CN109768584B - Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization - Google Patents
Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization Download PDFInfo
- Publication number
- CN109768584B CN109768584B CN201811586508.9A CN201811586508A CN109768584B CN 109768584 B CN109768584 B CN 109768584B CN 201811586508 A CN201811586508 A CN 201811586508A CN 109768584 B CN109768584 B CN 109768584B
- Authority
- CN
- China
- Prior art keywords
- value
- particle
- particles
- frequency
- immune
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- 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
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Feedback Control In General (AREA)
Abstract
The invention discloses a micro-grid self-control frequency and pressure regulating method and device based on an immune particle swarm algorithm, and aims to solve the technical problem of ensuring safe and reliable operation of a micro-grid. The method comprises the following steps: collecting current and voltage, detecting frequency, collecting an active power measured value and a reactive power measured value, calculating a fitness function, calculating an optimal droop control coefficient, and calculating an output active power instruction value and an output reactive power instruction value. The device provided by the invention is provided with a data acquisition module, a calculation module, a parameter optimization module, an instruction value acquisition module and a communication module. Compared with the prior art, the droop control parameters are optimized by adopting an immune particle swarm optimization algorithm, when the frequency and the voltage of the power system deviate or the microgrid mode is switched, the droop control parameters of the DG are regulated and controlled in real time by detecting the frequency and the voltage of the microgrid system, and the frequency and the voltage of the microgrid system are dynamically restored to rated values, so that the safe and reliable operation of the microgrid is ensured.
Description
Technical Field
The invention relates to a power control method and device, in particular to a microgrid control method and device.
Background
With the increasingly prominent energy and environmental problems, distributed power generation control methods and devices are continuously developed. The traditional power supply system has large electric energy loss and high pollution, and the requirements of users on the electric energy quality and safety are increasingly strict, so that renewable energy sources gradually play an important role in the sustainable development of social economy. The micro-grid can effectively utilize new energy and renewable energy to generate power, and is a powerful solution for large-scale application of distributed power generation. The development of the microgrid suitable for the basic national conditions of China is a necessary requirement for realizing transformation of a novel energy strategic structure in China, and is also an important power for promoting the sustainable economic development of China and improving the quality of life of people.
Usually, the micro grid is operated in parallel with the main grid, and the frequency and voltage of the grid are supported through rapid electric energy exchange with the grid. When a fault occurs somewhere in the grid, the microgrid needs to be disconnected from the grid and operated independently in order to provide uninterrupted power to the load. The principle of droop control (document 4: zhangmingrui, dumigoo, wang dynasty wave. droop control strategy and parameter selection research [ J ] electrician technical report 2014, 02: 136-144) in the microgrid is to simulate the static characteristic of the power frequency of the traditional synchronous generator, namely when the load power of the power system (system) is suddenly increased, the frequency and the voltage of a corresponding alternating current bus can be reduced; vice versa, when the system load power is suddenly reduced, the ac bus frequency and voltage will increase. Generally, there are two control methods for droop control, one is to adjust the output power by adjusting the voltage frequency, i.e., f-P and U-Q control, and the other is to adjust the output power to control the voltage and frequency. However, droop control has the disadvantages of power regulation at the expense of voltage or frequency, difficulty in maintaining voltage and frequency, and adverse effects on the safe and reliable operation of the microgrid.
Disclosure of Invention
The invention aims to provide a microgrid self-control frequency and pressure modulation method and device based on an immune particle swarm algorithm, and the technical problem to be solved is to ensure the safe and reliable operation of a microgrid.
The invention adopts the following technical scheme: a microgrid self-control frequency and pressure regulation method based on an immune particle swarm algorithm comprises the following steps:
collecting three-phase current and three-phase voltage of a common point, detecting power grid frequency, and collecting active power measured value P output by a distributed power supply in real timeiAnd a reactive power measurement QiCalculating a frequency difference value delta f (t) between the actual frequency and the standard frequency, and calculating a voltage difference value delta V (t) between the bus voltage at the common point and the standard voltage;
calculating a fitness function according to the frequency difference value and the voltage difference value:
in the formula (2), J represents a fitness function, t represents sampling time, t0 and t1 are respectively the starting time and the ending time of each control performance index, alpha is an adjusting factor, and alpha is 0.5;
iterative calculation of optimal droop control coefficient K of distributed power supply by adopting immune particle swarm optimizationviAnd Kfi；
Thirdly, using the frequency difference value, the voltage difference value, the actual active power value and the reactive power value and the optimal droop control coefficient KviAnd KfiCalculating load fluctuation in microgridOr active power instruction value P which needs to be output by distributed power supply in real time when power supply fluctuatesi *And a reactive power command value Qi *：
In formulae (3) and (4), KviAnd KfiThe droop control parameters of reactive power and active power corresponding to the ith DG are represented, i is 1-n, and n is at least 1;
fourthly, sending a power instruction value Pi *And a reactive power command value Qi *And executing by the micro-controller, and dynamically restoring the frequency and the voltage of the power system to rated values.
The standard frequency of the first step of the method is 50Hz, and the standard voltage is 0.4 kV.
