CN101816822B - Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm - Google Patents

Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm Download PDF

Info

Publication number
CN101816822B
CN101816822B CN2010101841330A CN201010184133A CN101816822B CN 101816822 B CN101816822 B CN 101816822B CN 2010101841330 A CN2010101841330 A CN 2010101841330A CN 201010184133 A CN201010184133 A CN 201010184133A CN 101816822 B CN101816822 B CN 101816822B
Authority
CN
China
Prior art keywords
hrv
variable
error
particle
pid
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN2010101841330A
Other languages
Chinese (zh)
Other versions
CN101816822A (en
Inventor
明东
张广举
邱爽
徐瑞
刘秀云
程龙龙
万柏坤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Yuxi Technology (Tianjin) Co.,Ltd.
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Priority to CN2010101841330A priority Critical patent/CN101816822B/en
Publication of CN101816822A publication Critical patent/CN101816822A/en
Application granted granted Critical
Publication of CN101816822B publication Critical patent/CN101816822B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention relates to the field of instruments for extremity rehabilitation by utilizing electric pulse stimulation and provides a setting method of a functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm. With the setting method, the current strength of an FES (Functional Electrical Stimulation) system can be accurately and stably controlled in real time and the accuracy and the stability of the FES system can be efficiently enhanced. The invention adopts the technical scheme that: firstly, a knee joint angle is predicted by utilizing a handle reaction vector (HRV) in the walk aid process; and secondly, a proportion calculus PID parameter is set by utilizing a chaos particle swarm algorithm, the FES current level strength is regulated in real time, and finally self-adaptive online setting of the proportion calculus PID parameter is realized, and the invention is also used for a functional electrical stimulation FES system. The invention is mainly applied to setting the PID parameter in functional electrical stimulation.

