Multi-objective balanced optimization method, system, equipment and medium for virtual power plant distribution network

Abstract

The present invention discloses a multi-objective balanced optimization method and system, equipment, and medium for a virtual power plant distribution network. In the multi-objective balanced optimization method of the present invention, the multi-objective function takes voltage safety as the optimization target, which ensures the voltage safety of the power system and the reliability and sustainability of the operation of the power system. In addition, after determining the mean of each optimization target, the weight parameters of each optimization target are optimized and solved to obtain the optimal solution of the multi-objective balance. Compared with the prior art that requires iterative optimization of the variable parameters contained in each optimization target, the iterative optimization solution process of each optimization target is omitted, which greatly improves the optimization solution efficiency, can realize automatic balanced optimization of three optimization targets, and can also adapt well to new operation requirements.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual power plant optimization operation, and in particular, to a multi-objective balanced optimization method and system for a virtual power plant distribution network, electronic equipment, and a computer-readable storage medium. Background Art

[0002] Virtual power plants integrate various energy resources, energy storage systems, and multiple energy management programs. They have a management framework for flexible load coordination. Renewable energy sources, such as wind turbines, photovoltaic power generation, and biomass waste units, are given priority in virtual power plants, which can reduce the emission of environmental pollutants. However, the power generation of these renewable energy sources is uncertain, and there will be differences between the day-ahead scheduling and real-time operation of virtual power plants based on renewable energy. This is particularly significant when the system flexibility is insufficient, which may lead to an imbalance between the real-time power generation and power consumption of the system. In order to solve this problem, it is necessary to use flexible resources (i.e., resources that can adjust active power), including electric vehicles, energy storage systems, demand response, and renewable energy in virtual power plants, to effectively manage system flexibility. Compared with fixed energy storage systems and renewable energy, mobile energy storage resources such as electric vehicles and demand response are easier to obtain and can be directly controlled by energy users. In addition, the use of fixed energy storage and renewable energy requires installation and construction costs, while electric vehicles and demand response can be combined with the operating goals of distribution system operators and guided through various incentive mechanisms.

[0003] The current virtual power plant optimization operation method usually adopts a multi-objective optimization method, which generally takes economic benefit maximization, operating cost minimization, operating risk minimization, optimal renewable energy consumption, optimal energy storage system benefits or carbon emission minimization as the optimization objectives, without considering the voltage safety of the power system during the optimization process, making it difficult to ensure the reliability and sustainability of the power system operation. In addition, when performing multi-objective optimization solutions, the existing technology needs to first iteratively optimize the variable parameters of each optimization objective, and then combine the set weight parameters to achieve multi-objective optimization solutions. The iterative optimization requires a large amount of calculation, and the optimization solution efficiency is low. In addition, the weight parameters of each optimization objective are usually set to fixed values ​​based on manual experience, making it difficult to achieve balanced optimization between multiple optimization objectives, and unable to adapt to new operating requirements. For example, Chinese patent application CN113919717A discloses a virtual power plant scheduling method for multi-objective synchronous optimization, which constructs a multi-objective function with minimization of operating costs, optimal consumption of renewable energy and optimal return of energy storage system as three optimization objectives. The weight of each optimization objective is set according to the decision maker's emphasis on each optimization objective. This scheme does not take into account the voltage safety of the power system, and it is difficult to ensure the reliability and sustainability of the power system operation. In addition, the weights of multiple objectives are set subjectively by humans, and it is impossible to achieve balanced optimization among multiple optimization objectives. Summary of the invention

[0004] The present invention provides a multi-objective balanced optimization method and system for a virtual power plant distribution network, an electronic device, and a computer-readable storage medium, which can ensure the voltage safety of the power system, improve the optimization solution efficiency, and realize automatic balanced optimization of three optimization objectives.

[0005] According to one aspect of the present invention, a multi-objective balanced optimization method for a virtual power plant distribution network is provided, comprising the following contents:

[0006] Taking energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints, a multi-objective function is constructed.

[0007] Determine the mean of each optimization objective in the multi-objective function;

[0008] The mean of each optimization objective is substituted into the multi-objective function, and the weight parameters of each optimization objective in the multi-objective function are optimized and solved to obtain the optimal solution of multi-objective equilibrium.