Step two of the invention adopts an immune particle swarm algorithm to iteratively calculate the optimal droop control coefficient K of the distributed power supplyviAnd KfiThe method comprises the following steps:
initializing a particle swarm, randomly initializing the speed and the position of N particles, and taking the initialized position of each particle as the optimal position P of each particlebestSetting the optimal position of the initial particle as the optimal position G of the populationbest. Substituting the position of each particle into a designed fitness function to calculate the fitness of each particle, and measuring the quality of the particles according to the fitness value to obtain an individual extreme value and a global extreme value of an initial particle swarm, wherein N is 40;
generating an immune vaccine, and taking the global optimal particles as immune particles;
and thirdly, updating the position and the speed of the particle, and updating the position and the speed of the particle according to a search vector equation:
vk+1＝wvk+c1r1(Pbest-xk)+c2r2(Gbest-xk)
xk+1＝xk+vk+1 (5)
in formula (5), v is a velocity vector of the particle, x is a position vector of the particle, w is an inertia factor, and is a value between [0.3 and 0.5], c1 and c2 are learning factors, c1 is a random number with c2 being 2, r1 and r2 being [0 and 1], and k is a particle number;
updating the population, calculating the fitness of the updated particles according to the fitness function, if the fitness is better than the fitness value before updating, retaining, otherwise, abandoning the particles to generate a new population consisting of N particles,
generating a new population, randomly generating M new particles and N particles in the step four to form N + M particle groups, calculating the selection probability of the generated (N + M) particles according to the following formula, and selecting N new particles according to the probability to form the new population, wherein M is 40;
in the formula (6), YiIs the selection probability of a particle, xiAnd J (x)i) Respectively representing the ith particle and a fitness function value thereof;
sixthly, immunization and immune selection, wherein the immune 'vaccine' generated in the step two is used for carrying out vaccine inoculation on R particles in the population, the particle fitness value after the vaccine is inoculated is calculated according to the fitness function, and immune selection is carried out according to the fitness value; if the fitness value decreases, the particle is retained, otherwise the immune particle is cancelled, thus forming a new generation of population with a particle number N, R being 5;
eighthly, updating the individual extreme value and the global extreme value of the group, substituting the position vector of the particles in the updated group into a fitness calculation function to calculate the fitness of each particle, and updating the individual extreme value and the global extreme value of the group according to the magnitude of the adaptive value;
checking an algorithm termination condition, terminating iteration if the iteration number of the operation does not exceed the maximum allowable iteration number of 100, and obtaining an optimal droop control coefficient of the distributed power supply;
in the iterative calculation of the immune particle swarm optimization, droop control coefficients of a distributed power supply are used as particles, the speed of the particles represents the adjusting direction of a droop control curve of the distributed power supply, the positions of the particles are the droop control coefficients, the optimal positions represent the optimal droop coefficients capable of dynamically adjusting the frequency error and the voltage error with the minimum integration, the population of the particles represents a set formed by different droop control coefficients during the adjustment of the droop curve, an immune vaccine represents the optimal particles obtained in the process of one-time parameter optimization iterative calculation, individual extreme values represent the optimal positions where the initial one particle can dynamically adjust the frequency error and the voltage error with the minimum integration when the iteration changes, and global extreme values represent the optimal positions where all different droop control coefficients can dynamically adjust the frequency error and the voltage error with the minimum integration when the iteration changes.
The iteration number of the operation in the step nine of the invention exceeds the maximum allowable iteration number, and the operation returns to the step two.
A microgrid autonomous frequency and pressure regulating device based on an immune particle swarm algorithm is provided with a data acquisition module, a calculation module, a parameter optimization module, an instruction value acquisition module and a communication module which are sequentially connected;
the data acquisition module acquires three-phase current and three-phase voltage of a common point, detects the frequency of a power grid, and acquires an active power measurement value Pi and a reactive power measurement value Qi output by a distributed power supply in real time;
the calculation module is used for calculating the frequency difference value between the actual frequency and the standard frequency and the voltage difference value between the common point bus voltage and the standard voltage in real time;
the parameter optimization module adopts an immune particle swarm algorithm to iteratively calculate an optimal droop control coefficient;
calculating a fitness function according to the frequency difference value and the voltage difference value:
in the formula (2), J represents a fitness function, t represents sampling time, t0 and t1 are respectively the starting time and the ending time of each control performance index, alpha is an adjusting factor, and alpha is 0.5;
iterative calculation of optimal droop control coefficient K of distributed power supply by adopting immune particle swarm optimizationviAnd Kfi；
Thirdly, using the frequency difference value, the voltage difference value, the actual active power value and the reactive power value and the optimal droop control coefficient KviAnd KfiCalculating the active power instruction value P required to be output by the distributed power supply in real time when the load fluctuation or the power supply fluctuation in the microgrid is calculatedi *And a reactive power command value Qi *：
In formulae (3) and (4), KviAnd KfiThe droop control parameters of reactive power and active power corresponding to the ith DG are represented, i is 1-n, and n is at least 1;
the instruction value acquisition module is used for calculating an active power instruction value P which needs to be output by the distributed power supply in real time when the frequency and the voltage fluctuatei *And a reactive power command value Qi *；
The communication module is used for communicating the microgrid self-control frequency and voltage regulation device based on the immune particle swarm algorithm with the distributed power supply controller and sending active and reactive control instructions to the distributed power supply controller for execution.
