Planning Methods and Systems for Energy Storage Equipment in Medium- and Low-Voltage AC-DC-Storage Hybrid Distribution Networks

Abstract

The present invention belongs to the technical field of energy storage equipment planning, and specifically relates to a method and system for energy storage equipment planning of a medium and low voltage AC / DC hybrid distribution network. The method includes, in the planning layer, taking an annual cycle and the minimum comprehensive planning cost as the goal, deciding on a solution for converter capacity, energy storage equipment capacity and line model, and transmitting the solution to the operation layer, clustering based on the historical data of DG output and load to obtain multiple typical scenarios of "source-load", and optimizing the results with the goal of minimizing the operating cost of each typical scenario. The planning layer multiplies the normal operating loss of the distribution network of each typical scenario by the corresponding number of days and then adds them to obtain the distribution network operating cost, and determines the feasibility level of the solution. The present invention solves the problem that the energy storage planning in the traditional hybrid power grid is not accurate enough and cannot cope with complex multi-objective optimization requirements, and helps to ensure that the energy storage system can maintain stable and efficient operation under a variety of complex conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy storage equipment planning, and in particular relates to an energy storage equipment planning method and system for a medium- and low-voltage AC / DC hybrid distribution network. Background Art

[0002] The widespread access to renewable energy such as wind and solar energy has brought great challenges to the traditional power system. Renewable energy is intermittent and volatile, and it is difficult to match power production with load demand in real time, resulting in grid frequency fluctuations, supply and demand imbalances, and power quality problems. In order to cope with these fluctuations, energy storage equipment is regarded as a key technology that can store excess electricity when power demand is low and release energy when demand is high, thereby balancing supply and demand and optimizing grid operation. As the power system develops towards intelligence, the distribution network is gradually changing from the traditional one-way power transmission mode to a two-way power flow and diversified power source coexistence mode. As an important part of the smart grid, the energy storage system can respond to real-time control instructions and improve the stability and reliability of the power grid. In the AC / DC hybrid power grid, energy storage equipment can be used as a regulating device for the AC power grid and can also be used for energy balance in the DC distribution network to help achieve coordinated scheduling of AC / DC power systems.

[0003] There are still some deficiencies in the existing research on the planning of energy storage equipment involved in medium and low voltage AC / DC hybrid distribution networks. Specifically, the traditional AC distribution network does not fully consider the energy loss and efficiency issues in the AC / DC conversion process, resulting in insufficient accuracy in the energy storage planning in the hybrid power grid. In addition, the traditional energy storage equipment planning method usually adopts a single objective for optimization and cannot cope with complex multi-objective optimization requirements. There are many technical constraints involved in the energy storage planning process. The traditional method cannot handle these multi-objective and multi-constraint problems, and the planning results are not very operational in practical applications. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method and system for planning energy storage equipment for a medium and low voltage AC / DC hybrid distribution network, which can effectively solve the problems involved in the above-mentioned background technology and help ensure that the energy storage system can maintain stable and efficient operation under a variety of complex conditions.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] The energy storage equipment planning method for a medium and low voltage AC / DC hybrid distribution network includes the following steps:

[0007] In the planning layer, the converter capacity, energy storage device capacity and line model are decided on with an annual cycle and the minimum comprehensive planning cost as the goal, and the plan is transmitted to the operation layer;

[0008] The operation layer clusters the historical data of DG output and load to obtain multiple typical "source-load" scenarios, and optimizes the results to minimize the operating cost of each typical scenario. The results are the charging and discharging power of the energy storage device and the conversion power of the converter at different times, and the results are passed to the planning layer.

[0009] The planning layer multiplies the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then adds them together to obtain the distribution network operation cost;

[0010] The scheme is applied to the power system simulation software, the simulated operating status of the energy storage equipment in the power system simulation software is evaluated, and the feasibility level of the scheme is determined in combination with the minimum comprehensive planning cost.

[0011] Preferably, the minimum comprehensive planning cost is expressed as follows:

[0012] ;

[0013] In the formula, is the minimum comprehensive planning cost, is the annual cost of investment, operation and maintenance, and decommissioning of the voltage source converter VSC, The annual cost of investment, operation and maintenance, and decommissioning of energy storage devices. The annual cost of the construction of distribution network lines, The operating cost of the distribution network;

[0014] ;

[0015] ;

[0016] In the formula, is the annual equivalent cost of the converter investment, is the annual equivalent cost of converter operation and maintenance costs, is the equivalent annual cost of the converter scrapping cost, is the equivalent annual cost of energy storage investment, is the annual cost of energy storage operation and maintenance costs, The equivalent annual value of the energy storage scrapping cost;

[0017] ;

[0018] Where b is the discount rate, x is the full life cycle of the line, is the number of lines, is the construction cost per unit length of the ith line, is the planned length of the i-th line;

[0019] ;

[0020] In the formula, is the number of distribution network operation scenarios obtained by clustering, is the probability of the kth scene, For electricity price, is the normal operation loss of the distribution network in the kth scenario;

[0021] Constraints:

[0022] ;

[0023] ;

[0024] In the formula, is the installed capacity of VSC, is the installed capacity of the energy storage device, and are all positive integers, For the unit installation capacity of VSC, The unit installed capacity of the energy storage device.

[0025] Preferably, the decision-making process for determining converter capacity, energy storage device capacity and line model is:

[0026] Step 1.1: Set the maximum number of iterations and the size of the population of the particle swarm algorithm, initialize the particle code and the position and velocity of each particle;

[0027] Step 1.2: Calculate the fitness of each particle and update the individual optimal value of each particle and the global optimal value of the entire particle group. The fitness is expressed as the minimum comprehensive planning cost.