[0009] Furthermore, the multi-objective function is:

[0010] ;

[0011] Among them, ob1, ob2 and ob3 represent the three indicators of energy cost, energy loss and voltage safety index respectively. They represent the weight parameters of energy cost, energy loss and voltage safety index respectively, subscript v represents the virtual power plant, subscript w represents the wth power consumption scenario, subscript t represents the current operation period, subscript b represents the power system node, subscript s represents the balance node, represents the probability of the wth electricity consumption scenario occurring, represents the energy price of the wth electricity consumption scenario in the tth operation period, P S b=s,t,w A represents the active power when the node where the distribution tower is located is a balanced node in the wth power consumption scenario and the tth operation period. V v，b represents the association matrix between power system node b and virtual power plant v, P V v,t,w represents the active power of the virtual power plant in the wth power consumption scenario and the tth operation period, P C b,t,w WSI represents the active power consumed by the passive user at the power system node b in the tth operation period and the wth power consumption scenario. t,w It represents the worst safety index of the weak bus in the distribution network in the tth operating period and the wth power consumption scenario.

[0012] Furthermore, the power flow constraint of the distribution network is:

[0013] ;

[0014] ;

[0015] ;

[0016] ;

[0017] ;

[0018] ;

[0019] Where, subscript r represents the index of other nodes connected to power system node b, subscript L represents the distribution line, and A L b,r represents the correlation matrix between power system node b and other connected nodes, P L Indicates the active power on the distribution line, Q S Represents the reactive power on the distribution tower, Q L Indicates the reactive power on the distribution line, Q V represents the reactive power of the virtual power plant, Q C Represents the reactive power consumed by passive users, g L and b L They represent the conductance and susceptance on the distribution line respectively, V represents the voltage amplitude, and α represents the voltage-current angle.

[0020] Furthermore, the calculation model of the worst safety index is:

[0021] ;

[0022] Where, the subscript p represents the weak busbar, the subscript p-1 represents the adjacent busbar connected to the weak busbar, and X L Represents line reactance, R L Indicates line impedance.

[0023] Furthermore, the process of determining the mean of each optimization objective in the multi-objective function includes the following contents:

[0024] Generate 2n+1 samples according to the number n of uncertainty parameters contained in each optimization objective, select a set of specific values ​​for these n uncertainty parameters in each sample, and calculate the mean and covariance of each uncertainty parameter;

[0025] Calculate the weight coefficient for each sample;

[0026] The specific value data of 2n+1 samples are respectively input into the objective function of the optimization target to obtain 2n+1 objective function values;

[0027] The mean of the objective function is obtained by weighted summation based on the 2n+1 objective function values ​​and the weight coefficient of each sample.

[0028] Furthermore, the process of optimizing and solving the weight parameters of each optimization objective in the multi-objective function to obtain the optimal solution of the multi-objective balance includes the following contents:

[0029] Calculate the weighted sum of the multi-objective function when the three weight parameters are 1 respectively, and record the minimum and maximum values;

[0030] Select a set of random values ​​for the three weight parameters and calculate the weighted and random values ​​of the multi-objective function at this time;

[0031] The weighted and random values ​​are compared with the minimum and maximum values ​​recorded, and the fuzzy values ​​of the weighted and random values ​​are determined and recorded according to the comparison results;

[0032] The above process is repeated until the maximum number of members specified by the Pareto front is reached, and the weight parameter scheme corresponding to the Pareto front point with the maximum fuzzy value is selected as the optimal solution.

[0033] Furthermore, if the weighted and random value is less than the minimum value recorded, the fuzzy value of the weighted and random value is set to 1; if the weighted and random value is greater than the maximum value recorded, the fuzzy value of the weighted and random value is set to 0; if the weighted and random value is between the minimum and maximum values ​​recorded, the fuzzy value of the weighted and random value is set to ; where F represents the weighted and random value, F min and F max Indicates the minimum and maximum values ​​recorded.

[0034] In addition, the present invention also provides a multi-objective balancing optimization system for a virtual power plant distribution network, comprising:

[0035] A multi-objective function building module is used to build a multi-objective function with energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints;

[0036] An optimization target mean determination module, used to determine the mean of each optimization target in the multi-objective function;

[0037] The weight parameter optimization solution module is used to substitute the mean value of each optimization target into the multi-objective function, optimize and solve the weight parameters of each optimization target in the multi-objective function, and obtain the optimal solution of multi-objective equilibrium.

[0038] In addition, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0039] In addition, the present invention also provides a computer-readable storage medium for storing a computer program for performing multi-objective balanced optimization on a virtual power plant distribution network, wherein the computer program executes the steps of the method described above when running on a computer.