The optimal droop control coefficient K of the distributed power supply is iteratively calculated by adopting the immune particle swarm optimizationviAnd KfiThe method comprises the following steps:
firstly, initializing a particle swarm, randomly initializing the speed and the position of N particles, and determining the speed and the position of each particleInitializing the position as the optimal position P for each particlebestSetting the optimal position of the initial particle as the optimal position G of the populationbest. Substituting the position of each particle into a designed fitness function to calculate the fitness of each particle, and measuring the quality of the particles according to the fitness value to obtain an individual extreme value and a global extreme value of an initial particle swarm, wherein N is 40;
generating an immune vaccine, and taking the global optimal particles as immune particles;
and thirdly, updating the position and the speed of the particle, and updating the position and the speed of the particle according to a search vector equation:
vk+1＝wvk+c1r1(Pbest-xk)+c2r2(Gbest-xk)
xk+1＝xk+vk+1 (5)
in formula (5), v is a velocity vector of the particle, x is a position vector of the particle, w is an inertia factor, and is a value between [0.3 and 0.5], c1 and c2 are learning factors, c1 is a random number with c2 being 2, r1 and r2 being [0 and 1], and k is a particle number;
updating the population, calculating the fitness of the updated particles according to the fitness function, if the fitness is better than the fitness value before updating, retaining, otherwise, abandoning the particles to generate a new population consisting of N particles,
generating a new population, randomly generating M new particles and N particles in the step four to form N + M particle groups, calculating the selection probability of the generated (N + M) particles according to the following formula, and selecting N new particles according to the probability to form the new population, wherein M is 40;
in the formula (6), YiIs the selection probability of a particle, xiAnd J (x)i) Respectively representing the ith particle and a fitness function value thereof;
sixthly, immunization and immune selection, wherein the immune 'vaccine' generated in the step two is used for carrying out vaccine inoculation on R particles in the population, the particle fitness value after the vaccine is inoculated is calculated according to the fitness function, and immune selection is carried out according to the fitness value; if the fitness value decreases, the particle is retained, otherwise the immune particle is cancelled, thus forming a new generation of population with a particle number N, R being 5;
eighthly, updating the individual extreme value and the global extreme value of the group, substituting the position vector of the particles in the updated group into a fitness calculation function to calculate the fitness of each particle, and updating the individual extreme value and the global extreme value of the group according to the magnitude of the adaptive value;
checking an algorithm termination condition, terminating iteration if the iteration number of the operation does not exceed the maximum allowable iteration number of 100, and obtaining an optimal droop control coefficient of the distributed power supply;
in the iterative calculation of the immune particle swarm optimization, droop control coefficients of a distributed power supply are used as particles, the speed of the particles represents the adjusting direction of a droop control curve of the distributed power supply, the positions of the particles are the droop control coefficients, the optimal positions represent the optimal droop coefficients capable of dynamically adjusting the frequency error and the voltage error with the minimum integration, the population of the particles represents a set formed by different droop control coefficients during the adjustment of the droop curve, an immune vaccine represents the optimal particles obtained in the process of one-time parameter optimization iterative calculation, individual extreme values represent the optimal positions where the initial one particle can dynamically adjust the frequency error and the voltage error with the minimum integration when the iteration changes, and global extreme values represent the optimal positions where all different droop control coefficients can dynamically adjust the frequency error and the voltage error with the minimum integration when the iteration changes.
The iteration number of the operation in the step nine of the invention exceeds the maximum allowable iteration number, and the operation returns to the step two.
Compared with the prior art, the droop control parameters are optimized by adopting an immune particle swarm optimization algorithm, when the frequency and the voltage of the power system deviate or the microgrid mode is switched, the droop control parameters of the DG are regulated and controlled in real time by detecting the frequency and the voltage of the microgrid system, and the frequency and the voltage of the microgrid system are dynamically restored to rated values, so that the safe and reliable operation of the microgrid is ensured.
Drawings
FIG. 1 is a schematic diagram of an apparatus according to an embodiment of the present invention.
Fig. 2 is a flow chart of a method of an embodiment of the present invention.
FIG. 3 is a flowchart of an immune particle swarm algorithm according to an embodiment of the present invention.
Fig. 4 is a block diagram of a droop control system according to an embodiment of the present invention.
Fig. 5 is a block diagram of an apparatus according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The invention relates to a microgrid self-control frequency and pressure regulation method and device based on an immune particle swarm algorithm, which are used for solving an optimal droop coefficient (coefficient) through an optimization algorithm so as to coordinate the output power of at least 1 Distributed Generation (DG) in a microgrid and improve the control performance of the microgrid. When the frequency and the voltage of the power system (system) deviate or the microgrid mode is switched, the frequency and the voltage of the system are detected, droop control parameters of each distributed power supply DG are regulated and controlled in real time, and the frequency and the voltage of the system are dynamically restored to rated values, so that the safe and reliable operation of the microgrid is guaranteed.
The microgrid system is an alternating current microgrid with a voltage level of 0.4kV and is connected with at least 1 distributed generation unit DG and at least 1 load. The micro-grid system consists of distributed micro-sources, energy storage units, loads and power electronic interface inverter devices, and is connected with a large power grid through a grid-connected breaker. When the grid-connected circuit breaker is in a closed state, the micro-grid system is connected with the large power grid and works in a grid-connected mode, and the large power grid and the distributed micro-source provide energy for a load together; when the grid-connected circuit breaker is in a disconnected state, the microgrid system is isolated from the large power grid and works in an island mode, and only the micro source provides energy for the load to support the load to work.
As shown in fig. 5, the microgrid autonomous frequency and pressure regulating device (apparatus) based on the immune particle swarm optimization of the present invention is provided with a data acquisition module, a calculation module, a parameter optimization module, an instruction value acquisition module and a communication module, which are connected in sequence. The instruction value acquisition module is connected with a droop controller.
The droop controller is provided with a droop control system (document 4), and as shown in fig. 4, the droop control system is provided with droop parameters optimized by an immune particle swarm algorithm and used for calculating the active power and the reactive power values which need to be output in real time by each distributed DG during load fluctuation, power fluctuation or mode switching in the microgrid. The microgrid voltage-modulated power command comprises an active power command value P output by each distributed power supply DGi *And a reactive power reference value Qi *。
As shown in fig. 1, the apparatus of the present invention establishes communication connections with micro controllers MC1, MC2, MC3,. and MCn via control network communication lines, communicating in an IEC-61850/Goose point-to-point fiber optic communication configuration. All DGs are uniformly coordinated with power output by a micro-grid monitoring center (monitoring center) connected with a control network, the monitoring center sends power reference signals to a micro-controller of each DG through the control network to adjust voltage and system frequency, and each DG micro-source uploads real-time voltage and frequency information to the monitoring center through control network communication and then is uploaded to the device provided by the invention through the monitoring center.
The data acquisition module acquires three-phase current and three-phase voltage of PCC through a hard cable, detects the frequency of a power grid, and acquires an active power measurement value Pi and a reactive power measurement value Qi output by each DG in real time.
The calculation module is used for calculating the difference value between the actual frequency and the standard frequency and the difference value between the common point bus voltage and the standard voltage in real time.
And the parameter optimization module adopts an immune particle swarm algorithm to iteratively calculate the optimal droop control coefficient.
