WO2018103294A1 - 基于双鱼群算法的电力无功优化系统及方法 - Google Patents

基于双鱼群算法的电力无功优化系统及方法 Download PDF

Info

Publication number
WO2018103294A1
WO2018103294A1 PCT/CN2017/088698 CN2017088698W WO2018103294A1 WO 2018103294 A1 WO2018103294 A1 WO 2018103294A1 CN 2017088698 W CN2017088698 W CN 2017088698W WO 2018103294 A1 WO2018103294 A1 WO 2018103294A1
Authority
WO
WIPO (PCT)
Prior art keywords
fish
reactive power
grid
individual
value
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.)
Ceased
Application number
PCT/CN2017/088698
Other languages
English (en)
French (fr)
Inventor
张化光
杨珺
孙秋野
吴飞业
刘鑫蕊
杨东升
王智良
冯健
黄博南
罗艳红
会国涛
刘振伟
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.)
Northeastern University China
Original Assignee
Northeastern University China
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 Northeastern University China filed Critical Northeastern University China
Priority to US15/577,127 priority Critical patent/US10439398B2/en
Publication of WO2018103294A1 publication Critical patent/WO2018103294A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • H02J3/18—Arrangements for adjusting, eliminating or compensating reactive power in networks
    • H02J3/1885—Arrangements for adjusting, eliminating or compensating reactive power in networks using rotating AC generators, e.g. synchronous generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • H02J3/18—Arrangements for adjusting, eliminating or compensating reactive power in networks
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05F—SYSTEMS FOR REGULATING ELECTRIC OR MAGNETIC VARIABLES
    • G05F1/00—Automatic systems in which deviations of an electric quantity from one or more predetermined values are detected at the output of the system and fed back to a device within the system to restore the detected quantity to its predetermined value or values, i.e. retroactive systems
    • G05F1/10—Regulating voltage or current 
    • G05F1/12—Regulating voltage or current  wherein the variable actually regulated by the final control device is AC
    • G05F1/14—Regulating voltage or current  wherein the variable actually regulated by the final control device is AC using tap transformers or tap changing inductors as final control devices
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • H02J3/12—Arrangements for adjusting voltage in AC networks by changing a characteristic of the network load
    • H02J3/16—Arrangements for adjusting voltage in AC networks by changing a characteristic of the network load by adjustment of reactive power
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for AC mains or AC distribution networks
    • H02J3/18—Arrangements for adjusting, eliminating or compensating reactive power in networks
    • H02J3/1878—Arrangements for adjusting, eliminating or compensating reactive power in networks using tap changing or phase shifting transformers
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—ELECTRIC POWER NETWORKS; CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2103/00—Details of circuit arrangements for mains or AC distribution networks
    • H02J2103/30—Simulating, planning, modelling, reliability check or computer assisted design [CAD] of electric power networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
    • Y02E40/30—Reactive power compensation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
    • Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S40/00—Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them
    • Y04S40/20—Information technology specific aspects, e.g. CAD, simulation, modelling, system security

