Offshore wind power plant reactive power optimization method and system based on genetic algorithm

Abstract

The invention relates to the technical field of reactive power optimization, and discloses an offshore wind power plant reactive power optimization method and system based on a genetic algorithm, and the method comprises the steps: constructing a reactive power optimization model of an offshore wind power plant, and setting a target function of the reactive power optimization model, the objective function comprises a load node voltage deviation sub-objective, an active power loss sub-objective and an adjustment time sub-objective; calculating a grid-connected point voltage deviation value based on the output power of the offshore wind plant and the parameters of the grid-connected transformer, and adaptively adjusting the weight coefficient of each sub-target based on the grid-connected point voltage deviation value to obtain an adjusted target function; and solving the adjusted target function through the reactive power optimization model based on a genetic algorithm, and carrying out reactive power optimization on the offshore wind plant based on a solving result of the adjusted target function. The priority of each optimization target is adaptively adjusted according to the working condition of the power grid so as to meet the real-time requirement of the power grid and ensure the safety, stability and economical efficiency of the novel power system.

Description

Technical Field

[0001] The present application relates to the technical field of reactive power optimization, and in particular to a reactive power optimization method and system for an offshore wind farm based on a genetic algorithm. Background Art

[0002] Wind power has developed rapidly as a clean energy source. With the continuous increase in the scale of wind power grid connection, the reactive power active support capability of traditional synchronous power sources cannot guarantee the safe and stable operation of the power system, especially as offshore wind farms are in a large-scale, centralized development stage. It is imperative to improve the reactive power active support capability of offshore wind farms.

[0003] The optimization objective function in the relevant technology is a fixed weight coefficient. As the proportion of renewable energy power generation in the power grid gradually increases, the operating conditions of the new power system will change frequently with the fluctuations in the output of renewable energy power generation. The weight coefficients of each optimization objective are fixed values, which will lead to the inability to adjust in real time according to the power grid conditions, affecting the safe and economical operation of the power system. Summary of the invention

[0004] The present application provides a reactive power optimization method and system for offshore wind farms based on genetic algorithms, which solves the technical problem that the power grid operating conditions cannot be adjusted in real time. When dealing with complex and changeable operating conditions, appropriate control decisions can be made based on real-time operating conditions, thereby achieving coordinated optimization of offshore wind farms.

[0005] In order to achieve the above objectives, the main technical solutions adopted in this application include:

[0006] In a first aspect, an embodiment of the present application provides a reactive power optimization method for an offshore wind farm based on a genetic algorithm, the method comprising:

[0007] Constructing a reactive power optimization model for an offshore wind farm, and setting an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective, and a regulation time sub-objective;

[0008] The grid connection point voltage deviation is calculated based on the output power of the offshore wind farm and the grid-connected transformer parameters, and the weight coefficient of each sub-objective is adaptively adjusted based on the grid connection point voltage deviation to obtain an adjusted objective function;

[0009] The adjusted objective function is solved by the reactive power optimization model based on a genetic algorithm, and reactive power optimization is performed on the offshore wind farm based on the solution result of the adjusted objective function.

[0010] The present application combines the real-time grid conditions, calculates the grid connection point voltage deviation through the output power of the offshore wind farm and the grid-connected transformer parameters, and then determines the optimization direction of the reactive power optimization model, and adaptively adjusts the weight coefficient corresponding to each optimization sub-objective, that is, adaptively adjusts the priority of each optimization objective, overcomes the technical problem in the related technology that the weight coefficient of each optimization objective is a fixed value and cannot be adjusted in real time, realizes the coordinated optimization of the active network loss and the grid connection point voltage of the offshore wind farm, and improves the operating efficiency and voltage stability of the offshore wind farm.

[0011] Optionally, the adaptively adjusting the weight coefficient of each sub-objective based on the grid connection point voltage deviation to obtain an adjusted objective function includes:

[0012] When the grid connection point voltage deviation is greater than a preset deviation range, the weight coefficient of the load node voltage deviation sub-target is adaptively increased to obtain an objective function containing the increased weight coefficient of the load node voltage deviation sub-target;

[0013] When the grid connection point voltage deviation is less than a preset deviation range, adaptively increasing the weight coefficients of the active grid loss sub-goal and the adjustment time sub-goal;

[0014] To obtain an objective function containing the weight coefficients of the active network loss sub-objective and the adjustment time sub-objective after each of them is increased.

[0015] Optionally, solving the adjusted objective function by using the reactive power optimization model based on a genetic algorithm, and performing reactive power optimization on the offshore wind farm based on a solution result of the adjusted objective function, comprises:

[0016] The initial population is randomly generated, and SVG and transformer ratio are set as control variables;

[0017] Calculating the fitness of each individual in the initial population based on the control variable, and iterating the initial population based on a genetic algorithm to obtain an iterated population after iteration;

[0018] If it is determined that the number of iterations reaches a preset number of iterations, the individual with the largest fitness in the iterative population is output as a solution result, and reactive power optimization is performed on the offshore wind farm based on the solution result of the adjusted objective function.

[0019] In the process of solving the reactive power optimization control strategy, an improved genetic algorithm is used to seek the optimal solution, which realizes the effective control of active network loss and voltage deviation of offshore wind farms.

[0020] Optionally, if it is determined that the number of iterations has not reached a preset number of iterations, the iteration step is repeated, wherein the iteration step is to iteratively process the initial population based on a genetic algorithm to obtain an iterated iterative population.

[0021] Optionally, the iterative processing of the initial population based on a genetic algorithm includes:

[0022] At least one of selection, crossover and mutation is performed on the initial population based on a fast non-dominated sorting genetic algorithm.

[0023] Optionally, setting the SVG and the transformer ratio as control variables includes:

[0024] The control variables are coded using a mixed encoding of real numbers and integers, wherein the SVG is coded using real numbers and the transformer ratio is coded using integers.

[0025] Optionally, the adjusted objective function is calculated according to the following formula:

[0026] minF=(α1F1+α2F2+α1F3)

[0027] Where F is the objective function, F1 is the load node voltage deviation sub-target, F2 is the active network loss sub-target, F3 is the regulation time sub-target, α1, α2, α3 are the adjusted weight coefficients of the load node voltage deviation sub-target, active network loss sub-target and regulation time sub-target, respectively.

[0028] Optionally, the adjusted load node voltage deviation sub-target is calculated according to the following formula:

[0029] minF1=|U POI -U ref |

[0030] Where U POI is the measured value of the grid connection point voltage, U ref is the grid connection point voltage reference value;

[0031] The adjusted active network loss sub-target is calculated according to the following formula:

[0032]

[0033] Where N b is the total number of wind farm branches, i and j are the node numbers at both ends of the branch, G b(i,j) is the conductance of the bth branch, U i , U j Represent the voltage values ​​of load nodes i and j respectively, δ ij Represents the voltage phase angle difference between nodes i and j;

[0034] The adjusted adjustment time sub-goal is calculated according to the following formula:

[0035] minF3=t

[0036] Where, t is the adjustment time;

[0037] The adjusted objective function is calculated according to the following formula:

[0038]

[0039] Where F is the objective function.

[0040] Optionally, before solving the adjusted objective function based on the genetic algorithm, the method further comprises: setting a constraint condition set of the reactive power optimization model, the constraint condition set comprising equality constraints and inequality constraints;

[0041] The step of solving the adjusted objective function based on a genetic algorithm comprises:

[0042] The adjusted objective function is solved according to the equality constraint and the inequality constraint based on the genetic algorithm.

[0043] This application constructs a complete set of reactive power optimization constraint conditions, which include but are not limited to the reactive power output range of the wind turbine, the voltage stability requirements of the wind farm, the reactive power balance of the system, etc. These constraints ensure that the optimization process of the reactive power optimization model can be executed in the actual operating environment and maintain the stability and efficient operation of the system under various operating conditions.

[0044] In a second aspect, an embodiment of the present application provides an offshore wind farm reactive power optimization system based on a genetic algorithm, the system comprising a construction module, an adjustment module and an optimization module;

[0045] The construction module is used to construct a reactive power optimization model for an offshore wind farm and set an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective, and an adjustment time sub-objective;

[0046] The adjustment module is used to calculate the grid connection point voltage deviation based on the output power of the offshore wind farm and the grid-connected transformer parameters, and adaptively adjust the weight coefficient of each sub-objective based on the grid connection point voltage deviation to obtain an adjusted objective function;

[0047] The optimization module is used to solve the adjusted objective function based on a genetic algorithm through the reactive power optimization model, and perform reactive power optimization on the offshore wind farm based on the solution result of the adjusted objective function. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0049] Figure 1 This is a flow chart of a reactive power optimization method for an offshore wind farm based on a genetic algorithm in an embodiment of the present application;

[0050] Figure 2 A diagram showing the steps of solving the adjusted objective function based on a genetic algorithm in an embodiment of the present application;

[0051] Figure 3 This is a schematic diagram of a reactive power optimization system for an offshore wind farm based on a genetic algorithm in an embodiment of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0053] Wind power has developed rapidly as a clean energy source. With the continuous increase in the scale of wind power grid connection, the reactive power active support capability of traditional synchronous power sources cannot guarantee the safe and stable operation of the power system, especially as offshore wind farms are in a large-scale, centralized development stage. It is imperative to improve the reactive power active support capability of offshore wind farms.

[0054] The optimization objective function in the relevant technology is a fixed weight coefficient. As the proportion of renewable energy power generation in the power grid gradually increases, the operating conditions of the new power system will change frequently with the fluctuations in the output of renewable energy power generation. The weight coefficients of each optimization objective are fixed values, which will lead to the inability to adjust in real time according to the power grid conditions, affecting the safe and economical operation of the power system.

[0055] This application proposes a reactive power optimization method for offshore wind farms based on genetic algorithms, which adaptively adjusts the priority of each optimization target according to the grid conditions to meet the real-time needs of the grid and ensure the safety, stability and economy of the new power system.

[0056] The embodiment of the present application provides a reactive power optimization method for an offshore wind farm based on a genetic algorithm, referring to Figure 1 , Figure 1This is a flow chart of a reactive power optimization method for an offshore wind farm based on a genetic algorithm in an embodiment of the present application, comprising:

[0057] S100, constructing a reactive power optimization model for an offshore wind farm, and setting an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective, and an adjustment time sub-objective.

[0058] Among them, the reactive power optimization model can be based on deep learning to adjust the reactive working conditions of the offshore wind farm according to historical reactive power optimization data. The load node voltage deviation sub-goal can be to optimize the voltage level or reduce the voltage adjustment cost. Optimizing the voltage level means reducing the voltage conditions of excessive high or low voltage. The active network loss sub-goal can be to optimize power distribution or improve transmission efficiency. Optimizing power distribution means adjusting power generation and load distribution to reduce the active power loss generated during voltage transmission. The adjustment time sub-goal can be to reduce the load node voltage deviation adjustment and active network loss adjustment response time.

[0059] Specifically, a reactive power optimization model of the offshore wind farm is constructed based on the actual number, type, operating parameters and power generation parameters of the wind turbines in the offshore wind farm, and an objective function is set based on the operating conditions of the offshore wind farm. The objective function can be any one of the sub-goals of considering the voltage deviation of the load node, the active network loss sub-goal or the adjustment time sub-goal.

[0060] S200, calculating the grid connection point voltage deviation based on the output power of the offshore wind farm and the grid-connected transformer parameters, and adaptively adjusting the weight coefficient of each sub-goal based on the grid connection point voltage deviation to obtain an adjusted objective function.

[0061] Among them, the grid-connected transformer parameters may include parameters such as rated capacity, rated voltage, impedance voltage, loss, insulation level, short-circuit impedance and cooling method. By detecting the grid-connected point voltage in different time periods, the grid-connected point voltage deviation between the current moment and the previous moment can be obtained. When the grid-connected point voltage deviation is greater than or equal to the preset deviation point value, the weight coefficient of the load node voltage deviation sub-target is increased; when the grid-connected point voltage deviation is less than the preset deviation point value, the weight coefficients of the active network loss sub-target and the adjustment time sub-target are increased.

[0062] The objective function is the weighted sum of the load node voltage deviation sub-goal, the active network loss sub-goal and the regulation time sub-goal. The objective function is calculated by the following formula:

[0063] minF=(α1F1+α2F2+α1F3)

[0064] Where F1 is the load node voltage deviation sub-target, F2 is the active network loss sub-target, F3 is the regulation time sub-target, α1, α2, α3 are the adjusted weight coefficients of the load node voltage deviation sub-target, active network loss sub-target and regulation time sub-target, respectively.

[0065] Further, S210, based on the output power of the offshore wind farm and the grid-connected transformer parameters, a grid-connected point voltage reference value of the offshore wind farm is calculated:

[0066] U ref =U s -I·(R line +jX line +R t +jX t )

[0067]

[0068] Z t =R t +jX t

[0069] Z line =R line +jX line

[0070] Where P is the output active power of the offshore wind farm (MW), Q is the output reactive power of the offshore wind farm (MVar), and U is s is the bus voltage at the grid connection point (kv), Z t is the grid-connected voltage device impedance, Z line is the line impedance between the grid connection point and the power grid, I is the current flowing through the grid connection point, U ref It is the grid connection point voltage reference value.

[0071] S220, the grid connection point voltage deviation is calculated based on the grid connection point voltage reference value and the grid connection point voltage measured value:

[0072] minF1=|U POI -U ref |

[0073] Where, F1 is the minimum grid connection point voltage deviation, U POI is the measured value of the grid connection point voltage, U ref It is the grid connection point voltage reference value, which is issued by the AVC master station.

[0074] S230, adaptively adjusting the weight coefficient of each sub-goal based on the grid connection point voltage deviation, adaptively adjusting the weight coefficient of each sub-goal based on the grid connection point voltage deviation, to obtain an adjusted objective function.

[0075] When the voltage deviation fluctuation at the grid connection point is within the preset deviation range, it meets the requirements, and the preset deviation range is -0.05~0.05pu; when the voltage deviation fluctuation at the grid connection point is greater than the preset deviation range of 0.05pu, it exceeds the safety range. At this time, the optimization goal of the reactive power optimization model of the offshore wind farm is to reduce the voltage deviation at the grid connection point to within the specified range, that is, to increase the weight coefficient of the load node voltage deviation sub-goal; when the voltage deviation fluctuation at the grid connection point is less than the preset deviation range of 0.03pu, the optimization goal of the reactive power optimization model of the offshore wind farm is to make the active network loss smaller and the adjustment time faster, that is, to increase the weight coefficients of the active network loss sub-goal and the adjustment time sub-goal.

[0076] The load node is the terminal point of power consumption, while the grid connection point is the starting point of power transmission. The main function of the load node is to consume power, while the function of the grid connection point is to transmit power from the power generation facility to the power grid. The demand of the load node directly affects the dispatch and supply strategy of the power system, while the configuration of the grid connection point affects the power transmission efficiency and the stability of the power grid.

[0077] S300, solving the adjusted objective function based on the genetic algorithm through the reactive power optimization model, and performing reactive power optimization on the offshore wind farm based on the solution result of the adjusted objective function.

[0078] Among them, Genetic Algorithm (GA) is an optimization method that simulates the biological evolution process.

[0079] Furthermore, reactive power optimization may be to re-determine the compensation capacitor capacity and the position of the adjustable transformer tap under real-time working conditions to perform reactive power optimization; and adjust the settings and control parameters of the reactive power compensation equipment based on the solution results to perform reactive power optimization.

[0080] In order to enhance the local random search capability of the genetic algorithm and enable the algorithm to obtain the optimal solution at the fastest convergence speed, the arithmetic crossover operator is introduced into the reactive power optimization model, which greatly improves the search capability of the operator in the genetic crossover process and speeds up the solution.

[0081] The present application combines the real-time grid conditions, calculates the grid connection point voltage deviation through the output power of the offshore wind farm and the grid-connected transformer parameters, and then determines the optimization direction of the reactive power optimization model, and adaptively adjusts the weight coefficient corresponding to each optimization sub-objective, that is, adaptively adjusts the priority of each optimization objective, overcomes the technical problem in the related technology that the weight coefficient of each optimization objective is a fixed value and cannot be adjusted in real time, realizes the coordinated optimization of the active network loss and the grid connection point voltage of the offshore wind farm, and improves the operating efficiency and voltage stability of the offshore wind farm.

[0082] As an implementation manner of the present application, the weight coefficient of each sub-objective is adaptively adjusted based on the grid connection point voltage deviation to obtain an adjusted objective function, including:

[0083] When the grid connection point voltage deviation is greater than a preset deviation range, the weight coefficient of the load node voltage deviation sub-target is adaptively increased to obtain an objective function containing the increased weight coefficient of the load node voltage deviation sub-target;

[0084] When the grid connection point voltage deviation is less than a preset deviation range, the weight coefficients of the active network loss sub-target and the adjustment time sub-target are adaptively increased to obtain an objective function containing the increased weight coefficients of the active network loss sub-target and the adjustment time sub-target.

[0085] As an implementation manner of the present application, solving the adjusted objective function based on a genetic algorithm through the reactive power optimization model, and performing reactive power optimization on the offshore wind farm based on the solution result of the adjusted objective function, includes:

[0086] Reference Figure 2 , Figure 2 This is a step diagram for solving the adjusted objective function based on the genetic algorithm in an embodiment of the present application, S310, inputting the initial data of the offshore wind farm into the reactive power optimization model, the initial data including the number of offshore wind turbines, the operating parameters of the wind turbines and the wind speed data of the wind turbines.

[0087] S320, the reactive power optimization model performs power flow calculation according to the initial data to obtain the active network loss of the wind farm and the voltage of each load node.

[0088] S330. Determine a reactive compensation node from multiple load nodes based on a voltage stability modal analysis method, and specifically select a node with a smaller value as the reactive compensation node.

[0089] S340, randomly generate an initial population, and set SVG and transformer ratio as control variables. cl 、T i As a control variable, a mixed coding of real numbers and integers is used, that is, Where Q cl Using real number coding, T i Integer encoding is used. Transformer ratio T i The calculation of T i =rank(-N2,+N2), which is a randomly generated integer within the transformer ratio adjustment range (-N2,+N2); the SVG compensation capacity is Q Ci =1+f c r and / 10, where fc Indicates a randomly generated +1 or -1, r and Represents a random real number in the interval (0,1).

[0090] S350, calculating the fitness of each individual in the initial population based on the control variable, and iterating the initial population based on the genetic algorithm to obtain an iterated population after iteration. The iterative processing includes at least one of selection, crossover, and mutation, wherein the mutation processing is to prevent the algorithm from falling into a local optimal solution.

[0091] S360. If it is determined that the number of iterations reaches a preset number of iterations, that is, the maximum number of iterations, the individual with the largest fitness in the iterative population is output as the solution result. The solution result is the reactive power optimization control strategy output by the reactive power optimization model, and the offshore wind farm is reactively optimized based on the solution result of the adjusted objective function.

[0092] In the process of solving the reactive power optimization control strategy, an improved genetic algorithm is used to find the optimal solution, which effectively controls the active network loss and voltage deviation of the offshore wind farm. The improved genetic algorithm gradually optimizes the quality of the solution and can handle complex optimization problems, especially when the solution space is large and irregular.

[0093] As an implementation mode of the present application, if it is determined that the number of iterations does not reach the preset number of iterations, the iteration step is repeated, and the iteration step is to iteratively process the initial population based on the genetic algorithm to obtain an iterated iterative population.

[0094] As an implementation manner of the present application, the iterative processing of the initial population based on a genetic algorithm includes:

[0095] At least one of selection, crossover and mutation is performed on the initial population based on a fast non-dominated sorting genetic algorithm.

[0096] As an implementation manner of the present application, setting the SVG and the transformer ratio as control variables includes:

[0097] The control variables are coded using a mixed encoding of real numbers and integers, wherein the SVG is coded using real numbers and the transformer ratio is coded using integers.

[0098] As an implementation manner of the present application, the adjusted objective function is calculated according to the following formula:

[0099] minF=(α1F1+α2F2+α1F3)

[0100] Where F is the objective function, F1 is the load node voltage deviation sub-target, F2 is the active network loss sub-target, F3 is the regulation time sub-target, α1, α2, α3 are the adjusted weight coefficients of the load node voltage deviation sub-target, active network loss sub-target and regulation time sub-target, respectively.

[0101] As an implementation manner of the present application, the adjusted load node voltage deviation sub-target is calculated according to the following formula:

[0102] minF1=|U POI -U ref |

[0103] α1=sin(yπ / 2)

[0104]

[0105] Where minF1 is the minimum voltage deviation of the load node, U POI is the measured value of the grid connection point voltage, U ref is the grid-connected point voltage reference value, the first weight α1 is adaptively adjusted using the sin function, ΔU max The difference in the maximum deviation range for the fan to operate without being disconnected from the grid is 0.1pu.

[0106]

[0107] The calibration method of the second weight is:

[0108] α2=(1-α1)×0.7

[0109] Where minF2 is the minimum active network loss sub-target of the offshore wind farm, N b is the total number of wind farm branches, i and j are the node numbers at both ends of the branch, G b(i,j) is the conductance of the bth branch, U i and U j Represent the voltage values ​​of load nodes i and j respectively, δ ij Represents the voltage phase angle difference between load nodes i and j.

[0110] minF3=t

[0111] The calibration method of the third weight is:

[0112] α3=(1-α1)×0.3

[0113] Where minF3 is the minimum sub-goal of adjustment time.

[0114]

[0115] Where F is the objective function.

[0116] As an implementation mode of the present application, before solving the adjusted objective function based on the genetic algorithm, the method further includes: setting a constraint condition set of the reactive power optimization model based on the output characteristics of the wind turbine unit and the operating characteristics of the offshore wind farm, wherein the constraint condition set includes equality constraints and inequality constraints;

[0117] The step of solving the adjusted objective function based on a genetic algorithm comprises:

[0118] The adjusted objective function is solved according to the equality constraint and the inequality constraint based on the genetic algorithm.

[0119] The set of constraints includes equality constraints and inequality constraints, including:

[0120] The equality constraints are calculated according to the following formula:

[0121]

[0122] Where P i , Q i is the active power and reactive power input to load node i; U i is the voltage of load node i, U j is the voltage of load node j, G i,j , B i.j are the real and imaginary parts of each element in the load node admittance matrix; φ i,j is the phase angle difference between load node i and load node j;

[0123] The inequality constraints are calculated according to the following formula:

[0124] Q clmin ≤Q cl ≤Q clmax

[0125] T pmin ≤T p ≤T pmax

[0126] P Gimin ≤P Gi ≤P Gimax

[0127] Q Gimin ≤Q Gi ≤Q Gimax

[0128] U imin ≤U i ≤U imax

[0129] Where P Gi , QGi are the active power output and reactive power output of the i-th wind turbine respectively; P G,i,min , P G,i,max , Q G,i,min and Q G,i,max are the minimum and maximum values ​​of active power output and reactive power output of the i-th wind turbine respectively; Q clmin , Q clmax are the minimum and maximum reactive output values ​​of the first SVG, Q cl is the reactive power output value of the first SVG; T pmin and T pmax are the maximum and minimum values ​​of the pth adjustable transformer tap, T p is the tap of the pth adjustable transformer; U imin and U imax are the minimum and maximum voltages of the i-th node, U i is the voltage value of the i-th node.

[0130] This application constructs a complete set of reactive power optimization constraint conditions, which include but are not limited to the reactive power output range of the wind turbine, the voltage stability requirements of the wind farm, the reactive power balance of the system, etc. These constraints ensure that the optimization process of the reactive power optimization model can be executed in the actual operating environment and maintain the stability and efficient operation of the system under various operating conditions.

[0131] This application combines the real-time grid conditions, calculates the grid connection point voltage deviation through the output power of the offshore wind farm and the grid-connected transformer parameters, and then determines the optimization direction of the reactive power optimization model, adaptively adjusts the weight coefficient corresponding to each optimization sub-goal, that is, adaptively adjusts the priority of each optimization goal, overcomes the technical problem that the weight coefficient of each optimization goal in the related technology is a fixed value and cannot be adjusted in real time, realizes the coordinated optimization of the active network loss and grid connection point voltage of the offshore wind farm, and improves the operating efficiency and voltage stability of the offshore wind farm. In the process of solving the reactive power optimization control strategy, an improved genetic algorithm is used to seek the optimal solution, realizing the effective control of the active network loss and voltage deviation of the offshore wind farm.

[0132] The embodiment of the present application provides a reactive power optimization system for offshore wind farms based on a genetic algorithm, referring to Figure 3 , Figure 3 This is a schematic diagram of a reactive power optimization system for an offshore wind farm based on a genetic algorithm in an embodiment of the present application. The system includes a construction module 100, an adjustment module 200 and an optimization module 300.

[0133] The construction module 100 is used to construct a reactive power optimization model for an offshore wind farm and set an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective and an adjustment time sub-objective.

[0134] The adjustment module 200 is used to calculate the grid connection point voltage deviation based on the output power of the offshore wind farm and the grid-connected transformer parameters, and adaptively adjust the weight coefficient of each sub-objective based on the grid connection point voltage deviation to obtain an adjusted objective function.

[0135] The optimization module 300 is used to solve the adjusted objective function based on the genetic algorithm through the reactive power optimization model, and perform reactive power optimization on the offshore wind farm based on the solution result of the adjusted objective function.

[0136] As an implementation of the present application, the adjustment module 200 includes:

[0137] When the grid connection point voltage deviation is greater than a preset deviation range, the weight coefficient of the load node voltage deviation sub-target is adaptively increased to obtain an objective function containing the increased weight coefficient of the load node voltage deviation sub-target.

[0138] When the voltage deviation at the grid connection point is less than the preset deviation range, the weight coefficients of the active network loss sub-goal and the adjustment time sub-goal are adaptively increased.

[0139] In order to obtain the objective function containing the weight coefficients of the active network loss sub-goal and the adjustment time sub-goal after their respective weight coefficients are increased.

[0140] As an implementation of the present application, the optimization module 300 includes:

[0141] The initial population is randomly generated, and SVG and transformer ratio are set as control variables.

[0142] The fitness of each individual in the initial population is calculated based on the control variables, and the initial population is iteratively processed based on the genetic algorithm to obtain an iterative population after iteration.

[0143] If it is determined that the number of iterations reaches the preset number of iterations, the individual with the largest fitness in the iterative population is output as the solution result, and the reactive power optimization of the offshore wind farm is performed based on the solution result of the adjusted objective function.

[0144] As an implementation mode of the present application, if it is determined that the number of iterations does not reach the preset number of iterations, the iteration step is repeated, and the iteration step is to iteratively process the initial population based on the genetic algorithm to obtain the iterated iterative population.

[0145] As an implementation of the present application, in the optimization module 300, the initial population is iteratively processed based on the genetic algorithm, including:

[0146] At least one of selection, crossover and mutation is performed on the initial population based on a fast non-dominated sorting genetic algorithm.

[0147] As an implementation mode of the present application, in the optimization module 300, SVG and transformer ratio are set as control variables, including:

[0148] The control variables are coded with mixed real and integer numbers, where SVG is coded with real numbers and transformer ratio is coded with integers.

[0149] As an implementation method of the present application, the adjusted objective function is calculated according to the following formula:

[0150] minF=(α1F1+α2F2+α1F3)

[0151] Where F is the objective function, F1 is the load node voltage deviation sub-target, F2 is the active network loss sub-target, F3 is the regulation time sub-target, α1, α2, α3 are the adjusted weight coefficients of the load node voltage deviation sub-target, active network loss sub-target and regulation time sub-target, respectively.

[0152] As an implementation method of the present application, the adjusted load node voltage deviation sub-target is calculated according to the following formula:

[0153] minF1=|U POI -U ref |

[0154] Where U POI is the measured value of the grid connection point voltage, U ref is the grid connection point voltage reference value;

[0155] The adjusted active network loss sub-target is calculated according to the following formula:

[0156]

[0157] Where N b is the total number of wind farm branches, i and j are the node numbers at both ends of the branch, G b(i,j) is the conductance of the bth branch, U i , U j Represent the voltage values ​​of load nodes i and j respectively, δ ij Represents the voltage phase angle difference between nodes i and j;

[0158] The adjusted adjustment time sub-goal is calculated according to the following formula:

[0159] minF3=t

[0160] Where, t is the adjustment time;

[0161] The adjusted objective function is calculated according to the following formula:

[0162]

[0163] Where F is the objective function.

[0164] As an implementation mode of the present application, before the optimization module 300, a constraint module is also included, which is used to set a constraint condition set of the reactive power optimization model, and the constraint condition set includes equality constraints and inequality constraints.

[0165] The optimization module 300 further includes:

[0166] The adjusted objective function is solved based on the equality and inequality constraints based on the genetic algorithm.

[0167] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

[0168] Although the embodiments of the present application have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present application, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A reactive power optimization method for offshore wind farms based on genetic algorithm, characterized in that: The method comprises: Constructing a reactive power optimization model for an offshore wind farm, and setting an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective, and a regulation time sub-objective; The grid connection point voltage deviation is calculated based on the output power of the offshore wind farm and the grid-connected transformer parameters, and the weight coefficient of each sub-objective is adaptively adjusted based on the grid connection point voltage deviation to obtain an adjusted objective function; The adjusted objective function is solved by the reactive power optimization model based on a genetic algorithm, and reactive power optimization is performed on the offshore wind farm based on the solution result of the adjusted objective function.

2. The method according to claim 1, characterized in that The adaptively adjusting the weight coefficient of each sub-objective based on the grid connection point voltage deviation to obtain an adjusted objective function includes: When the grid connection point voltage deviation is greater than a preset deviation range, the weight coefficient of the load node voltage deviation sub-target is adaptively increased to obtain an objective function containing the increased weight coefficient of the load node voltage deviation sub-target; When the grid connection point voltage deviation is less than a preset deviation range, adaptively increasing the weight coefficients of the active grid loss sub-goal and the adjustment time sub-goal; To obtain an objective function containing the weight coefficients of the active network loss sub-objective and the adjustment time sub-objective after each of them is increased.

3. The method according to claim 1, characterized in that Solving the adjusted objective function based on a genetic algorithm by using the reactive power optimization model, and performing reactive power optimization on the offshore wind farm based on the solution of the adjusted objective function, comprises: The initial population is randomly generated, and SVG and transformer ratio are set as control variables; Calculating the fitness of each individual in the initial population based on the control variable, and iterating the initial population based on a genetic algorithm to obtain an iterated population after iteration; If it is determined that the number of iterations reaches a preset number of iterations, the individual with the largest fitness in the iterative population is output as a solution result, and reactive power optimization is performed on the offshore wind farm based on the solution result of the adjusted objective function.

4. The method according to claim 3, characterized in that If it is determined that the number of iterations does not reach the preset number of iterations, the iteration step is repeated, wherein the iteration step is to iteratively process the initial population based on the genetic algorithm to obtain an iterated iterative population.

5. The method according to claim 3, characterized in that: The iterative processing of the initial population based on the genetic algorithm comprises: At least one of selection, crossover and mutation is performed on the initial population based on a fast non-dominated sorting genetic algorithm.

6. The method according to claim 3, characterized in that: The step of setting the SVG and the transformer ratio as control variables includes: The control variables are coded using a mixed encoding of real numbers and integers, wherein the SVG is coded using real numbers and the transformer ratio is coded using integers.

7. The method according to claim 1 or 2, characterized in that: The adjusted objective function is calculated according to the following formula: minF=(α1F1+α2F2+α1F3) Where F is the objective function, F1 is the load node voltage deviation sub-target, F2 is the active network loss sub-target, F3 is the regulation time sub-target, α1, α2, α3 are the adjusted weight coefficients of the load node voltage deviation sub-target, active network loss sub-target and regulation time sub-target, respectively.

8. The method according to claim 7, characterized in that The adjusted load node voltage deviation sub-target is calculated according to the following formula: minF1=|U POI -YOU ref | Where U POI is the measured value of the grid connection point voltage, U ref is the grid connection point voltage reference value; The adjusted active network loss sub-target is calculated according to the following formula: Where N b is the total number of wind farm branches, i and j are the node numbers at both ends of the branch, G b(i,j) is the conductance of the bth branch, U i , U j Represent the voltage values ​​of load nodes i and j respectively, δ ij Represents the voltage phase angle difference between nodes i and j; The adjusted adjustment time sub-goal is calculated according to the following formula: minF3=t Where, t is the adjustment time; The adjusted objective function is calculated according to the following formula: Where F is the objective function.

9. The method according to claim 1, characterized in that: Before solving the adjusted objective function based on the genetic algorithm, the method further includes: setting a constraint condition set of the reactive power optimization model, wherein the constraint condition set includes equality constraints and inequality constraints; The step of solving the adjusted objective function based on a genetic algorithm comprises: The adjusted objective function is solved according to the equality constraint and the inequality constraint based on the genetic algorithm.

10. A reactive power optimization system for offshore wind farms based on genetic algorithm, characterized in that: The system includes a construction module, an adjustment module and an optimization module; The construction module is used to construct a reactive power optimization model for an offshore wind farm and set an objective function of the reactive power optimization model, wherein the objective function includes a load node voltage deviation sub-objective, an active power network loss sub-objective, and an adjustment time sub-objective; The adjustment module is used to calculate the grid connection point voltage deviation based on the output power of the offshore wind farm and the grid-connected transformer parameters, and adaptively adjust the weight coefficient of each sub-objective based on the grid connection point voltage deviation to obtain an adjusted objective function; The optimization module is used to solve the adjusted objective function based on a genetic algorithm through the reactive power optimization model, and perform reactive power optimization on the offshore wind farm based on the solution result of the adjusted objective function.

Images