[0040] The present invention has the following beneficial effects:

[0041] The multi-objective balanced optimization method of the virtual power plant distribution network of the present invention uses voltage safety as the optimization target in the multi-objective function, which ensures the voltage safety of the power system and the reliability and sustainability of the power system operation. In addition, by determining the mean of each optimization target and optimizing the weight parameters of each optimization target, the optimal solution of multi-objective balance is obtained. Compared with the prior art that requires iterative optimization of the variable parameters contained in each optimization target, the iterative optimization solution process of each optimization target is omitted, which greatly improves the optimization solution efficiency, can realize automatic balanced optimization of the three optimization targets, and can also adapt well to new operation requirements.

[0042] In addition, the multi-objective balanced optimization system of the virtual power plant distribution network of the present invention also has the above advantages.

[0043] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0045] Figure 1 It is a flow chart of a multi-objective balancing optimization method for a virtual power plant distribution network according to a preferred embodiment of the present application;

[0046] Figure 2 yes Figure 1 Schematic diagram of the sub-process of step S2;

[0047] Figure 3 yes Figure 1 Schematic diagram of the sub-process of step S3;

[0048] Figure 4It is a schematic diagram of the module structure of a multi-objective balanced optimization system of a virtual power plant distribution network in another embodiment of the present application. DETAILED DESCRIPTION

[0049] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] Reference Figure 1 The preferred embodiment of the present application provides a multi-objective balanced optimization method for a virtual power plant distribution network, comprising the following contents:

[0051] Step S1: Taking energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints, a multi-objective function is constructed;

[0052] Step S2: determining the mean of each optimization objective in the multi-objective function;

[0053] Step S3: Substitute the mean of each optimization objective into the multi-objective function, optimize and solve the weight parameters of each optimization objective in the multi-objective function, and obtain the optimal solution of multi-objective equilibrium.

[0054] It can be understood that the multi-objective balanced optimization method of the virtual power plant distribution network in this embodiment uses voltage safety as the optimization target in the multi-objective function, which ensures the voltage safety of the power system and the reliability and sustainability of the power system operation. The weight parameters of each optimization target are optimized and solved after determining the mean of each optimization target to obtain the optimal solution for the multi-objective balance. Compared with the prior art that requires iterative optimization of the variable parameters contained in each optimization target, the iterative optimization solution process of each optimization target is eliminated, which greatly improves the optimization solution efficiency, can realize automatic balanced optimization of the three optimization targets, and can also adapt well to new operating requirements.

[0055] It can be understood that in step S1, for a virtual power plant based on renewable energy, considering that it has electric vehicles and demand response users that can perform load management, a multi-objective function of a smart distribution network is constructed based on the three optimization goals of distribution network economy, operation and voltage safety, with distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints. Among them, the multi-objective function is:

[0056] ;

[0057] Among them, ob1, ob2 and ob3 represent the three indicators of energy cost, energy loss and voltage safety index respectively. They represent the weight parameters of energy cost, energy loss and voltage safety index respectively. The value of the weight parameter is between 0 and 1, and the sum of the three weight parameters is always 1. The subscript v represents the virtual power plant, the subscript w represents the wth power consumption scenario, the subscript t represents the current operating period, the subscript b represents the power system node, and the subscript s represents the balance node. represents the probability of the wth electricity consumption scenario occurring, represents the energy price of the wth electricity consumption scenario in the tth operation period, P S b=s,t,w A represents the active power when the node where the distribution tower is located is a balanced node in the wth power consumption scenario and the tth operation period. V v，b represents the association matrix between power system node b and virtual power plant v, P V v,t,w represents the active power of the virtual power plant in the wth power consumption scenario and the tth operation period, P C b,t,w WSI represents the active power consumed by the passive user at the power system node b in the tth operation period and the wth power consumption scenario. t,w It represents the worst safety index of the weak bus in the distribution network in the tth operating period and the wth power consumption scenario.

[0058] It can be seen that the energy cost ob1 is the cost of the distribution network to purchase energy from the upstream network every hour, which is equal to the product of the active power of the distribution station located at the reference bus and the energy price. The energy loss ob2 is the sum of the active power of the distribution station and the virtual power plant minus the active power of the passive users in the network, where the passive users refer to users who only consume electricity and do not participate in other activities of the power grid. The voltage security index ob3 uses the worst security index (WSI) to quantify and evaluate the safety margin of the power system under the most unfavorable conditions. The WSI value varies from 0 to 1, where 0 indicates network voltage collapse and 1 indicates network no-load (i.e., the network with the smallest voltage drop). Therefore, in order to minimize the multi-objective function F, it is necessary to minimize the energy cost ob1 and energy loss ob2 and maximize the voltage security index ob3.