The instruction value acquisition module is used for calculating an active power instruction value P which needs to be output by each DG in real time when the frequency and the voltage fluctuate by adopting the droop controlleri *And do not haveWork power command value Qi *。
The communication module is used for communicating the devices with each DG micro-controller, the devices send active and reactive control instructions to the micro-controller for execution, and the system frequency and voltage are dynamically restored to rated values, so that safe and reliable operation of the micro-grid is guaranteed.
A data acquisition module of the device acquires the PCC three-phase current and three-phase voltage of a public node through a hard cable, detects the frequency of a power grid, and acquires an active power measured value P output by each DG in real timeiAnd a reactive power measurement QiThe calculation module calculates the difference value between the actual frequency and the standard frequency and the difference value between the common point bus voltage and the standard voltage; the parameter optimization module calculates a fitness function according to the frequency deviation and the voltage deviation, and the parameter optimization module adopts an immune particle swarm algorithm to iteratively calculate the optimal droop control coefficient of each DG; the instruction value obtaining module inputs the frequency difference value, the voltage difference value, the actual active power value and the reactive power value into the droop controller, and calculates the active power instruction value P required to be output by each DG in real time when the load fluctuation or the power supply fluctuation in the microgridi *And a reactive power command value Qi *(ii) a The communication module issues active and reactive control instructions to each DG micro-controller for execution, and the system frequency and voltage are dynamically restored to rated values, so that safe and reliable operation of the micro-grid is guaranteed.
According to the microgrid automatic control frequency and pressure regulation method (method) based on the immune particle swarm algorithm, the droop control parameters of the microgrid are optimized by the immune particle swarm algorithm, coordination among a plurality of DGs in the microgrid is realized, the numerical value of power matching is automatically calculated so that the microgrid is fast and stable, the stability of a power supply system under different operating conditions of the power system is realized, and the power quality is improved. The method comprises the following steps:
the data acquisition module acquires three-phase current and three-phase voltage of PCC through a hard cable, detects power grid frequency and acquires active power measured value P output by each DG in real timeiAnd a reactive power measurement QiCalculating the difference value delta f (t) (frequency deviation, frequency difference value) between the actual frequency and the standard frequency of 50Hz, and calculating the common point bus voltageDifference Δ V (t) (voltage deviation, voltage difference) of standard voltage 0.4 kV.
And secondly, calculating a fitness function by the calculating module according to the frequency deviation and the voltage deviation, and iteratively calculating the optimal droop control coefficient of each DG by the parameter optimizing module by adopting an immune particle swarm algorithm.
The cost function is minimized by using time-weighted absolute error integral ITAE [ document 1: ziming, sons shich, Qinchang [ J ] fractional order PID parameter tuning based on optimal osustaloup [ J ] control engineering, 2012, 19 (2): 283-285 ] is a fitness function of an immune algorithm:
the frequency deviation and the voltage deviation of the microgrid system are used as the optimizing performance indexes of each distributed DG droop coefficient, and the formula (1) can be designed as follows:
in the step (2), J represents a fitness function which is used as the fitness function of the immune algorithm, and the smaller the fitness function value is, the higher the fitness is, and the more excellent the parameters are; t represents a sampling time, t0 and t1 are respectively a start time and an end time for calculating each control performance index, the control performance index is a product of time t and a system absolute error e (t), and e (t) is e (t) ═ α | Δ f (t) | + (1- α) | Δ v (t) |, wherein α is an adjusting factor, the weight of frequency and voltage approximation can be adjusted, and is generally equal to 0.5; Δ f (t) represents a frequency deviation, and Δ v (t) represents a voltage deviation.
Iterative computation of optimal droop control coefficients K of all DGs by adopting immune particle swarm optimizationviAnd Kfi，KviAnd KfiAnd the droop control parameters represent the reactive power and the active power corresponding to the ith DG.
Thirdly, the instruction value obtaining module inputs the frequency difference value, the voltage difference value, the actual active power value and the reactive power value into the droop controller, namely the frequency difference value and the electricityDifferential pressure value, actual active power value and reactive power value, and optimal droop control coefficient KviAnd KfiCalculating the active power instruction value P which needs to be output by each DG in real time when the load fluctuation or the power supply fluctuation in the microgrid is carried outi *And a reactive power command value Qi *. Active power command value Pi *And a reactive power command value Qi *The calculation formulas are respectively as follows:
in formulae (3) and (4), KviAnd KfiThe droop control parameters of the reactive power and the active power corresponding to the ith DG are expressed, i is 1-n, n is at least 1, PiActive power measurement, Q, output in real time for DGiThe reactive power measurement is output in real time for DG.
Fourthly, sending power instruction value P under the communication modulei *And a reactive power command value Qi *And executing by the micro-controller, and dynamically restoring the system frequency and the voltage to rated values.
The immune particle swarm algorithm has a superior effect on the control performance index compared with the simple particle swarm algorithm (document 2: wangwei, old, control of polymerization temperature based on PID tuning of immune particle swarm algorithm [ J ] information communication, 2017 (4): 9-10), using the droop control parameters of each DG micro-source as particles of a particle swarm optimization algorithm, optimizing the droop control parameters by adopting the particle swarm optimization algorithm based on immune selection, using the frequency error and the voltage error of the micro-grid system as performance indexes, designing a fitness function of the performance indexes of the comprehensive evaluation system for evaluating the advantages and disadvantages of the particles, continuously searching the optimal particles with the highest fitness by adopting an iterative algorithm, searching the optimal droop control parameters when the system frequency and the voltage fluctuate or the micro-grid is switched, ensuring the system performance indexes to be optimal, namely the frequency error and the voltage error are comprehensively minimum, and dynamically restoring the system frequency and the voltage to rated values, thereby ensuring the safe and reliable operation of the micro-grid.