Definitions

  • the invention belongs to the field of power system reactive power optimization technology, and particularly relates to a power reactive power optimization system and method based on a Pisces group algorithm.
  • Reactive power optimization of power system refers to the determination of the value of some control variables in the system by optimizing the calculation of the system's active load, active power supply and power flow distribution to find the system under the premise that all constraints are met.
  • One or more performance indicators (such as minimum active network loss, optimal voltage quality, and minimum annual expenditure) to achieve optimal operation.
  • the reasonableness of the reactive power distribution of the system is directly related to the safety and stability of the power system, and is closely related to economic benefits.
  • the system's reactive power is insufficient, the voltage level will be low, some factories and household electrical appliances will not operate normally, and once the system is disturbed, the voltage may be lower than the threshold voltage, causing voltage collapse, resulting in loss of the system. A catastrophic accident that is simultaneously disintegrated.
  • the reactive power optimization methods mainly include two kinds of optimization algorithms: conventional mathematical methods and intelligent optimization algorithms.
  • Conventional mathematical methods have faster calculation speeds, but they have higher requirements for the calculation of continuity, non-convexity and differentiability of optimization functions, and it also has the disadvantages of being easy to fall into local optimal solutions.
  • the intelligent optimization algorithm shows a strong ability to optimize in dealing with nonlinear, multivariable, discontinuous, non-convex optimization problems.
  • artificial fish swarm algorithm has been applied to time-varying system online identification, robust PID parameter tuning and optimization of forward neural network, and achieved good results.
  • the algorithm has good ability to obtain global extremum, and has many advantages such as initial value, insensitivity to parameter selection, robustness, simplicity (only using objective function values), and easy implementation.
  • the artificial fish swarm algorithm still has some shortcomings, mainly: the efficiency of the solution is not high, the solution time is increased, the convergence speed will be slowed down, and it is easy to fall into the local optimal solution, and it is difficult to obtain an accurate optimal solution.
  • the present invention provides a power reactive power optimization system and method based on the Pisces group algorithm.
  • An electric power reactive optimization system based on a Pisces group algorithm comprising: a grid state acquisition module, a grid reactive power adjustment module, and a grid reactive execution module;
  • the grid state acquisition module includes a grid state collector and a relay transmitter
  • the grid reactive power adjustment module is a control terminal
  • the grid reactive power execution module includes each generator terminal voltage regulator, each transformer tap adjuster and each reactive power compensation regulator;
  • the input end of the grid state collector is connected to the power grid, the output end of the grid state collector is connected to the input end of the relay transmitter, and the output end of the relay transmitter is connected to the input end of the control terminal, the control terminal
  • the output end is connected to the input end of each motor terminal voltage regulator, the input end of each transformer tap adjuster and the input end of each reactive power compensation regulator, and the output end of the generator terminal voltage regulator is connected to the power grid.
  • a generator the output of the transformer tap adjuster is connected to a transformer in the power grid, and the output of the reactive power compensation regulator is connected to a reactive power compensation device in the power grid;
  • the grid state collector is configured to collect current network information of the power grid, and determine whether the current network information meets an optimal state required by the power grid. If the current network information cannot meet the optimal state required by the power grid, the current network information is transmitted.
  • the current network information includes: grid node information, branch information, generator information, transformer information, and reactive power compensation device information;
  • the relay transmitter is configured to transmit, to the control, the power grid node information, the branch road information, the generator information, the transformer information, and the reactive power compensation device information required for reactive power optimization in the network information collected by the grid state collector. terminal;
  • the grid reactive power adjustment module comprises: a parameter acquisition unit, a reactive power optimization unit based on a Pisces group algorithm, and an optimization decision control unit;
  • the parameter obtaining unit is configured to acquire grid node information, branch information, generator information, transformer information, and reactive power compensation device information transmitted by the relay transmitter, as initial data to be optimized by the current network;
  • the reactive power optimization unit based on the Pisces group algorithm is used to establish a mathematical model of reactive power optimization of the power system, and the optimal data of the current network to be optimized is optimized based on the Pisces group algorithm, and the optimized values of the control variables in the power grid are obtained.
  • the control variables include a generator terminal voltage amplitude, a transformer adjustable ratio, and a reactive power capacity of the reactive power compensation device;
  • the optimization decision control unit is configured to transmit an optimized value of each control variable to a grid reactive execution module
  • the generator terminal voltage regulator is configured to adjust the generator terminal voltage according to the optimized value of the generator terminal voltage obtained by the grid reactive power adjustment module;
  • the transformer tap adjuster is configured to optimize the value adjustment transformer tap according to the transformer adjustable ratio obtained by the grid reactive power adjustment module;
  • the reactive power compensation regulator is configured to adjust the compensation capacity of the reactive power compensator according to the reactive power capacity of the reactive power compensation device obtained by the power grid reactive power adjustment module.
  • F is the objective function of the mathematical model of power reactive power optimization. 1 ⁇ h ⁇ nl, nl is the total number of branches of the grid system, G lv is the conductance of the connecting branch lv, ⁇ v is the phase angle of the node v, ⁇ l is the phase angle of the node l, and ⁇ 1 is the minimum target of the network loss
  • the weight coefficient of the function ⁇ 2 is the weight coefficient of the voltage optimal objective function
  • u l is the voltage of node l
  • u v is the voltage of node v, 1 ⁇ v ⁇ N d
  • N d is the total number of load nodes in the grid system
  • 1 ⁇ k ⁇ N g , N g is the number of generators in the grid
  • V v is the node v voltage
  • For the voltage reference of node v For the voltage deviation maximum of node v, V v.max is the node v voltage upper limit, V v.min is the node v voltage lower limit
  • the specific process of optimizing the current network to be optimized based on the two-fishing group algorithm is obtained by using the above-mentioned two-fishing group algorithm, and the specific process of obtaining the optimized values of the control variables in the power grid is as follows:
  • Initialize the current optimal food concentration value into the bulletin board use the reciprocal 1/F of the objective function of the power reactive optimization mathematical model as the food concentration value FC, and calculate the food concentration value of the current network, each small fish and each fierce fish under the initial data conditions.
  • the maximum value of the food concentration value is taken as the current optimal food concentration value to enter the bulletin board, and the state and the current FC value are saved;
  • Action on small fish populations According to the distance between each small fish individual and the fierce fish individual, and the distance between the small fish individuals of small fish groups, the small fish individuals are subjected to foraging behavior, rear-end behavior and protective gathering. Group behavior, update bulletin board Y1;
  • Acting on fierce fish According to the distance between each fierce fish individual and the small fish individual, and the distance between the ferocious fish individual of the fierce fish group, the predation behavior, tracking behavior and clustering behavior of each fierce fish individual, Update bulletin board Y2;
  • the food concentration of the bulletin board Y1 and Y2 and its state are taken as the optimization results, and the optimized values of the control variables are obtained, that is, the generator terminal voltage amplitude and the transformer adjustable ratio And the optimized value of the reactive capacity of the reactive power compensation device.
  • a power reactive power optimization system based on a Pisces group algorithm for power reactive power optimization includes the following steps:
  • Step 1 Collect the current network information of the power grid through the grid state collector to determine whether the current network information meets the optimal state required by the power grid. If the current network information cannot meet the optimal state required by the power grid, the current network information is transmitted to Relay transmitter
  • Step 2 The grid node information, the branch information, the generator information, the transformer information, and the reactive power compensation device information required for reactive power optimization in the network information collected by the grid state collector are transmitted to the control terminal through the relay transmitter. ;
  • Step 3 obtaining, by the control terminal, the grid node information, the branch information, the generator information, the transformer information, and the reactive power compensation device information transmitted by the relay transmitter, as the initial data to be optimized by the current network;
  • Step 4 Establish a mathematical model of reactive power optimization of the power system through the control terminal, optimize the initial data to be optimized based on the Pisces group algorithm, obtain the optimized values of the control variables in the power grid, and optimize the control variables. The value is transmitted to the grid reactive execution module;
  • Step 5 Each generator terminal voltage regulator optimizes the value of the generator terminal voltage according to the amplitude value of the generator terminal voltage, and each transformer tap adjuster optimizes the value adjustment transformer tap according to the adjustable ratio of the transformer, and each reactive power compensation The regulator adjusts the compensation capacity of the reactive power compensator according to the reactive power capacity of the reactive power compensation device.
  • the step 4 includes the following steps:
  • Step 4.1 Taking the minimum network loss and the optimal voltage level as the optimization objectives, the ⁇ -method is used to establish the mathematical model of reactive power optimization of the power system:
  • Step 4.2 Set the parameters of the Pisces algorithm: small fish size N1, fierce fish size N2, small fish group perception range Visual1, fierce fish group perception range Visual2, small fish group moving step Step1, fierce fish group moving step size Step 2.
  • Small fish population congestion factor ⁇ 1 fierce fish population congestion factor ⁇ 2 , maximum iteration number K>0;
  • Step 4.3 Randomly generate small fish groups according to the upper limit of the values of the control variables to be optimized and the lower limit of the values to be optimized;
  • Step 4.4 Randomly generate a fierce fish group according to the upper limit of the value of each control variable to be optimized and the lower limit of the value of the control variable to be optimized;
  • Step 4.6 Action on small fish stocks: According to the distance between each small fish individual and the fierce fish individual, and the distance between the small fish individuals of small fish groups, the small fish individuals are subjected to foraging behavior, rear-end behavior and Protection clustering behavior, update bulletin board Y1;
  • Step 4.7 Acting on fierce fish: According to the distance between each fierce fish and the small fish individual, and the distance between the ferocious fish individual of the fierce fish, predation behavior, tracking behavior and cluster behavior of each fierce fish individual , update the bulletin board Y2;
  • Step 4.9 Take the maximum food concentration of the bulletin board Y1 and Y2 and its state as the optimization result, and obtain the optimized value of each control variable, that is, the generator terminal voltage amplitude, the transformer adjustable ratio and the reactive power compensation device reactive power Optimized value of capacity;
  • Step 4.10 Transfer the optimized values of each control variable to the grid reactive execution module.
  • the step 4.3 includes the following steps:
  • Step 4.3.2 randomly generate small fish individuals according to the range of values of the control variables to be optimized by the grid:
  • X is the sequence of control variables for small fish individuals
  • x s x s min + rand(g) ⁇ (x s max -x s min )
  • rand(g) is a random number in the interval (0, 1)
  • x s min, x s max respectively correspond to the upper limit value and the lower limit value of the control variable
  • U G1g, K T1t, Q C1c generator terminal voltage magnitude are small artificial fish in individual randomly generated
  • the transformer can be The modulation ratio and the reactive capacity of the reactive power compensation device
  • x s is the control variable of the randomly generated small fish individual
  • S N g + N t + N c
  • N g is the generator in the grid Number
  • N t is the number of transformers in the power grid
  • N c is the number of reactive power compensation devices in the power grid;
  • Step 4.3.4 If t 1 ⁇ N1, the current small fish population is obtained, and step 4.4 is performed; otherwise, return to step 4.3.2;
  • the step 4.4 includes the following steps:
  • Step 4.4.2 randomly generate fierce fish individuals according to the range of values of the control variables to be optimized by the grid:
  • the generator terminal voltage amplitude, the transformer adjustable ratio and the reactive capacity of the reactive power compensation device in the individual fish, w s is the control variable of the randomly generated fierce fish individual;
  • Step 4.4.4 If t 2 ⁇ N2, then obtain the current fierce fish group, perform step 4.5, otherwise, return to step 4.4.2;
  • the step 4.6 includes the following steps:
  • Step 4.6.2 If the distance between the current i-th small fish individual X i and each fierce fish individual W j D ij ⁇ Visual1, perform step 4.6.3, otherwise, perform step 4.6.5;
  • Step 4.6.3 determining the total number n of small individual fish within the i-th small sensing area X i fish individuals, in order to determine the aggregate security guard position X safe, i-th individual fish small clusters X i for protection behavior, to give the i A small fish individual carries out the protective clustering behavior update value X inext1 ;
  • X safe (X i1 +X i2 +...+X in ) ⁇ /n;
  • p is the total number of fierce fish in the small fish individual X i sensing range Visual1
  • X in is the small fish individual within the sensing range of the small fish individual X i ;
  • the calculation formula of the protection clustering behavior update value X inext1 of the i-th small fish individual is as follows:
  • X inext1 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X safe -X i )/
  • Step 4.6.4 Determine whether the current small fish individual performs the protection clustering update value X inext1 to satisfy its constraint condition and the power flow converges. If yes, record the state of the updated value X inext1 , calculate the food concentration, and update the bulletin board Y1. , go to step 4.7, otherwise, return to step 4.6.3;
  • Step 4.6.5 determining the i-th individual fish small sensing area X i fish food in small concentrations FC j 'is small, the maximum individual fish X j', i-th small fish individuals following behavior for X i, to give The i-th small fish individual X i performs rear-end behavior update value X inext2 ;
  • X inext2 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X j' -X i )/
  • Step 4.6.6 Determine the total number n of small fish individuals within the i i sensing range of the i-th small fish individual, thereby determining the position X c1 of the foraging group, and conducting the foraging phenomenon of the i-th small fish individual X i .
  • the i-th small fish individual was given the update value of foraging group behavior X inext3 ;
  • X c1 (X i1 +X i2 +...+X in )/n;
  • X inext3 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X c1 -X i )/
  • Step 4.6.7 Determine whether the current small fish individual performs the rear-end behavior update value X inext2 and the small fish individual to perform the foraging behavior update value X inext3 to satisfy the constraint condition and the power flow convergence, and if so, record the update value X inext2 The status and update value of X inext3 state, and calculate its food concentration, update bulletin board Y1, perform step 4.7, otherwise, return to step 4.6.5.
  • the step 4.7 includes the following steps:
  • Step 4.7.2 If the current distance between the i'th fierce artificial fish individual W i ' and each small artificial fish individual X j′ ′ L i′j′′ ⁇ Visual 2, perform step 4.7.3, otherwise, perform step 4.7. 6;
  • Step 4.7.3 determining a current of the i 'th individual fish ferocious W is i' perception of the total number of individual small fish in the range of m, of the current i 'th individual fish ferocious W is i' perception within the range of a small fish as the current center position X c2
  • X c2 (X i'1 +X i'2 +...+X i'm )/m;
  • Xi'm is a small fish within the range of perception of the ferocious fish individual W i ' ;
  • Step 4.7.4 the current of the i 'th individual fish ferocious W is i' perception within the range of the center position X c2 small fish food as this maximum concentration, the first i 'th individual fish ferocious W is i' for tracking behavior, to give The updated value of the tracking behavior of the i'th fierce fish individual W i ' W i'next2 ;
  • W i'next2 W 'i + rand(g) ⁇ Step2 ⁇ ⁇ 2 ⁇ (X c2 - W i' ) /
  • Step 4.7.5 Analyzing this ferocious gudgeon W i 'is updated value W i'next1 ferocious fish and predatory behavior of individual W i' is updated to track the behavior of the value W i'next2 which satisfy the constraints and flow convergence, If yes, update bulletin board Y2, perform step 4.7.6, otherwise, return to step 4.7.3;
  • Step 4.7.6 Determine the total number r of fierce fish individuals within the i i fierce fish individual W i′ , thereby determining the cluster center position W c and clustering the i'th fierce fish individual W i′ Obtaining an updated value W i'next3 of the clustering behavior of the i'th fierce fish individual W i ' ;
  • W c (W i'1 +W i'2 +...+W i'r )/r;
  • W i'r is the fierce fish individual within the sensing range of the i'th fierce fish individual W i ' ;
  • W i'next1 W i' + rand(g) ⁇ Step2 ⁇ ⁇ 2 ⁇ (W c - W i' ) /
  • Step 4.7.7 Determine whether the current fierce fish individual W i' performs an update value of the clustering behavior W i'next3 satisfies its constraint condition and the power flow converges, and if so, records the state of the updated value W i'next3 and calculates its Food concentration, update bulletin board Y2, perform step 4.8, otherwise, return to step 4.7.6.
  • the invention provides a power reactive power optimization system and method based on the Pisces group algorithm.
  • the invention performs modular processing on the system, is easy to control and implements a reactive power optimization scheme, and introduces a new fish group based on the reactive power optimization method of the Pisces group-
  • the reactive power optimization ability is to optimize the distribution network to get a more reasonable reactive power flow distribution.
  • FIG. 2 is a structural block diagram of a power reactive power optimization system based on a Pisces group algorithm according to an embodiment of the present invention
  • FIG. 3 is a flowchart of a power reactive power optimization method based on a Pisces group algorithm according to an embodiment of the present invention
  • FIG. 4 is a flow chart of a process for optimizing initial data to be optimized based on a squid group algorithm according to a specific embodiment of the present invention, and obtaining optimized values of control variables in the power grid.
  • the simulation is performed using the IEEE 10-node system as shown in FIG. 1.