[0059] Among them, the power flow constraint of the distribution network is:

[0060] ;

[0061] ;

[0062] ;

[0063] ;

[0064] ;

[0065] ;

[0066] Where, subscript r represents the index of other nodes connected to power system node b, subscript L represents the distribution line, and P S b,t,w A represents the active power of the node where the distribution tower is located (i.e., power system node b) in the wth power consumption scenario and the tth operating period. L b,r represents the correlation matrix between power system node b and other connected nodes, P L Indicates the active power on the distribution line, P L b,r,t,w represents the active power of the distribution line between the power system node b and other connected nodes r in the wth power consumption scenario and the tth operation period, Q S Represents the reactive power on the distribution tower, Q S b,t,w represents the reactive power of power system node b in the wth power consumption scenario and the tth operation period, Q V represents the reactive power of the virtual power plant, and Q V v,t,w represents the reactive power of the virtual power plant in the wth power consumption scenario and the tth operation period, Q L Represents the reactive power on the distribution line, Q L b,r,t,w represents the reactive power of the distribution line between the power system node b and other connected nodes r in the wth power consumption scenario and the tth operation period, Q C Represents the reactive power consumed by passive users, g L and b L They represent the conductance and susceptance on the distribution line, both of which are standard values. L b,r represents the conductance of the distribution line between the power system node b and other connected nodes r, b L b,r represents the susceptance on the distribution line between the power system node b and other connected nodes r, V represents the voltage amplitude, V b,t,w represents the voltage amplitude of the power system node b in the wth power consumption scenario and the tth operation period, V r,t,w represents the voltage amplitude of other nodes r connected to the power system node b in the wth power consumption scenario and the tth operation period, α represents the voltage-current angle, α b,t,w represents the voltage and current angle of the power system node b in the wth power consumption scenario and the tth operation period, α r,t,w It represents the voltage and current angle of other nodes r connected to the power system node b in the wth power consumption scenario and the tth operating period.

[0067] It can be seen that the first and second formulas in the distribution network flow constraint conditions represent the active power balance and reactive power balance on the bus, the third and fourth formulas represent the calculation method of active power and reactive power on the distribution line, and the fifth and sixth formulas represent the voltage-current angle and the expected voltage amplitude on the reference bus.

[0068] In addition, the distribution network operation constraints are:

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] in, Indicates the maximum apparent power of the distribution line, Indicates the maximum apparent power of the distribution tower. and Represent the upper and lower limits of the voltage amplitude respectively. It can be seen that the first formula in the distribution network operation constraint limits the capacity of the distribution line, the second formula limits the substation capacity, and the third formula constrains the range of the voltage amplitude. The lower limit prevents the smart distribution network from shutting down under severe voltage drop conditions, and the upper limit prevents the insulation damage of the smart distribution network equipment due to overvoltage. The fourth formula simulates the power factor limit of the distribution column, where the power factor is equal to the ratio of active power to apparent power.

[0074] In addition, the bus with the lowest voltage amplitude in the smart distribution network is defined as a weak bus, and the voltage security constraint is evaluated based on the weak bus. The voltage security index is evaluated by calculating the WSI index of the weak bus. The calculation model of the WSI index is:

[0075] ;

[0076] Where, subscript p represents the weak busbar, subscript p-1 represents the adjacent busbar connected to the weak busbar, V p-1 t,w R represents the voltage amplitude of the adjacent bus connected to the weak bus in the wth power consumption scenario and the tth operation period, L Represents the line impedance, R L p-1,p It represents the line impedance of the distribution line between the weak busbar and its adjacent busbar, P L p-1,p,t,w represents the active power of the distribution line between the weak busbar and its adjacent busbar in the wth power consumption scenario and the tth operation period, X L Indicates line reactance, X L p-1,p It represents the reactance of the distribution line between the weak busbar and its adjacent busbar, Q L p-1,p,t,w It represents the reactive power of the distribution line between the weak bus and its adjacent bus in the wth power consumption scenario and the tth operating period.