[0028] Step 1.3: Introduce adaptive weights and shrinkage factors to update the position and velocity of particles:

[0029] Adaptive Weights:

[0030] ;

[0031] In the formula, is the adaptive weight, the value range is [0.5, 1], rand is a random number between 0 and 1, is the maximum value of the adaptive weight, is the minimum value of the adaptive weight;

[0032] Shrinkage Factor:

[0033] ;

[0034] In the formula, is the shrinkage factor, for , The sum of the two learning factors is greater than 4;

[0035] During each iteration, the particle updates its velocity and position according to the following formulas;

[0036] ;

[0037] ;

[0038] In the formula, is the velocity of the oth particle at the m+1th iteration, is the velocity of the oth particle at the mth iteration, m is the number of iterations, is the first learning factor, is the second learning factor, , is a random number between [0, 1]. is the position of the oth particle at the mth iteration, is the position of the oth particle at the m+1th iteration, is the optimal position of the oth particle itself at the mth iteration, is the global optimal position of the particle group at the mth iteration;

[0039] Step 1.4: Determine whether the maximum number of iterations has been reached. If so, output the solution corresponding to the optimal particle. Otherwise, return to step 1.2 to continue iterating and optimizing.

[0040] Preferably, the operation layer obtains multiple typical “source-load” scenarios by clustering historical data of DG output and load.

[0041] Step 2.1: Select the characteristic quantities of DG daily output curve and daily load curve. The selected characteristic quantities are:

[0042] ;

[0043] In the formula, is the average daily output of DG, is the DG output during period t, is the variance of DG daily force, is the average daily load, is the load power in period t, is the variance of daily load; T is the total number of time periods;

[0044] Step 2.2: Set the range of the number of clusters to w belongs to , select w initial cluster centers, traverse the 365 groups of "source-load" sample data throughout the year, calculate the Euclidean distance between the rth sample and each cluster center, and assign the rth sample to the cluster center that minimizes the Euclidean distance. The calculation formula of the Euclidean distance between the rth sample and each cluster center is as follows:

[0045] ;

[0046] In the formula, is the Euclidean distance between the rth sample and the wth cluster center, is the hth eigenvector of the rth sample, is the hth eigenvector of the wth cluster center;

[0047] Step 2.3: Count the number of samples in each class, update w cluster centers according to the mean of the sample data, and repeat steps 2.2 and 2.3 until the cluster center no longer changes;

[0048] Step 2.4: Calculate clustering effectiveness index based on clustering results , which is used to evaluate the quality of clustering results. The specific calculation formula is:

[0049] ;

[0050] In the formula, is the sum of squares of inter-class deviations when the number of clusters is w, reflecting the separation between classes. is the sum of squares of intra-class deviations when the number of clusters is w, reflecting the compactness within the class. N is the total number of clustered samples. As the number of clusters w increases, tends to decline, It tends to rise and reaches a maximum value at a certain w, which is the optimal number of clusters;

[0051] Step 2.5: Determine whether all cluster analyses within the cluster number range are completed. If completed, proceed to the next step; otherwise, return to step 2.2;

[0052] Step 2.6: Compare the results for different w values index, The w value corresponding to the maximum index value is the optimal number of clusters, and the "source-load" scenario and its probability corresponding to the number of clusters are output.

[0053] Preferably, the optimization results are obtained with the goal of minimizing the operating cost of each typical scenario.

[0054] The operation layer optimizes each typical scenario according to the scheme delivered by the planning layer, with the goal of minimizing the total loss of normal operation of the distribution network. The objective function is:

[0055] ;

[0056] In the formula, is the normal operation loss of the distribution network in the kth scenario, is the number of distribution network nodes, The number of VSC connections. is the injected active power of node f in the kth scenario, is the transmission power of the g-th VSC in the k-th scenario, is the weight coefficient of VSC transmission power;

[0057] The constraints include:

[0058] Power flow constraints:

[0059] ;

[0060] In the formula, The active power injected into node f is The reactive power injected into node f, is the voltage increase at node f, is the real part of the node admittance matrix, is the imaginary part of the node admittance matrix, is the phase angle difference between nodes f and d, is the current between nodes f and d, represents the set of nodes connected to node f, is the voltage increase at node d;

[0061] Node voltage constraints and line current carrying capacity constraints:

[0062] ;

[0063] ;

[0064] In the formula, is the lower limit of the voltage amplitude at node f, is the upper limit of the voltage amplitude at node f, is the current amplitude of the ith line, is the upper limit of the current amplitude of the ith line;

[0065] VSC operation constraints:

[0066] ;

[0067] In the formula, is the active power flowing into the g-th VSC, is the reactive power flowing into the g-th VSC, is the installed capacity of the g-th VSC;

[0068] Energy storage device operation constraints:

[0069] ;

[0070] ;

[0071] ;

[0072] In the formula, is a 0-1 variable for the charging state, is a 0-1 variable for the discharge state, is the maximum charge and discharge power, , are the storage capacity of the energy storage device in time period t and time period t+1 respectively, is the maximum value of the power that can be stored. For charging efficiency, is the discharge efficiency, is the charging power of the energy storage device in time period t, is the discharge power of the energy storage device in time period t.

[0073] Preferably, the power flow constraint is converted into a mixed integer second-order cone rule model for solution.