  • the system is a closed loop system with 6 lines, 3 generator nodes, 3 load nodes, and 3 on-load tap-changers.
  • node 1 acts as the balance node
  • the other two generator nodes are PV nodes.
  • the other nodes in the system are PQ nodes, two reactive compensation nodes (nodes are 5, 10), and the adjustable transformer branch It is 1-4, 2-8, 3-6, 4-10.
  • the voltage rating is 220kV and the total system load is 315+j255MVA.
  • a power reactive power optimization system based on the Pisces group algorithm includes: a grid state acquisition module, a grid reactive power adjustment module, and a grid reactive power execution module.
  • the grid state acquisition module includes a grid state collector and a relay transmitter.
  • the grid reactive power adjustment module is a control terminal.
  • the grid reactive power execution module includes each generator terminal voltage regulator, each transformer tap adjuster and each reactive power compensation regulator.
  • the input end of the grid state collector is connected to the grid
  • the output of the grid state collector is connected to the input end of the relay transmitter
  • the output end of the relay transmitter is connected to the input end of the control terminal
  • the output end of the control terminal is connected to each motor end.
  • the input of the voltage regulator, the input of each transformer tap adjuster and the input of each reactive compensation regulator, the output of the generator terminal voltage regulator is connected to the generator in the grid
  • the output of the transformer tap adjuster The terminal is connected to the transformer in the power grid, and the output of the reactive power compensation regulator is connected to the reactive power compensation device in the power grid.
  • the model of the power grid state collector is DCZL23-CL156N, which is used to collect current network information of the power grid, and determine whether the current network information meets the optimal state required by the power grid. If the current network information cannot meet the most demanded by the power grid, The optimal state transmits the current network information to the relay transmitter, and the current network information includes: grid node information, branch information, generator information, transformer information, and reactive power compensation device information.
  • the transformer information and the corresponding branch information obtained according to the IEEE 10-node system of FIG. 1 are as shown in Tables 1 and 2, the load parameters are as shown in Table 3, and the branch parameters are as shown in Table 4.
  • R is a resistance and X is a reactance.
  • Table 2 Three winding transformer information and corresponding branch information
  • P load is the active power of the load
  • Q load is the reactive power of the load
  • the model of the relay transmitter is RD980, which is used to perform grid node information, branch information, generator information, transformer information, and none required for reactive power optimization in the network information collected by the grid state collector.
  • the power compensation device information is transmitted to the control terminal.
  • Grid reactive power adjustment module including parameter acquisition unit, reactive power optimization unit based on Pisces group algorithm, and optimization decision control unit.
  • the parameter obtaining unit is configured to obtain grid node information, branch information, generator information, transformer information, and reactive power compensation device information transmitted by the relay transmitter, as the initial data to be optimized of the current network.
  • the Pisces Group algorithm reactive power optimization unit Based on the Pisces Group algorithm reactive power optimization unit, it is used to establish the mathematical model of power system reactive power optimization.
  • the Pisces group algorithm is used to optimize the current network to be optimized initial data, and the optimized values of the control variables in the grid are obtained.
  • the control variables include the generator terminal voltage amplitude, the transformer adjustable ratio, and the reactive capacity of the reactive power compensation device.
  • the minimum network loss and the voltage level are optimized as the optimization target, and the mathematical model of the reactive power optimization of the power system is established by using the ⁇ -method in the form of a penalty function as shown in the formula (1):
  • F is the objective function of the mathematical model of power reactive power optimization. 1 ⁇ h ⁇ nl, nl is the total number of branches of the grid system, G lv is the conductance of the connecting branch lv, ⁇ v is the phase angle of the node v, ⁇ l is the phase angle of the node l, and ⁇ 1 is the minimum target of the network loss
  • the weight coefficient of the function ⁇ 2 is the weight coefficient of the voltage optimal objective function
  • u l is the voltage of node l
  • u v is the voltage of node v, 1 ⁇ v ⁇ N d
  • N d is the total number of load nodes in the grid system
  • 1 ⁇ k ⁇ N g , N g is the number of generators in the grid
  • V v is the node v voltage
  • For the voltage reference of node v For the voltage deviation maximum of node v, V v.max is the node v voltage upper limit, V v.min is the node v voltage lower limit
  • the current data to be optimized for the current network to be optimized is optimized based on the Pisces group algorithm, and the specific process of obtaining the optimized values of the control variables in the power grid is as follows:
  • control variables include: the generator terminal voltage amplitude, the transformer adjustable ratio, and the reactive capacity of the reactive power compensation device;
  • the fierce fish population is randomly generated according to the upper limit of the value of each control variable to be optimized and the lower limit of the value of the control variable to be optimized;
  • Initialize the current optimal food concentration value into the bulletin board use the reciprocal of the objective function of the power reactive power optimization mathematical model as the food concentration value FC, calculate the food concentration value under the initial data condition of the current network to be optimized, and the individual artificial fish under the individual state
  • the food concentration value, the food concentration value of each fierce artificial fish individual state take the maximum value of the food concentration value as the current optimal food concentration value into the bulletin board, save its state and the current optimal food concentration value FC;
  • the calculation formula of the food concentration value FC is as shown in the formula (2):
  • Acting on fierce fish based on the distance between each fierce fish and small fish, and the distance between the fierce fish individual and the individual fierce fish, predation behavior, tracking behavior and cluster behavior of each ferocious artificial fish, updated Y2;
  • the food concentration of the bulletin board Y1 and Y2 is taken as the optimization result, and the optimized value of each control variable is obtained, that is, the generator terminal voltage amplitude, the transformer adjustable ratio, and the reactive power compensation.
  • the optimum value of the reactive capacity of the device is obtained.
  • the optimization decision control unit is configured to transmit the optimized value of each control variable to the grid reactive execution module.
  • the model of the generator terminal voltage regulator is EQ1512, which is used for adjusting the generator terminal voltage according to the optimized value of the generator terminal voltage obtained by the grid reactive power adjustment module.
  • the transformer tap adjuster model is TSGC2J, which is used to optimize the value adjustment transformer tap according to the transformer adjustable ratio obtained by the grid reactive power adjustment module.
  • the reactive power compensation regulator model is GLSSC, and is used for adjusting the compensation capacity of the reactive power compensator according to the reactive power capacity of the reactive power compensation device obtained by the power grid reactive power adjustment module.
  • a power reactive power optimization system based on a Pisces group algorithm for power reactive power optimization, as shown in FIG. 3, includes the following steps:
  • Step 1 Collect the current network information of the power grid through the grid state collector to determine whether the current network information meets the optimal state required by the power grid. If the current network information cannot meet the optimal state required by the power grid, the current network information is transmitted to Relay transmitter.
  • Step 2 The grid node information, the branch information, the generator information, the transformer information, and the reactive power compensation device information required for reactive power optimization in the network information collected by the grid state collector are transmitted to the control terminal through the relay transmitter. .
  • Step 3 Obtain the grid node information, the branch information, the generator information, the transformer information, and the reactive power compensation device information transmitted by the relay transmitter through the control terminal, as the initial data to be optimized of the current network.
  • the initial data to be optimized of the current network is node information, branch information, the number of generators, the upper and lower limits of the voltage amplitude U G of each generator terminal, the number of transformers, and the adjustable ratio K t of each transformer.
  • Step 4 Establish a mathematical model of reactive power optimization of the power system through the control terminal, using the Pisces group algorithm based on the acquired The current network to be optimized initial data is optimized, and the optimized values of the control variables in the grid are obtained, and the optimized values of the control variables are transmitted to the grid reactive execution module, as shown in FIG. 4 .
  • Step 4.1 Taking the minimum network loss and the optimal voltage level as the optimization objectives, the mathematical model of reactive power optimization of the power system is established by the ⁇ -method in the weighted sum method.
  • Step 4.2 Set the parameters of the Pisces algorithm: small fish size N1, fierce fish size N2, small fish group perception range Visual1, fierce fish group perception range Visual2, small fish group moving step Step1, fierce fish group moving step size Step 2.
  • Small fish population congestion factor ⁇ 1 fierce fish population congestion factor ⁇ 2 , small fish population attempt number try_number1, fierce fish group trial number try_number2, maximum iteration number K>0.
  • the maximum number of iterations K is 100.
  • Step 4.3 randomly generate small fish groups according to the upper limit of the values of the control variables to be optimized and the lower limit of the values.
  • Step 4.3.2 Randomly generate small artificial fish according to the upper limit and lower limit of each control variable to be optimized by the grid, as shown in equation (3):
  • X is the sequence of control variables for small artificial fish
  • x s x s min +rand(g) ⁇ (x s max -x s min )
  • rand(g) is a random interval of (0,1)
  • x s min is the lower limit of the corresponding control variable
  • x s max is the upper limit of the corresponding control variable
  • U G1g is the amplitude of the generator terminal voltage in the randomly generated small artificial fish
  • K T1t is randomly generated.
  • Step 4.3.4 If the current number of small fishes t 1 reaches the size of the small fish group N1, the current small fish population is obtained, and step 4.4 is performed; otherwise, return to step 4.3.2.
  • Step 4.4 Randomly generate a fierce fish group according to the upper limit of the value of each control variable to be optimized and the lower limit of the value of the control variable to be optimized.
  • Step 4.4.2 Randomly generate fierce artificial fish according to the upper limit and lower limit of each control variable to be optimized by the grid, as shown in formula (4):
  • Step 4.4.4 If the current number of fierce fish t 2 reaches the fierce fish size N2, then the current fierce fish group is obtained, proceed to step 4.5, otherwise, return to step 4.4.2.
  • the food concentration value under the piece, the food concentration value of each small artificial fish individual state, the food concentration value of each fierce artificial fish individual state, and the maximum value of the food concentration value is taken as the current optimal food concentration value to enter the bulletin board, and the preservation is performed. Its state and current optimal food concentration value FC.
  • Step 4.6 Action on small fish stocks: For small fish individuals, foraging behavior, rear-end behavior and protective clustering behavior, update bulletin board Y1.
  • Step 4.6.1 Determine the distance between the current i-th small artificial fish individual X i and each fierce artificial fish individual W j as shown in equation (5):
  • N2 is the number of fierce artificial fish individuals.
  • Step 4.6.2 If the distance between the current i-th small artificial fish individual X i and each fierce artificial fish individual W j D ij ⁇ Visual1, perform step 4.6.3, otherwise, perform step 4.6.5.
  • Step 4.6.3 Determine the total number n of small fish in the i-small size of the i-th small fish X i , and determine the safe location of the protective gathering X safe , and conduct protective clustering behavior on the i-th small fish X i to obtain the i-th small The fish conducts a protective clustering behavior update value X inext1 .
  • the small fish within the sensing range of the i-th small fish X i is determined, that is, the small fish individual X in the formula (6) is satisfied:
  • the calculation formula of the protection aggregation safe position X safe is as shown in the formula (7):
  • X safe (X i1 +X i2 +...+X in ) ⁇ /n(7)
  • p is the total number of fierce fish in the small fish individual X i sensing range Visual1
  • X in is the small fish individual within the sensing range of the small fish individual X i , introducing the escape factor
  • the purpose of ⁇ is to allow small fish groups to gather as close as possible to a safe location far from the fierce fish.
  • the calculation formula of the protection clustering behavior update value X inext1 of the i-th small fish individual is as shown in the formula (8):
  • X inext1 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X safe -X i )/
  • Step 4.6.4 Determine whether the current small artificial fish individual performs the protection clustering update value X inext1 to satisfy its constraint condition and the power flow converges. If yes, record the state of the updated value X inext1 , calculate the food concentration, and update the bulletin board. Y1, go to step 4.7, otherwise, return to step 4.6.3.
  • each constraint condition is the number of corresponding generators, the upper and lower limits of the generator voltage amplitude U G , the number of transformers, and the upper and lower limits of the adjustable transformer ratio K t of each transformer, and the reactive power compensation device.
  • Step 4.6.5 determining the i-th small sensing area X i fish fish food in small concentrations FC j 'is small, the largest fish X j', i-th small fish following behavior for X i, the i-th give The small fish X i performs the rear-end behavior update value X inext2 .
  • the calculation formula of the updated value X inext2 of the rear-end behavior of the i-th small artificial fish individual X i is as shown in the formula (9):
  • X inext2 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X j' -X i )/
  • Step 4.6.6 Determine the total number of small fish n in the sensing range of the i-th small fish X i , thereby determining the position of the foraging group X c1 , and conducting the foraging phenomenon of the i-th small fish X i to obtain the i-th A small fish is updated for the behavior of the foraging group X inext3 .
  • the calculation formula of the update index X inext3 of the foraging constellation behavior of the i-th small artificial fish individual is as shown in the formula (11):
  • X inext3 X i +rand(g) ⁇ Step1 ⁇ 1 ⁇ (X c1 -X i )/
  • Step 4.6.7 Determine whether the current small-scale artificial fish individual update value X inext2 and the small artificial fish individual carry out the feeding group behavior update value X inext3 to satisfy the constraint condition and the power flow convergence, and if so, record the update value The state of X inext2 and the state of the updated value X inext3 , and calculate its food concentration, update bulletin board Y1, perform step 4.7, otherwise, return to step 4.6.5.
  • Step 4.7 Acting fierce fish: According to the distance between each fierce fish individual and the small fish individual in the fierce fish group, and the distance between the individual fierce fish, predation behavior, tracking behavior and gathering of each fierce fish individual Group behavior, and update the bulletin board Y2 after the action of the fierce fish.
  • Step 4.7.1 Determine the distance between the current i'th fierce fish W i' and each small fish X j " ′ as shown in equation (12):
  • L i'j′′
  • N1 is the number of small fish individuals.
  • Step 4.7.2 If L i'j" ⁇ Visual2, perform step 4.7.3, otherwise, perform step 4.7.6.
  • X c2 (X i'1 +X i'2 +...+X i'm )/m (13)
  • X i'm is a small artificial fish within the range of the fierce fish W i ' .
  • Step 4.7.4 the current of the i 'th ferocious fish W is i' perception within the range of the center position X c2 small fish food as this maximum concentration, the first i 'th ferocious fish W is i' tracking behavior, to give the i 'A fierce fish W i' performs an updated value of the tracking behavior W i'next2 .
  • the calculation formula of the updated value W i'next2 of the tracking behavior of the i'th fierce fish W i ' is as shown in the formula (14):
  • W i'next2 W 'i + rand(g) ⁇ Step2 ⁇ ⁇ 2 ⁇ (X c2 - W i' ) /
  • Step 4.7.5 Analyzing this ferocious gudgeon W i 'is updated value W i'next1 ferocious fish and predatory behavior of individual W i' is updated to track the behavior of the value W i'next2 which satisfy the constraints and flow convergence, if the updated value W i'next1 and a state update value W i'next2 is recorded, and calculating the concentration of the food, the bulletin board update Y2, step 4.7.6, otherwise, it returns to step 4.7.3.
  • Step 4.7.6 Determine the total number r of fierce fish individuals within the sensing range of the i' fierce fish W i' , thereby determining the cluster center position W c and clustering the i'th fierce fish W i ' The i'th fierce fish W i' performs an updated value of the clustering behavior W i'next3 .
  • W i'r is the fierce fish individual within the range of i i fierce fish individual W i ' .
  • W i'next1 W i' +rand(g) ⁇ Step2 ⁇ 2 ⁇ (W c -W i ')/
  • Step 4.7.7 Determine whether the current fierce fish individual W i' performs an update value of the clustering behavior W i'next3 satisfies its constraint condition and the power flow converges, and if so, records the state of the updated value W i'next3 and calculates its Food concentration, update bulletin board Y2, perform step 4.8, otherwise, return to step 4.7.6.
  • Step 4.9 Take the food concentration of the bulletin board Y1 and Y2 and its state as the optimization result, and obtain the optimized value of each control variable, that is, the generator terminal voltage amplitude, the transformer adjustable ratio, and the reactive power compensation device. The optimized value of the work capacity.
  • Step 4.10 Transfer the optimized values of each control variable to the grid reactive execution module.
  • Step 5 Each generator terminal voltage regulator optimizes the value of the generator terminal voltage according to the amplitude value of the generator terminal voltage, and each transformer tap adjuster optimizes the value adjustment transformer tap according to the adjustable ratio of the transformer, and each reactive power compensation The regulator adjusts the compensation capacity of the reactive power compensator according to the reactive power capacity of the reactive power compensation device.
  • the reactive power optimization simulation is performed on the system by using the particle swarm algorithm, the genetic algorithm and the Pisces group algorithm in the present invention respectively.
  • the three algorithms are iterated 100 times, and all of the 10 simulation experiments are performed.
  • the comparison of the simulation results is shown in the table. 5 and Table 6:
  • Table 5 uses the particle swarm optimization algorithm, the genetic algorithm and the Pisces group algorithm in the present invention to optimize the node voltage comparison before and after.
  • Table 6 uses the particle swarm optimization algorithm, the genetic algorithm and the Pisces group algorithm in the present invention to optimize the network loss before and after optimization.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Automation & Control Theory (AREA)
  • Supply And Distribution Of Alternating Current (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Feedback Control In General (AREA)

Abstract

本发明提出基于双鱼群算法的电力无功优化系统及方法，该系统包括电网状态采集模块、电网无功调节模块和电网无功执行模块；电网状态采集模块包括电网状态采集器和中继传输器；电网无功调节模块为控制终端；电网无功执行模块包括各发电机端电压调节器、各变压器分接头调节器和各无功补偿调节器；该方法为获取当前网络待优化初始数据；采用基于双鱼群算法对当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值；电网无功执行模块根据得到电网中各控制变量的优化取值调节发电机端电压、变压器分接头、无功补偿器的补偿容量。该方法能够有效的提高电网的无功优化寻优能力是待优化配电网能够得到更为合理的无功潮流分布。

Description

基于双鱼群算法的电力无功优化系统及方法 技术领域
本发明属于电力系统无功优化技术领域，具体涉及基于双鱼群算法的电力无功优化系统及方法。
背景技术
电力系统无功优化，是指系统有功负荷、有功电源及潮流分布已经给定的况下，通过优化计算确定系统中某些控制变量的值，以找到在满足所有约束条件的前提下，使系统的某一个或多个性能指标(如有功网损最小、电压质量最优、年支出费用最少)达到最优时的运行方式。系统无功分布的合理与否直接关系着电力系统的安全和稳定，并且和经济效益有着密切的联系。一方面，如果系统的无功不足，将使电压水平低下，一些工厂和家庭的电器不能正常运行，而且系统一旦发生扰动，就可能使电压低于临界电压，产生电压崩溃，从而导致系统因失去同步而瓦解的灾难性事故。另一方面，系统无功过剩会使电压过高，危害系统和设备的安全。另外，系统无功的不合理流动，会使线路的压降增大、线路的损耗增加、供电的经济性下降。总之，无功设备的合理配置和优化运行能有效地降低网损，改善电压质量和保证系统电压稳定性，从而提高电力系统运行的安全性和经济性。
目前所用的无功优化方法主要有常规的数学方法和智能优化算法这两大类优化算法。常规的数学方法具有较快的计算速度，但它对优化函数的连续性、非凸性、可微性的计算具有较高要求，而且它还存在易于陷入局部最优解等缺点。智能优化算法在处理非线性、多变量、不连续、非凸等优化问题上体现出了很强的寻优能力。
人工鱼群算法作为一种新型的全局寻优策略，已应用于时变系统在线辨识、鲁棒PID的参数整定和优化前向神经网络中，取得了较好的效果。该算法具有良好的求取全局极值能力，并具有对初值、参数选择不敏感、鲁棒性强、简单(只使用目标函数值)、易实现等诸多优点。不过人工鱼群算法仍然存在一些不足，主要是：求解的效率不高，求解时间增多，后期收敛的速度将减慢，容易陷入局部最优解，难得到精确的最优解等。
发明内容
针对现有技术的不足，本发明提出一种基于双鱼群算法的电力无功优化系统及方法。
本发明技术方案如下：
一种基于双鱼群算法的电力无功优化系统，其特征在于，包括：电网状态采集模块、电网无功调节模块和电网无功执行模块；
所述电网状态采集模块，包括电网状态采集器和中继传输器；
所述电网无功调节模块为控制终端；
所述电网无功执行模块，包括各发电机端电压调节器、各变压器分接头调节器和各无功补偿调节器；
所述电网状态采集器的输入端连接电网，所述电网状态采集器的输出端连接中继传输器的输入端，所述中继传输器的输出端连接控制终端的输入端，所述控制终端的输出端连接发各电机端电压调节器的输入端、各变压器分接头调节器的输入端和各无功补偿调节器的输入端，所述发电机端电压调节器的输出端连接电网中的发电机，所述变压器分接头调节器的输出端连接电网中的变压器，所述无功补偿调节器的输出端连接电网中的无功补偿装置；
所述电网状态采集器，用于采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器，所述当前网络信息包括：电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息；
所述中继传输器，用于将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端；
所述电网无功调节模块：包括参数获取单元、基于双鱼群算法无功优化单元、优化决策控制单元；
所述参数获取单元，用于获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据；
所述基于双鱼群算法无功优化单元，用于建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，所述各控制变量包括发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量；
所述优化决策控制单元，用于将各控制变量的优化取值传送至电网无功执行模块；
所述发电机端电压调节器，用于根据电网无功调节模块得到的发电机端电压幅值优化取值调节发电机端电压；
所述变压器分接头调节器，用于根据电网无功调节模块得到的变压器可调变比优化取值调节变压器分接头；
所述无功补偿调节器，用于根据电网无功调节模块得到的无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
所述建立电力系统无功优化的数学模型如下所示：
Figure PCTCN2017088698-appb-000001
其中，F为电力无功优化数学模型目标函数，
Figure PCTCN2017088698-appb-000002
Figure PCTCN2017088698-appb-000003
1＜h＜nl，nl为电网系统总支路数，Glv为连接支路l-v的电导，θv为节点v的相角，θl为节点l的相角，α1为网损最小目标函数的权重系数，α2为电压最优目标函数的权重系数，ul为节点l的电压，uv为节点v的电压，1＜v＜Nd，Nd为电网系统中负荷节点总数，1＜k＜Ng，Ng为电网中发电机个数，Vv为节点v电压，
Figure PCTCN2017088698-appb-000004
为节点v的电压给定值，
Figure PCTCN2017088698-appb-000005
为节点v的电压偏差最大值，Vv.max为节点v电压上限值，Vv.min为节点v电压下限值，λu为负荷节点电压越界惩罚系数，λq发电机无功出力越界惩罚系数，Qk为发电机节点k的无功出力，Qk.max为发电机节点k的无功出力的上限值，Qk.min为发电机节点k的无功出力的下限值；
所述采用上述基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值的具体过程如下：
根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群和凶猛鱼群；
初始化当前最优食物浓度值进入公告板：以电力无功优化数学模型目标函数的倒数1/F作为食物浓度值FC，计算初始数据条件下当前网络、各小型鱼及各凶猛鱼的食物浓度值，取食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前FC值；
对小型鱼群进行行动：根据各小型鱼个体与凶猛鱼个体之间的距离，以及小型鱼群各小型鱼个体之间的距离，对小型鱼个体进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1；
对凶猛鱼群进行行动：根据各凶猛鱼个体与小型鱼个体之间的距离，以及凶猛鱼群各凶猛鱼个体之间的距离，对各凶猛鱼个体进行捕食行为、追踪行为和聚群行为，更新公告板Y2；
当迭代次数达到了最大迭代次数时，取公告板Y1和Y2中食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比及无功补偿装置的无功容量的优化取值。
采用基于双鱼群算法的电力无功优化系统进行电力无功优化的方法，包括以下步骤：
步骤1：通过电网状态采集器采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器；
步骤2：通过中继传输器将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端；
步骤3：通过控制终端获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据；
步骤4：通过控制终端建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，将各控制变量的优化取值传送至电网无功执行模块；
步骤5：各发电机端电压调节器根据发电机端电压幅值优化取值调节发电机端电压，各变压器分接头调节器根据变压器可调变比优化取值调节变压器分接头，各无功补偿调节器根据无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
所述步骤4包括以下步骤：
步骤4.1：以网损最小和电压水平最优为优化目标，采用α-法建立电力系统无功优化的数学模型：
Figure PCTCN2017088698-appb-000006
步骤4.2：设定双鱼群算法参数：小型鱼群规模N1、凶猛鱼群规模N2、小型鱼群感知范围Visual1、凶猛鱼群感知范围Visual2、小型鱼群移动步长Step1、凶猛鱼群移动步长Step2、小型鱼群拥挤度因子δ1、凶猛鱼群拥挤度因子δ2、最大迭代次数K＞0；
步骤4.3：根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群；
步骤4.4：根据电网待优化的各控制变量的取值上限及取值下限随机生成凶猛鱼群；
步骤4.5：初始化当前迭代次数k′＝0，初始化当前最优食物浓度值进入公告板：以1/F作为食物浓度值FC，计算初始数据条件下当前网络、各小型鱼及各凶猛鱼的食物浓度值，取 食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前FC值；
步骤4.6：小型鱼群进行行动：根据各小型鱼个体与凶猛鱼个体之间的距离，以及小型鱼群各小型鱼个体之间的距离，对小型鱼个体进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1；
步骤4.7：凶猛鱼群进行行动：根据各凶猛鱼与小型鱼个体之间的距离，以及凶猛鱼群各凶猛鱼个体之间的距离，对各凶猛鱼个体进行捕食行为、追踪行为和聚群行为，更新公告板Y2；
步骤4.8：判断当前迭代次数k′是否达到了最大迭代次数K，若是执行步骤4.9，否则，令k′＝k′+1，返回步骤4.6；
步骤4.9：取公告板Y1和Y2食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比及无功补偿装置的无功容量的优化取值；
步骤4.10：将各控制变量的优化取值传送至电网无功执行模块。
所述步骤4.3包括以下步骤：
步骤4.3.1：初始化生成小型鱼个数t1＝0；
步骤4.3.2：根据电网待优化的各控制变量的取值范围随机生成小型鱼个体：
X＝[UG11，UG12，...UG1g，KT11，KT12，...，KT1t，QC11，QC12，...，QC1c]＝[x1，x2，...，xs，...，xS]；
其中，X为小型鱼个体的控制变量取值序列，xs＝xs min+rand(g)×(xs max-xs min)，rand(g)为(0，1)区间的随机数，xs min、xs max分别为对应控制变量的取值下限和取值上限，UG1g、KT1t、QC1c分别为随机生成的小型人工鱼个体中的发电机端电压幅值、变压器可调变比和无功补偿装置的无功容量，xs为随机生成的小型鱼个体的控制变量，1≤s≤S，S＝Ng+Nt+Nc，Ng为电网中发电机个数，Nt为电网中变压器个数，Nc为电网中无功补偿装置的个数；
步骤4.3.3：对步骤4.3.2中随机生成的小型鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前小型鱼个体的潮流值收敛，则保留该小型鱼个体，令当前生成小型鱼个数t1＝t1+1，执行步骤4.3.4，否则，不保留该小型鱼个体，返回步骤4.3.2；
步骤4.3.4：若t1≥N1，则得到当前小型鱼群，执行步骤4.4，否则，返回步骤4.3.2；
所述步骤4.4包括以下步骤：
步骤4.4.1：初始化生成凶猛鱼个数t2＝0；
步骤4.4.2：根据电网待优化的各控制变量的取值范围随机生成凶猛鱼个体：
W＝[UG21，UG22，...UG2g，KT21，KT22，...，KT2t，QC21，QC22，...，QC2c]＝[w1，w2，...，ws，...，wS]；
其中，W为凶猛鱼个体的控制变量取值序列，ws＝xs max+rand(g)×(xs max-xs min)，UG2g、KT2t、QC2c分别为随机生成的凶猛鱼个体中的发电机端电压幅值、变压器可调变比和无功补偿装置的无功容量，ws为随机生成的凶猛鱼个体的控制变量；
步骤4.4.3：对步骤4.4.2中随机生成的凶猛鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前凶猛鱼个体的潮流值收敛，则保留该凶猛鱼个体，令当前生成凶猛鱼个数t2＝t2+1，执行步骤4.4.4，否则，不保留该凶猛鱼个体，返回步骤4.4.2；
步骤4.4.4：若t2≥N2，则得到当前凶猛鱼群，执行步骤4.5，否则，返回步骤4.4.2；
所述步骤4.6包括以下步骤：
步骤4.6.1：确定当前第i个小型鱼个体Xi与各凶猛鱼个体Wj之间的距离D＝{Di1，Di2，...，Dij，...，DiN2}，其中，Dij＝||Xi-Wj||，1＜j＜N2，N2为凶猛鱼个体数；
步骤4.6.2：若当前第i个小型鱼个体Xi与各凶猛鱼个体Wj之间的距离Dij≤Visual1，执行步骤4.6.3，否则，执行步骤4.6.5；
步骤4.6.3：确定第i个小型鱼个体Xi感知范围内小型鱼个体总数n，从而确定防护聚集安全位置Xsafe，对第i个小型鱼个体Xi进行防护聚群行为，得到第i个小型鱼个体进行防护聚群行为更新值Xinext1；
所述防护聚集安全位置Xsafe的计算公式如下：
Xsafe＝(Xi1+Xi2+...+Xin)×λ/n；
其中，λ＝(n+p)/n为逃离因子，p为小型鱼个体Xi感知范围Visual1内凶猛鱼个体总数，Xin为小型鱼个体Xi感知范围内的小型鱼个体；
所述第i个小型鱼个体进行防护聚群行为更新值Xinext1的计算公式如下：
Xinext1＝Xi+rand(g)×Step1×δ1×(Xsafe-Xi)/||Xsafe-Xi||；
步骤4.6.4：判断当前小型鱼个体进行防护聚群行为更新值Xinext1是否满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext1的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.3；
步骤4.6.5：确定第i个小型鱼个体Xi感知范围内小型鱼群中食物浓度FCj′为最大的小型鱼个体Xj′，对第i个小型鱼个体Xi进行追尾行为，得到第i个小型鱼个体Xi进行追尾行为更新值Xinext2；
所述第i个小型鱼个体Xi进行追尾行为的更新值Xinext2的计算公式如下：
Xinext2＝Xi+rand(g)×Step1×δ1×(Xj′-Xi)/||Xj′-Xi||；
步骤4.6.6：确定第i个小型鱼个体Xi感知范围内小型鱼个体总数n，从而确定觅食聚群中心位置Xc1，对第i个小型鱼个体Xi进行觅食聚群行为，得到第i个小型鱼个体进行觅食聚群行为更新值Xinext3；
所述觅食聚群中心位置Xc1的计算公式如下：
Xc1＝(Xi1+Xi2+...+Xin)/n；
所述第i个小型鱼个体进行觅食聚群行为更新值Xinext3的计算公式如下：
Xinext3＝Xi+rand(g)×Step1×δ1×(Xc1-Xi)/||Xc1-Xi||；
步骤4.6.7：判断当前小型鱼个体进行追尾行为更新值Xinext2和小型鱼个体进行觅食聚群行为更新值Xinext3是否均满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext2的状态和更新值Xinext3的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.5。
所述步骤4.7包括以下步骤：
步骤4.7.1：确定当前第i′个凶猛鱼个体Wi′与各小型鱼个体Xj′之间的距离L＝{Li′1，Li′2，...，Li′j″，...，Li′N1}，其中，Li′j″＝||Wi′-Xj″||，1＜j″＜N1，N1为小型鱼个体个数；
步骤4.7.2：若当前第i′个凶猛人工鱼个体Wi′与各小型人工鱼个体Xj″之间的距离 Li′j″≤Visual2执行步骤4.7.3，否则，执行步骤4.7.6；
步骤4.7.3：确定当前第i′个凶猛鱼个体Wi′感知范围内小型鱼个体总数m，以当前第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼个体Wi′进行捕食行为，将第i′个凶猛鱼个体Wi′更新为小型鱼群中心位置Xc2，得到第i′个凶猛鱼个体Wi′进行捕食行为的更新值Wi′next1＝Xc2；
所述第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2的计算公式如下：
Xc2＝(Xi′1+Xi′2+...+Xi′m)/m；
其中，Xi′m为凶猛鱼个体Wi′感知范围内小型鱼；
步骤4.7.4：以当前第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼个体Wi′进行追踪行为，得到第i′个凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2；
所述第i′个凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2的计算公式如下：
Wi′next2＝W′i+rand(g)×Step2×δ2×(Xc2-Wi′)/||Xc2-Wi′||；
步骤4.7.5：判断当前凶猛鱼个体Wi′进行捕食行为的更新值Wi′next1和凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2是否均满足其约束条件并且潮流收敛，若是，更新公告板Y2，执行步骤4.7.6，否则，返回步骤4.7.3；
步骤4.7.6：确定第i′个凶猛鱼个体Wi′感知范围内凶猛鱼个体总数r，从而确定聚群行为中心位置Wc，对第i′个凶猛鱼个体Wi′进行聚群行为，得到第i′个凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3；
所述聚群行为中心位置Wc的计算公式如下：
Wc＝(Wi′1+Wi′2+...+Wi′r)/r；
其中，Wi′r为第i′个凶猛鱼个体Wi′感知范围内凶猛鱼个体；
所述第i′个凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3的计算公式如下：
Wi′next1＝Wi′+rand(g)×Step2×δ2×(Wc-Wi′)/||Wc-Wi′||；
步骤4.7.7：判断当前凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3是否满足其约束条件并且潮流收敛，若是，则记录该更新值Wi′next3的状态，并计算其食物浓度，更新公告板Y2，执行步骤4.8，否则，返回步骤4.7.6。
本发明的有益效果：
本发明提出一种基于双鱼群算法的电力无功优化系统及方法，本发明对系统进行模块化处理，易于控制和执行无功优化方案，基于双鱼群的无功优化方法，引入新鱼群-凶猛鱼群，增大搜索范围，很大改善了基本鱼群算法容易陷入局部最优解、难得到精确最优解的缺点，提高了其在计算过程中的收敛性，能够有效的提高电网的无功优化寻优能力是待优化配电网能够得到更为合理的无功潮流分布。
附图说明
图1为本发明具体实施方式中IEEE10节点系统；
图2为本发明具体实施方式中基于双鱼群算法的电力无功优化系统的结构框图；
图3为本发明具体实施方式中基于双鱼群算法的电力无功优化方法的流程图；
图4为本发明具体实施方式中采用基于双鱼群算法对当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值的过程流程图。
具体实施方式
下面结合附图对本发明具体实施方式加以详细的说明。
本实施方式中，采用如图1所示的IEEE10节点系统进行仿真。该系统是一个闭环系统，有6条线路，3个发电机节点，3个负荷节点，3台有载调压变压器。3个发电机节点中，节点1作为平衡节点，其余2个发电机节点为PV节点，系统中其它节点为PQ节点，2个无功补偿节点(节点为5，10)，可调变压器支路为1-4，2-8，3-6，4-10。电压等级是220kV，系统总负荷为315+j255MVA。
一种基于双鱼群算法的电力无功优化系统，如图2所示，包括：电网状态采集模块、电网无功调节模块和电网无功执行模块。
电网状态采集模块，包括电网状态采集器和中继传输器。
电网无功调节模块为控制终端。
电网无功执行模块，包括各发电机端电压调节器、各变压器分接头调节器和各无功补偿调节器。
电网状态采集器的输入端连接电网，电网状态采集器的输出端连接中继传输器的输入端，中继传输器的输出端连接控制终端的输入端，控制终端的输出端连接发各电机端电压调节器的输入端、各变压器分接头调节器的输入端和各无功补偿调节器的输入端，发电机端电压调节器的输出端连接电网中的发电机，变压器分接头调节器的输出端连接电网中的变压器，无功补偿调节器的输出端连接电网中的无功补偿装置。
本实施方式中，电网状态采集器的型号为DCZL23-CL156N，用于采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器，当前网络信息包括：电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息。
本实施方式中，根据图1中IEEE10节点系统得到的变压器信息和对应支路信息如表1和表2所示，负荷参数如表3所示，支路参数如表4所示。
表1双绕组变压器信息和对应支路信息
变压器支路 R X 变比
2-6 0 0.0625 1.0
3-8 0 0.0586 1.0
表中，R为电阻，X为电抗。
表2三绕组变压器信息和对应支路信息
变压器支路 高压阻抗(R/X) 中压阻抗(R/X) 低压阻抗(R/X)
1-4-10 0.0011  0.085 0.0037  -0.015 0.0015  0.095
表3负荷参数
节点号 Pload Qload
5 1.25 0.80
7 1.00 0.55
9 0.90 1.20
表中，Pload为负载的有功功率，Qload为负载的无功功率。
表4支路参数
支路号 首末节点号 R X B/2
1 4-5 0.0100 0.085 0.0440
2 5-6 0.0320 0.161 0.0765
3 6-7 0.0085 0.072 0.0373
4 7-8 0.0119 0.1008 0.0523
5 8-9 0.0390 0.170 0.0895
6 9-4 0.017 0.092 0.0395
本实施方式中，中继传输器的型号为RD980，用于将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端。
电网无功调节模块：包括参数获取单元、基于双鱼群算法无功优化单元、优化决策控制单元。
参数获取单元，用于获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据。
基于双鱼群算法无功优化单元，用于建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，所述各控制变量包括发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量。
本实施方式中，以网损最小和电压水平最优为优化目标，采用α-法，以罚函数的形式，建立电力系统无功优化的数学模型如式(1)所示：
Figure PCTCN2017088698-appb-000007
其中，F为电力无功优化数学模型目标函数，
Figure PCTCN2017088698-appb-000008
Figure PCTCN2017088698-appb-000009
1＜h＜nl，nl为电网系统总支路数，Glv为连接支路l-v的电导，θv为节点v的相角，θl为节点l的相角，α1为网损最小目标函数的权重系数，α2为电压最优目标函数的权重系数，ul为节点l的电压，uv为节点v的电压，1＜v＜Nd，Nd为电网系统中负荷节点总数，1＜k＜Ng，Ng为电网中发电机个数，Vv为节点v电压，
Figure PCTCN2017088698-appb-000010
为节点v的电压给定值，
Figure PCTCN2017088698-appb-000011
为节点v的电压偏差最大值，Vv.max为节点v电压上限值，Vv.min为节点v电压下限值，λu为负荷节点电压越界惩罚系数，λq发电机无功出力越界惩罚系数，Qk为发电机节点k的无功出力，qk.max为发电机节点k的无功出力的上限值，Qk.min为发电机节点k的无功出力的下限值。
本实施方式中，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值的具体过程如下：
根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群，各控制变量包括：发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量；
根据电网待优化的各控制变量的取值上限及取值下限随机生成凶猛鱼群；
初始化当前最优食物浓度值进入公告板：以电力无功优化数学模型目标函数的倒数作为食物浓度值FC，通过计算当前网络待优化初始数据条件下的食物浓度值、各小型人工鱼个体状态下的食物浓度值、各凶猛人工鱼个体状态下的食物浓度值，取食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前最优食物浓度值FC；
本实施方式中，食物浓度值FC的计算公式如式(2)所示：
FC＝1/F            (2)
对小型鱼群进行行动：根据各小型鱼与凶猛鱼之间的距离，以及小型鱼群各小型鱼个体之间的距离，对小型鱼进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1；
对凶猛鱼群进行行动：根据各凶猛鱼与小型鱼之间的距离，以及凶猛鱼群各凶猛鱼个体之间的距离，对各凶猛人工鱼个体进行捕食行为、追踪行为和聚群行为，更新Y2；
当达到终止条件时，取公告板Y1和Y2的食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量的优化取值。
优化决策控制单元，用于将各控制变量的优化取值传送至电网无功执行模块。
本实施方式中，发电机端电压调节器型号为EQ1512，用于根据电网无功调节模块得到的发电机端电压幅值优化取值调节发电机端电压。
本实施方式中，变压器分接头调节器型号为TSGC2J，用于根据电网无功调节模块得到的变压器可调变比优化取值调节变压器分接头。