In the iterative calculation by adopting an immune particle swarm algorithm, droop control coefficients of each distributed DG are taken as particles, the speed of each particle represents the direction of regulating a droop control curve of a DG power supply, the speed of each particle is changed, namely the regulating direction of the droop control coefficient is changed, the position of each particle is a corresponding droop control coefficient, the optimal position represents the optimal droop control coefficient with minimum comprehensive dynamic regulation frequency error and voltage error (minimum system absolute error), the population of the particles represents a set formed by different droop control coefficients when the droop curve is regulated, the immune vaccine represents the optimal particle obtained in the process of one-time parameter optimization iterative calculation, namely the optimal droop control coefficient, an individual extreme value represents the optimal position with minimum comprehensive dynamic regulation frequency error and voltage error when an initial particle (droop control coefficient) is subjected to iterative variation, the global extreme value represents the optimal position where all different droop control coefficients can dynamically adjust the frequency error and the voltage error to be minimum in synthesis (the system absolute error is minimum) when the droop control coefficients are changed in an iteration mode.
As shown in fig. 3, an immune particle swarm algorithm was used [ document 3: eagle, xieli immune particle swarm optimization algorithm [ J ] computer engineering and application, 2004, 41 (6): 4-6, iteratively calculating the optimal droop control coefficient, comprising the following steps:
initializing a particle swarm, randomly initializing the speed and the position of N particles, and taking the initialized position of each particle as the optimal position P of each particlebestSetting the optimal position of the initial particle as the optimal position G of the populationbest. Substituting the position of each particle into a designed fitness function to calculate the fitness of each particle, and measuring the quality of the particles according to the fitness value to obtain an individual extreme value and a global extreme value of the initial particle swarm. N is 40.
And secondly, generating an immune vaccine, and taking the globally optimal (highest fitness) particles as immune particles.
And thirdly, updating the position and the speed of the particle, and updating the position and the speed of the particle according to a search vector equation:
vk+1＝wvk+c1r1(Pbest-xk)+c2r2(Gbest-xk)
xk+1＝xk+vk+1 (5)
in formula (5), v is a velocity vector of the particle; x is the position vector of the particle; w is an inertia factor and takes a value between [0.3 and 0.5 ]; c1 and c2 are learning factors, and c1 is generally equal to c2 is equal to 2; r1 and r2 are random numbers of [0,1 ]; k is the number of particles.
And fourthly, updating the population, calculating the fitness of the updated particles according to the fitness function, if the fitness is better than the fitness value before updating, retaining, and otherwise, discarding the particles, so that a new population consisting of N particles can be generated.
Fifthly, generating a new population: randomly generating M new particles and N particles in the fourth step to form N + M particle groups, calculating the selection probability of the generated (N + M) particles according to the following formula, and selecting N new particles according to the probability to form a new group. M is 40.
In the formula (6), YiIs the selection probability of a particle, xiAnd J (x)i) Respectively representing the ith particle and the fitness function value thereof.
And sixthly, immunization and immune selection, wherein the immune 'vaccine' generated in the step two is used for carrying out vaccination (correction) on R particles in the population, the particle fitness value after the vaccination is calculated according to the fitness function, and the immune selection is carried out according to the fitness value. If the fitness value decreases, the particle is retained, otherwise the immune particle is cancelled, thus forming a new generation population with a particle number N. R is 5.
And eighthly, updating the individual extreme value and the global extreme value of the group, substituting the position vector of the particles in the updated group into a fitness calculation function to calculate the fitness of each particle, and updating the individual extreme value and the global extreme value of the group according to the magnitude of the adaptive value.
And ninthly, checking an algorithm termination condition, and terminating iteration if the iteration number of the operation does not exceed the maximum allowable iteration number to obtain the optimal droop control coefficient of the DG. And if the iteration number of the operation exceeds the maximum allowable iteration number, returning to the step two to continue the iterative computation, wherein in the embodiment, the maximum allowable iteration number is 100.
The particles consist of droop control coefficients of distributed DGs of the microgridvi，Kfi]，KviAnd KfiIndicating the corresponding droop parameter for the ith DG.
The particle swarm optimization is a quick and effective optimization algorithm, has the advantages of simplicity, easiness in implementation, high convergence speed and few parameters needing to be adjusted, but in the process of continuously learning and evolving particles, the convergence speed of the particles is higher and higher, the particles are easy to fall into local optimum, and the early-maturing condition occurs.
According to the method, droop control parameters are optimized by adopting an immune particle swarm optimization algorithm, when the frequency and the voltage of a power system deviate or a microgrid mode is switched, the droop control parameters of a DG power supply are regulated and controlled in real time by detecting the frequency and the voltage of the system, and the frequency and the voltage of the system are dynamically restored to rated values, so that the safe and reliable operation of a microgrid is ensured.
The particle swarm PSO algorithm is an algorithm for performing global optimal search on a space. Its advantages are less parameters, high convergence speed and high global optimization power. However, the PSO algorithm has the disadvantages that the convergence speed is slow and local optimization easily occurs at the late stage of optimization. The immune algorithm has the advantages of immunity and heredity, so that the defects of the PSO algorithm can be overcome by combining the immune algorithm with the PSO algorithm. And taking the antigen as a problem to be solved, taking the antibody as particles of the particle swarm as a solution of the problem, and constructing a fitness function to search for the optimal antibody. Immune particle swarm algorithm [ reference 2: wangwei, old, control of polymerization temperature based on PID tuning of immune particle swarm algorithm [ J ] information communication, 2017 (4): 9-10. Document 3: eagle, xieli immune particle swarm optimization algorithm [ J ] computer engineering and application, 2004, 41 (6): 4-6, the diversity of the particle group can be maintained through immunological memory and self-regulation in the immune algorithm, and the antibody can evolve towards one direction through the PSO algorithm, thereby converging towards the optimal solution in a few generations. Therefore, the PSO algorithm and the immune algorithm can be combined to be used for optimizing and solving the droop control parameters, when the voltage and the frequency of a power grid fluctuate or the microgrid mode is switched, the optimal droop control parameters are quickly solved through the immune particle swarm optimization algorithm, the droop curve parameters of the micro source are adjusted and controlled in real time, and the frequency and the voltage of the system are quickly recovered to a rated value, so that the safe and reliable operation of the microgrid is guaranteed.
