CN104298876A - Partition recovery method for fault of large-scale power distribution network - Google Patents

Abstract

The invention relates to a partition recovery method for fault of a large-scale power distribution network. The partition recovery method includes the steps of (1) partitioning a power failure region by a lap-joint type partitioning manner, repartitioning when node charge of the power distribution network and capacity margin of key branches change so as to realize real-time dynamic partitioning, and adopting the last partitioning scheme when the power distribution network is in fault; (2) calculating the fault recovery with maximum power failure charge through a mathematical model and adopting a recovery scheme with low system loss when the power distribution network is in fault; and (3) recovering the fault of the power distribution network through a multi-agent system according to the partitioning scheme. By the partition recovery method for fault of the large-scale power distribution network, problems about large amount of agency and large amount of agent access information during recovery are solved, the optimal fault recovery scheme is provided for dispatchers in time, and loss of power consumers is lowered.

Description

Large-scale distribution network failure partition recovery method

Technical field

The present invention relates to a kind of distribution network failure restoration methods, especially large-scale distribution network failure partition recovery method.

Background technology

After large-scale power distribution network breaks down, Utilities Electric Co. needs to restore electricity to dead electricity user rapidly, shortens power off time as far as possible, reduces the impact on dead electricity user, to ensure the electricity consumption reliability of power consumer.For large-scale distribution network failure, tradition is joined fault recovery method and mainly be there is following problem:

(1) required time is long, can not meet the requirement of user to electricity consumption reliability;

(2) large-scale Complicated Distribution Network is not suitable for;

(3) search volume is too large, can not form recovery scheme fast;

(4) act on behalf of substantial amounts, there is coordinate synchronization problem when handling failure is recovered.

Summary of the invention

For the deficiencies in the prior art, the object of the present invention is to provide a kind of large-scale distribution network failure partition recovery method, to provide optimum fault recovery scheme in time, reduce the loss of power consumer.

In order to achieve the above object, the invention provides a kind of large-scale distribution network failure partition recovery method, method step is as follows:

(1) bridging type partition method is utilized to carry out a point Division to dead electricity district, when the Capacity Margin of power distribution network node load and crucial branch road changes, repartition region, realize dynamic realtime and divide, when power distribution network breaks down, adopt last partition scheme;

(2) after power distribution network breaks down, by the fault recovery that calculated with mathematical model dead electricity load is maximum, adopt system losses little recovery scheme;

(3) distribution network failure recovery is carried out according to partition scheme and use multi-agent system.

In the present invention, the bridging type partition method in described step (1) has two constraint conditions:

1. crucial tributary capacity nargin and region internal loading will balance;

2. in region, key node number is constrained to 1.

In the present invention, in described step (2), mathematical model is:

$\min f_1=\underset{i\∈N}{\mathrm{\Σ}}{\mathrm{\λ}}_i{L}_{\text{i}}$

In formula, N be 1,2,3 ..., Li is i-th recovery charge, and λ i assignment is for being defaulted as 1, and when load restoration, λ i assignment is 0.

In the present invention, in described step (3), multi-agent system comprises:

1. data collection agent, acquisition node load capacity and tributary capacity nargin data;

2. subregion calculates agency, carries out real time partitioned to electrical network;

3. agency is optimized at center, to the recovery and optimization in non-faulting dead electricity district;

4. evaluate alternatives agency, evaluates all feasible schemes, determines optimized database restore system schema;

5. node load agency, sends to data collection agent by each information on load;

6. distributed power source agency, is sent to subregion and enters to calculate agency by generated energy.

In the present invention, the coordination process in described step (3) between multi-agent system is:

The real-time load data of each node is sent to blackboard by node load agency, data collection agent is to the load data of each node in blackboard request receiving power distribution network, blackboard sends the data to data collection agent by the coordination system, subregion calculates the data acted on behalf of and collect according to data collection agent, adopts above-mentioned bridging type zone method that power distribution network is carried out subregion; According to the partitioning scenario before fault occurs, center is optimized agency and is adopted ant group algorithm to complete can the search of restoration path, search for rear transmission message to act on behalf of to evaluate alternatives, evaluate alternatives agency evaluate recovery scheme according to target and constraint condition recovery scheme, if do not meet, send message and optimize agency till meeting to center.

The invention has the beneficial effects as follows: propose a kind of partition recovery method for large-scale distribution network fault, the method adopts bridging type partition method to carry out real time partitioned to power distribution network each bar branch road before power distribution network breaks down, after occurring when breaking down, employing ant group algorithm completes the search to key node in all subregion, finally draw recovery scheme, arrange function porxy, each agency completes respective task simultaneously, solves and coordinates nonsynchronous problem between agency in the past.

Accompanying drawing explanation

Fig. 1 is the process flow diagram of distribution network failure partition recovery of the present invention.

Fig. 2 is the schematic diagram of bridging type zone method of the present invention.

Fig. 3 coordinates schematic diagram between multi-agent system of the present invention.

Embodiment

Below by embodiment, the invention will be further described by reference to the accompanying drawings.

As shown in Figure 1, 2, the invention provides a kind of large-scale distribution network failure partition recovery method, method step is as follows:

(1) bridging type partition method is utilized to carry out a point Division to dead electricity district, when the Capacity Margin of power distribution network node load and crucial branch road changes, repartition region, realize dynamic realtime and divide, when power distribution network breaks down, adopt last partition scheme;

(2) after power distribution network breaks down, by the fault recovery that calculated with mathematical model dead electricity load is maximum, adopt system losses little recovery scheme;

(3) distribution network failure recovery is carried out according to partition scheme and use multi-agent system.

In the present embodiment, the bridging type partition method in described step (1) has two constraint conditions:

1. crucial tributary capacity nargin and region internal loading will balance;

2. in region, key node number is constrained to 1.

In the present embodiment, in described step (2), mathematical model is:

$\min f_1=\underset{i\∈N}{\mathrm{\Σ}}{\mathrm{\λ}}_i{L}_{\text{i}}$

In formula, N be 1,2,3 ..., Li is i-th recovery charge, and λ i assignment is for being defaulted as 1, and when load restoration, λ i assignment is 0.

In the present embodiment, in described step (3), multi-agent system comprises:

1. data collection agent, acquisition node load capacity and tributary capacity nargin data;

2. subregion calculates agency, carries out real time partitioned to electrical network;

3. agency is optimized at center, to the recovery and optimization in non-faulting dead electricity district;

4. evaluate alternatives agency, evaluates all feasible schemes, determines optimized database restore system schema;

5. node load agency, sends to data collection agent by each information on load;

6. distributed power source agency, is sent to subregion and enters to calculate agency by generated energy.

In the present embodiment, the coordination process in described step (3) between multi-agent system is:

As shown in Figure 3, the real-time load data of each node is sent to blackboard by node load agency, data collection agent is to the load data of each node in blackboard request receiving power distribution network, blackboard sends the data to data collection agent by the coordination system, subregion calculates the data acted on behalf of and collect according to data collection agent, adopts above-mentioned bridging type zone method that power distribution network is carried out subregion; According to the partitioning scenario before fault occurs, center is optimized agency and is adopted ant group algorithm to complete can the search of restoration path, search for rear transmission message to act on behalf of to evaluate alternatives, evaluate alternatives agency evaluate recovery scheme according to target and constraint condition recovery scheme, if do not meet, send message and optimize agency till meeting to center.

Claims (5)

1. large-scale distribution network failure partition recovery method, it is characterized in that, method step is as follows:

(1) bridging type partition method is utilized to carry out a point Division to dead electricity district, when the Capacity Margin of power distribution network node load and crucial branch road changes, repartition region, realize dynamic realtime and divide, when power distribution network breaks down, adopt last partition scheme;

(2) after power distribution network breaks down, by the fault recovery that calculated with mathematical model dead electricity load is maximum, adopt system losses little recovery scheme;

(3) distribution network failure recovery is carried out according to partition scheme and use multi-agent system.

2. large-scale distribution network failure partition recovery method according to claim 1, it is characterized in that, the bridging type partition method in described step (1) has two constraint conditions:

1. crucial tributary capacity nargin and region internal loading will balance;

2. in region, key node number is constrained to 1.

3. large-scale distribution network failure partition recovery method according to claim 1, is characterized in that, in described step (2), mathematical model is:

min $f_1=\underset{i\∈N}{\mathrm{\Σ}}{\mathrm{\λ}}_i{L}_i$

In formula, N be 1,2,3 ..., L _ibe i-th recovery charge, λ _iassignment for being defaulted as 1, the λ when load restoration _iassignment is 0.

4. large-scale distribution network failure partition recovery method according to claim 1, is characterized in that, in described step (3), multi-agent system comprises:

1. data collection agent, acquisition node load capacity and tributary capacity nargin data;

2. subregion calculates agency, carries out real time partitioned to electrical network;

3. agency is optimized at center, to the recovery and optimization in non-faulting dead electricity district;

4. evaluate alternatives agency, evaluates all feasible schemes, determines optimized database restore system schema;

5. node load agency, sends to data collection agent by each information on load;

6. distributed power source agency, is sent to subregion and enters to calculate agency by generated energy.

5. the large-scale distribution network failure partition recovery method according to claim 1,4, is characterized in that, the coordination process in described step (3) between multi-agent system is:

The real-time load data of each node is sent to blackboard by node load agency, data collection agent is to the load data of each node in blackboard request receiving power distribution network, blackboard sends the data to data collection agent by the coordination system, subregion calculates the data acted on behalf of and collect according to data collection agent, adopts above-mentioned bridging type zone method that power distribution network is carried out subregion; According to the partitioning scenario before fault occurs, center is optimized agency and is adopted ant group algorithm to complete can the search of restoration path, search for rear transmission message to act on behalf of to evaluate alternatives, evaluate alternatives agency evaluate recovery scheme according to target and constraint condition recovery scheme, if do not meet, send message and optimize agency till meeting to center.