本实施方式中，无功补偿调节器型号为GLTSC，用于根据电网无功调节模块得到的无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
采用基于双鱼群算法的电力无功优化系统进行电力无功优化的方法，如图3所示，包括以下步骤：
步骤1：通过电网状态采集器采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器。
步骤2：通过中继传输器将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端。
步骤3：通过控制终端获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据。
本实施方式中，当前网络待优化初始数据为节点信息、支路信息、发电机个数及各发电机端电压幅值UG的上下限、变压器个数及各变压器可调变比Kt的上下限、无功补偿装置的投切组数及各无功补偿装置无功容量Qc的上下限。
步骤4：通过控制终端建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的 当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，将各控制变量的优化取值传送至电网无功执行模块，如图4所示。
步骤4.1：以网损最小和电压水平最优为优化目标，采用加权和法中的α-法建立电力系统无功优化的数学模型。
步骤4.2：设定双鱼群算法参数：小型鱼群规模N1、凶猛鱼群规模N2、小型鱼群感知范围Visual1、凶猛鱼群感知范围Visual2、小型鱼群移动步长Step1、凶猛鱼群移动步长Step2、小型鱼群拥挤度因子δ1、凶猛鱼群拥挤度因子δ2、小型鱼群尝试次数try_number1、凶猛鱼群尝试次数try_number2、最大迭代次数K＞0。
本实施方式中，最大迭代次数K为100。
步骤4.3：根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群。
步骤4.3.1：初始化生成小型鱼个数t1＝0。
步骤4.3.2：根据电网待优化的各控制变量的取值上限及取值下限随机生成小型人工鱼个体如式(3)所示：
X＝[UG11，UG12，...UG1g，KT11，KT12，...，KT1t，QC11，QC12，...，QC1c]＝[x1，x2，...，xs，...，xS]    (3)
其中，X为小型人工鱼个体的控制变量取值序列，xs＝xs min+rand(g)×(xs max-xs min)，rand(g)为(0，1)区间的随机数，xs min为对应控制变量的取值下限，xs max为对应控制变量的取值上限，UG1g为随机生成的小型人工鱼个体中发电机端电压幅值，KT1t为随机生成的小型人工鱼个体中变压器可调变比，QC1c为随机生成的小型人工鱼个体中无功补偿装置的无功容量，xs为随机生成的小型人工鱼个体的控制变量，1≤s≤S，S＝Ng+Nt+Nc，Ng为电网中发电机个数，Nt为电网中变压器个数，Nc为电网中无功补偿装置的个数。
步骤4.3.3：对步骤4.3.2中随机生成的小型鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前小型鱼个体的潮流值收敛，则保留该小型鱼个体，令t1＝t1+1，执行步骤4.3.4，否则，不保留该小型鱼个体，返回步骤4.3.2。
步骤4.3.4：若当前生成小型鱼个数t1达到小型鱼群规模N1，则得到当前小型鱼群，执行步骤4.4，否则，返回步骤4.3.2。
步骤4.4：根据电网待优化的各控制变量的取值上限及取值下限随机生成凶猛鱼群。
步骤4.4.1：初始化生成凶猛鱼个数t2＝0。
步骤4.4.2：根据电网待优化的各控制变量的取值上限及取值下限随机生成凶猛人工鱼个体如式(4)所示：
W＝[UG21，UG22，...UG2g，KT21，KT22，...，KT2t，QC21，QC22，...，QC2c]＝[w1，w2，...，ws，...，wS]    (4)
步骤4.4.3：对步骤4.4.2中随机生成的凶猛人工鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前凶猛鱼个体的潮流值收敛，则保留该凶猛鱼个体，令t2＝t2+1，执行步骤4.4.4，否则，不保留该凶猛鱼个体，返回步骤4.4.2。
步骤4.4.4：若当前凶猛鱼个数t2达到凶猛鱼群规模N2，则得到当前凶猛鱼群，执行步骤4.5，否则，返回步骤4.4.2。
步骤4.5：初始化当前迭代次数k′＝0，初始化当前最优食物浓度值进入公告板：以电力无功优化数学模型目标函数的倒数作为食物浓度值FC，通过计算当前网络待优化初始数据条 件下的食物浓度值、各小型人工鱼个体状态下的食物浓度值、各凶猛人工鱼个体状态下的食物浓度值，取食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前最优食物浓度值FC。步骤4.6：小型鱼群进行行动：对小型鱼个体进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1。
步骤4.6.1：确定当前第i个小型人工鱼个体Xi与各凶猛人工鱼个体Wj之间的距离如式(5)所示：
D＝{Di1，Di2，...，Dij，...，DiN2}(5)
其中，Dij＝||Xi-Wj||，1＜j＜N2，N2为凶猛人工鱼个体数。
步骤4.6.2：若当前第i个小型人工鱼个体Xi与各凶猛人工鱼个体Wj之间的距离Dij≤Visual1，执行步骤4.6.3，否则，执行步骤4.6.5。
步骤4.6.3：确定第i个小型鱼Xi感知范围内小型鱼个体总数n，从而确定防护聚集安全位置Xsafe，对第i个小型鱼Xi进行防护聚群行为，得到第i个小型鱼进行防护聚群行为更新值Xinext1。
本实施方式中，确定第i个小型鱼Xi感知范围内小型鱼，即满足公式(6)所示的小型鱼个体Xin：
diin＝||Xi-Xin||≤Visual1(6)
本实施方式中，防护聚集安全位置Xsafe的计算公式如式(7)所示：
Xsafe＝(Xi1+Xi2+...+Xin)×λ/n(7)
其中，λ＝(n+p)/n为逃离因子，p为小型鱼个体Xi感知范围Visual1内凶猛鱼个体总数，Xin为小型鱼个体Xi感知范围内的小型鱼个体，引入逃离因子λ的目的是使小型鱼群进行防护聚群行为时，尽可能的聚集到离凶猛鱼群较远的安全位置。
本实施方式中，第i个小型鱼个体进行防护聚群行为更新值Xinext1的计算公式如式(8)所示：
Xinext1＝Xi+rand(g)×Step1×δ1×(Xsafe-Xi)/||Xsafe-Xi||     (8)
步骤4.6.4：判断当前小型人工鱼个体进行防护聚群行为更新值Xinext1是否满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext1的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.3。
本实施方式中，各约束条件为对应的发电机个数及各发电机端电压幅值UG的上下限、变压器个数及各变压器可调变比Kt的上下限、无功补偿装置的投切组数及各无功补偿装置无功容量Qc的上下限。
步骤4.6.5：确定第i个小型鱼Xi感知范围内小型鱼群中食物浓度FCj′为最大的小型鱼Xj′，对第i个小型鱼Xi进行追尾行为，得到第i个小型鱼Xi进行追尾行为更新值Xinext2。
本实施方式中，第i个小型人工鱼个体Xi进行追尾行为的更新值Xinext2的计算公式如式(9)所示：
Xinext2＝Xi+rand(g)×Step1×δ1×(Xj′-Xi)/||Xj′-Xi||     (9)
步骤4.6.6：确定第i个小型鱼Xi感知范围内小型鱼总数n，从而确定觅食聚群中心位置 Xc1，对第i个小型鱼Xi进行觅食聚群行为，得到第i个小型鱼进行觅食聚群行为更新值Xinext3。
本实施方式中，觅食聚群中心位置Xc1的计算公式如式(10)所示：
Xc1＝(Xi1+Xi2+...+Xin)/n       (10)
本实施方式中，第i个小型人工鱼个体进行觅食聚群行为更新值Xinext3的计算公式如式(11)所示：
Xinext3＝Xi+rand(g)×Step1×δ1×(Xc1-Xi)/||Xc1-Xi||       (11)
步骤4.6.7：判断当前小型人工鱼个体进行追尾行为更新值Xinext2和小型人工鱼个体进行觅食聚群行为更新值Xinext3是否均满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext2的状态和更新值Xinext3的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.5。
步骤4.7：凶猛鱼群进行行动：根据凶猛鱼群中各凶猛鱼个体与小型鱼个体之间的距离，以及各凶猛鱼个体之间的距离，对各凶猛鱼个体进行捕食行为、追踪行为和聚群行为，并对凶猛鱼群进行行动后的公告板Y2进行更新。
步骤4.7.1：确定当前第i′个凶猛鱼Wi′与各小型鱼Xj″之间的距离如式(12)所示：
L＝{Li′1，Li′2，...，Li′j″，...，Li′N1}         (12)
其中，Li′j″＝||Wi′-Xj″||，1＜j″＜N1，N1为小型鱼个体个数。
步骤4.7.2：若Li′j″≤Visual2，执行步骤4.7.3，否则，执行步骤4.7.6。
步骤4.7.3：确定当前第i′个凶猛鱼Wi′感知范围内小型鱼总数m，以当前第i′个凶猛鱼Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼Wi′进行捕食行为，将第i′个凶猛鱼Wi′更新为小型鱼群中心位置Xc2，得到第i′个凶猛鱼Wi′进行捕食行为的更新值Wi′next1＝Xc2。
本实施方式中，第i′个凶猛鱼Wi′感知范围内小型鱼群中心位置Xc2的计算公式如式(13)所示：
Xc2＝(Xi′1+Xi′2+...+Xi′m)/m        (13)
其中，Xi′m为凶猛鱼Wi′感知范围内小型人工鱼。
步骤4.7.4：以当前第i′个凶猛鱼Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼Wi′进行追踪行为，得到第i′个凶猛鱼Wi′进行追踪行为的更新值Wi′next2。
本实施方式中，第i′个凶猛鱼Wi′进行追踪行为的更新值Wi′next2的计算公式如式(14)所示：
Wi′next2＝W′i+rand(g)×Step2×δ2×(Xc2-Wi′)/||Xc2-Wi′||       (14)
步骤4.7.5：判断当前凶猛鱼个体Wi′进行捕食行为的更新值Wi′next1和凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2是否均满足其约束条件并且潮流收敛，若是，则记录该更新值Wi′next1的状态和更新值Wi′next2的状态，并计算其食物浓度，更新公告板Y2，执行步骤4.7.6，否则，返回步骤4.7.3。
步骤4.7.6：确定第i′个凶猛鱼Wi′感知范围内凶猛鱼个体总数r，从而确定聚群行为中心位置Wc，对第i′个凶猛鱼Wi′进行聚群行为，得到第i′个凶猛鱼Wi′进行聚群行为的更新值Wi′next3。
本实施方式中，聚群行为中心位置Wc的计算公式如式(15)所示：
Wc＝(Wi′1+Wi′2+...+Wi′r)/r(15)
其中，Wi′r为第i′个凶猛鱼个体Wi′感知范围内凶猛鱼个体。
本实施方式中，第i′个凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3的计算公式如式(16)所示：
Wi′next1＝Wi′+rand(g)×Step2×δ2×(Wc-Wi′)/||Wc-Wi′||     (16)
步骤4.7.7：判断当前凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3是否满足其约束条件并且潮流收敛，若是，则记录该更新值Wi′next3的状态，并计算其食物浓度，更新公告板Y2，执行步骤4.8，否则，返回步骤4.7.6。
步骤4.8：判断当前迭代次数k′是否达到了最大迭代次数K，若是执行步骤4.9，否则，令k′＝k′+1，返回步骤4.6。
步骤4.9：取公告板Y1和Y2的食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量的优化取值。
步骤4.10：将各控制变量的优化取值传送至电网无功执行模块。
步骤5：各发电机端电压调节器根据发电机端电压幅值优化取值调节发电机端电压，各变压器分接头调节器根据变压器可调变比优化取值调节变压器分接头，各无功补偿调节器根据无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
本实施方式中，分别采用粒子群算法、遗传算法和本发明中双鱼群算法对该系统进行无功优化仿真，三种算法均迭代100次，均通过10次仿真实验，仿真结果对比情况如表5和表6所示：
表5采用粒子群算法、遗传算法和本发明中双鱼群算法优化前后节点电压对比
Figure PCTCN2017088698-appb-000012
表6采用粒子群算法、遗传算法和本发明中双鱼群算法优化前后系统网损对比
Figure PCTCN2017088698-appb-000013
根据表5和表6可知，通过补偿后的对比分析，可知双鱼群算法相比于粒子群算法和遗传算法，在无功优化过程中，取得更优的效果，电压质量提高明显，网损下降最大，使得无功分布得到改善，保证了系统的安全性和稳定性，其结果证明该算法具有很好的可行性和实用性。

Claims (9)

  1. 一种基于双鱼群算法的电力无功优化系统，其特征在于，包括：电网状态采集模块、电网无功调节模块和电网无功执行模块；
    所述电网状态采集模块，包括电网状态采集器和中继传输器；
    所述电网无功调节模块为控制终端；
    所述电网无功执行模块，包括各发电机端电压调节器、各变压器分接头调节器和各无功补偿调节器；
    所述电网状态采集器的输入端连接电网，所述电网状态采集器的输出端连接中继传输器的输入端，所述中继传输器的输出端连接控制终端的输入端，所述控制终端的输出端连接发各电机端电压调节器的输入端、各变压器分接头调节器的输入端和各无功补偿调节器的输入端，所述发电机端电压调节器的输出端连接电网中的发电机，所述变压器分接头调节器的输出端连接电网中的变压器，所述无功补偿调节器的输出端连接电网中的无功补偿装置；
    所述电网状态采集器，用于采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器，所述当前网络信息包括：电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息；
    所述中继传输器，用于将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端；
    所述电网无功调节模块：包括参数获取单元、基于双鱼群算法无功优化单元、优化决策控制单元；
    所述参数获取单元，用于获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据；
    所述基于双鱼群算法无功优化单元，用于建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，所述各控制变量包括发电机端电压幅值、变压器可调变比、无功补偿装置的无功容量；
    所述优化决策控制单元，用于将各控制变量的优化取值传送至电网无功执行模块；
    所述发电机端电压调节器，用于根据电网无功调节模块得到的发电机端电压幅值优化取值调节发电机端电压；
    所述变压器分接头调节器，用于根据电网无功调节模块得到的变压器可调变比优化取值调节变压器分接头；
    所述无功补偿调节器，用于根据电网无功调节模块得到的无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
  2. 根据权利要求1所述的基于双鱼群算法的电力无功优化系统，其特征在于，所述建立电力系统无功优化的数学模型如下所示：
    Figure PCTCN2017088698-appb-100001
    其中，F为电力无功优化数学模型目标函数，
    Figure PCTCN2017088698-appb-100002
    Figure PCTCN2017088698-appb-100003
    1＜h＜nl，nl为电网系统总支路数，Glv为连接支路l-v 的电导，θv为节点v的相角，θl为节点l的相角，α1为网损最小目标函数的权重系数，α2为电压最优目标函数的权重系数，ul为节点l的电压，uv为节点v的电压，1＜v＜Nd，Nd为电网系统中负荷节点总数，1＜k＜Ng，Ng为电网中发电机个数，Vv为节点v电压，
    Figure PCTCN2017088698-appb-100004
    为节点v的电压给定值，ΔVv max为节点v的电压偏差最大值，Vv.max为节点v电压上限值，Vv.min为节点v电压下限值，λu为负荷节点电压越界惩罚系数，λq发电机无功出力越界惩罚系数，Qk为发电机节点k的无功出力，Qk.max为发电机节点k的无功出力的上限值，Qk.min为发电机节点k的无功出力的下限值。
  3. 根据权利要求1或2所述的基于双鱼群算法的电力无功优化系统，其特征在于，所述采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值的具体过程如下：
    根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群和凶猛鱼群；
    初始化当前最优食物浓度值进入公告板：以电力无功优化数学模型目标函数的倒数1/F作为食物浓度值FC，计算初始数据条件下当前网络、各小型鱼及各凶猛鱼的食物浓度值，取食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前FC值；
    对小型鱼群进行行动：根据各小型鱼个体与凶猛鱼个体之间的距离，以及小型鱼群各小型鱼个体之间的距离，对小型鱼个体进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1；
    对凶猛鱼群进行行动：根据各凶猛鱼个体与小型鱼个体之间的距离，以及凶猛鱼群各凶猛鱼个体之间的距离，对各凶猛鱼个体进行捕食行为、追踪行为和聚群行为，更新公告板Y2；
    当迭代次数达到了最大迭代次数时，取公告板Y1和Y2中食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比及无功补偿装置的无功容量的优化取值。
  4. 采用权利要求2或3所述的基于双鱼群算法的电力无功优化系统进行电力无功优化的方法，其特征在于，包括以下步骤：
    步骤1：通过电网状态采集器采集电网的当前网络信息，判断当前网络信息是否满足电网所需求的最优状态，若当前网络信息不能满足电网所需求的最优状态，则将当前网络信息传输至中继传输器；
    步骤2：通过中继传输器将电网状态采集器所采集的网络信息中进行无功优化所需要的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息传输至控制终端；
    步骤3：通过控制终端获取中继传输器传输的电网节点信息、支路信息、发电机信息、变压器信息和无功补偿装置信息，作为当前网络待优化初始数据；
    步骤4：通过控制终端建立电力系统无功优化的数学模型，采用基于双鱼群算法对获取的当前网络待优化初始数据进行优化，得到电网中各控制变量的优化取值，将各控制变量的优化取值传送至电网无功执行模块；
    步骤5：各发电机端电压调节器根据发电机端电压幅值优化取值调节发电机端电压，各变压器分接头调节器根据变压器可调变比优化取值调节变压器分接头，各无功补偿调节器根据无功补偿装置的无功容量优化取值调节无功补偿器的补偿容量。
  5. 根据权利要求4所述的基于双鱼群算法的电力无功优化的方法，其特征在于，所述步骤4包括以下步骤：
    步骤4.1：以网损最小和电压水平最优为优化目标，采用α-法建立电力系统无功优化的数学模型：
    Figure PCTCN2017088698-appb-100005
    步骤4.2：设定双鱼群算法参数：小型鱼群规模N1、凶猛鱼群规模N2、小型鱼群感知范围Visual1、凶猛鱼群感知范围Visual2、小型鱼群移动步长Step1、凶猛鱼群移动步长Step2、小型鱼群拥挤度因子δ1、凶猛鱼群拥挤度因子δ2、最大迭代次数K＞0；
    步骤4.3：根据电网待优化的各控制变量的取值上限及取值下限随机生成小型鱼群；
    步骤4.4：根据电网待优化的各控制变量的取值上限及取值下限随机生成凶猛鱼群；
    步骤4.5：初始化当前迭代次数k′＝0，初始化当前最优食物浓度值进入公告板：以1/F作为食物浓度值FC，计算初始数据条件下当前网络、各小型鱼及各凶猛鱼的食物浓度值，取食物浓度值的最大值作为当前最优食物浓度值进入公告板，保存其状态及当前FC值；
    步骤4.6：小型鱼群进行行动：根据各小型鱼个体与凶猛鱼个体之间的距离，以及小型鱼群各小型鱼个体之间的距离，对小型鱼个体进行觅食聚群行为、追尾行为和防护聚群行为，更新公告板Y1；
    步骤4.7：凶猛鱼群进行行动：根据各凶猛鱼与小型鱼个体之间的距离，以及凶猛鱼群各凶猛鱼个体之间的距离，对各凶猛鱼个体进行捕食行为、追踪行为和聚群行为，更新公告板Y2；
    步骤4.8：判断当前迭代次数k′是否达到了最大迭代次数K，若是执行步骤4.9，否则，令k′＝k′+1，返回步骤4.6；
    步骤4.9：取公告板Y1和Y2食物浓度最大者及其状态作为优化结果，得到各控制变量的优化取值，即发电机端电压幅值、变压器可调变比及无功补偿装置的无功容量的优化取值；
    步骤4.10：将各控制变量的优化取值传送至电网无功执行模块。
  6. 根据权利要求5所述的基于双鱼群算法的电力无功优化的方法，其特征在于，所述步骤4.3包括以下步骤：
    步骤4.3.1：初始化生成小型鱼个数t1＝0；
    步骤4.3.2：根据电网待优化的各控制变量的取值范围随机生成小型鱼个体：
    X＝[UG11，UG12，...UG1g，KT11，KT12，...，KT1t，QC11，QC12，...，QC1c]＝[x1，x2，...，xs，...，xS]；
    其中，X为小型鱼个体的控制变量取值序列，xs＝xs min+rand(g)×(xs max-xs min)，rand(g)为(0，1)区间的随机数，xs min、xs max分别为对应控制变量的取值下限和取值上限，UG1g、KT1t、QC1c分别为随机生成的小型人工鱼个体中的发电机端电压幅值、变压器可调变比和无功补偿装置的无功容量，xs为随机生成的小型鱼个体的控制变量，1≤s≤S，S＝Ng+Nt+Nc，Ng为电网中发电机个数，Nt为电网中变压器个数，Nc为电网中无功补偿装置的个数；
    步骤4.3.3：对步骤4.3.2中随机生成的小型鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前小型鱼个体的潮流值收敛，则保留该小型鱼个体，令当前生成小型鱼个数t1＝t1+1，执行步骤4.3.4，否则，不保留该小型鱼个体，返回步骤4.3.2；
    步骤4.3.4：若t1≥N1，则得到当前小型鱼群，执行步骤4.4，否则，返回步骤4.3.2。
  7. 根据权利要求5所述的基于双鱼群算法的电力无功优化的方法，其特征在于，所述步骤4.4包括以下步骤：
    步骤4.4.1：初始化生成凶猛鱼个数t2＝0；
    步骤4.4.2：根据电网待优化的各控制变量的取值范围随机生成凶猛鱼个体：
    W＝[UG21，UG22，...UG2g，KT21，KT22，...，KT2t，QC21，QC22，...，QC2c]＝[w1，w2，...，ws，...，wS]；
    其中，W为凶猛鱼个体的控制变量取值序列，ws＝xs max+rand(g)×(xs max-xs min)，UG2g、KT2t、QC2c分别为随机生成的凶猛鱼个体中的发电机端电压幅值、变压器可调变比和无功补偿装置的无功容量，ws为随机生成的凶猛鱼个体的控制变量；
    步骤4.4.3：对步骤4.4.2中随机生成的凶猛鱼个体中各控制变量采用P-Q分解法进行潮流计算，若当前凶猛鱼个体的潮流值收敛，则保留该凶猛鱼个体，令当前生成凶猛鱼个数t2＝t2+1，执行步骤4.4.4，否则，不保留该凶猛鱼个体，返回步骤4.4.2；
    步骤4.4.4：若t2≥N2，则得到当前凶猛鱼群，执行步骤4.5，否则，返回步骤4.4.2。
  8. 根据权利要求5所述的基于双鱼群算法的电力无功优化的方法，其特征在于，所述步骤4.6包括以下步骤：
    步骤4.6.1：确定当前第i个小型鱼个体Xi与各凶猛鱼个体Wj之间的距离D＝{Di1，Di2，...，Dij，...，DiN2}，其中，Dij＝||Xi-Wj||，1＜j＜N2，N2为凶猛鱼个体数；
    步骤4.6.2：若当前第i个小型鱼个体Xi与各凶猛鱼个体Wj之间的距离Dij≤Visual1，执行步骤4.6.3，否则，执行步骤4.6.5；
    步骤4.6.3：确定第i个小型鱼个体Xi感知范围内小型鱼个体总数n，从而确定防护聚集安全位置Xsafe，对第i个小型鱼个体Xi进行防护聚群行为，得到第i个小型鱼个体进行防护聚群行为更新值Xinext1；
    所述防护聚集安全位置Xsafe的计算公式如下：
    Xsafe＝(Xi1+Xi2+...+Xin)×λ/n；
    其中，λ＝(n+p)/n为逃离因子，p为小型鱼个体Xi感知范围Visual1内凶猛鱼个体总数，Xin为小型鱼个体Xi感知范围内的小型鱼个体；
    所述第i个小型鱼个体进行防护聚群行为更新值Xinext1的计算公式如下：
    Xinext1＝Xi+rand(g)×Step1×δ1×(Xsafe-Xi)/||Xsafe-Xi||；
    步骤4.6.4：判断当前小型鱼个体进行防护聚群行为更新值Xinext1是否满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext1的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.3；
    步骤4.6.5：确定第i个小型鱼个体Xi感知范围内小型鱼群中食物浓度FCj′为最大的小型鱼个体Xj′，对第i个小型鱼个体Xi进行追尾行为，得到第i个小型鱼个体Xi进行追尾行为更新值Xinext2；
    所述第i个小型鱼个体Xi进行追尾行为的更新值Xinext2的计算公式如下：
    Xinext2＝Xi+rand(g)×Step1×δ1×(Xj′-Xi)/||Xj′-Xi||；
    步骤4.6.6：确定第i个小型鱼个体Xi感知范围内小型鱼个体总数n，从而确定觅食聚群中心位置Xc1，对第i个小型鱼个体Xi进行觅食聚群行为，得到第i个小型鱼个体进行觅食聚群行为更新值Xinext3；
    所述觅食聚群中心位置Xc1的计算公式如下：
    Xc1＝(Xi1+Xi2+...+Xin)/n；
    所述第i个小型鱼个体进行觅食聚群行为更新值Xinext3的计算公式如下：
    Xinext3＝Xi+rand(g)×Step1×δ1×(Xc1-Xi)/||Xc1-Xi||；
    步骤4.6.7：判断当前小型鱼个体进行追尾行为更新值Xinext2和小型鱼个体进行觅食聚群行为更新值Xinext3是否均满足其约束条件并且潮流收敛，若是，则记录该更新值Xinext2的状态和更新值Xinext3的状态，并计算其食物浓度，更新公告板Y1，执行步骤4.7，否则，返回步骤4.6.5。
  9. 根据权利要求5所述的基于双鱼群算法的电力无功优化的方法，其特征在于，所述步骤4.7包括以下步骤：
    步骤4.7.1：确定当前第i′个凶猛鱼个体Wi′与各小型鱼个体Xj″之间的距离L＝{Li′1，Li′2，...，Li′j″，...，Li′N1}，其中，Li′j″＝||Wi′-Xj″||，1＜j″＜N1，N1为小型鱼个体个数；
    步骤4.7.2：若当前第i′个凶猛人工鱼个体Wi′与各小型人工鱼个体Xj″之间的距离Li′j″≤Visual2执行步骤4.7.3，否则，执行步骤4.7.6；
    步骤4.7.3：确定当前第i′个凶猛鱼个体Wi′感知范围内小型鱼个体总数m，以当前第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼个体Wi′进行捕食行为，将第i′个凶猛鱼个体Wi′更新为小型鱼群中心位置Xc2，得到第i′个凶猛鱼个体Wi′进行捕食行为的更新值Wi′next1＝Xc2；
    所述第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2的计算公式如下：
    Xc2＝(Xi′1+Xi′2+...+Xi′m)/m；
    其中，Xi′m为凶猛鱼个体Wi′感知范围内小型鱼；
    步骤4.7.4：以当前第i′个凶猛鱼个体Wi′感知范围内小型鱼群中心位置Xc2作为当前食物浓度最大值，对第i′个凶猛鱼个体Wi′进行追踪行为，得到第i′个凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2；
    所述第i′个凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2的计算公式如下：
    Wi′next2＝W′i+rand(g)×Step2×δ2×(Xc2-Wi′)/||Xc2-Wi′||；
    步骤4.7.5：判断当前凶猛鱼个体Wi′进行捕食行为的更新值Wi′next1和凶猛鱼个体Wi′进行追踪行为的更新值Wi′next2是否均满足其约束条件并且潮流收敛，若是，更新公告板Y2，执行步骤4.7.6，否则，返回步骤4.7.3；
    步骤4.7.6：确定第i′个凶猛鱼个体Wi′感知范围内凶猛鱼个体总数r，从而确定聚群行为中心位置Wc，对第i′个凶猛鱼个体Wi′进行聚群行为，得到第i′个凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3；
    所述聚群行为中心位置Wc的计算公式如下：
    Wc＝(Wi′1+Wi′2+...+Wi′r)/r；
    其中，Wi′r为第i′个凶猛鱼个体Wi′感知范围内凶猛鱼个体；
    所述第i′个凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3的计算公式如下：
    Wi′next1＝Wi′+rand(g)×Step2×δ2×(Wc-Wi′)/||Wc-Wi′||；
    步骤4.7.7：判断当前凶猛鱼个体Wi′进行聚群行为的更新值Wi′next3是否满足其约束条件并且潮流收敛，若是，则记录该更新值Wi′next3的状态，并计算其食物浓度，更新公告板Y2，执行步骤4.8，否则，返回步骤4.7.6。
PCT/CN2017/088698 2016-12-07 2017-06-16 基于双鱼群算法的电力无功优化系统及方法 Ceased WO2018103294A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/577,127 US10439398B2 (en) 2016-12-07 2017-06-16 Reactive power optimization system and method of power grid based on the double-fish-swarm algorithm

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611114232.5 2016-12-07
CN201611114232.5A CN106549394B (zh) 2016-12-07 2016-12-07 基于双鱼群算法的电力无功优化系统及方法

Publications (1)

Publication Number Publication Date
WO2018103294A1 true WO2018103294A1 (zh) 2018-06-14

Family

ID=58396256

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/088698 Ceased WO2018103294A1 (zh) 2016-12-07 2017-06-16 基于双鱼群算法的电力无功优化系统及方法

Country Status (3)

Country Link
US (1) US10439398B2 (zh)
CN (1) CN106549394B (zh)
WO (1) WO2018103294A1 (zh)

Cited By (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111864690A (zh) * 2020-07-24 2020-10-30 贵州电网有限责任公司 一种基于差分-花粉算法的反时限过流定值优化方法
CN112165102A (zh) * 2020-09-25 2021-01-01 国网青海省电力公司果洛供电公司 一种应用电力电子调压器延长稀疏地区配电网供电半径的方法
CN113067347A (zh) * 2021-05-12 2021-07-02 国网江苏省电力有限公司经济技术研究院 一种交直流电网分布式动态无功补偿优化方法及系统
CN113239623A (zh) * 2021-05-17 2021-08-10 上海交通大学 一种适用于电力装备的故障定位方法
CN113364058A (zh) * 2020-03-05 2021-09-07 中国电力科学研究院有限公司 一种配电网无功控制方法和系统
CN113806539A (zh) * 2021-09-17 2021-12-17 平安科技（深圳）有限公司 一种文本数据增强系统、方法、设备及介质
CN114024321A (zh) * 2021-11-15 2022-02-08 云南电网有限责任公司电力科学研究院 一种配电网无功电压分层自组织控制模型及方法
CN114047372A (zh) * 2021-11-16 2022-02-15 国网福建省电力有限公司营销服务中心 一种基于电压特征的台区拓扑辨识系统
CN114142483A (zh) * 2021-11-23 2022-03-04 国网辽宁省电力有限公司本溪供电公司 一种基于改进粒子群算法的avc差分控制系统及方法
CN114389277A (zh) * 2022-01-12 2022-04-22 福州大学 基于蜉蝣算法的配电网无功优化方法
CN114548467A (zh) * 2020-11-24 2022-05-27 国网冀北电力有限公司经济技术研究院 基于图论和能量平衡的配电网网格划分方法及装置
CN115333117A (zh) * 2022-08-09 2022-11-11 中国电力科学研究院有限公司 城市配电网磁控多功能变压器协同控制方法、设备及终端
CN118429137A (zh) * 2024-05-27 2024-08-02 淮阴工学院 一种林场建筑物运行系统及方法
CN120370677A (zh) * 2025-01-07 2025-07-25 山东大学 基于智能优化的模糊电液集成阀控制方法及系统

Families Citing this family (33)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106549394B (zh) * 2016-12-07 2018-11-27 东北大学 基于双鱼群算法的电力无功优化系统及方法
CN107026458A (zh) * 2017-03-30 2017-08-08 国网上海市电力公司 一种变电站无功投切优化仿真系统及其方法
CN109886446B (zh) * 2018-12-14 2023-04-21 贵州电网有限责任公司 基于改进混沌粒子群算法的电力系统动态经济调度方法
CN109768573B (zh) * 2019-01-29 2022-05-17 三峡大学 基于多目标差分灰狼算法的配电网无功优化方法
CN110086350B (zh) * 2019-05-30 2021-01-15 南京邮电大学 一种基于模拟退火-爬山混合算法的隔离型双向dc-dc效率优化方法
CN110401197B (zh) * 2019-06-27 2021-09-21 国网上海市电力公司 计及分布式电源随机出力的配电网多目标无功优化方法
CN110323764B (zh) * 2019-07-30 2020-12-25 中冶南方都市环保工程技术股份有限公司 基于储能单元与负荷控制的孤网系统稳定运行控制方法
CN110796352B (zh) * 2019-10-18 2021-11-02 广东电网有限责任公司广州供电局 电网网架设计方法、计算机设备和存储介质
CN110994704B (zh) * 2019-12-19 2021-05-14 国电南瑞科技股份有限公司 一种暂态稳定预防控制辅助决策方法、装置及电子设备
CN111525547B (zh) * 2020-03-24 2023-06-16 云南电网有限责任公司临沧供电局 一种基于最优无功补偿的低电压智能治理方法
CN111668852B (zh) * 2020-06-19 2024-10-25 中国石油化工股份有限公司 一种供电系统动态电压稳定优化调控方法
CN111754039B (zh) * 2020-06-23 2023-10-17 北京交通大学 纯电动公交线网综合集成优化设计的方法
CN112183199B (zh) * 2020-08-24 2023-01-06 中国科学院信息工程研究所 一种基于Fish-SVC的2.4GHz频段信号识别方法及系统
CN112332407A (zh) * 2020-10-21 2021-02-05 国网青海省电力公司西宁供电公司 一种10kV电压偏高的优化方法
CN112564138A (zh) * 2020-11-18 2021-03-26 国网山西省电力公司晋城供电公司 一种三相不平衡无功优化方法及其系统
CN112560239B (zh) * 2020-12-03 2022-01-21 广东电网有限责任公司云浮供电局 一种台区线路阻抗的计算方法、计算系统及计算机可读存储介质
CN113344426B (zh) * 2021-06-28 2023-06-06 杭州市电力设计院有限公司余杭分公司 一种储能站规划方法、装置及设备
CN113472015B (zh) * 2021-06-30 2024-04-26 广东电网有限责任公司 配电网自愈网架建模方法、系统、终端设备及存储介质
CN113533946B (zh) * 2021-07-09 2024-05-14 桂林电子科技大学 一种基于kl距离的板级电路测点选择方法
CN113742999B (zh) * 2021-08-12 2023-09-26 中国船舶重工集团公司第七一九研究所 印刷电路板式换热器设计方法及装置
CN113676134B (zh) * 2021-08-16 2024-04-09 阳光智维科技股份有限公司 光伏电站avc指令异常判断方法和装置
CN114386323B (zh) * 2021-12-24 2025-04-25 国网辽宁省电力有限公司电力科学研究院 基于改进蝴蝶算法的含分布式电源的配电网无功优化方法
CN114564807B (zh) * 2022-01-28 2025-05-27 国网河北省电力有限公司衡水供电分公司 主动配电网的优化方法、装置及电子设备
CN115117892B (zh) * 2022-02-22 2024-10-25 中铁第一勘察设计院集团有限公司 一种电气化铁路电力贯通线无功补偿方案优化设计方法
CN114792066B (zh) * 2022-03-15 2024-06-14 湖北工业大学 一种改进麻雀搜索算法的无线充电系统补偿网络优化方法
CN114997013B (zh) * 2022-06-14 2024-07-05 辽宁大学 一种吸能结构智能识别优化方法
CN115271384A (zh) * 2022-07-11 2022-11-01 广西电网有限责任公司 基于蝠鲼觅食优化算法的配电网新能源承载力评估方法
CN115224728A (zh) * 2022-07-19 2022-10-21 贵州大学 一种风力发电系统模型辨识方法和系统
CN115271443A (zh) * 2022-07-28 2022-11-01 长沙理工大学 一种基于模型-数据混合驱动的综合能源系统区间状态估计方法
CN115965288B (zh) * 2022-12-29 2023-11-03 国网湖北省电力有限公司经济技术研究院 基于IWO优化BiLSTM的有源配电网频率安全评估方法
CN117074940B (zh) * 2023-08-22 2024-08-27 湖南铁道职业技术学院 一种轨道交通电机测试系统及方法
CN119051465B (zh) * 2024-08-31 2025-04-08 青岛青电电气集团有限公司 用于超低频介质损耗测量装置的可控整流变压器控制方法
CN120810612B (zh) * 2025-09-12 2025-11-25 安徽网华信息科技有限公司 一种配网运行调控系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102684207A (zh) * 2012-05-23 2012-09-19 甘肃省电力公司电力科学研究院 基于改进人工鱼群混合优化算法的大规模风电并网无功电压优化方法
CN103580023A (zh) * 2013-11-28 2014-02-12 云南电力试验研究院（集团）有限公司电力研究院 一种基于多目标聚类鱼群算法配电网无功优化方法
US20160301210A1 (en) * 2013-11-26 2016-10-13 Siemens Aktiengesellschaft Method for the computer-assisted configuration of an electrical power grid
CN106549394A (zh) * 2016-12-07 2017-03-29 东北大学 基于双鱼群算法的电力无功优化系统及方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100217452A1 (en) * 2009-02-26 2010-08-26 Mccord Alan Overlay packet data network for managing energy and method for using same
US20100217549A1 (en) * 2009-02-26 2010-08-26 Galvin Brian R System and method for fractional smart metering

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102684207A (zh) * 2012-05-23 2012-09-19 甘肃省电力公司电力科学研究院 基于改进人工鱼群混合优化算法的大规模风电并网无功电压优化方法
US20160301210A1 (en) * 2013-11-26 2016-10-13 Siemens Aktiengesellschaft Method for the computer-assisted configuration of an electrical power grid
CN103580023A (zh) * 2013-11-28 2014-02-12 云南电力试验研究院（集团）有限公司电力研究院 一种基于多目标聚类鱼群算法配电网无功优化方法
CN106549394A (zh) * 2016-12-07 2017-03-29 东北大学 基于双鱼群算法的电力无功优化系统及方法

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113364058A (zh) * 2020-03-05 2021-09-07 中国电力科学研究院有限公司 一种配电网无功控制方法和系统
CN111864690A (zh) * 2020-07-24 2020-10-30 贵州电网有限责任公司 一种基于差分-花粉算法的反时限过流定值优化方法
CN112165102A (zh) * 2020-09-25 2021-01-01 国网青海省电力公司果洛供电公司 一种应用电力电子调压器延长稀疏地区配电网供电半径的方法
CN112165102B (zh) * 2020-09-25 2022-08-19 国网青海省电力公司果洛供电公司 一种应用电力电子调压器延长稀疏地区配电网供电半径的方法
CN114548467A (zh) * 2020-11-24 2022-05-27 国网冀北电力有限公司经济技术研究院 基于图论和能量平衡的配电网网格划分方法及装置
CN113067347A (zh) * 2021-05-12 2021-07-02 国网江苏省电力有限公司经济技术研究院 一种交直流电网分布式动态无功补偿优化方法及系统
CN113067347B (zh) * 2021-05-12 2022-09-06 国网江苏省电力有限公司经济技术研究院 一种交直流电网分布式动态无功补偿优化方法及系统
CN113239623A (zh) * 2021-05-17 2021-08-10 上海交通大学 一种适用于电力装备的故障定位方法
CN113806539A (zh) * 2021-09-17 2021-12-17 平安科技（深圳）有限公司 一种文本数据增强系统、方法、设备及介质
CN113806539B (zh) * 2021-09-17 2023-07-07 平安科技（深圳）有限公司 一种文本数据增强系统、方法、设备及介质
CN114024321A (zh) * 2021-11-15 2022-02-08 云南电网有限责任公司电力科学研究院 一种配电网无功电压分层自组织控制模型及方法
CN114024321B (zh) * 2021-11-15 2023-11-10 云南电网有限责任公司电力科学研究院 一种配电网无功电压分层自组织控制模型及方法
CN114047372A (zh) * 2021-11-16 2022-02-15 国网福建省电力有限公司营销服务中心 一种基于电压特征的台区拓扑辨识系统
CN114047372B (zh) * 2021-11-16 2024-03-12 国网福建省电力有限公司营销服务中心 一种基于电压特征的台区拓扑辨识系统
CN114142483A (zh) * 2021-11-23 2022-03-04 国网辽宁省电力有限公司本溪供电公司 一种基于改进粒子群算法的avc差分控制系统及方法
CN114142483B (zh) * 2021-11-23 2023-10-13 国网辽宁省电力有限公司本溪供电公司 一种基于改进粒子群算法的avc差分控制系统及方法
CN114389277A (zh) * 2022-01-12 2022-04-22 福州大学 基于蜉蝣算法的配电网无功优化方法
CN115333117A (zh) * 2022-08-09 2022-11-11 中国电力科学研究院有限公司 城市配电网磁控多功能变压器协同控制方法、设备及终端
CN118429137A (zh) * 2024-05-27 2024-08-02 淮阴工学院 一种林场建筑物运行系统及方法
CN120370677A (zh) * 2025-01-07 2025-07-25 山东大学 基于智能优化的模糊电液集成阀控制方法及系统

Also Published As

Publication number Publication date
US10439398B2 (en) 2019-10-08
CN106549394A (zh) 2017-03-29
US20180309294A1 (en) 2018-10-25
CN106549394B (zh) 2018-11-27

Similar Documents

Publication Publication Date Title
WO2018103294A1 (zh) 基于双鱼群算法的电力无功优化系统及方法
CN110518575B (zh) 基于区域划分的多时间尺度主动配电网电压优化控制方法
CN105574238B (zh) 一种用于在线稳定裕度评估的断面有功控制策略分析方法
CN108199404B (zh) 高渗透率分布式能源系统的谱聚类集群划分方法
EP4304041B1 (de) Verfahren, vorrichtung und elektronische ausrüstung zur blindleistungseinstellung einer integrierten feldstation für eine windkraftanlage, eine photovoltaik und eine energiespeichereinheit
CN105512745A (zh) 基于粒子群—bp神经网络的风电功率区间预测方法
CN105279608A (zh) 基于序关系法-熵权法的县域电网评估方法
CN108270216B (zh) 一种考虑多目标的复杂配电网故障恢复系统及方法
CN113300379B (zh) 一种基于深度学习的电力系统无功电压控制方法及系统
CN104112237A (zh) 一种基于wams的改进遗传算法的电网无功容量优化配置方法
CN112952829B (zh) 电力系统节点运行安全性评估方法、装置、电力系统
CN104578091B (zh) 一种含多源电网的无延时最优无功协调控制系统及方法
CN110783913A (zh) 基于群的计及预想事故集最佳电网拓扑在线优化方法
CN105429134A (zh) 一种基于电力大数据的电网电压稳定性预测方法
CN108364084A (zh) 一种基于信息融合理论的bp神经网络光伏功率预测方法
CN107123983A (zh) 一种基于安全域的变电站接入方案辅助评估方法
CN103618315A (zh) 一种基于bart算法和超吸收壁的电网电压无功优化方法
CN120749726A (zh) 一种基于低压电网多源数据融合的线损治理方法及系统
CN113629788A (zh) 一种电力系统自动电压控制方法
CN110661264A (zh) 基于带惯性权重粒子群算法的安全约束最优潮流计算方法
CN105914752A (zh) 基于快速搜索与密度峰值聚类的主导节点选择方法
CN109888844B (zh) 一种风电场无功电压控制方法
CN117200248A (zh) 基于自适应遗传算法的变电站无功优化配置方法及系统
CN117709640A (zh) 一种基于混合型多属性群决策法的电网规划建设评价方法
CN108075467A (zh) 一种基于多源证据融合的配电网低电压预测方法

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 15577127

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17877481

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 30/10/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 17877481

Country of ref document: EP

Kind code of ref document: A1