Claims (7)
1. A microgrid self-control frequency and pressure regulation method based on an immune particle swarm algorithm comprises the following steps:
the method comprises the steps of collecting three-phase current and three-phase voltage of a common Point (PCC), detecting the frequency of a power grid, and collecting active power measured value P output by a distributed power supply (DG) in real timeiAnd a reactive power measurement QiCalculating a frequency difference value delta f (t) between the actual frequency and the standard frequency, and calculating a voltage difference value delta V (t) between the bus voltage at the common point and the standard voltage;
calculating a fitness function according to the frequency difference value and the voltage difference value:
in the formula (2), J represents a fitness function, t represents a sampling time, t0 and t1 are respectively a start time and an end time of each control performance index, a is an adjustment factor, and a is 0.5;
iterative computation of optimal droop control coefficients for Distributed Generators (DGs) using immune particle swarm optimizationKviAnd Kfi；
Thirdly, using the frequency difference value, the voltage difference value, the real-time output active power measured value and the real-time output reactive power measured value, and the optimal droop control coefficient KviAnd KfiCalculating an active power instruction value P required to be output by a Distributed Generation (DG) in real time when load fluctuation or power supply fluctuation in a microgrid is carried outi *And a reactive power command value Qi *：
In formulae (3) and (4), KviAnd KfiThe droop control parameter is used for representing a reactive power instruction value and an active power instruction value corresponding to the ith distributed power supply (DG), i is 1-n, and n is at least 1;
fourthly, issuing an active power instruction value Pi *And a reactive power command value Qi *And a micro-controller to the distributed power supply executes to dynamically restore the frequency and the voltage of the power system to rated values.
2. The microgrid self-control frequency and voltage regulation method based on the immune particle swarm optimization algorithm, which is characterized in that: the standard frequency of the first step is 50Hz, and the standard voltage is 0.4 kV.
3. The microgrid self-control frequency and voltage regulation method based on the immune particle swarm optimization algorithm, which is characterized in that: the second step adopts an immune particle swarm algorithm to iteratively calculate the optimal droop control coefficient K of the Distributed Generation (DG)viAnd KfiThe method comprises the following steps:
initializing a particle group, randomly initializing the speed and the position of N particles, and setting the initialized position of each particle as eachOptimum position P of individual particlebestSetting the optimal position of the initial particle as the optimal position G of the populationbestSubstituting the position of each particle into a designed fitness function to calculate the fitness value of each particle, and measuring the quality of the particles according to the fitness value to obtain an individual extreme value and a global extreme value of an initial particle swarm, wherein N is 40;
(II) generating an immune vaccine, and taking the global optimal particles as immune particles;
and (III) updating the position and the speed of the particle, and updating the position and the speed of the particle according to a search vector equation:
vk+1＝wvk+c1r1(Pbest-xk)+c2r2(Gbest-xk)
xk+1＝xk+vk+1 (5)
in formula (5), v is a velocity vector of the particle, x is a position vector of the particle, w is an inertia factor, and is a value between [0.3 and 0.5], c1 and c2 are learning factors, c1 is a random number with c2 being 2, r1 and r2 being [0 and 1], and k is a particle number;
(IV) updating the population, calculating the fitness value of the updated particles according to the fitness function, if the fitness value is superior to the fitness value before updating, reserving the particles, otherwise, abandoning the particles to generate a new population consisting of N particles,
generating a new population, randomly generating M new particles and N particles in the step (IV) to form N + M particle groups, calculating the selection probability of the generated N + M particles according to the following formula, and selecting N new particles according to the probability to form the new population, wherein M is 40;
in the formula (6), YiIs the selection probability of a particle, xiAnd J (x)i) Respectively representing the ith particle and a fitness function value thereof;
(VI), immunization and immune selection, wherein the immune 'vaccine' generated in the step (II) is used for carrying out vaccination on R particles in the population, the particle fitness value after the vaccination is calculated according to the fitness function, and immune selection is carried out according to the fitness value; if the fitness value decreases, the particle is retained, otherwise the immune particle is cancelled, thus forming a new generation of population with a particle number N, R being 5;
updating the individual extreme value and the global extreme value of the group, substituting the position vector of the particles in the updated group into a fitness function to calculate the fitness value of each particle, and updating the individual extreme value and the global extreme value of the group according to the fitness value;
checking an algorithm termination condition, and terminating iteration if the iteration number of the operation does not exceed the maximum allowable iteration number of 100 to obtain an optimal droop control coefficient of the Distributed Generation (DG);
in the iterative calculation of the immune particle swarm algorithm, droop control coefficients of a Distributed Generator (DG) are used as particles, the speed of the particles represents the adjusting direction of a droop control curve of the Distributed Generator (DG), the positions of the particles are the droop control coefficients, the optimal positions represent the optimal droop control coefficients capable of dynamically adjusting the minimum frequency error and voltage error, the population of the particles represents a set formed by different droop control coefficients during the adjustment of the droop curve, an immune vaccine represents the optimal particles obtained in the process of one-time parameter optimization iterative calculation, individual extreme values represent the optimal positions where an initial particle can dynamically adjust the minimum frequency error and voltage error during the iterative change, and global extreme values represent the optimal positions where all different droop control coefficients can dynamically adjust the minimum frequency error and voltage error during the iterative change.
4. The microgrid self-control frequency and voltage regulation method based on the immune particle swarm optimization algorithm, which is characterized in that: and (5) the iteration number of the operation in the step (eight) exceeds the maximum allowable iteration number, and the step (two) is returned.