[0077] In addition, the operating constraints of the virtual power plant based on renewable energy, considering that it has electric vehicles and demand response users for load management, are as follows:

[0078] ;

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] ;

[0089] ;

[0090] Among them, P V Represents the active power of the virtual power plant, P PV , P WT and P BU Represent the active power of photovoltaic power generation, wind power generation and biomass power generation, P DR represents the active power of demand response, P CH and P DCH They represent the active power of the electric vehicle in the charging and discharging states after it is connected to the charging pile, P C Represents the active power consumed by passive users; P V v,t,w represents the active power of the virtual power plant in the wth power consumption scenario and the tth operation period, P PV v,t,w represents the active power of photovoltaic power generation in the wth power consumption scenario and the tth operation period, P WT v,t,w represents the active power of wind power generation in the wth power consumption scenario and the tth operation period, P BU v,t,w represents the active power of biomass power generation in the wth power consumption scenario and the tth operation period, P DR v,t,w represents the active power of the demand response in the wth power consumption scenario and the tth operation period, P DCH v,t,wP represents the active power of an electric vehicle in the discharge state, in the wth power usage scenario, and in the tth operating period after the electric vehicle is connected to the charging pile. CH v,t,w P represents the active power of an electric vehicle in the charging state, in the wth power usage scenario, and in the tth operating period after the electric vehicle is connected to the charging pile. C v,t,w represents the active power consumed by the passive user in the wth power consumption scenario and the tth operation period; Q V represents the reactive power of the virtual power plant, Q PV , Q WT and Q BU Respectively represent the reactive power of photovoltaic power generation, wind power generation and biomass power generation, Q EV Indicates the reactive power of the charging pile, Q C Represents the reactive power consumed by passive users; Q V v,t,w represents the reactive power of the virtual power plant in the wth power consumption scenario and the tth operation period, Q PV v,t,w , Q WT v,t,w and Q BU v,t,w They represent the reactive power of photovoltaic power generation, wind power generation and biomass power generation in the wth power consumption scenario and the tth operation period, respectively. EV v,t,w represents the reactive power of the charging pile in the wth power consumption scenario and the tth operation period, Q C v,t,w represents the reactive power consumed by the passive user in the wth power consumption scenario and the tth operation period; , , and They represent the maximum apparent power of photovoltaic power generation, wind power generation, biomass power generation and charging piles respectively. represents the response rate of the contracted users in the virtual power plant demand response; CR represents the total charging power of electric vehicles connected to the virtual power plant, and DR represents the total discharge power of electric vehicles connected to the virtual power plant; E A represents the initial energy of the electric vehicles newly connected to the virtual power plant during this period, E D is the energy consumed by electric vehicles leaving the virtual power plant, represents the charging efficiency of electric vehicles, Represents the discharge efficiency of electric vehicles, E EV Represents the energy of an electric vehicle battery.

[0091] It can be seen that among the operating constraints of the virtual power plant, the first and second formulas represent the calculation of the active power and reactive power of the virtual power plant from the perspective of the smart distribution network. The third to fifth formulas represent the apparent power capacity limits of controllable photovoltaic, wind power and biomass power generation. These constraints also represent the capacity curve of renewable energy. The sixth and seventh formulas model the operating model of demand response. The sixth formula represents the active power control limit of the contracted users participating in demand response. The seventh formula ensures that all energy consumed by consumers participating in demand response during operation is provided by the smart distribution network. The eighth to twelfth formulas represent the modeling of the operating mode of electric vehicles. The eighth formula represents the energy storage in the electric vehicle battery. The ninth and tenth formulas represent the charging limit and discharging limit of the electric vehicle battery, respectively. The eleventh formula indicates that the charging and discharging of electric vehicles will not occur at the same time. The twelfth formula represents the apparent power limit of the electric vehicle charger.

[0092] It can be understood that in step S1, a multi-objective function and four constraints have been constructed. In the actual operation process, the P C , Q C , P WT , P PV , P BU ,γ,CR,DR,E A and E D These ten parameters are unknown and have uncertainty. If the explicit calculation method is adopted to iteratively solve the value of each uncertainty parameter, and then optimize the weight parameters of each optimization target after calculating the value of each optimization target, the amount of calculation is large and the calculation complexity is very high. Therefore, the present invention takes into account that these uncertainty parameters can be modeled as a certain distribution to represent, such as Gaussian distribution, and each objective function is a nonlinear function. The uncertainty parameters will cause complex changes in the output distribution after passing through the nonlinear function. Therefore, the present invention adopts an untraceable transformation algorithm to capture the statistical characteristics after nonlinear transformation, and determines the mean of each optimization target in the multi-objective function through the untraceable transformation algorithm, and then substitutes the mean into the multi-objective function to optimize and solve the weight parameters of each optimization target.

[0093] Among them, Figure 2 As shown, the process of determining the mean value of each optimization objective in the multi-objective function includes the following contents:

[0094] Step S21: generating 2n+1 samples according to the number n of uncertainty parameters included in each optimization objective, selecting a set of specific values ​​for the n uncertainty parameters in each sample, and calculating the mean and covariance of each uncertainty parameter;

[0095] Step S22: Calculate the weight coefficient of each sample;

[0096] Step S23: inputting the specific value data of 2n+1 samples into the objective function of the optimization target respectively to obtain 2n+1 objective function values;

[0097] Step S24: Perform weighted summation based on the 2n+1 objective function values ​​and the weight coefficient of each sample to obtain the mean of the objective function.

