4's follower frequency converter 84. A detector (not shown) such as a light cell triggers a picture execution command. Alternatively, another smart camera 88 can be installed to take pictures of the finished products and this camera is connected to the master frequency converter 83. [044] The smart camera 87 has software that is specifically meant for this type of picturing situation, ldealistically the smart cameras have a wide library of software to choose an appropriate program for the industrial application in question. When the camera is taken into use, its software is initialized and the appropriate parameter values are chosen. During initialization, the analysis area's size and location are chosen as well as analysis type (for example, gray scale average, filtered gray scale average), lighting, interval timing and number of pictures to be taken if the average of many consecutive pictures is being used. The analysis program is such that the smart camera uses the picture it has taken to analyze if the dispenser has dispensed the jam, if the jam is in the centre of the disk and if not, by how much is the jam off centre. These analyses can be carried out using gray scale values. In production, every time the camera takes a picture, it carries out the aforementioned analysis and forms an analysis message. If the analysis results in information that indicates that the quality of the measured product is compromised, for example in many dough disks, this can be and indication that the AC motor's conveyor speed is too high. In this case, the analysis message can contain a speed reference command to reduce the speed of the conveyor. The smart camera can also receive a message from the frequency converter, changing some of the quality parameters so that it includes in its analysis those products that have a lower quality than the new minimum quality parameters will allow. Then the camera forms an analysis message where it asks the frequency converter to increase the conveyor's speed. This way, the level of quality is consciously calculated, maintaining acceptable quality all the while.

[045] In practice, it is still a better choice to place only the desired values in the message, like what is the percentage of flawless products out of all products, what are the average limits of error in the faulty products, etc. In other words, the message does not contain any commands to change the speed. This way, the frequency converter's control unit uses this information in its calculation of whether or not the line speed needs to change.

[046] In both cases the follower frequency converter 84, which has the smart camera connected, does not change the speed of its AC motor but asks the master frequency converter permission to change its speed. The decision making hierarchy is then always with the master frequency converter, which has a better picture of the whole situation than the follower frequency converter does. The situation could be so that if the master frequency converter accepts the request, it could lead to speed changes in other follower frequency converters as well. [047] For that reason, it is economical to add another smart camera 88, which is connected to the master frequency converter 83, to monitor the finished product. This smart camera's software is made to analyze the quality of the final product. Quality criteria could be the percentage of faulty products out of all those produced. The measurements, which determine whether a product is faulty or not, are set by the user. This camera sends the master frequency converter information about the final product's quality, based on the figures used in the analysis. The software in the master frequency converter can decide based on these figures if some step in the process has to be slowed down so that the quality stays at the desired level. This way it can refuse the follower frequency converter's request to change speed because the change would not fit with the speeds of the other motors.

[048] Figure 9 shows the previous example's events in the smart camera. During the camera's initialization, the analysis task is set and the acceptance limits for jam dispensing are set, stage 90. When the camera receives a trigger pulse from the light sensor, stage 92, it takes a picture, stage 93. Immediately after taking the picture, the camera's analysis software analyzes the picture and checks if the product's jam portion is within acceptable limits, stage 94. The results are stored on a statistics list in the memory for calculations, stage 95. If there are enough samples, estimates on whether line speed has to be increased or decreased, stage 96, are calculated from the statistics recorded in the camera's memory. After this, the analysis results are sent to the follower frequency converter, stage 97.

[049] Figure 10 shows phases in the follower frequency converter. Once the results are received, it brings them to its logic, stage 100, and forms a new potential speed reference command, stage 101. However, it does not have the authority to send the speed reference command to its inverter. It sends a permission request to the master frequency converter, stage 102.

[050] Figure 11 shows phases in the master frequency converter. Once it has received the request to change speed, the master frequency converter checks if the requested speed change is possible, taking the current speeds of the other frequency converters' AC motors into consideration, stage 110. Because it is the master to all of the other frequency converters it obviously is aware of their situations. If the change is possible, the master forms and sends a new speed reference command to the follower frequency converter, stage 111. The command can be either an increase or decrease in speed. After this, it sends new instructions to the follower, stage 112. If it is not possible to change speed in stage 110, the master sends the follower a "permission denied" message.

[051] In practice, it is more usual that if the request by a follower frequency converter for a change in speed is accepted, the master frequency converter has to change the speeds of the other follower frequency converters. If the line speed as a whole can be increased, then the master frequency converter gladly requests that all of the follower frequency converters use a higher speed because it means a larger amount of production. It is also possible that the request is accepted even though the speeds of other steps in the process are not changed. More often than not, the reason then is that the master frequency converter has information that, even though an increase in speed will most likely reduce quality, the user has given permission for the quality to be decreased for one reason or another. [052] The follower frequency converter receives a message from the master, stage 120, figure 12, and checks if the speed request has been accepted, stage 121. If not, it continues to receive information from the camera. If the speed request has been accepted, the follower frequency converter sends the new speed reference command values to the inverter, stage 123, from which the changed voltage and frequency is sent to the AC motor operating stage 4's conveyor.

Third application example of the invention

[053] This example is related to using servo drives. Servo drives are widely used for moving and positioning various industrial operating devices. AC motors (three-phase motors) work well as servo drives because of their good dynamics (small moment of inertia, large torque) as well as their minimal need for maintenance. Servo drives always need frequency converters, controlled by the control logic, to produce the controlled voltage and frequency for the motor. [054] Figure 13 shows the invention applied as a servo drive. In this example is a line used to find faulty sections in veneer 1300, such as knots. The servo driven locating devices 1313, 1314 and 1315 remove the faults and replace them with a veneer patch. The patching devices move along a path across the width of the veneer using a toothed rack, belt or other mechanical transmission, each on its own AC motor 1310, 1311, and 1312. Each AC motor is controlled by its own frequency converter 131, 132 and 133. The frequency converters function as follower frequency converters so each of them is connected by optical cable 1317 to the master frequency converter 130, which gives the operating voltage to be used by the main AC motor 1316.

[055] In the invention, the patches are detected using a smart camera, in this case, multiple matrix cameras set up in a parallel 134, 135, 136 and 137, that can, in one shot, take pictures of thin strips across the whole width of the veneer. The cameras are mechanically attached above the line to bars running perpendicular to the line and each camera is electronically attached to the bus 138, by which communication with the master frequency converter occurs. The bus can be an Ethernet connection, whose connection to the bus is made via the frequency converter's Ethernet connector. At the edge of the line is an edge detector 1318 that gives a signal when the front edge of the incoming sheet of veneer is in line with the detector. The same happens with the back edge of the veneer sheet. The detector is connected to the master frequency converter's I/O input so that this frequency converter always gets a pulse when an edge is in line with a detector. The line also has an absolute sensor 1319 measuring the line's vertical movement. Unlike an incremental sensor, the absolute sensor does not need a calculator because it always remembers its location. The information from the absolute sensor is sent to the master frequency converter as well as the operating devices' frequency converters 131, 132 and 133. Whenever the edge sensor detects an edge and sends a pulse to the master frequency converter, it takes a reading from the absolute sensor into its memory. Because the distance between the edge sensor and the smart cameras is known in the direction that the conveyor is moving, the master frequency converter knows at every moment how far the veneer's front edge is from the smart cameras, based on the edge sensor's pulse and the absolute sensor's readings. [056] As the veneer proceeds along the conveyor operated by the AC motor 1316, the smart cameras take pictures both at once and in intervals depending on the master frequency converter's commands. The time to take consecutive pictures is when the picture of one thin strip does not go on top of the previous and following strip in the predicted way. Each camera analyzes the picture it has taken for knots in the picture area, which are seen as abnormal darker areas in the environment. If that kind of area or areas is found, the camera calculates the coordinates of the dark area's centre point. The coordinates are numeric values that indicate how many pixels to count in the direction of both edges in order to get to the darker area's centre point. In practice, the camera has been initialized so that it is known how many pixels correspond to one millimetre of the photographed area. This way, the camera knows how to scale its calculated centre point coordinates directly in millimetres. Once the picturing period has ended, the master frequency converter asks the cameras for the results of their analyses. Only those cameras that have found an area to be patched inform the master frequency converter of the coordinates of the faulty area in millimetre scale.

[057] Because the master frequency converter has information about the veneer's front edge arrival time and readings from the absolute sensor, it knows exactly where the centre point of the faulty area is, calculated along the length of the veneer from the front edge and along the width from the side edge. Now it can inform the cross coordinates to the follower frequency converters, which then activate their AC motors to position a patching device in the right place perpendicularly to the direction of motion of the veneer. When the areas to be patched arrive at the points where patching devices have already been correctly positioned, the master frequency converter stops the line for a moment, gives the patch execution command to the frequency converter which controls the patching device in question, which then removes the faulty area and replaces it with a patch. This action is carried out at each patching device's location.

[058] Figure 14 describes the events that occur in each smart camera. Before being taken into use, the camera is initialized, stage 141. During initialization, the appropriate software for the application is chosen from the software library, if the camera has many different programs. For this task, a program that is meant to recognize different tones from a surface is chosen. This program is suitable for this purpose because the area to be identified is a knot, which shows up darker in grey scale pictures. After this, the dimensions of the pictures taken by the camera are scaled into millimetres. This is done using the reference point, whose measurements are exactly known. Then, the areas to be taken up for analysis are determined. The smallest faulty area can be given as the parameters for recognition; a minimal trace of knot may not always be reason enough to execute a patch for example. Parameters can also be given as the largest faulty area to be recognized; if the area is very large, there may be other things than just knots in question. Frequency converters are also initialized at this point.

[059] In production, cameras receive their picture execution command via fieldbus from the master frequency converter, stage 142. Cameras take simultaneous pictures of the veneer moving underneath them, stage 143, after which each camera's analysis program analyzes the pictures and seeks out potential patch areas, stage 144. If faulty areas are found, the analysis program calculates their centre point's coordinates in millimetre scale, records them in its memory and waits for the results request from the master frequency converter. When the request arrives, only those smart cameras that have recognized an area for patching send the coordinates of the area to the master frequency converter, stage 147.

[060] According to figure 15, master frequency converter receives the centre point coordinates from the cameras. They are sent in millimetres from the pictures edge, from which the frequency converter calculates the real coordinates, which are the distances from the veneer's front edge and side edge, stage 153. Finally, the coordinates of the areas to be patched, which are sent to the servo frequency converter, are sent to each frequency converter, whose servo device (patching device) has been commanded to remove and patch a faulty area, stage 153. [061] When the follower frequency converter receives the coordinates, it sets its AC motor in motion and positions the patching device at the location indicated, and then waits for the patch command, stage 162. The master frequency converter calculates when the faulty area has arrived at the location of the patching device. When this occurs, it stops the line for a moment, during which the patching device cuts out the faulty area and installs a patch, stage 162. This occurs at each frequency controller controlling a patching device.

[062] The aforementioned examples are just some of the possible areas where the invention can be applied. It is clear to professionals that the invention can be used in a variety of environments, from the most singular to complex process automation systems.

Claims

1. A method for controlling AC motor drive in a continuous production process, whose operation is maintained by at least one electrical drive comprising a frequency converter that converts a supply network's constant input voltage affecting in its inlet connectors to output voltage affecting in its outlet connectors and that has a control circuit for controlling the output voltage and its frequency according to a given speed reference command, an AC motor, whose supply voltage is the frequency converter's output voltage, wherein the frequency converter controls the AC motor's rotational speed and thus an industrial process, characterized by the steps of: taking with at least one smart camera consecutive pictures of products moving along in the production process, analyzing each picture in the smart camera by handling its pixel values according to a predetermined task, generating in the smart camera an analysis message containing analysis results, sending an analysis message from the smart camera to a control unit of the frequency converter, processing the analysis results in the control unit in order to determine whether there is a need to change the AC motor's speed reference value, and forming by the control unit a new speed reference command, if necessary, to alter the AC motor's voltage, rotational speed, and/or torque.

2. The method as in claim 1, characterized in that the analysis message is sent in response to a request message sent by the frequency converter.

3. The method as in claim 1 , characterized in that analysis messages are sent to the frequency converter that is a master frequency converter in a set of multiple frequency converters coupled together, and if required, the control unit of the master frequency converter forms speed reference commands for the AC motor connected to it as well as AC motors of follower frequency converters to alter their voltage, rotational speed, and/or torque.

4. The method as in claim 1 , characterized in that analysis messages are sent to the frequency converter that is a follower frequency converter connected to the set of frequency converters, one of which is the master frequency converter, and the follower frequency converter sends a request to the master frequency converter to use the new speed reference command, the new speed reference command is used only when the master frequency converter allows it.

5. The method as in claim 1, characterized by taking pictures with several smart cameras, each of which independently analyzes the pictures taken by it and forms an analysis message.

6. The method as in claim 1 , characterized by calculating a ratio of rejected products to all products in the predetermined task, placing the calculated ratio in the analysis message containing the analysis results, changing the AC motor's speed reference command in the control unit so that a portion of the rejected products is within a desired limit.

7. An electrical drive comprising a frequency converter which converts a supply network's constant input voltage affecting in its inlet connectors to output voltage affecting in its outlet connectors and which has a control circuit for controlling frequency of the output voltage according to a given speed reference command, an AC motor, whose supply voltage is the frequency converter's output voltage, at which time the frequency converter controls the AC motor's rotational speed and thus an industrial process, characterized in that the electrical drive further comprises: a smart camera connected to the frequency converter's control circuit via communication connection, analysis means in the smart camera for analyzing pictures taken by the camera according to a predetermined task and for sending analysis results to the frequency converter's control circuit via the communication connection, means in the frequency converter's control circuit for processing the analysis results received via the communication connection and for forming a new speed reference command as a result of the handling, so that the AC motor's rotational speed and/or torque is altered according to the new speed reference command.

8. The electrical drive as in claim 7, characterized in that the frequency converter's control circuit is configured to send a request to the smart camera and to obtain the analysis results via the communication connection.

9. The electrical drive as in claim 7, characterized in that the smart camera is configured to take continuously pictures of an object moving along in a production process and to separately analyze each picture.

10. The electrical drive as in claim 7, characterized in that the frequency converter is a master frequency converter in a set of multiple frequency converters having a communcation connection with the master frequency converter and, on the basis a new speed reference, it sends speed reference commands also to follower frequency converters.

11. The electrical drive as in claim 7, characterized in that the frequency converter is a follower frequency converter in a set of multiple frequency converters, one of which is a master frequency converter with which the follower frequency converters have a communication connection, and that the follower frequency converter is configured to send to the master frequency converter a request to use the new speed reference command.

12. The electrical drive as in claim 7, characterized in that a plurality of smart cameras are connected to the control circuit of the same frequency conver through the communication connection.

13. The electrical drive as in claim 12, characterized in that the data transfer connection is a common bus.

14. A servo drive comprising a frequency converter having conversion means for converting a supply network's constant voltage to output voltage affecting in its outlet connectors, and a control circuit for controlling frequency of output voltage, an AC motor, whose operational voltage is the frequency converter's output voltage, and which under the frequency converter's control positions through mechanical transmission an actuator handling a product of an industrial process, characterized in that it further comprises: at least one smart camera which has a communication connection with a master frequency converter and takes consecutively pictures of a product moving along in an industrial process, analysis means of a smart camera for analyzing its pixel values by handling each picture according to a predetermined task and a data transfer equipment for sending analysis results via the data transfer connection to the master frequency converter's control circuit, equipment located in the control circuit to process results of a received task and on the basis of it to give a speed reference command to a frequency converter that controls the AC motor using the actuator.

15. The servo drive as in patent claim 14 is characterized in that there are several actuators and the frequency converters controlling respective AC motors are configured to be follower frequency converters having a communication connection to the master frequency converter.

16. The servo drive as in patent claim 14 or 15 is characterized in that the master frequency converter controls the AC motor that brings products available to the actuator

17. The servo drive as in patent claim 14 or 15 is characterized in that there are several smart cameras and they are connected to the master frequency converter via a common data transfer bus.

18. The servo drive as in patent claim 15 is characterized in that the master frequency converter gives to each follower frequency converter an individual speed reference command.