5. The utility model provides a microgrid is from controlling frequency modulation pressure regulating device based on immune particle swarm algorithm which characterized in that: the microgrid self-control frequency and pressure regulating device based on the immune particle swarm algorithm is provided with a data acquisition module, a calculation module, a parameter optimization module, an instruction value acquisition module and a communication module which are sequentially connected;
the data acquisition module acquires three-phase current and three-phase voltage of a common Point (PCC), detects the frequency of a power grid, and acquires an active power measurement value Pi and a reactive power measurement value Qi output by a distributed power supply (DG) in real time;
the calculation module is used for calculating the frequency difference value between the actual frequency and the standard frequency and the voltage difference value between the common point bus voltage and the standard voltage in real time;
the parameter optimization module adopts an immune particle swarm algorithm to iteratively calculate an optimal droop control coefficient;
calculating a fitness function according to the frequency difference value and the voltage difference value:
in the formula (2), J represents a fitness function, t represents a sampling time, t0 and t1 are respectively a start time and an end time of each control performance index, a is an adjustment factor, a is 0.5, Δ f (t) is a frequency difference value between an actual frequency and a standard frequency, and Δ v (t) is a voltage difference value between a common-point bus voltage and a standard voltage;
iterative computation of optimal droop control coefficient K of Distributed Generation (DG) by adopting immune particle swarm optimizationviAnd Kfi；
Using frequency difference value, voltage difference value, real-time output active power measured value and reactive power measured value, optimum droop control coefficient KviAnd KfiCalculating an active power instruction value P required to be output by a Distributed Generation (DG) in real time when load fluctuation or power supply fluctuation in a microgrid is carried outi *And a reactive power command value Qi *：
In formulae (3) and (4), KviAnd KfiThe droop control parameter is used for representing a reactive power instruction value and an active power instruction value corresponding to the ith distributed power supply (DG), i is 1-n, and n is at least 1;
the instruction value acquisition module is used for calculating an active power instruction value P which needs to be output by a Distributed Generation (DG) in real time when frequency and voltage fluctuatei *And a reactive power command value Qi *；
The communication module is used for communicating the microgrid autonomous frequency and voltage modulation and regulation device based on the immune particle swarm algorithm with a Distributed Generation (DG) controller and sending an active power instruction value and a reactive power instruction value to the Distributed Generation (DG) controller for execution.
6. The microgrid self-control frequency and voltage regulation device based on the immune particle swarm optimization algorithm is characterized in that: the optimal droop control coefficient K of the Distributed Generation (DG) is calculated by adopting an immune particle swarm algorithm in an iterative mannerviAnd KfiThe method comprises the following steps:
initializing a particle group, randomly initializing the speed and the position of N particles, and taking the initialized position of each particle as the optimal position P of each particlebestSetting the optimal position of the initial particle as the optimal position G of the populationbestSubstituting the position of each particle into a designed fitness function to calculate the fitness value of each particle, and measuring the quality of the particles according to the fitness value to obtain an individual extreme value and a global extreme value of an initial particle swarm, wherein N is 40;
(II) generating an immune vaccine, and taking the global optimal particles as immune particles;
and (III) updating the position and the speed of the particle, and updating the position and the speed of the particle according to a search vector equation:
vk+1＝wvk+c1r1(Pbest-xk)+c2r2(Gbest-xk)
xk+1＝xk+vk+1 (5)
in formula (5), v is a velocity vector of the particle, x is a position vector of the particle, w is an inertia factor, and is a value between [0.3 and 0.5], c1 and c2 are learning factors, c1 is a random number with c2 being 2, r1 and r2 being [0 and 1], and k is a particle number;
(IV) updating the population, calculating the fitness value of the updated particles according to the fitness function, if the fitness value is superior to the fitness value before updating, reserving the particles, otherwise, abandoning the particles to generate a new population consisting of N particles,
generating a new population, randomly generating M new particles and N particles in the step (IV) to form N + M particle groups, calculating the selection probability of the generated N + M particles according to the following formula, and selecting N new particles according to the probability to form the new population, wherein M is 40;
in the formula (6), YiIs the selection probability of a particle, xiAnd J (x)i) Respectively representing the ith particle and a fitness function value thereof;
(VI), immunization and immune selection, wherein the immune 'vaccine' generated in the step (II) is used for carrying out vaccination on R particles in the population, the particle fitness value after the vaccination is calculated according to the fitness function, and immune selection is carried out according to the fitness value; if the fitness value decreases, the particle is retained, otherwise the immune particle is cancelled, thus forming a new generation of population with a particle number N, R being 5;
updating the individual extreme value and the global extreme value of the group, substituting the position vector of the particles in the updated group into a fitness function to calculate the fitness value of each particle, and updating the individual extreme value and the global extreme value of the group according to the fitness value;
checking an algorithm termination condition, and terminating iteration if the iteration number of the operation does not exceed the maximum allowable iteration number of 100 to obtain an optimal droop control coefficient of the Distributed Generation (DG);
in the iterative calculation of the immune particle swarm algorithm, droop control coefficients of a Distributed Generator (DG) are used as particles, the speed of the particles represents the adjusting direction of a droop control curve of the Distributed Generator (DG), the positions of the particles are the droop control coefficients, the optimal positions represent the optimal droop control coefficients capable of dynamically adjusting the minimum frequency error and voltage error, the population of the particles represents a set formed by different droop control coefficients during the adjustment of the droop curve, an immune vaccine represents the optimal particles obtained in the process of one-time parameter optimization iterative calculation, individual extreme values represent the optimal positions where an initial particle can dynamically adjust the minimum frequency error and voltage error during the iterative change, and global extreme values represent the optimal positions where all different droop control coefficients can dynamically adjust the minimum frequency error and voltage error during the iterative change.