Description

Functional electrostimulation pid parameter double source characteristic fusion particle swarm setting method
Technical field
The present invention relates to carry out the instrument field of limb rehabilitating, especially the double source Feature Fusion chaos particle swarm setting method of pid parameter in the functional electrostimulation with electric pulse stimulation.
Background technology
(Functional Electrical Stimulation is to stimulate limb motion muscle group and peripheral nerve thereof through current pulse sequence FES) to functional electrostimulation, recovers or rebuild the technology of the componental movement function of paralytic patient effectively.According to statistics, because the spinal cord regeneration ability is faint, to the spinal cord injury paralysed patient, the effective treatment method that can directly repair damage is not arranged as yet at present, implementing function rehabilitation training is effective measures.Spinal cord injury paralysed patient number increases year by year, and function rehabilitation training is a technology of demanding demand urgently.The sixties in 20th century, Liberson successfully utilizes the electro photoluminescence nervus peronaeus to correct the gait of hemiplegic patient's drop foot first, has started the new way that functional electrostimulation is used to move and Sensory rehabilitation is treated.At present, FES has become the componental movement function of recovering or rebuilding paralytic patient, is important rehabilitation means.Yet how accurate triggering sequential and the pulse current intensity of controlling FES can accurately be accomplished the key problem in technology that the intended function action is still FES with assurance electro photoluminescence action effect.According to statistics; The mode of the triggering of FES control is at present studied still few; And according to action effect and predetermined action deviation; Come adjustment FES stimulus intensity and time sequence parameter automatically with closed-loop control, thereby improved the accuracy and the stability of FES system greatly, but present effectively control method is still among exploring.
Handle retroaction vector (handle reactions vector; HRV) be that in fact the effectiveness that walker offers the patient can be divided into clear and definite independently 3 parts in the process of standing and walking under helping according to walker: sagittal trying hard to recommend into, about to dynamic balance and the power support of upward and downward; This also can be regarded as the patient in fact and keeps the new ideas that the required to external world additional mechanics demand of self normal stand walking proposes; Promptly be that the patient is reduced to centre-point load to the effect of walker is synthetic in the walking process of standing, represent with two mechanics vectors at handle mid point xsect centre of form place respectively, as shown in Figure 1; Vector is at x; Y, the size of making a concerted effort of the durection component on the z axle can characterize the patient respectively and tries hard to recommend into dynamic balance and power support level by what walker obtained.Wherein, the x axle forward that sets of definition coordinate system is patient's dextrad, and y axle forward be patient's a forward direction, z axle forward be the patient on to.Like this, the defined formula of HRV also can be written as:
[HRV]=[HRV 1,HRV r] T=[F lx,F ly,F lz,F rx,F ry,F rz] T (1)
At present, the situation when HRV is widely used in supervision patient walks in the electro photoluminescence process prevents that then patient from falling down, and causes the secondary injury.This patent proposes to utilize this parameter prediction knee joint angle, the accurate then levels of current intensity of controlling the FES system, and assurance electro photoluminescence action effect can accurately be accomplished the intended function action, and prevents muscular fatigue.
Ratio infinitesimal analysis (proportional-integral-differential; PID) be a kind of very practical feedback regulation algorithm; It detects according to system or the operation deviation; Proportion of utilization, integration, the required regulated quantity of acquisition of differentiating are widely used in engineering practice so that system is carried out FEEDBACK CONTROL because of it is easy to operate.Especially indeterminate or when being difficult to timely on-line determination, safe closed-loop control can be adopted the PID setting algorithm when the controlled system characterisitic parameter.In the face of the complicacy and the time variation operating environment of muscle, because good stability, the reliable operation of PID have still obtained in the functional electrostimulation field using widely at present.The PID core technology is accurate confirm wherein ratio, integration, differential coefficient, especially in the FES field, system stability is required very strictness, so select particularly important to pid parameter.PID control will obtain controls effect preferably, must adjust ratio, integration and three kinds of control actions of differential, forms in the controlled quentity controlled variable not only to cooperatively interact but also the relation of mutual restriction.
Summary of the invention
For overcoming the deficiency of prior art; The double source Feature Fusion chaos particle swarm setting method of pid parameter in a kind of functional electrostimulation is provided; Can accurately stablize and control systematically strength of current of FES in real time; Improve FES system accuracy and stability effectively, and obtain considerable social benefit and economic benefit.For achieving the above object, the technical scheme that the present invention adopts is: the double source Feature Fusion chaos particle swarm setting method of pid parameter in the functional electrostimulation comprises:
At first, utilize the handle retroaction vector HRV forecasting knee joint angle of walk help process;
Secondly, utilize the chaos particle swarm optimization ratio infinitesimal analysis pid parameter of adjusting, real-time monitoring FES levels of current intensity, the flow process of adjusting is: at first according to three decision variable K of ratio infinitesimal analysis PID p, K iAnd K dThe bound of span; Confirm parameters such as population population size, search volume dimension, and the speed and the position of initialization particle colony, the fitness value that calculates each particle in the population through the corresponding relation of actual joint angles and muscle model output joint angles as appropriate evaluation function utilized then; And its fitness and optimum position fitness value itself made comparisons; And with it as the particle typical value, then in adjustment particle's velocity and other parameters, change the optimum position of particle; Till stable, calculate the K that final best position promptly gets ratio infinitesimal analysis PID p, K iAnd K dThree coefficients; Computing system output yout under the new ratio infinitesimal analysis PID coefficient and with the deviation of muscle model output joint angles after get into the self study and the weighting coefficient self-adjusting of next step neural network again; This process repeatedly; The self-adaptation on-line tuning of final realization ratio infinitesimal analysis pid control parameter, and be used for functional electrostimulation FES system.
Said muscle model output joint angles is the method that adopts PLS, that is:
Be provided with m HRV variable HRV1 ..., HRVm, p M variable, M1 ..., Mp, common i (i=1 ..., the n) data set of individual observed reading, T, U are respectively the composition that from HRV variable and M variable, extracts, and are called the offset minimum binary factor,
Concentrate the linear combination of extracting first couple of composition T1, U1 to be from original variable:
T 1=ω 11HRV 1+…+ω 1mHRV m=ω 1′HRV (4)
U 1=v 11M 1+…+v 1pM p=v 1′M (5)
ω wherein 1=(ω 11..., ω 1m) ' be model effect weight, v 1=(v 11..., v 1p) ' be M variable weight is converted into the requirement of said extracted first composition and asks constrained extremal problem:
Figure GDA0000021781110000021
Wherein t1, u1 are the score vector of first pair of composition of being tried to achieve by sample, and HRV0, M0 are initializaing variable, utilize method of Lagrange multipliers, and the problems referred to above are converted into asks vector of unit length ω 1And v 1, make θ 11' HRV 0' M 0V 1Maximum is promptly asked matrix H RV 0' M 0M 0' HRV 0Eigenwert and proper vector, its eigenvalue of maximum is θ 1 2, corresponding unit character vector is exactly the ω that separates that is asked 1, and v 1By formula
Figure GDA0000021781110000022
Obtain;
Next sets up the equation of initializaing variable to T1
HRV 0 = t 1 α 1 ′ + E 1 M 0 = t 1 β 1 ′ + F 1 - - - ( 7 )
T wherein 1Meaning is the same, α 1'=(α 11..., α 1m), β 1'=(β 11..., β 1p) be the parameter vector when only a M measures t1, E1, F1 are respectively n * m and n * p residual error battle array, can try to achieve coefficient vector α according to common least square method 1And β 1, α wherein 1Become model effect load capacity;
Can not reach the precision of regression model like first composition that extracts, utilization residual error battle array E1, F1 replace X0, Y0, repeat to extract composition, and the like, supposing finally to have extracted r composition, HRV0, M0 to the regression equation of r composition are:
HRV 0 = t 1 α 1 ′ + . . . + t r α r ′ + E r M 0 = t 1 β 1 ′ + . . . + t r β r ′ + F r - - - ( 8 )
The first step analyze extract in the gained HRV amount composition Tk (k=1 ..., r) regression equation that the M amount is set up r composition, i.e. t are brought in linear combination into rK1HRV 1+ ... + ω KmHRV mSubstitution M j=t 1β 1j+ ... + t rβ Rj(j=1 ..., p), promptly get the regression equation M of standardized variable jJ1HRV 1+ ... + α JmHRV m
At last according to formula L=M * HRV -1, can obtain L, M representes knee joint angle, and HRV representes that the user is applied to the handle retroaction vector of power on the walker, and L representes the relation between HRV and the M.
The said chaos particle swarm optimization tuning PID parameter of utilizing further is refined as:
Ratio infinitesimal analysis PID adopts ratio unit P, integral unit I and differentiation element D three parts to form, according to the error of system, through the K that sets p, K iAnd K dThree parameters are controlled system:
yout ( t ) = K p error ( t ) + K i Σ j = 0 t error ( j ) + K d [ error ( t ) - error ( t - 1 ) ] - - - ( 9 )
K wherein pBe scale-up factor, K iBe integral coefficient, K dBe differential coefficient, error is the deviation of preset output with actual output, and u (t) is the output of PID, is again the input of controlled system simultaneously, by PID output formula Yout ( t ) = K p Error ( t ) + K i Σ j = 0 t Error ( j ) + K d [ Error ( t ) - Error ( t - 1 ) ] Can obtain
u ( t - 1 ) = K p error ( t - 1 ) + K i Σ j = 0 t - 1 error ( j ) + K d [ error ( t - 1 ) - error ( t - 2 ) ] - - - ( 10 )
According to:
Δu(t)=u(t)-u(t-1)
=K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))
…………………………………………………………… (11)
Have:
u(t)=Δu(t)+u(t-1)=
u(t-1)+K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))
………………(12)
Adopt the chaos particle swarm optimization to carry out the adaptive optimization of ratio infinitesimal analysis pid control parameter; Selecting to receive the rope space is 3 dimensions, promptly is respectively three parameters of PID controller, chooses population size m=20; The initial velocity of colony and position produce in certain spatial dimension at random, are expressed as respectively: v i=(v I1, v I2, v I3), x i=(x I1, x I2, x I3), remember that it is p that i particle searches optimal location so far i=(p I1, p I2, p I3), whole population searches to such an extent that optimal location is p up to now Gi=(p Gi1, p Gi2, p Gi3), wherein, i=1,2 ..., 20, particle swarm optimization algorithm adopts following formula that population is operated,
v id←v id+c 1r 1(p id-x id)+c 2r 2(p gid-x id)+c 3r 3(q id-x id) (13)
x id←x id+v id (14)
Wherein, i=1,2 ..., 20; Study factor c 1, c 2And c 3Be nonnegative number, general value is 0.5; r 1, r 2And r 3Be the random number between [0,1], q IdIt is the picked at random particle position;
The concrete steps that realize are:
1, confirms parameter: study factor c 1, c 2And c 3And the scale N of colony, evolution number of times and chaos optimizing number of times;
2, producing N particle at random operates;
3, by formula operate particle (13) and (14);
4, to optimal location p Gi=(p Gi1, p Gi2, p Gi3) carry out chaos optimization, with p Gid(i=1,2 ..., 20), be mapped to
Logistic equation z I+1=μ z i(1-z i) i=0,1,2 ... Field of definition [0,1]; I=1,2 ..., 20, then, carry out iteration with the Logistic equation and produce Chaos Variable
Figure GDA0000021781110000042
M=1,2 ... Again the Chaos Variable sequence that produces
Figure GDA0000021781110000043
(m=1,2 ...) through inverse mapping
Figure GDA0000021781110000044
(m=1,2 ...) turn back to former solution space, get
Figure GDA0000021781110000045
(m=1,2 ...)
In former solution space each feasible solution to the Chaos Variable experience
Figure GDA0000021781110000046
(m=1,2 ...) calculate its adaptive value, the feasible solution p that retention property is best *
5, a particle of from current colony, selecting is at random used p *Replace;
6, if reach maximum algebraically or obtain satisfactory solution, then optimizing process finishes, otherwise returns step 3.
Characteristics of the present invention are: utilize the HRV variation prediction knee joint angle of walk helper to change; Optimize scale-up factor, differential coefficient and the integral coefficient of PID then through the chaos particle cluster algorithm; Then control the current impulse intensity of FES system, improved FES system accuracy and stability effectively.
Description of drawings
Fig. 1 handle retroaction vector (HRV) definition synoptic diagram.
Fig. 2 is based on the FES system architecture diagram of HRV.
The structured flowchart of Fig. 3 chaos particle cluster algorithm tuning PID parameter control method.
Manikin in Fig. 4 walk-aiding functional electric stimulation.
Fig. 5 experiment scene.
The result is followed the trail of in the PID control that Fig. 6 chaos particle cluster algorithm is adjusted.
Input pass angle and the actual relative error of exporting are preset in the control of Fig. 7 chaos particle cluster algorithm tuning PID parameter down.
Embodiment
Structure based on the control of the precision in the functional electrostimulation walk help of HRV The application of new technique is as shown in Figure 2; Its workflow is: at first, utilize the HRV forecasting knee joint angle of walk help process, secondly; Utilize chaos particle swarm optimization tuning PID parameter, real-time monitoring FES levels of current intensity.Its structural representation of adjusting is as shown in Figure 3, for: at first according to three decision variable K of PID p, K iAnd K dThe bound of span; Confirm parameters such as population population size, search volume dimension, and the speed and the position of initialization particle colony, the fitness value that calculates each particle in the population through the corresponding relation of actual joint angles and muscle model output joint angles as appropriate evaluation function utilized then; And its fitness and optimum position fitness value itself made comparisons; And with it as the particle typical value, then in other parameters such as adjustment particle's velocity etc., change the optimum position of particle; Till stable, calculate the K that final best position promptly gets PID p, K iAnd K dThree coefficients.Computing system output yout under the new PID coefficient and with the deviation of muscle model after get into the self study and the weighting coefficient self-adjusting of next step neural network again.This process finally realizes the self-adaptation on-line tuning of pid control parameter repeatedly, and is used for the FES system.
One, HRV forecasting knee joint angle model
In the walk help process, when the user under the functional electrostimulation effect, when lifting leg and taking a step; In order to support body steadiness, user's applied force on walker is then different, because varying in size of joint can make the gravity center of human body be in diverse location; It is also different then to overcome the gravity applied force; The residing planimetric position of human body also changes to some extent simultaneously, and applied force also changes to some extent for the position is tumbled then in the plane, therefore; Joint angles and user have certain relation to the walker applied force, and be as shown in Figure 4.
M=L·HRV+wPW (1)
Wherein, M representes knee joint angle, and HRV representes that the user is applied to the handle retroaction vector of power on the walker, and L representes the relation between HRV and the M, and w representes coefficient, and W representes the center of gravity of upper arm, trunk and lower limb, and P representes the relation between three centers of gravity and the M.
In the reality, because the effect of walker, the gravity center of human body moves less, and knee joint angle then can be expressed as
M=L·HRV (2)
Wherein, M representes knee joint angle, and HRV representes that the user is applied to the handle retroaction vector of power on the walker, and L representes the relation between HRV and the M.Shown in formula 2, confirm that L just can utilize HRV to take out the knee joint angle in the corresponding moment.
L=M□HRV -1 (3)
When this patent is found the solution L, adopted the method for PLS.
Be provided with m HRV variable HRV1 ..., HRVm, p M variable, M1 ..., Mp, common i (i=1 ..., the n) data set of individual observed reading.T, U are respectively the composition that from HRV variable and M variable, extracts, and the composition that extracts here is commonly referred to the offset minimum binary factor.
Concentrate the linear combination of extracting first couple of composition T1, U1 to be from original variable:
T 1=ω 11HRV 1+…+ω 1mHRV m=ω 1′?HRV (4)
U 1=v 11M 1+…+v 1pM p=v 1′M (5)
ω wherein 1=(ω 11..., ω 1m) ' be model effect weight, v 1=(v 11..., v 1p) be M variable weight.For guaranteeing that T1, U1 extract the variation information of place set of variables separately as much as possible; Guarantee that simultaneously degree of correlation between the two reaches maximum; According to the character that the covariance of composition can be calculated by the inner product of the score vector of corresponding composition, the requirement of said extracted first composition is converted into asks conditional extremum to ask.
Figure GDA0000021781110000061
Wherein t1, u1 are the score vector of first pair of composition of being tried to achieve by sample, and HRV0, M0 are initializaing variable.Utilize method of Lagrange multipliers, the problems referred to above are converted into asks vector of unit length ω 1And v 1, make θ 11HRV 0' M 0v 1Maximum is promptly asked matrix H RV 0' M 0M 0' HRV 0Eigenwert and proper vector, its eigenvalue of maximum is θ 1 2, corresponding unit character vector is exactly the ω that separates that is asked 1, and v 1By formula
Figure GDA0000021781110000062
Obtain.
Next sets up the equation of initializaing variable to T1
HRV 0 = t 1 α 1 ′ + E 1 M 0 = t 1 β 1 ′ + F 1 - - - ( 7 )
Wherein the t1 meaning is the same, α 1'=(α 11..., α 1m), β 1'=(β 11..., β 1p) be the parameter vector when only a M measures t1, E1, F1 are respectively n * m and n * p residual error battle array.Can try to achieve coefficient vector α according to common least square method 1And β 1, α wherein 1Become model effect load capacity.
Can not reach the precision of regression model like first composition that extracts, utilization residual error battle array E1, F1 replace X0, Y0, repeat to extract composition, and the like.Suppose finally to have extracted r composition, HRV0, M0 to the regression equation of r composition are:
HRV 0 = t 1 α 1 ′ + . . . + t r α r ′ + E r M 0 = t 1 β 1 ′ + . . . + t r β r ′ + F r - - - ( 8 )
The first step analyze extract in the gained HRV amount composition Tk (k=1 ..., r) regression equation that the M amount is set up r composition, i.e. t are brought in linear combination into rK1HRV 1+ ... + ω KmHRV mSubstitution M j=t 1β 1j+ ... + t rβ Rj(j=1 ..., p), promptly get the regression equation M of standardized variable jJ1HRV 1+ ... + α JmHRV m
Based on formula 3, can obtain L at last.
Two, chaos particle cluster algorithm tuning PID parameter control
PID is made up of ratio unit P, integral unit I and differentiation element D three parts, according to the error of system, through the K that sets p, K iAnd K dThree parameters are controlled system.
yout ( t ) = K p error ( t ) + K i Σ j = 0 t error ( j ) + K d [ error ( t ) - error ( t - 1 ) ] - - - ( 9 )
K wherein pBe scale-up factor, K iBe integral coefficient, K dBe differential coefficient, error is the deviation of preset output with actual output, and u (t) is the output of PID, is again the input of controlled system simultaneously.
Can obtain by PID output formula (1)
u ( t - 1 ) = K p error ( t - 1 ) + K i Σ j = 0 t - 1 error ( j ) + K d [ error ( t - 1 ) - error ( t - 2 ) ] - - - ( 10 )
According to:
Δu(t)=u(t)-u(t-1)
=K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))
……………………………………………………………(11)
Have:
u(t)=Δu(t)+u(t-1)=
u(t-1)+K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))
………………(12)
The present invention adopts chaos particle swarm algorithm to carry out the adaptive optimization of pid control parameter; Selecting to receive the rope space is 3 dimensions, promptly is respectively three parameters of PID controller, chooses population size m=20; The initial velocity of colony and position produce in certain spatial dimension at random, are expressed as respectively: v i=(v I1, v I2, v I3), x i=(x I1, x I2, x I3), remember that it is p that i particle searches optimal location so far i=(p I1, p I2, p I3), whole population searches to such an extent that optimal location is p up to now Gi=(p Gi1, p Gi2, p Gi3).Wherein, i=1,2 ..., 20, particle swarm optimization algorithm adopts following formula that population is operated.
v id←v id+c 1r 1(p id-x id)+c 2r 2(p gid-x id)+c 3r 3(q id-x id) (13)
x id←x id+v id (14)
Wherein, i=1,2 ..., 20; Study factor c 1, c 2And c 3Be nonnegative number, general value is 0.5; r 1, r 2And r 3Be the random number between [0,1], q IdIt is the picked at random particle position.
The concrete steps of its realization are:
1, confirms parameter: study factor c 1, c 2And c 3And the scale N of colony, evolution number of times and chaos optimizing number of times;
2, producing N particle at random operates;
3, by formula operate particle (13) and (14);
4, to optimal location p Gi=(p Gi1, p Gi2, p Gi3) carry out chaos optimization, with p Gid(i=1,2 ..., 20), be mapped to Logistic equation z I+1=μ z i(1-z i) i=0,1,2 ... Field of definition [0,1];
Figure GDA0000021781110000073
Figure GDA0000021781110000074
I=1,2 ..., 20, then, carry out iteration with the Logistic equation and produce Chaos Variable M=1,2 ..., again the Chaos Variable sequence that produces (m=1,2 ...) through inverse mapping
Figure GDA0000021781110000081
(m=1,2 ...) turn back to former solution space, get
Figure GDA0000021781110000082
(m=1,2 ...)
In former solution space each feasible solution to the Chaos Variable experience
Figure GDA0000021781110000083
(m=1,2 ...) calculate its adaptive value, the feasible solution p that retention property is best *
5, a particle of from current colony, selecting is at random used p *Replace.
6, if reach maximum algebraically or obtain satisfactory solution, then optimizing process finishes, otherwise returns step 3.
Three, experimental program
Experimental provision adopts the walker system of wireless transmission and the Parastep functional electric stimulation system that U.S. SIGMEDICS company produces, and this system comprises microprocessor and boost pulse generation circuit, contains six stimulation channels, powered battery.Experiment content is: utilize the FES system that the relevant muscle group of lower limb is stimulated; Make the experimenter according to predetermined actions, record is applied to HRV on the walker at first through being installed in voltage signal and the knee joint angle movement locus that foil gauge (BX3506AA) network of electrical bridge changes into that lead of 12 on the walker simultaneously.Require the experimenter healthy, no lower limb muscles, bone illness, impassivity illness and severe cardiac pulmonary disease.The experimenter sits idly before walker during experiment, and stimulating electrode is fixed in corresponding position, and when not applying electro photoluminescence, it is light that the experimenter keeps.The FES experiment scene is as shown in Figure 5.The electric stimulation pulse sequence adopts classical Lilly waveform, and pulsed frequency is 25Hz, pulsewidth 150 μ s, and pulse current is adjustable in 0~120m scope.In the experiment, write down HRV in real time and can adjust stimulus intensity to change the knee joint angle that produces by stimulating through changing the pulse current size.Before the experiment, set the knee joint angle movement locus of expectation, utilize the measurement of angle meter to detect the knee joint subtended angle in real time in the experiment and change.The experimental data sampling rate is 128Hz, and the data recording duration is 60s.
Beneficial effect
The new algorithm of chaos particle swarm algorithm tuning PID parameter is calculated the FES pulse current amplitude and is adjusted, the knee joint angle that the FES effect is produced move the movement locus of expection.The result is followed the trail of in the PID control that Fig. 6 adjusts for chaos particle swarm algorithm.Red line representes that desired movement track, blue line are actual output joint angles among the figure.The X axle is the time, and the Y axle is the motion of knee joint angle.For more clearly observing the departure of chaos particle cluster algorithm tuning PID; Shown in the relative error of preset input knee joint angle and actual knee joint angle under Fig. 7 chaos particle swarm Tuning PID Controller; Can find out that then error all within 5%, can reach accurate control.
Purport of the present invention is the precision control method that proposes a kind of new FES; Utilize the error of knee joint angle and the joint angles of actual knee joint angle prediction of the HRV parameter prediction of walker; Through scale-up factor, integral coefficient and the differential coefficient of chaos swarm optimization algorithm PID, the accurately stable then systematically strength of current of FES of controlling in real time.This invention can improve FES system accuracy and stability effectively, and obtains considerable social benefit and economic benefit.Optimum implementation is intended and is adopted patent transfer, technological cooperation or product development.

Claims (2)

1. a functional electrostimulation pid parameter double source characteristic fusion particle swarm setting method is characterized in that, comprises the following steps:
At first, utilize the handle retroaction vector HRV forecasting knee joint angle of walk help process;
Secondly, utilize the chaos particle swarm optimization ratio infinitesimal analysis pid parameter of adjusting, real-time monitoring FES levels of current intensity, the flow process of adjusting is: at first according to three decision variable K of ratio infinitesimal analysis pid parameter p, K iAnd K dThe bound of span; Confirm to comprise the parameter of population population size, search volume dimension; And the speed and the position of initialization particle colony; Utilize then through actual joint angles and muscle model and export the fitness value of the corresponding relation of joint angles as each particle in the appropriate evaluation function calculating population, corresponding relation is L=M * HRV -1, wherein, M representes knee joint angle; HRV representes that the user is applied to the handle retroaction vector of power on the walker, and L representes the relation between HRV and the M, adopts the method for PLS to confirm muscle model output joint angles; And the fitness value and the optimum position fitness value of this particle made comparisons; And as the particle typical value, and then adjustment particle's velocity and other parameters change the optimum position of particle with the optimum position of the fitness value of this particle and this particle itself; Till stable, calculate three decision variable K that final best position promptly gets ratio infinitesimal analysis pid parameter p, K iAnd K d, computing function property electro photoluminescence FES system output yout under the new ratio infinitesimal analysis pid parameter and with the deviation of muscle model output joint angles after get into the self study of next step chaos particle swarm optimization neural network and three decision variable K of pid parameter at that time again p, K iAnd K dThe self-adjusting of weighting coefficient, this process finally realizes the self-adaptation on-line tuning of ratio infinitesimal analysis pid parameter repeatedly, and is used for functional electrostimulation FES system.
2. a kind of functional electrostimulation pid parameter double source characteristic fusion particle swarm setting method according to claim 1 is characterized in that, muscle model output joint angles is the method that adopts PLS, that is:
Be provided with m HRV variable HRV1 ..., HRVm, p M variable, M1 ..., Mp, the data set of common i observed reading, i=1 ..., n; T, U are respectively the composition that from HRV variable and M variable, extracts, and are called the offset minimum binary factor, concentrate the linear combination of extracting first couple of composition T1, U1 to be from original variable:
T 1=ω 11HRV 1+…+ω 1mHRV m=ω′ 1HRV (4)
U 1=v 11M 1+…+v 1pM p=v′ 1M (5)
ω wherein 1=(ω 11..., ω 1m) ' be model effect weight, v 1=(v 11..., v 1p) ' be M variable weight is converted into the requirement of first pair of composition of said extracted and asks constrained extremal problem:
Figure FDA00001724155900011
Wherein t1, u1 are the score vector of first pair of composition of being tried to achieve by sample, and HRV0, M0 are initializaing variable, utilize method of Lagrange multipliers, and the problems referred to above are converted into asks vector of unit length ω 1And v 1, make θ 1=ω ' 1HRV ' 0M ' 0v 1' maximum is promptly asked matrix H RV ' 0M 0M ' 0HRV 0Eigenwert and proper vector, its eigenvalue of maximum is θ 1 2, corresponding unit character vector is exactly the ω that separates that is asked 1, and v 1By formula Obtain;
Next sets up the equation of initializaing variable to T1
Wherein the t1 meaning is the same, α ' 1=(α 11..., α 1m), β ' 1=(β 11..., β 1p) be the parameter vector during a M variable t1 only, E1, F1 are respectively n * m and n * p residual error battle array, try to achieve coefficient vector α according to common least square method 1And β 1, α wherein 1Be called model effect load capacity;
Can not reach the precision of regression model like the first pair of composition that extracts, utilization residual error battle array E1, F1 replace HRV 0, M 0, repeat to extract composition, and the like, supposing finally to have extracted r composition, HRV0, M0 to the regression equation of r composition are:
Figure FDA00001724155900022
The first step analyze extract in the gained HRV amount composition Tk (k=1 ..., r) regression equation that the M variable is set up r composition, i.e. t are brought in linear combination into rK1HRV 1+ ... + ω KmHRV mSubstitution M j=t 1β 1j+ ... T rβ Rj(j=1 ..., p), promptly get the regression equation M of standardized variable jJ1HRV 1+ ... + α JmHRV m
At last according to formula L=M * HRV -1, can obtain L.
3 .A kind of functional electrostimulation pid parameter double source characteristic fusion particle swarm setting method according to claim 1 is characterized in that, utilizes chaos particle swarm optimization tuning PID parameter, further is refined as:
Ratio infinitesimal analysis PID adopts ratio unit P, integral unit I and differentiation element D three parts to form, according to the error of functional electric stimulation system, through the K that sets p, K iAnd K dThree parameters are controlled functional electric stimulation system:
Figure FDA00001724155900023
K wherein pBe scale-up factor, K iBe integral coefficient, K dBe differential coefficient, error is the deviation of preset output with actual output, and u (t) is the output of PID, is again the input of controlled system simultaneously, by PID output formula
Figure FDA00001724155900024
can obtain
Figure FDA00001724155900025
According to:
Δu(t)=u(t)-u(t-1)
=K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))……………………………………………………………(11)
Have:
u(t)=Δu(t)+u(t-1)=
u(t-1)+K p(error(t)-error(t-1))+K ierror(t)+K d(error(t)-2error(t-1)+error(t-2))
………………(12)
Adopt the chaos particle swarm optimization to carry out the adaptive optimization of ratio infinitesimal analysis pid control parameter; Selecting the search volume is 3 dimensions, promptly is respectively three parameters of PID controller, chooses population size m=20; The initial velocity of colony and position produce in certain spatial dimension at random, are expressed as respectively: v i=(v I1, v I2, v I3), x i=(x I1, x I2, x I3), remember that it is p that i particle searches optimal location so far i=(p I1, p I2, p I3), whole population searches to such an extent that optimal location is p up to now Gi=(p Gi1, p Gi2, p Gi3), wherein, i=1,2 ..., 20, particle swarm optimization algorithm adopts following formula population to be operated v Id← v Id+ c 1r 1(p Id-x Id)+c 2r 2(p Gid-x Id)+c 3r 3(q Id-x Id) (13)
x id←x id+v id (14)
Wherein, study factor c 1, c 2And c 3Be nonnegative number, value is 0.5; r 1, r 2And r 3Be the random number between [0,1], q IdBe the picked at random particle position, d representes the coding method of ion, d=10;
The concrete steps that realize are:
1, confirms parameter: study factor c 1, c 2And c 3And the scale N of colony, evolution number of times and chaos optimizing number of times;
2, producing N particle at random operates;
3, by formula operate particle (13) and (14);
4, to optimal location p Gi=(p Gi1, p Gi2, p Gi3) carry out chaos optimization, with p GidBe mapped to Logistic equation z I+1=μ z i(1-z i), z I+1Be the value of Chaos Variable, its field of definition is [0,1], and μ is the Logistic controlled variable, the control chaos state;
Figure FDA00001724155900031
Then, carry out iteration with the Logistic equation and produce Chaos Variable
Figure FDA00001724155900032
M=1,2 ..., again the Chaos Variable sequence that produces
Figure FDA00001724155900033
Through inverse mapping
Figure FDA00001724155900034
Turn back to former solution space,
Figure FDA00001724155900035
a IdEach minimum parameter in the expression pid parameter, b IdEach maximum parameter in the expression pid parameter;
In former solution space each feasible solution to the Chaos Variable experience
Figure FDA00001724155900036
Calculate its adaptive value, the feasible solution p that retention property is best *
5, a particle of from current colony, selecting is at random used p *Replace;
6, if reach maximum iteration time or obtain satisfactory solution, then optimizing process finishes, otherwise returns step 3.
CN2010101841330A 2010-05-27 2010-05-27 Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm Active CN101816822B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010101841330A CN101816822B (en) 2010-05-27 2010-05-27 Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010101841330A CN101816822B (en) 2010-05-27 2010-05-27 Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm

Publications (2)

Publication Number Publication Date
CN101816822A CN101816822A (en) 2010-09-01
CN101816822B true CN101816822B (en) 2012-11-28

Family

ID=42652190

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010101841330A Active CN101816822B (en) 2010-05-27 2010-05-27 Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm

Country Status (1)

Country Link
CN (1) CN101816822B (en)

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
BR112013024491A2 (en) * 2011-03-24 2017-03-21 California Inst Of Techn neurostimulator.
CN103472731B (en) * 2013-09-24 2016-05-25 南方电网科学研究院有限责任公司 The method that the analysis of a kind of micro-electrical network small-signal stability parameter coordination are adjusted
CN104090490B (en) * 2014-07-04 2018-11-02 北京工业大学 A kind of input shaper closed loop control method based on Chaos particle swarm optimization algorithm
CN104834215B (en) * 2015-03-24 2018-02-09 浙江师范大学 A kind of BP neural network pid control algorithm of mutation particle swarm optimization
CN106647247B (en) * 2016-12-29 2019-08-20 西安工程大学 A kind of control algolithm suitable for servo controller
CN109491349B (en) * 2018-12-18 2020-06-23 江南大学 Batch running track and space adjusting method applied to continuous stirring reaction kettle
CN109848990B (en) * 2019-01-28 2022-01-11 南京理工大学 PSO-based knee joint exoskeleton gain variable model-free angle control method
CN110404168B (en) * 2019-09-11 2023-06-13 中山大学 Self-adaptive electric stimulation training system
CN110768558A (en) * 2019-09-24 2020-02-07 山东电工电气集团新能科技有限公司 Inverter midpoint voltage balancing method based on time distribution factor method
CN111166294B (en) * 2020-01-29 2021-09-14 北京交通大学 Automatic sleep apnea detection method and device based on inter-heartbeat period
CN111643321B (en) * 2020-04-30 2023-05-12 北京精密机电控制设备研究所 Exoskeleton joint angle prediction method and system based on sEMG signals
CN113941090B (en) * 2021-09-18 2023-01-17 复旦大学 Self-adaptive closed-loop deep brain stimulation method and device and electronic equipment
WO2023040521A1 (en) * 2021-09-18 2023-03-23 复旦大学 Adaptive closed-loop deep brain stimulation method and apparatus, and electronic device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6171239B1 (en) * 1998-08-17 2001-01-09 Emory University Systems, methods, and devices for controlling external devices by signals derived directly from the nervous system
CN101596338A (en) * 2009-04-29 2009-12-09 天津大学 Functional electric stimulation precision control method based on BP neural network tuned proportion integration differentiation PID

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6171239B1 (en) * 1998-08-17 2001-01-09 Emory University Systems, methods, and devices for controlling external devices by signals derived directly from the nervous system
CN101596338A (en) * 2009-04-29 2009-12-09 天津大学 Functional electric stimulation precision control method based on BP neural network tuned proportion integration differentiation PID

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Abolfazl Jalilvand.Advanced Particle Swarm Optimization-Based PIDController Parameters Tuning.《Proceedings of the 12th IEEE International Multitopic Conference》.2008,430-435. *
Longlong Cheng et al.Radial Basis Function Neural Network-based PID Model Electrical Stimulation System Control.《31st Annual International Conference of the IEEE EMBS Minneapolis》.2009,3481-3484. *
栾丽君等.基于粒子群算法的PID参数自整定.《2006中国控制与决策学术年会论文集》.2006,476-482. *

Also Published As

Publication number Publication date
CN101816822A (en) 2010-09-01

Similar Documents

Publication Publication Date Title
CN101816822B (en) Setting method of functional electrical stimulation PID (Proportion Integration Differentiation) parameter double source characteristic fusion particle swarm
CN101596338A (en) Functional electric stimulation precision control method based on BP neural network tuned proportion integration differentiation PID
CN101794114B (en) Method for tuning control parameter in walk-aiding functional electric stimulation system by utilizing genetic algorithm
Dao From deep learning to transfer learning for the prediction of skeletal muscle forces
Li et al. A two-joint human posture control model with realistic neural delays
CN109589247A (en) It is a kind of based on brain-machine-flesh information loop assistant robot system
CN105615890A (en) Angle and myoelectricity continuous decoding method for human body lower limb walking joint
Makssoud et al. Mathematical muscle model for functional electrical stimulation control strategies
CN101816821B (en) Walking aid functional electrical stimulation precision control method based on ant colony fuzzy controller
CN101846977B (en) Genetic fuzzy control method of joint angles by functional electrical stimulation
CN102274581A (en) Precise control method for functional electric stimulation
CN102488964A (en) Functional electro stimulation closed loop fuzzy proportional integral derivative (PID) control method
CN101837165B (en) Walking aid electrostimulation fine control method based on genetic-ant colony fusion fuzzy controller
CN101837164B (en) Double source feature fusion ant colony tuning method for PID (Proportion Integration Differention) parameter in functional electro-stimulation
Ren et al. Computational models to synthesize human walking
Ibrahim et al. Fuzzy modelling of knee joint with genetic optimization
Li et al. Electrotactile Feedback-Based Muscle Fatigue Alleviation for Hand Manipulation
Huang et al. Human-in-the-loop optimization of knee-joint biomechanical energy harvester to maximize power generation with minimal user effort
Heidlauf et al. An integrated model of the neuromuscular system
Hou et al. Prediction of EMG signals of trunk muscles in manual lifting using a neural network model
Wan et al. Cyclic Anaerobic Exercise Performance and Neuromuscular Activity Based on Artificial Intelligence Genetic Algorithm
Chéron et al. Toward an integrative dynamic recurrent neural network for sensorimotor coordination dynamics
CN105769186A (en) Complex network based electromyographic signal acquisition position selecting method
Ibrahim et al. Development of fuzzy muscle contraction and activation model using multi-objective optimisation
Ibrahim et al. An Approach for Dynamic Characterisation of Passive Viscoelasticity and Estimation of Anthropometric Inertia Parameters of Paraplegic’s Knee Joint

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20210301

Address after: Room 50, room 03, 8 / F, phase I, zhikongjian Plaza, Southeast of the intersection of Huizhi North Road and Huizhi Ring Road, Dongli Lake, Tianjin

Patentee after: Yuxi Technology (Tianjin) Co.,Ltd.

Address before: 300072 Tianjin City, Nankai District Wei Jin Road No. 92

Patentee before: Tianjin University

TR01 Transfer of patent right