[0098] Specifically, define the output vector y∈R r is an uncertain output vector with r elements, z∈R n Represents an uncertain input vector. The relationship between the input and output vectors is expressed as a function y=f(z). The first objective function in the multi-objective function can be expressed as: , μ z and σ z is the mean and covariance of the input vector z, and the mean μ of the output vector y needs to be determined by the unscented transformation method y and covariance σ y The specific process is:

[0099] First, generate 2n+1 samples according to the number of uncertainty parameters n contained in the first objective function, select a set of specific values ​​for these n uncertainty parameters in each sample, and calculate the mean μ of each uncertainty parameter z and covariance σ z , then the center point of 2n+1 samples can be expressed as: , the positive offset point can be expressed as: , the negative offset point can be expressed as: , z s represents the input vector of the sth offset sample point, z0 represents the input vector of the center point, represents the weight of the center point z0, represents the adjustment parameter, which determines the distribution range of the sampling points, affects the distance between the sampling points and the mean of the input variable, and is related to the number n of uncertainty parameters. Represents the expansion factor, usually a small positive value (such as 0.001), which determines the degree of deviation of the sampling point from the mean.

[0100] Then, the weight coefficient of each sample point is calculated using the following formula:

[0101] ;

[0102] ;

[0103] ;

[0104] ;

[0105] in, is the weight coefficient of the sth positive offset sample point, is the weight coefficient of the s+nth negative offset sample point.

[0106] Next, the input vectors of the 2n+1 sample points are input into the objective function In this paper, we get 2n+1 objective function values, y s represents the output vector of the sth sample point, .

[0107] Finally, a weighted sum is performed based on the 2n+1 objective function values ​​and the weight coefficient of each sample to obtain the mean of the objective function. The calculation formula is: .

[0108] For the other two objective functions and , the same unscented transformation process is used to obtain the mean and .

[0109] It can be understood that the present invention adopts an untraceable transformation algorithm, which can determine the mean of each optimization objective in the multi-objective function through only a small number of sample points, which greatly reduces the computational complexity compared to the existing explicit calculation method.

[0110] In addition, in step S2, the covariance of each optimization target can also be calculated, and the calculation formula is: After calculating the mean and covariance of each optimization target, it can also be used to optimize the constraints. Assuming that the output of the uncertainty parameter is y and the mean is μ y , the covariance is σ y , the output y is not a single value, but a distribution, so the mean μ can be used y and covariance σ y To construct confidence intervals, we can verify the robustness of constraints through statistical probability. For example, for the output y, at a given confidence level k (for example, k = 1.96 corresponds to a 95% confidence interval), the output interval is: If the constraint is an upper limit, such as g(y)≤b (such as voltage upper limit, equipment capacity), then you need to ensure ; If the constraint is a lower limit, such as g(y) ≥ a (such as a voltage lower limit), you need to ensure that: Therefore, constructing the confidence interval through the mean and covariance not only takes into account the worst case of uncertainty parameters, but also improves the robustness of the constraints.

[0111] It can be understood that in step S3, the mean value of each optimization objective is substituted into the multi-objective function, and the multi-objective function can be expressed as: , and then the fuzzy decision algorithm is used to optimize the weight parameters of each optimization objective, and the compromise solution of multi-objective equilibrium is obtained as the optimal solution.

[0112] Among them, Figure 3 As shown, the process of optimizing and solving the weight parameters of each optimization objective in the multi-objective function to obtain the optimal solution of the multi-objective balance includes the following contents:

[0113] Step S31: Calculate the weighted sum of the multi-objective function when the three weight parameters are 1 respectively, and record the minimum and maximum values;

[0114] Step S32: selecting a set of random values ​​for the three weight parameters, and calculating the weighted and random values ​​of the multi-objective function at this time;

[0115] Step S33: comparing the weighted and random values ​​with the recorded minimum and maximum values, determining and recording the fuzzy value of the weighted and random values ​​according to the comparison result;

[0116] Step S34: Repeat the above process until the maximum number of members specified by the Pareto front is reached, and select the weight parameter scheme corresponding to the Pareto front point with the maximum fuzzy value as the optimal solution.

[0117] Specifically, for multi-objective functions: , respectively calculated When a weight parameter is 1, the other two weight parameters are 0, and the minimum value F is recorded. min and the maximum value F max Then, for , and Choose a set of random values, , and calculate the weighted sum random value F of the multi-objective function at this time. Then, the weighted sum random value F is compared with the recorded minimum value F min and the maximum value F max Compare and determine the fuzzy value φ of the weighted and random value according to the comparison result and record it. If the weighted and random value F is less than the minimum value F recorded, min , then the fuzzy value of the weighted and random value is set to 1, that is, φ=1; if the weighted and random value F is greater than the maximum value F recorded max , then the fuzzy value of the weighted and random value is set to 0, that is, φ=0; if the weighted and random value is within the minimum value F recorded min and the maximum value F max , then the fuzzy value of the weighted and random value is set to Then, the above process is repeated until the maximum number of members specified by the Pareto front is reached. The more the maximum number of members, the more times the repetition is performed. The weight parameter scheme corresponding to the Pareto front point with the maximum value of φ can be defined according to actual needs and selected as the optimal solution.

[0118] It can be understood that the present invention selects the weight parameter scheme corresponding to the Pareto front point with the maximum value of φ as the optimal solution. On the one hand, it can avoid extreme solutions, because some points on the Pareto front may tend to minimize or maximize a single objective (for example, minimum energy consumption but poor voltage safety). On the other hand, it can achieve a balance between the various optimization objectives, avoid bias towards a single optimization objective, and ensure the best overall effect.

[0119] like Figure 4 As shown, another embodiment of the present invention further provides a multi-objective balanced optimization system for a virtual power plant distribution network, preferably using the multi-objective balanced optimization method for a virtual power plant distribution network as described above, including:

[0120] A multi-objective function building module is used to build a multi-objective function with energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints;

[0121] An optimization target mean determination module, used to determine the mean of each optimization target in the multi-objective function;

[0122] The weight parameter optimization solution module is used to substitute the mean value of each optimization target into the multi-objective function, optimize and solve the weight parameters of each optimization target in the multi-objective function, and obtain the optimal solution of multi-objective equilibrium.

[0123] It can be understood that in the multi-objective balanced optimization system of the virtual power plant distribution network of this embodiment, the multi-objective function takes voltage safety as the optimization target, which ensures the voltage safety of the power system and the reliability and sustainability of the power system operation. The weight parameters of each optimization target are optimized and solved after determining the mean of each optimization target to obtain the optimal solution of the multi-objective balance. Compared with the prior art that requires iterative optimization of the variable parameters contained in each optimization target, the iterative optimization solution process of each optimization target is eliminated, which greatly improves the optimization solution efficiency, can realize automatic balanced optimization of the three optimization targets, and can also adapt well to new operating requirements.

[0124] In addition, another embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor executes the steps of the above method by calling the computer program stored in the memory.

[0125] In addition, another embodiment of the present invention further provides a computer-readable storage medium for storing a computer program for performing multi-objective balanced optimization on a virtual power plant distribution network, wherein the computer program executes the steps of the method described above when running on a computer.

[0126] Common computer-readable storage media include: floppy disks, flexible disks, hard disks, magnetic tapes, any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tapes, any other physical media with patterns of holes, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash-erasable programmable read-only memory (FLASH-EPROM), any other memory chip or cartridge, or any other medium that can be read by a computer. Instructions can further be transmitted or received by a transmission medium. The term transmission medium may include any tangible or intangible medium that can be used to store, encode or carry instructions for execution by a machine, and includes digital or analog communication signals or intangible media that facilitate communication of the above instructions. Transmission media include coaxial cables, copper wire, and optical fiber, including the wires of a bus used to transmit a computer data signal.

[0127] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present application may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0128] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0129] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0130] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0131] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0132] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0133] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-objective balanced optimization method for a virtual power plant distribution network, characterized in that: Includes the following: Taking energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints, a multi-objective function is constructed. Determine the mean of each optimization objective in the multi-objective function; Substitute the mean value of each optimization objective into the multi-objective function, optimize and solve the weight parameters of each optimization objective in the multi-objective function, and obtain the optimal solution of multi-objective equilibrium; The process of determining the mean value of each optimization objective in the multi-objective function includes the following: According to the number of uncertain parameters contained in each optimization objective n Build 2 n +1 sample, in each sample for this n A set of specific values ​​are selected for each uncertainty parameter, and the mean and covariance of each uncertainty parameter are calculated; Calculate the weight coefficient for each sample; 2 n +1 sample specific value data are input into the objective function of the optimization target, and 2 n +1 objective function value; Based on 2 n +1 objective function value and the weight coefficient of each sample are weighted summed to obtain the mean of the objective function; The process of optimizing and solving the weight parameters of each optimization objective in the multi-objective function to obtain the optimal solution of the multi-objective balance includes the following contents: Calculate the weighted sum of the multi-objective function when the three weight parameters are 1 respectively, and record the minimum and maximum values; Select a set of random values ​​for the three weight parameters and calculate the weighted and random values ​​of the multi-objective function at this time; The weighted and random values ​​are compared with the minimum and maximum values ​​recorded, and the fuzzy values ​​of the weighted and random values ​​are determined and recorded according to the comparison results; The above process is repeated until the maximum number of members specified by the Pareto front is reached, and the weight parameter scheme corresponding to the Pareto front point with the maximum fuzzy value is selected as the optimal solution.

2. The multi-objective balanced optimization method for a virtual power plant distribution network according to claim 1, characterized in that: The multi-objective function is: ； in, ob 1. ob 2 and ob 3 They represent three indicators: energy cost, energy loss and voltage safety index. represent the weight parameters of energy cost, energy loss and voltage safety index respectively, and the subscript v represents a virtual power plant, the subscript w Indicates w Power usage scenarios, subscript t Indicates the current running time period, subscript b represents a power system node, the subscript s represents a balance node, Indicates w The probability of a power usage scenario occurring, Indicates w The electricity usage scenario is t Energy prices for the operating period, P S b=s,t,w Indicates w Power usage scenario, t The active power when the node where the distribution tower is located is a balanced node during the operation period. A V v，b Represents a power system node b Virtual Power Plant v The correlation matrix of P V v,t,w Indicates that the virtual power plant w The electricity usage scenario and t The active power of the operating period, P C b,t,w Represents a power system node b In the t The running time, w The active power consumed by passive users in each electricity usage scenario, WSI t,w Indicates that the weak busbar in the distribution network is t The running time, w The worst safety index in each electricity usage scenario.

3. The multi-objective balanced optimization method for a virtual power plant distribution network according to claim 2, characterized in that: The power flow constraint of the distribution network is: ； ； ； ； ； ； Among them, the subscript r Represents the power system node b Index of other connected nodes, subscript L Indicates the power distribution line. A L b,r represents the association matrix between power system node b and other connected nodes, P L Indicates the active power on the distribution line. Q S Represents the reactive power on the distribution tower. Q L Represents the reactive power on the distribution line. Q V represents the reactive power of the virtual power plant, Q C represents the reactive power consumed by passive users, g L and b L They represent the conductance and susceptance on the distribution line, respectively. V Indicates the voltage amplitude, α Represents the voltage-current angle.

4. The multi-objective balanced optimization method for a virtual power plant distribution network according to claim 3, characterized in that: The calculation model of the worst safety index is: ； Among them, the subscript p Indicates weak busbar, subscript p -1 indicates the adjacent busbar connected to the weak busbar, X L is the line reactance, R L Indicates line impedance.

5. The multi-objective balanced optimization method for a virtual power plant distribution network according to claim 1, characterized in that: If the weighted and random value is less than the minimum value recorded, the fuzzy value of the weighted and random value is set to 1; if the weighted and random value is greater than the maximum value recorded, the fuzzy value of the weighted and random value is set to 0; if the weighted and random value is between the minimum and maximum values ​​recorded, the fuzzy value of the weighted and random value is set to (( F − F max ) / ( F min − F max ));in, F represents a weighted and random value, F min and F max Indicates the minimum and maximum values ​​recorded.

6. A multi-objective balanced optimization system for a virtual power plant distribution network, using the multi-objective balanced optimization method for a virtual power plant distribution network as claimed in any one of claims 1 to 5, characterized in that: include: A multi-objective function building module is used to build a multi-objective function with energy cost, energy loss and voltage safety as optimization targets, and distribution network flow constraints, distribution network operation constraints, voltage safety constraints and virtual power plant operation constraints as constraints; An optimization target mean determination module, used to determine the mean of each optimization target in the multi-objective function; The weight parameter optimization solution module is used to substitute the mean value of each optimization target into the multi-objective function, optimize and solve the weight parameters of each optimization target in the multi-objective function, and obtain the optimal solution of multi-objective equilibrium.

7. An electronic device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor is used to execute the steps of the method according to any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium for storing a computer program for performing multi-objective balanced optimization on a virtual power plant distribution network, characterized in that: When the computer program is run on a computer, the steps of the method according to any one of claims 1 to 5 are executed.

Images