7. The microgrid self-control frequency and voltage regulation device based on the immune particle swarm optimization algorithm is characterized in that: and (5) the iteration number of the operation in the step (eight) exceeds the maximum allowable iteration number, and the step (two) is returned.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811586508.9A CN109768584B (en) | 2018-12-24 | 2018-12-24 | Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811586508.9A CN109768584B (en) | 2018-12-24 | 2018-12-24 | Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109768584A CN109768584A (en) | 2019-05-17 |
| CN109768584B true CN109768584B (en) | 2020-11-24 |
Family
ID=66450944
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811586508.9A Active CN109768584B (en) | 2018-12-24 | 2018-12-24 | Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109768584B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110350671B (en) * | 2019-06-14 | 2023-01-03 | 武汉大学 | Wireless induction link parameter real-time estimation method based on particle swarm optimization |
| CN110890768A (en) * | 2019-12-21 | 2020-03-17 | 东北电力大学 | Power distribution method under low-voltage alternating-current micro-grid island mode |
| CN113036774A (en) * | 2019-12-25 | 2021-06-25 | 中国电力科学研究院有限公司 | Power distribution network reactive compensation optimization method and device |
| CN111244935B (en) * | 2019-12-27 | 2021-12-10 | 国网山西省电力公司大同供电公司 | Energy efficiency optimization method based on fine adjustment of temperature set point of water heater |
| CN112612019B (en) * | 2020-11-30 | 2024-02-02 | 中国人民解放军空军工程大学 | Active anti-interference method based on frequency control array phase center |
| CN113505873A (en) * | 2021-05-31 | 2021-10-15 | 南昌大学 | Method for setting control parameters of double parallel inverters based on sailfish algorithm |
| CN113241780B (en) * | 2021-06-11 | 2023-01-20 | 云南电网有限责任公司电力科学研究院 | Power grid secondary frequency modulation control method and system based on bald eagle search algorithm |
| CN113359475B (en) * | 2021-07-07 | 2023-11-14 | 宁波奥克斯电气股份有限公司 | Servo control parameter optimization method and device and servo controller |
| CN114513017B (en) * | 2022-02-15 | 2022-08-12 | 山东大学 | Distributed tracking method and system for power distribution network instructions of alternating current-direct current micro-grid |
| CN114899844B (en) * | 2022-06-13 | 2023-06-20 | 华能国际电力股份有限公司德州电厂 | Primary frequency modulation control system of small-disturbance thermal power generating unit |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8606424B2 (en) * | 2011-04-05 | 2013-12-10 | King Fahd University Of Petroleum And Minerals | Particle swarm optimization system and method for microgrids |
| CN104333037B (en) * | 2014-11-02 | 2016-08-24 | 中国科学院电工研究所 | Wind accumulation participates in the cooperative control method of power system frequency modulation and voltage modulation |
| CN105515006A (en) * | 2016-01-20 | 2016-04-20 | 上海电力学院 | Micro-grid multi-master-slave mixed control method based on improved droop control |
| CN105610202B (en) * | 2016-02-04 | 2021-01-08 | 中国电力科学研究院 | Multi-agent system based active control method for autonomous AC/DC micro-grid |
| CN106487042B (en) * | 2016-11-22 | 2019-01-15 | 合肥工业大学 | A kind of Multiple Time Scales micro-capacitance sensor voltage power-less optimized controlling method |
| CN107069775A (en) * | 2017-04-11 | 2017-08-18 | 华北水利水电大学 | A kind of isolated island micro-capacitance sensor frequency control method based on random accelerated particle group's algorithm |
-
2018
- 2018-12-24 CN CN201811586508.9A patent/CN109768584B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109768584A (en) | 2019-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109768584B (en) | Micro-grid self-control frequency and pressure regulation method and device based on immune particle swarm optimization | |
| Zhang et al. | Data-driven control for interlinked AC/DC microgrids via model-free adaptive control and dual-droop control | |
| CN109586269A (en) | Consider the direct-current grid virtual inertia control method and system of parameter self-optimization | |
| CN108767869B (en) | Static reactive power compensator voltage adjusting method based on artificial neural network | |
| CN110265991B (en) | Distributed coordination control method for direct-current micro-grid | |
| CN106451418B (en) | Photovoltaic plant divides group's equivalent modeling method online | |
| CN110768239B (en) | Virtual direct current motor control method based on P-U droop characteristic | |
| CN104769802A (en) | Method for the computer-aided control of the power in an electrical grid | |
| Song et al. | Security-constrained line loss minimization in distribution systems with high penetration of renewable energy using UPFC | |
| CN113541146A (en) | Power flow calculation optimization method for power system considering distributed power supply | |
| Egbomwan et al. | Twin delayed deep deterministic policy gradient (TD3) based virtual inertia control for inverter-interfacing DGs in microgrids | |
| Colak et al. | Fuzzy logic and artificial neural network based grid-interactive systems for renewable energy sources: a review | |
| CN115764989A (en) | Virtual synchronous generator system based on network-building type current converter | |
| WO2020018313A1 (en) | Control for energy resources in a microgrid | |
| CN106950831B (en) | Reactive compensation method for off-line optimization/on-line switching | |
| Lu et al. | Low-Frequency Oscillation Analysis of Grid-Connected VSG System Considering Multi-Parameter Coupling. | |
| CN113162063B (en) | Design method of multi-direct-current coordination controller for inhibiting ultralow frequency oscillation | |
| Liu et al. | A Non-linear Decentralized Secondary Frequency Control of islanded Microgrid via Adaptive $ L_ {2} $ Gain Disturbance Attenuation | |
| CN115313510A (en) | Adaptive reactive compensation photovoltaic inverter control method and system | |
| Li et al. | Virtual inertia control of the virtual synchronous generator: A review | |
| CN108494017A (en) | A kind of Autonomous Model micro-grid system distributed and coordinated control method based on inverter | |
| Kandasamy et al. | Artificial neural network based intelligent controller design for grid-tied inverters of microgrid under load variation and disturbance | |
| Du et al. | Research on Autonomous Frequency Modulation and Voltage Regulation Method of Microgrid Based on Immunity Selection-particle Swarm Optimization (ISPSO) | |
| CN112467778A (en) | Method for improving robustness of multi-terminal flexible direct-current transmission control system | |
| Li et al. | Simulation Research on Virtual Synchronization Control Strategy of Optical Storage Integration Based on DIgSILENT Platform |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |