Micro-grid group optimization configuration method, device and equipment for supporting reactive compensation of distribution network and medium

Abstract

The invention relates to the technical field of distributed energy management control under the condition of large-scale access of a micro-grid group, in particular to a micro-grid group optimization configuration method, device and equipment for supporting reactive compensation of a distribution network and a medium, and the method comprises the steps: building a chance constraint reactive power optimization model by considering the random characteristics of grid connection of a large-scale micro-grid group; the optimal installation capacity of the reactive power compensation device is solved through a random simulation particle swarm algorithm; based on the optimal installation capacity, probability distribution of system node voltage and line power flow is calculated by applying a semi-invariant probabilistic power flow method, and confidence level constraint is set for a result of the probability distribution to ensure that a planning result meets a predetermined confidence interval requirement, so that the equivalent annual cost generated by system operation is reduced, and the planning efficiency is improved. The overall economic benefit of the system is improved, and the probabilistic load flow calculation method based on the semi-invariant method converts complex convolution operation into simple arithmetic operation between semi-invariants, so that the complexity of the calculation process is reduced.

Description

Technical Field

[0001] The present invention relates to the field of distributed energy management and control technology under the condition of large-scale access of microgrid groups, and in particular to a microgrid group optimization configuration method, device, equipment and medium supporting reactive power compensation of distribution networks. Background Art

[0002] The continuous occurrence of smog weather in many places across the country and the severe ecological pressure have forced people to re-examine the energy supply structure. The desire to use and develop pollution-free distributed renewable green energy is unprecedentedly strong. The development and utilization of renewable energy is one of the effective measures to solve global energy and environmental problems. my country also regards the use of renewable energy as an important measure of the energy sustainable development strategy and vigorously promotes the construction and development of renewable clean energy. In order to promote the development of renewable energy, the State Grid Corporation of China has connected large-scale microgrid groups to the power grid under the support of industrial policies.

[0003] After the microgrid group is connected to the grid, the uncertainty of its output has a significant impact on reactive power compensation and voltage regulation. In addition, the existence of other uncertain factors in the system, such as load and faults, greatly increases the complexity and difficulty of reactive power and voltage optimization of the power system. At the same time, the microgrid group has a certain reactive power output capacity, and its operation mode will affect the flow of the connected power grid. The role of the reactive power output of the microgrid group itself needs to be considered in the reactive power planning of the power system. Due to the large number of microgrid group sites and the different ways in which each microgrid group is connected to the regional power grid, the regional power grid containing the microgrid group covers multiple voltage levels and the power grid topology is complex. At the same time, the output of the microgrid group is affected by the weather, and the output is characterized by volatility, intermittency and randomness, and the problem of grid voltage stability is prominent.

[0004] The information disclosed in this background technology section is only intended to deepen the understanding of the overall background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0005] The present invention provides a method, device, equipment and medium for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network, thereby effectively solving the problems in the background technology.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is: a microgrid group optimization configuration method supporting reactive power compensation of distribution network, comprising the following steps:

[0007] S10: Considering the random characteristics of the large-scale microgrid cluster grid connection, a chance-constrained reactive power optimization model is established, wherein the chance-constrained reactive power optimization model includes an objective function, equality constraints, and inequality constraints;

[0008] S20: Based on the chance-constrained reactive power optimization model, solving the optimal installation capacity of reactive power compensation devices by using a random simulation particle swarm algorithm;

[0009] S30: Based on the optimal installed capacity, the probability distribution of system node voltage and line flow is calculated using a semi-invariant probabilistic power flow method, and confidence level constraints are set on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation.

[0010] Furthermore, in step S10, the objective function of the opportunity-constrained reactive power optimization model takes minimization of the system equivalent annual total expenditure as the optimization goal, and the model of the system equivalent annual total expenditure includes:

[0011] minf=μQ Cost +P Cost ;

[0012] In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate.

[0013] Furthermore, the investment in the new compensation equipment and supporting equipment is Cost Models include:

[0014]

[0015] Where N PQ Represents the system PQ node set; R Qi K represents the unit capacity cost of capacitor installed at node i; QC Indicates the unit price of capacitor; Q Ci represents the installed capacity of the capacitor at node i; N PV represents the set of reactive power compensation nodes of the photovoltaic power station in the system; j Represents a logical variable, 1 means that the node has SVG installed, 0 means it is not installed; R SVGj K represents the cost of installing SVG at node j; SVGC Indicates the unit price of SVG; Q SVGj Indicates the SVG installation capacity of node j.

[0016] Furthermore, the annual operating cost of the network loss P Cost Models include:

[0017] P Cost =ηP loss cτ;

[0018] Where, η represents the market electricity price; P lossrepresents the active network loss in the chance constrained programming model; τ represents the maximum load utilization hours; c represents the load utilization hours correction factor; cτ represents the load equivalent utilization hours in the chance constrained programming model.

[0019] Furthermore, in step S10, the equality constraint is used to limit the active power and reactive power of each node, and the constraint conditions of the active power and the reactive power are:

[0020]

[0021] Where P i , Q i 、V i Respectively represent the injected active power, reactive power and voltage at node i; G ij , B ij ,θ ij Respectively represent the conductance, susceptance and phase angle difference between nodes i and j; V j represents the voltage amplitude at node j.

[0022] Furthermore, in step S10, the inequality constraint is used to limit the control variable, and the inequality constraint condition of the control variable is:

[0023]

[0024] Where P Gk , They represent the active power and upper and lower limits of the generator respectively; Q Gk , They represent the reactive power and upper and lower limits of the generator respectively; Q ci , Respectively represent the reactive capacity and upper and lower limits of the reactive compensation device; T j , They represent the transformer ratio and upper and lower limits respectively.

[0025] Furthermore, in step S10, the inequality constraint is used to restrict the state variable, and the inequality constraint condition of the state variable adopts the form of chance constraint, including:

[0026]

[0027] In the formula, α 1 , α 2 Indicates the confidence level of the constraints given in advance; P k , They represent the active power transmitted by line k and its upper and lower limits respectively; b represents the number of transmission lines in the system; V i , They represent the allowed voltage and upper and lower limits of node i respectively; n represents the total number of nodes.

[0028] Furthermore, in step S10, the chance-constrained reactive power optimization model is based on the improvement of the objective function and the chance constraint condition, and is optimized by a solution method. The chance-constrained reactive power optimization model includes:

[0029]

[0030] In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate; l represents the penalty factor; Indicates that the control variable inequality violates the deterministic constraint; Indicates that the state variable inequality violates the chance constraint.

[0031] Furthermore, the control variable inequality violates the deterministic constraint and the state variable inequality violates the chance constraint Models include:

[0032]

[0033] Where P cer , P max , P min They represent the actual value, upper limit and lower limit of the control variable respectively; p(x) represents the probability distribution function of the state variable; x max 、x min They represent the upper and lower limits of the state variables respectively; α represents the confidence level of the constraints given in advance.

[0034] Furthermore, in step S20, the steps of the particle swarm algorithm include:

[0035] S21: Initialization: Randomly generate the positions and velocities of m particles in the N-dimensional problem space;

[0036] S22: Evaluate particles: calculate the target fitness value of each individual particle;

[0037] S23: Update the optimal: Update the optimal information in the particle swarm, the steps include:

[0038] S231: Individual optimal update: Compare the particle's current target fitness value and its individual optimal value P best , if the current target fitness value is better than the individual optimal value P best , then update the individual optimal value P bestis the current target fitness value;

[0039] S232: Group Optimal Update: Compare the current target fitness value of the particle with the optimal value of the particle group G best , if the current target fitness value is better than the group optimal value G best , then update the group optimal value G best is the current target fitness value;

[0040] S24: Update Particles: Update the velocity and position of particles according to the following model:

[0041]

[0042] In the formula, v id represents the d-th dimension of the velocity vector of particle i; x id represents the d-th dimension of the position vector of particle i; r 1 、r 2 Represents a random number between [0,1]; c 1 、c 2 represents the acceleration coefficient; w represents the inertia weight; The term represents the “epistemic term”, which is related to the particle’s own cognitive experience; The term represents the “social term”, which represents the information sharing and cooperation among particle groups;

[0043] S25: Stop condition: Determine whether the termination condition is met. The stopping condition is set as the current optimal fitness value meets the preset condition or reaches the maximum iteration number. If so, the iteration is terminated. If not, return to step S22 to continue the iteration.

[0044] Further, in step S30, the steps of the semi-invariant probabilistic power flow method include:

[0045] S31: Fit the probability distribution of random variables in the power system and obtain the semi-invariants of each order of the input random variables by calculating the central moment;

[0046] S32: Based on the semi-invariants of the input random variables, in combination with the Jacobian matrix and the sensitivity matrix, perform arithmetic operations on the semi-invariants of the input random variables to obtain semi-invariants of various orders of output variables;

[0047] S33: Combining expansion methods of different series, based on the semi-invariant of the output variable, deriving the probability density expansion equation of the output variable, thereby determining the probability distribution of the output random variable.

[0048] The present invention also includes a microgrid group optimization configuration device supporting reactive power compensation of distribution network, using the above method, including:

[0049] A model building module is used to consider the random characteristics of the large-scale microgrid cluster grid connection and establish a chance-constrained reactive power optimization model, wherein the chance-constrained reactive power optimization model includes an objective function, equality constraints, and inequality constraints;

[0050] A capacity solving module, used for solving the optimal installation capacity of reactive power compensation devices through a random simulation particle swarm algorithm based on the chance-constrained reactive power optimization model;

[0051] A probability distribution calculation module, used to calculate the probability distribution of system node voltage and line power flow using a semi-invariant probability power flow method based on the optimal installed capacity;

[0052] The confidence level constraint setting module is used to set confidence level constraints on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation.

[0053] The present invention also includes a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described above is implemented.

[0054] The present invention also includes a storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described above is implemented.

[0055] The beneficial effects of the present invention are:

[0056] Considering the random characteristics of large-scale microgrid cluster grid connection, a chance-constrained reactive power optimization model is established; based on the chance-constrained reactive power optimization model, the optimal installed capacity of reactive compensation devices is solved by using a random simulation particle swarm algorithm; based on the optimal installed capacity, the semi-invariant probabilistic power flow method is used to calculate the probability distribution of system node voltage and line power flow, and confidence level constraints are set on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation and improving the overall economic benefits of the system. The probabilistic power flow calculation method based on the semi-invariant method converts complex convolution operations into simple arithmetic operations between semi-invariants, thereby reducing the complexity of the calculation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 A flow chart of a method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network;

[0059] Figure 2 This is the flow chart of particle swarm algorithm;

[0060] Figure 3 This is the flow chart of the probability power flow calculation using the invariant method;

[0061] Figure 4 A schematic diagram of the structure of a device for optimizing the configuration of a microgrid group supporting reactive power compensation of a distribution network;

[0062] Figure 5 A schematic diagram of the structure of a computer device. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0064] like Figures 1 to 3 As shown: A method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network comprises the following steps:

[0065] S10: Considering the random characteristics of large-scale microgrid cluster grid connection, a chance-constrained reactive power optimization model is established. The chance-constrained reactive power optimization model includes objective function, equality constraint and inequality constraint to characterize the reactive power compensation demand of the system under various uncertain factors.

[0066] S20: Based on the chance-constrained reactive power optimization model, the optimal installation capacity of reactive power compensation devices is solved by random simulation particle swarm algorithm to meet the reactive power demand of the system;

[0067] S30: Based on the optimal installed capacity, the semi-invariant probability power flow method is used to calculate the probability distribution of system node voltage and line power flow, and confidence level constraints are set on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation and improving the overall economic benefits of the system.

[0068] The probability power flow calculation method based on the semi-invariant method converts complex convolution operations into simple arithmetic operations between semi-invariants, thereby reducing the complexity of the calculation process.

[0069] In the problem of reactive power optimization of power system, the objectives of reactive power optimization are usually divided into two categories: maximum economic benefit or minimum network loss. In the problem of power system planning, it is necessary to consider the system operation status within a certain time scale, so the maximum economic benefit is taken as the goal; in the problem of optimal operation analysis, because the economic investment has been determined, the optimization analysis is often carried out from the perspective of minimizing the system network loss.

[0070] By constructing a chance-constrained reactive power optimization model, the scheme can take into account the random factors brought about by the large-scale microgrid grid connection. The model contains objective functions and constraints, and can accurately characterize the reactive power compensation demand of the system under various uncertain factors.

[0071] The random simulation particle swarm algorithm is used to calculate the optimal reactive power compensation installation capacity to ensure that the reactive power demand of the system under different loads and uncertain factors is met. By reasonably configuring the compensation capacity, the scheme achieves the optimization of reactive power compensation, effectively reduces the equivalent annual operating cost of the system, and improves economic benefits.

[0072] The semi-invariant probabilistic power flow method transforms the originally complex convolution calculation into simple arithmetic operations between semi-invariants, which greatly reduces the complexity of the calculation process and improves the computational efficiency of reactive power optimization. This enables the scheme to quickly respond to the real-time operation needs of the power grid and reduces the consumption of computing resources. It is suitable for efficient reactive power compensation configuration in a large-scale microgrid grid-connected environment.

[0073] By imposing confidence level constraints on the probability distribution of node voltages and line flows, the technical solution ensures the voltage stability of the system under different operating conditions, improves the safety and reliability of the power system, and helps maintain stable operation of the power grid.

[0074] In this embodiment, in step S10, the objective function of the opportunity-constrained reactive power optimization model takes minimization of the system equivalent annual total expenditure as the optimization goal, including: the investment Q of the newly added compensation equipment and supporting equipment Cost , the annual operating cost of network loss P Cost , the model of the system equivalent annual total expenditure includes:

[0075] minf=μQ Cost +P Cost ;

[0076] In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate.

[0077] By incorporating "investment in new compensation equipment and supporting equipment" and "annual operating costs of network losses" into the objective function, minimizing the overall cost can effectively reduce the total expenditure of system operation, optimize the economic benefits of the power grid, and consider the cost of the entire life cycle, ensuring the balance between the return on investment and operating costs of the reactive compensation configuration method, thereby improving the overall economy.

[0078] Investment in new compensation equipment and supporting equipment Q Cost It includes two parts: reactive power compensation equipment fee (proportional to reactive power compensation capacity) and supporting installation fee, among which the equipment fee and installation fee of capacitor and SVG are different;

[0079] Among them, the investment in new compensation equipment and supporting equipment is Q Cost Models include:

[0080]

[0081] Where N PQ Represents the system PQ node set; R Qi K represents the unit capacity cost of capacitor installed at node i; QC Indicates the unit price of capacitor; Q Ci represents the installed capacity of the capacitor at node i; N PV represents the set of reactive power compensation nodes of the photovoltaic power station in the system; j Represents a logical variable, 1 means that the node has SVG installed, 0 means it is not installed; R SVGj K represents the cost of installing SVG at node j; SVGC Indicates the unit price of SVG; Q SVGj Indicates the SVG installation capacity of node j.

[0082] Among them, the annual operating cost of network loss is P Cost Models include:

[0083] P Cost =ηP loss cτ;

[0084] Where, η represents the market electricity price; P loss represents the active network loss in the chance constrained programming model; τ represents the maximum load utilization hours; c represents the load utilization hours correction factor; cτ represents the load equivalent utilization hours in the chance constrained programming model.

[0085] The model incorporates factors such as market electricity prices, active network losses and load utilization hours into the calculation of annual operating costs, making the operating cost estimate more in line with the actual market and grid operation conditions, thereby improving the accuracy of cost forecasts. Accurate cost estimates can help system managers better formulate operating budgets and optimization strategies.

[0086] As a preference of the above embodiment, in step S10, the equality constraint is used to limit the active power and reactive power of each node, and the constraint conditions of the active power and reactive power are:

[0087]

[0088] Where P i , Q i 、V i Respectively represent the injected active power, reactive power and voltage at node i; G ij , B ij ,θ ij Respectively represent the conductance, susceptance and phase angle difference between nodes i and j; V j represents the voltage amplitude at node j.

[0089] By accurately limiting the active and reactive power of each node, the equation constraint can effectively maintain the stability of the system voltage. Especially when large-scale distributed power sources are connected, the equation constraint helps control voltage fluctuations and imbalances, reduce voltage deviations of the power grid under different load conditions, and ensure the safe operation of the system.

[0090] In this embodiment, in step S10, the inequality constraint is used to limit the control variable, and the inequality constraint condition of the control variable is:

[0091]

[0092] Where P Gk , They represent the active power and upper and lower limits of the generator respectively; Q Gk , They represent the reactive power and upper and lower limits of the generator respectively; Q ci , Respectively represent the reactive capacity and upper and lower limits of the reactive compensation device; T j , They represent the transformer ratio and upper and lower limits respectively.

[0093] By setting upper and lower limits for the active and reactive power of the generator, power output beyond the safe range is avoided, ensuring that the power generation equipment operates within a reasonable power range. This limitation can effectively prevent the occurrence of overload and improve the safety and stability of the power system.

[0094] In step S10, the inequality constraint is used to restrict the state variable, and the inequality constraint condition of the state variable adopts the form of chance constraint, including:

[0095]

[0096] In the formula, α 1 , α 2 Indicates the confidence level of the constraints given in advance; P k , They represent the active power transmitted by line k and its upper and lower limits respectively; b represents the number of transmission lines in the system; V i , They represent the allowed voltage and upper and lower limits of node i respectively; n represents the total number of nodes.

[0097] The chance constraint form sets a confidence level for the state variables, ensuring that the system state remains within a safe range with high confidence, which helps to improve the stability and reliability of the system under various operating conditions, especially in the face of uncertain factors such as load fluctuations and distributed power generation fluctuations, and can effectively ensure the robustness of the system.

[0098] In the planning process, a penalty function is used to deal with violations of constraints. The out-of-bounds inequality constraints are added to the original objective function in the form of penalty terms, so that the objective function in this case will inevitably be eliminated.

[0099] As a preferred embodiment of the above embodiment, in step S10, the opportunity-constrained reactive power optimization model is based on the improvement of the objective function and the opportunity constraint condition, and adopts a solution method to optimize the configuration. The opportunity-constrained reactive power optimization model includes:

[0100]

[0101] In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate; l represents the penalty factor; Indicates that the control variable inequality violates the deterministic constraint; Indicates that the state variable inequality violates the chance constraint.

[0102] By introducing new compensation equipment investment, network loss costs and penalty factors into the objective function, the optimization model can comprehensively consider the investment and operating costs to ensure that the equivalent annual total expenditure of the system is minimized. It takes into account both the fixed investment of the system and the long-term operating costs, making the reactive power compensation configuration scheme more economically advantageous.

[0103] Among them, the control variable inequality violates the deterministic constraint and state variable inequalities violate chance constraints Models include:

[0104]

[0105] Where P cer , P max , P min They represent the actual value, upper limit and lower limit of the control variable respectively; p(x) represents the probability distribution function of the state variable; x max 、x min They represent the upper and lower limits of the state variables respectively; α represents the confidence level of the constraints given in advance.

[0106] By establishing a clear model to describe the violations of control variables and state variables, the scheme can flexibly respond to complex operating conditions and ensure that the system can still meet the main operating constraints when uncertain fluctuations occur, making the system more robust and stable in dynamic environments such as load fluctuations and renewable energy access.

[0107] The main idea of ​​the basic particle swarm optimization algorithm is to achieve information sharing and collaboration in the group through the self-experience learning and experience exchange of individuals in the group, dynamically change the speed of particles, and achieve group optimization. The algorithm model is established on the basis of self-learning, group learning, and optimal change. Suppose in an N-dimensional target search space, there are m particles forming a group, where the position X of the i-th particle in the N-dimensional space is i =(x i1 ,x i2 ,...,x iN ) T ,i=1,2,...,m; flight speed V i =(v i1 ,v i2 ,...,v iN ) T ,i=1,2,...,m; target fitness value fitness i =f(X i ), then P besti and are the maximum fitness value and the corresponding position that the i-th particle has ever reached, G best It is the best position experienced by all particles in the group, and its index number is g.

[0108] As a preferred embodiment of the above, Figure 2 As shown, in step S20, the steps of the particle swarm algorithm include:

[0109] S21: Initialization: Randomly generate the positions and velocities of m particles in the N-dimensional problem space;

[0110] S22: Evaluate particles: calculate the target fitness value of each individual particle;

[0111] S23: Update the optimal: Update the optimal information in the particle swarm, the steps include:

[0112] S231: Individual optimal update: Compare the particle's current target fitness value and its individual optimal value P best , if the current target fitness value is better than the individual optimal value P best , then update the individual optimal value P best is the current target fitness value;

[0113] S232: Group optimal update: Compare the particle's current target fitness value with the particle group's optimal value G best , if the current target fitness value is better than the group optimal value G best , then update the group optimal value G best is the current target fitness value;

[0114] S24: Update Particles: Update the velocity and position of particles according to the following model:

[0115]

[0116] In the formula, v id represents the d-th dimension of the velocity vector of particle i; x id represents the d-th dimension of the position vector of particle i; r 1 、r 2 Represents a random number between [0,1]; c 1 、c 2 represents the acceleration coefficient; w represents the inertia weight; The term represents the “epistemic term”, which is related to the particle’s own cognitive experience; The term represents the “social term”, which represents the information sharing and cooperation among particle groups;

[0117] S25: Stop condition: Determine whether the termination condition is met. The stopping condition is set to that the current optimal fitness value meets the preset condition or reaches the maximum iteration number. If so, the iteration is terminated. If not, return to step S22 to continue iteration.

[0118] The algorithm uses a dual update mechanism of individual optimality and group optimality, allowing particles to search based on their own experience (cognitive terms) and perform global optimization with the help of the group's best solution (social terms). The dual mechanism effectively balances the relationship between exploration and utilization, and significantly improves global search capabilities and optimization efficiency.

[0119] Compared with other intelligent algorithms, the particle swarm algorithm has the characteristics of simple algorithm theory and dynamically adjustable search capabilities. At the same time, according to the calculation model of the particle swarm algorithm, it can be seen that the particle swarm algorithm requires fewer parameters, and it is these limited number of parameters that determine the performance of the particle swarm algorithm.

[0120] Population size: The larger the particle swarm population size, the more particles there are in the corresponding particle swarm, so the search capability of the particle swarm algorithm can be better exerted; however, the increase in the number of particles will inevitably lead to a significant increase in computing time. Considering various factors, the population size in the general particle swarm algorithm is 20-40, and the number of particles can be more than 100 for special problems that are difficult to solve.

[0121] Inertia weight: As an important variable that controls the influence of the previous velocity of the particle swarm on the current velocity, the selection of its value has a great influence on the global and local search capabilities of the particles. In terms of function, the inertia weight plays a role in balancing the global search and local search capabilities: a larger inertia weight value means a stronger global search capability, which is conducive to searching new areas; a smaller inertia weight value means a higher local precision search capability, which is conducive to finding the local optimal solution.

[0122] Acceleration coefficient: The acceleration coefficient is also called the learning factor. It is a weight parameter for the particle to adjust its speed based on its own historical optimal experience and the group's optimal information. The acceleration coefficient and the inertia weight work together to achieve the purpose of balancing the global search and local search capabilities. However, it is worth noting that the larger the acceleration coefficient, the stronger the local search capability and the more conducive to algorithm convergence, which is exactly the opposite of the value of the inertia weight.

[0123] Maximum speed: The maximum speed of a particle indicates the maximum distance it moves in one iteration. If the maximum speed is large, the particle has a strong global exploration capability and is likely to fly across the boundary of feasible solutions. If the maximum speed is small, the particle can complete a refined local search and is likely to fall into a local optimal solution. Currently, in the processing method of the particle swarm optimization algorithm, the maximum speed is not specially processed and is only used as a simple constraint.

[0124] Stopping conditions: The stopping conditions of the particle swarm algorithm are generally divided into two categories: one is to meet the pre-set solution judgment conditions, which need to set corresponding criteria according to the nature of the problem and the randomness of the search; the other is to reach the maximum number of iterations. Usually, the two types of criteria together constitute the stopping conditions.

[0125] The semi-invariant method calculates the probability power flow, and the target result is divided into two parts: one is the probability distribution constraint of the node voltage; the other is the probability distribution constraint of the line power. Although the calculation methods of node voltage and line power flow are both based on the Newton-Raphson method, there are certain differences in the specific calculation process. The semi-invariant calculation methods of node voltage and line power flow are given separately.

[0126] As a preferred embodiment of the above, Figure 3 As shown, in step S30, the steps of the semi-invariant probability power flow method include:

[0127] S31: Fit the probability distribution of random variables in the power system and obtain the semi-invariants of each order of the input random variables by calculating the central moment;

[0128] S32: Based on the semi-invariants of the input random variables, in combination with the Jacobian matrix and the sensitivity matrix, arithmetic operations are performed on the semi-invariants of the input random variables to obtain semi-invariants of various orders of the output variables;

[0129] S33: Combining the expansion methods of different series, based on the semi-invariant of the output variable, the probability density expansion equation of the output variable is derived to determine the probability distribution of the output random variable.

[0130] By converting complex probability calculations into central moment and semi-invariant operations, the computational complexity is reduced and the cumbersome convolution calculations in traditional probabilistic power flow analysis are avoided. This not only simplifies the calculation process, but also significantly improves the solution efficiency, making it suitable for real-time computing needs in large-scale distribution network scenarios.

[0131] After the semi-invariant method completes the calculation, a reverse operation is required to obtain the probability density distribution of the original random variable from the semi-invariant parameters of each order. There are many ways to calculate the function value of the distribution of an unknown random variable through the moments of each order or the semi-invariants of each order. In the power system, there is the Gram-Charlier series expansion.

[0132] The semi-invariant method is used to calculate the probabilistic power flow. The target result is divided into two parts: one is the probability distribution constraint of the node voltage; the other is the probability distribution constraint of the line power. The calculation methods of the node voltage and the line power flow are both based on the Newton-Raphson method, but there are certain differences in the specific calculation process. The semi-invariant calculation methods of the node voltage and the line power flow are given respectively.

[0133] The power equation of the power system node is:

[0134] For a normal operating system state, W can be expressed as 0 =f(X 0 ,Y0 );

[0135] Where: W 0 Indicates the node active and reactive power injection vector under normal circumstances; X 0 represents the state vector composed of node voltage and phase angle under normal conditions; Y 0 Indicates the network parameters under normal conditions.

[0136] Assuming that the system power injection is disturbed by ΔW, or the network changes by ΔY, then the state quantity must change to a certain extent by ΔX. The above data satisfies W 0 +ΔW=f(X 0 +ΔX,Y 0 +ΔY);

[0137] Expanding the above formula according to Taylor series, we can get:

[0138]

[0139] When the disturbance and state change are not large, ignore (ΔX) 2 And its higher-order terms, since f(X,Y) is a linear function of Y, f″ yy (X,Y)=0, which simplifies to:

[0140] W 0 +ΔW=f(X 0 ,Y 0 )+f′ x (X 0 ,Y 0 )·ΔX+f′ y (X 0 ,Y 0 )·ΔY+f″ xy (X 0 ,Y 0 )ΔXΔY;

[0141] That is, ΔW = f′ x (X 0 ,Y 0 )·ΔX+f′ y (X 0 ,Y 0 )·ΔY+f″ xy (X 0 ,Y 0 )ΔXΔY;

[0142] From the above formula, it can be clearly seen that the linear relationship between the state variable, node power disturbance and network structure change is:

[0143] ΔX=[f x'(X 0 ,Y 0 )+f x ” y (X 0 ,Y 0 )ΔY] -1 [ΔW-f y '(X 0 ,Y 0 )·ΔY];

[0144] When the grid structure remains unchanged, ΔY=0, the above formula changes to:

[0145] ΔX=[f x '(X 0 ,Y 0 )] -1 ΔW=S 0 ΔW, where:

[0146]

[0147] Where: J 0 represents the Jacobian matrix at the end of the power flow iteration; S 0 Represents the sensitivity matrix. Since the Jacobian matrix has been calculated during the power flow calculation process, the sensitivity matrix is ​​easy to obtain.

[0148] Similarly, the linear relationship between the branch power flow equation and the injected power ΔW can be obtained through corresponding derivation. The branch power equation is:

[0149]

[0150] Where: t ij Indicates the per-unit value of branch transformation ratio; b ij0 Indicates half of the branch capacity.

[0151] The above formula can be expressed as Z = g(X), and the Taylor series expansion yields:

[0152] Z=Z 0 +ΔZ=g(X 0 +ΔX)=g(X 0 )+G 0 ΔX+···;

[0153] Because Z 0 is the calculated value of the power flow under normal conditions, and in the steady-state operation mode Z 0 =g(X 0 ), ignoring the higher-order terms, we can get ΔZ=G 0 ΔX, where G 0It is a 2b×2N-order matrix (b is the number of branches and N is the number of nodes).

[0154] In order to meet the requirements of convolution operation of independent random variables, the operation relationship of branch power flow on independent random variables can be obtained:

[0155] ΔZ=G 0 S 0 ΔW=T 0 ΔW;

[0156] That is the basic formula for solving the node voltage probability distribution and line power flow probability distribution in the probability power flow by the semi-invariant method, T 0 is the line power flow sensitivity matrix.

[0157] According to the properties of semi-invariants, the complex convolution calculation can be transformed into a matrix product operation by performing Taylor expansion and linearization on the system power flow constraint equation:

[0158]

[0159] Where: ΔX (k) , ΔZ (k) k-order semi-invariant representing the node voltage state vector and line power flow state vector; represents the matrix composed of the kth power of the elements in the node voltage sensitivity matrix and the line power flow sensitivity matrix respectively;

[0160] ΔW (k) represents a k-th order semi-invariant of the independently injected random variables.

[0161] Usually, in order to obtain the probability density function of a certain state variable, the corresponding probability density function is obtained by performing series expansion on the semi-invariants of each state variable.

[0162] The embodiment of the present invention supports the optimization configuration method of a microgrid group for supporting reactive compensation of a distribution network with efficient collaboration between the main and microgrids, fully considers the random characteristics of the grid connection of a large-scale microgrid cluster, establishes a chance-constrained reactive optimization model based on objective functions, equality constraints and inequality constraints, and applies a random simulation particle swarm algorithm to solve the installed capacity of reactive compensation devices; uses a semi-invariant probabilistic power flow method to solve the probability distribution of system node voltages and line power flows, and imposes confidence level constraints on them; a probabilistic power flow calculation method based on the semi-invariant method converts complex convolution operations into simple arithmetic operations between semi-invariants, thereby reducing the complexity of the calculation process.

[0163] The present invention also includes a microgrid group optimization configuration device supporting reactive power compensation of distribution network, such as Figure 4 As shown, using the method as described above, including:

[0164] The model building module is used to consider the random characteristics of the large-scale microgrid cluster grid connection and establish a chance-constrained reactive power optimization model. The chance-constrained reactive power optimization model includes an objective function, equality constraints, and inequality constraints.

[0165] The capacity solving module is used to solve the optimal installation capacity of reactive power compensation devices through random simulation particle swarm algorithm based on the chance-constrained reactive power optimization model;

[0166] The probability distribution calculation module is used to calculate the probability distribution of system node voltage and line power flow based on the optimal installed capacity using the semi-invariant probability power flow method;

[0167] The confidence level constraint setting module is used to set confidence level constraints on the results of probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation.

[0168] See also Figure 5 A computer device 400 provided in an embodiment of the present application includes: a processor 410 and a memory 420, wherein the memory 420 stores a computer program executable by the processor 410, and when the computer program is executed by the processor 410, the above method is executed.

[0169] The embodiment of the present application further provides a storage medium 430 on which a computer program is stored. When the computer program is run by the processor 410, the above method is executed.

[0170] Among them, the storage medium 430 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0171] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. "Multiple" means two or more, unless otherwise clearly and specifically defined.

[0172] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0173] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0174] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0175] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0176] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0177] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0178] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network, characterized in that: The steps include: S10: Considering the random characteristics of the large-scale microgrid cluster grid connection, a chance-constrained reactive power optimization model is established, wherein the chance-constrained reactive power optimization model includes an objective function, equality constraints, and inequality constraints; S20: Based on the chance-constrained reactive power optimization model, solving the optimal installation capacity of reactive power compensation devices by using a random simulation particle swarm algorithm; S30: Based on the optimal installed capacity, the probability distribution of system node voltage and line flow is calculated using a semi-invariant probabilistic power flow method, and confidence level constraints are set on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation.

2. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S10, the objective function of the opportunity-constrained reactive power optimization model takes minimization of the system equivalent annual total expenditure as the optimization goal, and the model of the system equivalent annual total expenditure includes: minf=μQ Cost +P Cost ; In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate.

3. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 2, characterized in that: The investment in the new compensation equipment and supporting equipment is Q Cost Models include: Where N PQ Represents the system PQ node set; R Qi K represents the unit capacity cost of capacitor installed at node i; QC Indicates the unit price of capacitor; Q Ci represents the installed capacity of the capacitor at node i; N PV represents the set of reactive power compensation nodes of the photovoltaic power station in the system; j Represents a logical variable, 1 means that the node has SVG installed, 0 means it is not installed; R SVGj K represents the cost of installing SVG at node j; SVGC Indicates the unit price of SVG; Q SVGj Indicates the SVG installation capacity of node j.

4. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 2, characterized in that: The annual operating cost of the network loss is P Cost Models include: P Cost =ηP loss cτ; Where η represents the market electricity price; P loss represents the active network loss in the chance constrained programming model; τ represents the maximum load utilization hours; c represents the load utilization hours correction factor; cτ represents the load equivalent utilization hours in the chance constrained programming model.

5. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S10, the equality constraint is used to limit the active power and reactive power of each node, and the constraint conditions of the active power and the reactive power are: Where P i , Q i 、V i Respectively represent the injected active power, reactive power and voltage at node i; G ij , B ij ,θ ij Respectively represent the conductance, susceptance and phase angle difference between nodes i and j; V j represents the voltage amplitude at node j.

6. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S10, the inequality constraint is used to limit the control variable, and the inequality constraint condition of the control variable is: Where P Gk , They represent the active power and upper and lower limits of the generator respectively; Q Gk , They represent the reactive power and upper and lower limits of the generator respectively; Q ci , Respectively represent the reactive capacity and upper and lower limits of the reactive compensation device; T j , They represent the transformer ratio and upper and lower limits respectively.

7. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S10, the inequality constraint is used to restrict the state variable, and the inequality constraint condition of the state variable is in the form of a chance constraint. include: In the formula, α1 and α2 represent the confidence levels of the constraints given in advance; P k , They represent the active power transmitted by line k and its upper and lower limits respectively; b represents the number of transmission lines in the system; V i , They represent the allowed voltage and upper and lower limits of node i respectively; n represents the total number of nodes.

8. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S10, the chance-constrained reactive power optimization model is configured based on the improvement of the objective function and the chance constraint condition, and adopts a solution method to optimize the configuration. The chance-constrained reactive power optimization model includes: In the formula, Q Cost Represents the investment in new compensation equipment and supporting equipment; P Cost represents the annual operating cost of network loss; μ represents the investment recovery rate; l represents the penalty factor; Indicates that the control variable inequality violates the deterministic constraint; Indicates that the state variable inequality violates the chance constraint.

9. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 8, characterized in that: The control variable inequality violates the deterministic constraint and the state variable inequality violates the chance constraint Models include: Where P cer , P max , P min They represent the actual value, upper limit and lower limit of the control variable respectively; p(x) represents the probability distribution function of the state variable; x max 、x min They represent the upper and lower limits of the state variables respectively; α represents the confidence level of the constraints given in advance.

10. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S20, the steps of the particle swarm algorithm include: S21: Initialization: Randomly generate the positions and velocities of m particles in the N-dimensional problem space; S22: Evaluate particles: calculate the target fitness value of each individual particle; S23: Update the optimal: Update the optimal information in the particle swarm, the steps include: S231: Individual optimal update: Compare the particle's current target fitness value and its individual optimal value P best , if the current target fitness value is better than the individual optimal value P best , then update the individual optimal value P best is the current target fitness value; S232: Group Optimal Update: Compare the current target fitness value of the particle with the optimal value of the particle group G best , if the current target fitness value is better than the group optimal value G best , then update the group optimal value G best is the current target fitness value; S24: Update Particles: Update the velocity and position of particles according to the following model: In the formula, v id represents the d-th dimension of the velocity vector of particle i; x id represents the d-th dimension component of the position vector of particle i; r1 and r2 represent random numbers between [0,1]; c1 and c2 represent acceleration coefficients; w represents inertia weight; The term represents the "cognitive term", which is related to the particle's own cognitive experience; The term represents the "social term", which represents the information sharing and cooperation between particle groups; S25: Stop condition: Determine whether the termination condition is met. The stopping condition is set as the current optimal fitness value meets the preset condition or reaches the maximum iteration number. If so, the iteration is terminated. If not, return to step S22 to continue the iteration.

11. The method for optimizing configuration of a microgrid group supporting reactive power compensation of a distribution network according to claim 1, characterized in that: In step S30, the steps of the semi-invariant probabilistic power flow method include: S31: Fit the probability distribution of random variables in the power system and obtain the semi-invariants of each order of the input random variables by calculating the central moment; S32: Based on the semi-invariants of the input random variables, in combination with the Jacobian matrix and the sensitivity matrix, perform arithmetic operations on the semi-invariants of the input random variables to obtain semi-invariants of various orders of output variables; S33: Combining expansion methods of different series, based on the semi-invariant of the output variable, deriving the probability density expansion equation of the output variable, thereby determining the probability distribution of the output random variable.

12. A microgrid group optimization configuration device supporting reactive power compensation of distribution network, characterized in that: Use of the method according to any one of claims 1 to 11, comprising: A model building module is used to consider the random characteristics of the large-scale microgrid cluster grid connection and establish a chance-constrained reactive power optimization model, wherein the chance-constrained reactive power optimization model includes an objective function, equality constraints, and inequality constraints; A capacity solving module, used for solving the optimal installed capacity of reactive power compensation devices through a random simulation particle swarm algorithm based on the chance-constrained reactive power optimization model; A probability distribution calculation module, used for calculating the probability distribution of system node voltage and line power flow using a semi-invariant probability power flow method based on the optimal installed capacity; The confidence level constraint setting module is used to set confidence level constraints on the results of the probability distribution to ensure that the planning results meet the predetermined confidence interval requirements, thereby reducing the equivalent annual cost generated by system operation.

13. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 11 is implemented.

14. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

Images