[0074] The nonlinear power flow constraint is converted into a linear power flow constraint by variable substitution. The substitution method is as follows:

[0075] ;

[0076] The following conversion formula is obtained:

[0077] ;

[0078] Introduce the rotating cone constraint:

[0079] ;

[0080] in, The node f is in phase The normalized representation of the square of the voltage under is the phase between node f and node d The cosine component of the voltage product under Indicates that node f is in phase The voltage amplitude under Indicates that node d is in phase The voltage amplitude under The phase between node f and node d is The phase angle difference is Indicates the phase relationship between node f and node d The sinusoidal component of the voltage product under The node f is in phase Active power injection under The node f is in phase Reactive power injection under The phase between node f and node d is The square of the current under Node f is in phase The real part of the self-admittance matrix under The phase between node f and node d is The real part of the admittance matrix under The phase between node f and node d is The imaginary part of the admittance matrix under The node f is in phase The imaginary part of the self-admittance matrix under For node d in phase The normalized representation of the square of the voltage under.

[0081] Preferably, the simulated operating state of the energy storage device in the power system simulation software is evaluated.

[0082] Obtaining a simulated operating state data set of an energy storage device in a power system simulation software, wherein the simulated operating state data set of an energy storage device in the power system simulation software specifically includes the instantaneous power fluctuation frequency of the energy storage device, the charging efficiency of the energy storage device, and the total energy loss of the energy storage device;

[0083] Based on the data set of simulated operation status of energy storage equipment obtained in the power system simulation software, a comprehensive analysis is performed to obtain the simulated operation status evaluation value of the energy storage equipment. The simulated operation status evaluation value of the energy storage equipment is used as the analysis basis for determining the feasibility level of the scheme.

[0084] Preferably, the energy storage device simulates the operating status evaluation value,

[0085] ;

[0086] In the formula, is the evaluation value of the simulated operating status of the energy storage device, bd is the instantaneous power fluctuation frequency of the energy storage device, xl is the charging efficiency of the energy storage device, sh is the total energy loss of the energy storage device, and e is a natural constant.

[0087] Preferably, the feasibility level of the scheme is determined by combining the minimum comprehensive planning cost.

[0088] The simulated operation status evaluation value of the energy storage device and the minimum comprehensive planning cost are stored as a designated tag, and the designated tag is compared with the feasibility characteristic value of the scheme corresponding to each designated tag stored in the database to obtain the feasibility characteristic value of the scheme corresponding to the designated tag;

[0089] Compare the feasibility characteristic value of the scheme with the feasibility level threshold of the scheme stored in the database;

[0090] If the feasibility characteristic value of the scheme is not lower than the feasibility level threshold of the scheme, the feasibility level of the scheme corresponding to the feasibility characteristic value of the scheme is level 2;

[0091] If the feasibility characteristic value of the scheme is lower than the feasibility level threshold of the scheme, the feasibility level of the scheme corresponding to the feasibility characteristic value of the scheme is level one.

[0092] The energy storage equipment planning system for medium and low voltage AC / DC hybrid distribution network is used to implement the above method, including a scheme decision module, a result optimization module, a distribution network operation cost determination module and a feasibility level determination module, wherein:

[0093] The solution decision module is used to determine the solution of converter capacity, energy storage equipment capacity and line model in the planning layer with an annual cycle and the minimum comprehensive planning cost as the goal, and transmit the solution to the operation layer;

[0094] The result optimization module is used to cluster the historical data of DG output and load in the operation layer to obtain multiple typical "source-load" scenarios. The goal is to minimize the operating cost of each typical scenario, and the results are optimized. The results are the charging and discharging power of the energy storage device and the conversion power of the converter at different times, and the results are passed to the planning layer.

[0095] The distribution network operation cost determination module is used to multiply the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then add them together to obtain the distribution network operation cost;

[0096] The feasibility level determination module is used to apply the scheme to the power system simulation software, evaluate the simulated operating status of the energy storage equipment in the power system simulation software, and determine the feasibility level of the scheme in combination with the minimum comprehensive planning cost.

[0097] The present invention has the following advantages or beneficial effects:

[0098] The present invention provides a method and system for planning energy storage equipment for a medium- and low-voltage AC / DC hybrid distribution network. By planning the converter capacity, energy storage equipment capacity and line model with the goal of minimizing the comprehensive planning cost at the planning layer, the construction and operation costs of the entire system can be effectively reduced. By optimizing the charging and discharging power of the energy storage system and the conversion power of the converter based on typical scenarios at the operation layer, the operation efficiency of the energy storage system can be further improved, and daily operation losses and power dispatching costs can be reduced. The two-level optimization ensures that the system is not only economical in the planning stage, but also can maintain low cost and high efficiency during actual operation. The typical scenario optimization based on "source-load" clustering at the operation layer can effectively simulate and respond to different power grid operating conditions, ensuring that the energy storage system can maintain stable and efficient operation under a variety of complex conditions.

[0099] The present invention uses an annual cycle and the minimum comprehensive planning cost as the goal in the planning layer to decide on the solutions for the converter capacity, energy storage device capacity and line model, and transmits the solutions to the operation layer. It conducts comprehensive planning on an annual cycle, optimizes the configuration of infrastructure such as converters, energy storage devices and line models, and can effectively reduce the construction cost, operation and maintenance cost and energy consumption cost of the system through full-year cost analysis. By performing global optimization at the planning layer, it can accurately decide on the reasonable capacity of the energy storage device and converter, avoid waste of resources due to over-design or operational instability due to insufficient capacity, and reasonably plan the capacity of the converter to ensure that the power conversion equipment can operate efficiently and avoid overload or excessive redundancy.

[0100] The present invention applies the scheme to power system simulation software, evaluates the simulated operating status of energy storage equipment in the power system simulation software, and determines the feasibility level of the scheme in combination with the minimum comprehensive planning cost. The power system simulation software can truly simulate the behavior and performance of energy storage equipment in actual operation, and compare and verify the theoretical planning scheme with the actual operating status, which can help discover potential problems in the planning. The operating data of the energy storage equipment in the simulation can be fed back to the planning layer to help make further optimization decisions in the planning stage. A large amount of operating data can be generated through simulation, including power output, charge and discharge cycle, and equipment efficiency. These data can provide a scientific basis for planning and operational decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 It is a schematic diagram of the method steps of the present invention;

[0102] Figure 2 It is a schematic diagram of system module connection of the present invention;

[0103] Figure 3 It is a curve diagram showing the changes in clustering effectiveness index and the number of clustering scenarios;

[0104] Figure 4 This is the load curve of electric vehicle charging pile after considering electric vehicle demand response. DETAILED DESCRIPTION

[0105] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0106] Example 1: Figure 1 As shown, the energy storage equipment planning method for medium and low voltage AC / DC hybrid distribution network includes: in the planning layer, with an annual cycle and the minimum comprehensive planning cost as the goal, deciding the converter capacity, energy storage equipment capacity and line model plan, and transmitting the plan to the operation layer.

[0107] The minimum comprehensive planning cost is expressed as follows:

[0108] ;

[0109] In the formula, is the minimum comprehensive planning cost, is the annual cost of investment, operation and maintenance, and decommissioning of the voltage source converter VSC, The annual cost of investment, operation and maintenance, and decommissioning of energy storage devices. The annual cost of the construction of distribution network lines, The operating cost of the distribution network;

[0110] ;

[0111] ;

[0112] In the formula, is the annual equivalent cost of the converter investment, is the annual equivalent cost of converter operation and maintenance costs, is the equivalent annual cost of the converter scrapping cost, is the equivalent annual cost of energy storage investment, is the annual cost of energy storage operation and maintenance costs, The equivalent annual value of the energy storage scrapping cost;

[0113] ;

[0114] Where b is the discount rate, For the entire life cycle of VSC, is the number of VSCs installed, is the investment cost per unit capacity of VSC, is the installed capacity of the vth VSC;

[0115] ;

[0116] In the formula, is the annual operation and maintenance cost coefficient of VSC;

[0117] ;

[0118] In the formula, The cost coefficient for post-processing of the converter equipment after it is scrapped;

[0119] ;

[0120] Where TQ is the full life cycle of the energy storage device, is the number of installed energy storage devices, is the investment cost per unit capacity of the energy storage device, is the capacity of the zth energy storage device;

[0121] ;

[0122] In the formula, is the annual operation and maintenance cost coefficient of the energy storage device;

[0123] ;

[0124] In the formula, The cost coefficient for post-processing of scrapped energy storage equipment;

[0125] ;

[0126] In the formula, For the entire life cycle of the line, is the number of lines, For the The construction cost per unit length of the line, is the planned length of the i-th line;

[0127] ;

[0128] In the formula, is the number of distribution network operation scenarios obtained by clustering, is the probability of the kth scene, For electricity price, is the normal operation loss of the distribution network in the kth scenario;

[0129] Constraints:

[0130] ;

[0131] ;

[0132] In the formula, is the installed capacity of VSC, is the installed capacity of the energy storage device, and are all positive integers, For the unit installation capacity of VSC, The unit installed capacity of the energy storage device.

[0133] By incorporating the annual investment costs of VSC and energy storage systems into comprehensive planning, the economic benefits of the project throughout its life cycle can be more comprehensively evaluated. Considering the investment cost helps ensure that the selected equipment provides the best cost-effectiveness in the years of use to come. The planning layer aims to minimize the annual comprehensive cost and can find the optimal balance between the technical performance and economy of VSC and energy storage equipment. In the comprehensive cost planning, combining the investment costs of VSC and energy storage equipment with their contribution to the system's operating efficiency helps select solutions that can reduce losses in the power transmission and conversion process in the distribution network. Energy storage equipment can reduce the peak-valley electricity price difference, and VSC can improve the voltage stability and power factor of the power grid, thereby improving the operating efficiency of the entire system.

[0134] Decide on the solution of converter capacity, energy storage device capacity and line model. Step 1.1: Set the maximum number of iterations of the particle swarm algorithm and the size of the population, initialize the particle code (the upper triangular matrix of NTS order can be used to represent the connection status or other related parameters between nodes in the network. Each element of the upper triangular matrix represents the specific attributes between a pair of nodes, such as whether they are connected, the cost of connection, capacity, etc., that is, the upper triangular matrix of NTS order corresponding to each particle represents the specific content of a solution, including the relevant information of converter capacity, energy storage device capacity and line model) and the position and speed of each particle. The code of an exemplary particle is as follows:

[0135] ;

[0136] Where X is an upper triangular matrix of NTS order, representing a specific solution;

[0137] Step 1.2: Calculate the fitness of each particle (which can be expressed using the minimum comprehensive planning cost above. The lower the cost, the better the solution represented by the particle and the higher the fitness), and update the individual optimal value (pbest) of each particle and the global optimal value (gbest) of the entire particle group; Step 1.3: Introduce adaptive weights and shrinkage factors to update the position and velocity of particles:

[0138] Adaptive Weights:

[0139] ;

[0140] In the formula, is the adaptive weight, the value range is [0.5, 1], rand is a random number between 0 and 1, is the maximum value of the adaptive weight, is the minimum value of the adaptive weight;

[0141] Shrinkage Factor:

[0142] ;

[0143] In the formula, is the shrinkage factor, for , The sum of the two learning factors is greater than 4;

[0144] During each iteration, the particle updates its velocity and position according to the following formulas;

[0145] ;

[0146] ;

[0147] In the formula, is the velocity of the oth particle at the m+1th iteration, is the velocity of the oth particle at the mth iteration, m is the number of iterations, is the first learning factor, is the second learning factor, , is a random number between [0, 1]. is the position of the oth particle at the mth iteration, is the position of the oth particle at the m+1th iteration, is the optimal position of the oth particle itself at the mth iteration, is the global optimal position of the particle group at the mth iteration; Step 1.4: Determine whether the maximum number of iterations has been reached. If so, output the solution corresponding to the optimal particle. Otherwise, return to step 1.2 to continue iterating and optimizing.

[0148] The introduction of adaptive weights and shrinkage factors can balance the global search and local search capabilities of the algorithm. The adaptive weights tend to be larger values ​​in the early stage of optimization, thereby enhancing the ability of global search and avoiding falling into the local optimal solution too early. In the later stage of the algorithm, the weights gradually decrease, enhancing the accuracy of local search, and helping to quickly converge to the global optimal solution. It can ensure that a better solution is found in complex power system optimization problems. By using the upper triangular matrix of the NTS order to represent the load access plan, the access and growth ratio of multiple loads can be considered at the same time, which is suitable for handling complex load change scenarios of large-scale distribution networks. It helps to find the optimal converter capacity, energy storage device capacity and line configuration plan under different load conditions, and optimize the overall operating performance of the power system. The distribution network planning problem involves a large number of decision variables, such as converter capacity, energy storage device capacity, line model, etc. Traditional optimization methods are difficult to effectively handle these high-dimensional problems. PSO can process multiple variables at the same time at a lower computing cost through the parallel search mechanism of particle swarms, improve the solution efficiency, and adapt to multi-objective optimization needs.

[0149] The operation layer clusters the historical data of DG output and load to obtain multiple typical "source-load" scenarios, and optimizes the results to minimize the operating cost of each typical scenario. The results are the charging and discharging power of the energy storage equipment and the conversion power of the converter at different times, and the results are passed to the planning layer.

[0150] The operation layer clusters the historical data of DG output and load to obtain multiple typical "source-load" scenarios. Step 2.1: Select the characteristic quantities of the DG daily output curve and daily load curve. The selected characteristic quantities are:

[0151] ;

[0152] In the formula, is the average daily output of DG, is the DG output during period t, is the variance of DG daily force, is the average daily load, is the load power in period t, is the variance of daily load; T is the total number of time periods;

[0153] Step 2.2: Set the range of the number of clusters to w belongs to , select w initial cluster centers, traverse the 365 groups of "source-load" sample data throughout the year, calculate the Euclidean distance between the rth sample and each cluster center, and assign the rth sample to the cluster center that minimizes the Euclidean distance. The calculation formula of the Euclidean distance between the rth sample and each cluster center is as follows:

[0154] ;

[0155] In the formula, is the Euclidean distance between the rth sample and the wth cluster center, is the hth eigenvector of the rth sample, is the hth eigenvector of the wth cluster center;

[0156] Step 2.3: Count the number of samples in each class, update w cluster centers according to the mean of the sample data, and repeat steps 2.2 and 2.3 until the cluster center no longer changes;

[0157] Step 2.4: Calculate clustering effectiveness index based on clustering results , which is used to evaluate the quality of clustering results. The specific calculation formula is:

[0158] ;

[0159] In the formula, is the sum of squares of inter-class deviations when the number of clusters is w, reflecting the separation between classes. is the sum of squares of intra-class deviations when the number of clusters is w, reflecting the compactness within the class. N is the total number of clustered samples. As the number of clusters w increases, tends to decline, It tends to rise and reaches a maximum value at a certain w, which is the optimal number of clusters;

[0160] Step 2.5: Determine whether all cluster analyses within the cluster number range are completed. If completed, proceed to the next step; otherwise, return to step 2.2;

[0161] Step 2.6: Compare the results for different w values index, The w value corresponding to the maximum index value is the optimal number of clusters, and the "source-load" scenario and its probability corresponding to the number of clusters are output.

[0162] The output and load of distributed generation (DG) in the distribution network vary in complexity. By performing k-means clustering on the characteristic quantities (mean and variance) of its daily output and load curves, 365 sets of data throughout the year can be summarized into several typical "source-load" scenarios, which can effectively reduce the complexity of the problem and compress complex historical data into a small number of representative typical scenarios, laying the foundation for subsequent power grid planning and optimization. By extracting multiple typical scenarios and calculating their occurrence probabilities, a reliable data basis can be provided for the planning and dispatching decisions of the power system. The pattern extraction method based on historical data can ensure that each typical scenario has a certain degree of representativeness, avoid making unreasonable decisions based on single or extreme situations, and improve the accuracy and feasibility of power grid optimization and dispatching decisions. The "source-load" typical scenarios obtained by cluster analysis can be directly used in the planning and optimization model of the power grid. Each typical scenario reflects a specific power generation output and load pattern. In the subsequent dispatch optimization, the operation of the power grid can be finely adjusted according to these typical scenarios. Optimization based on typical scenarios can significantly improve the economy and reliability of the power grid.

[0163] The operation cost of each typical scenario is minimized and the optimization result is obtained. The operation layer optimizes each typical scenario according to the scheme delivered by the planning layer and minimizes the total loss of normal operation of the distribution network. The objective function is:

[0164] ;

[0165] In the formula, is the normal operation loss of the distribution network in the kth scenario, is the number of distribution network nodes, The number of VSC connections. is the injected active power of node f in the kth scenario, is the transmission power of the g-th VSC in the k-th scenario, is the weight coefficient of VSC transmission power; the constraints include:

[0166] Power flow constraints:

[0167] ;

[0168] In the formula, The active power injected into node f is The reactive power injected into node f, is the voltage increase at node f, is the real part of the node admittance matrix, is the imaginary part of the node admittance matrix, is the phase angle difference between nodes f and d, is the current between nodes f and d, represents the set of nodes connected to node f, is the voltage increase at node d;

[0169] Node voltage constraints and line current carrying capacity constraints:

[0170] ;

[0171] ;

[0172] In the formula, is the lower limit of the voltage amplitude at node f, is the upper limit of the voltage amplitude at node f, is the current amplitude of the ith line, is the upper limit of the current amplitude of the ith line;

[0173] VSC operation constraints:

[0174] ;

[0175] In the formula, is the active power flowing into the g-th VSC, is the reactive power flowing into the g-th VSC, is the installed capacity of the g-th VSC;

[0176] Energy storage device operation constraints:

[0177] ;

[0178] ;

[0179] ;

[0180] In the formula, is a 0-1 variable for the charging status (1 means charging), is a 0-1 variable for the discharge state (1 means discharging), is the maximum charge and discharge power, , are the storage capacity of the energy storage device in time period t and time period t+1 respectively, is the maximum value of the power that can be stored. For charging efficiency, is the discharge efficiency, is the charging power of the energy storage device in time period t, is the discharge power of the energy storage device in time period t.

[0181] The core of the objective function is to minimize the total loss of the distribution network, including the node injection active power loss and the VSC (voltage source converter, also known as voltage source converter) power transmission loss, optimize the operation in each scenario, and ensure that the distribution network can maintain the minimum power loss under different load fluctuations, thereby improving the overall energy efficiency of the system. The flow constraint ensures that the flow of power in the distribution network conforms to the laws of physics. Through the calculation of the node admittance matrix and the equilibrium equation, the rationality of the power distribution of the power grid in different scenarios is guaranteed, and the power imbalance or system instability caused by unreasonable power dispatching is avoided. The voltage constraint and line current carrying capacity constraint ensure that the distribution network will not exceed the allowable safety range during operation, avoid overvoltage, undervoltage or line overload, and ensure that the distribution network is in a stable and safe operating state in any scenario, reduce the probability of accidents, and improve the effective utilization rate of VSC and energy storage systems. By optimizing the active and reactive power of each node and VSC, the power flow in the distribution network is ensured to be in a balanced state. Reasonable active and reactive power distribution can improve the operation efficiency of the power grid, reduce power loss, maintain voltage stability, and further improve the reliability of the power system.

[0182] Furthermore, the power flow constraint is converted into a mixed integer second-order cone rule model for solution. The nonlinear power flow constraint is converted into a linear power flow constraint by variable substitution. The substitution method is as follows:

[0183] ;

[0184] The following conversion formula is obtained:

[0185] ;

[0186] Introduce the rotating cone constraint:

[0187] ;

[0188] in, The node f is in phase The normalized representation of the square of the voltage under is the phase between node f and node d The cosine component of the voltage product under Indicates that node f is in phase The voltage amplitude under Indicates that node d is in phase The voltage amplitude under The phase between node f and node d is The phase angle difference is Indicates the phase relationship between node f and node d The sinusoidal component of the voltage product under The node f is in phase Active power injection under The node f is in phase Reactive power injection under The phase between node f and node d is The square of the current under Node f is in phase The real part of the self-admittance matrix under The phase between node f and node d is The real part of the admittance matrix under The phase between node f and node d is The imaginary part of the admittance matrix under The node f is in phase The imaginary part of the self-admittance matrix under For node d in phase The normalized representation of the square of the voltage under.

[0189] The traditional power flow constraint equation is in nonlinear form. Solving it directly will increase the complexity of calculation, especially in large-scale power systems. The solution efficiency is low and it is difficult to ensure the global optimal solution. By replacing variables (such as replacing the voltage variable with its square form), the original nonlinear constraint can be converted into a linear constraint, which greatly simplifies the complexity of the problem. After linearization, the originally complex nonlinear equation can be solved more efficiently, especially when the problem scale is large (such as multi-node distribution network). This transformation significantly improves the feasibility and efficiency of the solution. Through variable replacement and the introduction of cone constraints, although the power flow equation is mathematically transformed, its essential physical meaning is not changed. Variables such as voltage, current, active power and reactive power still maintain their original physical interpretation. Therefore, the model conversion will not affect the actual operation interpretation of the system or lead to physically unreasonable results. By optimizing the power flow and power distribution of the power grid, the risk of power grid overload or voltage fluctuation can be effectively reduced, the safety and reliability of the system can be improved, and the stability of power supply can be guaranteed.

[0190] The planning layer multiplies the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then adds them together to obtain the distribution network operation cost.

[0191] There may be large differences in operating losses under different typical scenarios. Directly relying on the cost of a certain scenario for global estimation may introduce deviations. By weighted aggregation of loss results from multiple scenarios, deviations caused by certain extreme situations or special operating conditions can be avoided, thereby obtaining a more objective distribution network operating cost assessment. Different typical scenarios represent the operation of the power grid under different load levels and power generation output conditions throughout the year. Combining the losses of these typical scenarios, we can more comprehensively assess the impact of load fluctuations throughout the year on the operation of the distribution network, and ensure that the distribution network operating cost assessment results cover changes in different operating conditions. Through accurate distribution network operating cost assessment, the planning layer can optimize the configuration of equipment capacity, such as the size of transformers, VSCs (voltage source converters), and energy storage devices, based on the loss results in each typical scenario, to help the power grid avoid over-design or insufficient capacity of equipment, thereby maximizing the utilization of system resources and enhancing the economy and stability of the power grid.

[0192] The historical data of photovoltaic power output throughout the year and the historical data of photovoltaic power and load throughout the year are clustered, and the curve of the change of clustering effectiveness index and the number of clustering scenarios is obtained as shown in the figure. Figure 3 As shown, when the number of clustering scenarios is equal to 10, the effectiveness index is the largest, so 10 is selected as the number of scenario clusters. The probability of each scenario is shown in Table 1. The VSC related parameters, line model parameters, distributed power supply location and capacity are shown in Tables 2-4:

[0193] Table 1 Scenario Probability

[0194]

[0195] Table 2 VSC related parameters

[0196]

[0197] Table 3 Parameters of the line models to be selected

[0198]

[0199] Table 4 Distributed power source location and capacity

[0200]

[0201] In order to verify the effectiveness of the joint optimization method proposed in this chapter, the following schemes are set for comparison:

[0202] Scenario 1: Plan the load of the existing substations. The initial AC load of substation 1 is 190kw, the load of the electric vehicle charging pile is 50kw, the AC load of substation 2 is 245kw, and the load of the 5G base station is 50kw.

[0203] Solution 1: Based on the traditional radial AC distribution system, the VSC capacity and energy storage capacity are planned by adopting the traditional DC source and load separate access method.

[0204] Solution 2: Based on the hand-in-hand AC / DC distribution system, the VSC capacity and energy storage capacity are planned by adopting the form of AC / DC line hybrid.

[0205] By dynamically adjusting the charging strategy of electric vehicle users, any net load value within the adjustable capacity range of net load demand response can be obtained. It is assumed that 30% of electric vehicle users participate in demand response. Considering the impact of the net load size on the planning results of the model proposed in this chapter, by changing the electric vehicle charging strategy, extreme scenarios considering electric vehicle demand response are obtained, namely, the scenario with the largest net load, i.e., the worst DC source-load complementary characteristics, and the scenario with the smallest net load, i.e., the best source-load complementary characteristics. Figure 4 This is the load curve of electric vehicle charging pile after considering electric vehicle demand response.

[0206] Scenario 1: The load participating in demand response is 30% of the original load, and the load from 12:00 to 16:00 is shifted to 17:00 to 21:00.

[0207] Scenario 2: The load participating in demand response is 0% of the original load, that is, the load of electric vehicle charging piles considering disorderly charging.

[0208] Scenario 3: The load participating in demand response is 30% of the original load, and the load from 12:00 to 16:00 is shifted to 23:00 to 2:00.

[0209] Corresponding to the three demand response scenarios, three sub-schemes are generated in Scheme 2.

[0210] Solution 2-1: The electric vehicle charging pile load has the load characteristics under the demand response of scenario 1.

[0211] Scheme 2-2: The electric vehicle charging pile load is the load characteristic under scenario 2 demand response.

[0212] Scheme 2-3: The electric vehicle charging pile load is the load characteristic under scenario 3 demand response.

[0213] In scheme 1, three sub-schemes are generated accordingly according to different access demand response strategies.

[0214] The first planning results are shown in Tables 5 to 10.

[0215] Table 5 Results of the first planning

[0216]

[0217] Table 6 Results of the first planning

[0218]

[0219] Table 7 Results of the first planning

[0220]

[0221] Table 8 Results of the first planning

[0222]

[0223] Table 9 Results of the first planning

[0224]

[0225] Table 10 Results of the first planning

[0226]

[0227] Comparing Scheme 1 and Scheme 2, the VSC capacity required for Scheme 2 is greatly reduced. This is because when the DC source load is large, compared with Scheme 1, Scheme 2 forms a DC line through two VSCs with larger capacity, and the DC source load can be directly connected to the distribution network without passing through the VSC. At the same time, the hand-in-hand wiring mode connects the two radial substations together, and the DC loads of the two substations have certain complementary characteristics. The peak load of the load characteristics of the DC load is reduced, which reduces the VSC capacity required when the DC load reaches the peak, and the construction and operation costs of the converter are reduced. Moreover, in the hand-in-hand wiring mode, the DC load can be powered by the AC power on both sides at the same time, which can balance the loads of the two substations and optimize the flow. The operating loss of the system is also reduced. When the network carries the same load, the annual comprehensive cost of the AC / DC hybrid distribution network in the hand-in-hand wiring mode after adding energy storage is greatly reduced, the flow of the entire system is also optimized, the loads of the two substations are more balanced, and the operating cost of the system is reduced. Therefore, the economy of Scheme 2 is better.

[0228] Scenario 2: Considering the load growth, the capacity of the existing substation is increased and modified. When the load of 5G base stations and electric vehicle charging piles both increase by 50%, two capacity increase and modification plans are considered.

[0229] Solution 1: Based on the traditional grid capacity expansion and transformation, the traditional radial distribution system is transformed by using the traditional new line expansion plan. The new line takes on part of the load. When the DC load increases by 50%, the VSC capacity, energy storage capacity and lines are planned.

[0230] Option 2: By adding energy storage and changing the grid structure of the distribution network, the VSC capacity, energy storage capacity and lines are planned when the load increases by 50Kw.

[0231] The second planning results are shown in Tables 11 to 16.

[0232] Table 11 Second planning results 1

[0233]

[0234] Table 12 Second planning results 2

[0235]

[0236] Table 13 Second planning results 3

[0237]

[0238] Table 14 Second planning results 4

[0239]

[0240] Table 15 Second planning results

[0241]

[0242] Table 16 Second planning results VI

[0243]

[0244] According to the planning results, the newly installed capacity of the converters in Scheme 1 is 50kVA, and the models of the newly built lines are all YJV22-4*70; the newly installed capacities of the converters in Schemes 2-1 / 2-2 / 2-3 are 43kVA, 33kVA, and 36kVA respectively, and the model of the newly built lines in Scheme 2-1 is YJV22-4*70, and the installed capacities of the energy storage are 0kwh, 10 / 8kwh / kw, and 5 / 2kwh / kw respectively.

[0245] When the load in the network increases by 50%, when the traditional capacity expansion solution based on the capacity-to-load ratio is adopted, no AC line can meet the load-bearing requirements, so one AC line needs to be expanded; the DC load is supplied by the upper power grid through VSC and distributed photovoltaics. When the DC load increases, a larger converter capacity is required to meet the peak power supply requirements of the DC load under the condition that the photovoltaic output remains unchanged. When the load in the network increases by 50%, the solution of adding energy storage and transforming the grid structure is adopted for capacity expansion. By adding energy storage, the peak load is reduced, the peak load in the system is reduced, and the cost increased by line expansion is avoided; the two substations are interconnected through VSC, and the two substations are connected into a power-assisted whole. Because the load characteristics of the two substations are complementary to a certain extent, after the two substations are together, the load characteristics on the DC line become smooth, and the load peak is greatly reduced, which reduces the capacity of the converter required when the load reaches the peak and improves the utilization rate of the converter.

[0246] From scenario 1, it can be seen that Scheme 2-1 is a scenario with poor DC source-load complementary characteristics, Scheme 2-2 is a scenario with good DC source-load complementary characteristics, and Scheme 2-3 is a scenario with very good DC source-load complementary characteristics. When the DC load increases by 50%, in Scheme 1, the net load of each AC substation is greater than the maximum capacity of the line, and it is necessary to expand the substation by building a new line in each of the two substations. In Scheme 2-1, the net load peak of the substation far exceeds the maximum capacity of the line. If no new line expansion is carried out, two large-capacity energy storages need to be configured to reduce the net load peak. The expansion method of building a new line L1 is more economical; in Schemes 2-2 and 2-3, the net load peak of the substation is slightly greater than the maximum capacity of the line. Energy storage can be added to reduce the peak and delay the expansion investment of the line. Scheme 2-3 has better DC source-load complementary characteristics. Therefore, only a small amount of energy storage needs to be configured without building a new line to achieve normal operation of the system.

[0247] It can be seen from Table 6 that the costs of Scheme 2-1 / 2-2 / 2-3 are RMB 0.557 / 0.538 / 0.498 million respectively. Compared with the traditional method of building new lines to increase the capacity of the power grid, Scheme 2 reduces the investment cost of power grid expansion, delays the expansion of the line, and reduces the annual comprehensive cost by 20.63%, 22.35% and 28.75% respectively.

[0248] The scheme is applied to the power system simulation software, the simulated operating status of the energy storage equipment in the power system simulation software is evaluated, and the feasibility level of the scheme is determined in combination with the minimum comprehensive planning cost.

[0249] The simulated operation status data set of the energy storage device in the power system simulation software is obtained. The simulated operation status data set of the energy storage device in the power system simulation software specifically includes the instantaneous power fluctuation frequency of the energy storage device, the charging efficiency of the energy storage device, and the total energy loss of the energy storage device. Based on the obtained simulated operation status data set of the energy storage device in the power system simulation software, a comprehensive analysis is performed to obtain an evaluation value of the simulated operation status of the energy storage device. The evaluation value of the simulated operation status of the energy storage device is used as an analysis basis for determining the feasibility level of the scheme.

[0250] Energy storage equipment simulated operation status assessment value,

[0251] ;

[0252] In the formula, is the evaluation value of the simulated operating status of the energy storage device, bd is the instantaneous power fluctuation frequency of the energy storage device, xl is the charging efficiency of the energy storage device, sh is the total energy loss of the energy storage device, and e is a natural constant.

[0253] Through key data such as the instantaneous power fluctuation frequency, charging efficiency and total energy loss of energy storage equipment, the performance of energy storage equipment under different operating conditions can be quantified, reflecting the load conditions, energy conversion efficiency and energy loss of the system in actual work, helping to understand the performance of the equipment in a real environment. The performance of energy storage equipment is not only reflected in a single parameter, but is the comprehensive result of multiple factors. The multi-dimensional data obtained in the simulation (such as power fluctuation frequency, charging efficiency, total loss, etc.) can comprehensively evaluate the performance of the equipment. The comprehensive evaluation value integrates data from various dimensions to make the evaluation result more scientific. The simulated operating status evaluation value of the energy storage equipment simplifies multiple complex performance data into a comprehensive evaluation indicator. The intuitive indicator provides an easy-to-understand basis for the decision-maker, making the comparison and selection between different schemes more efficient and helping to assess the feasibility level of the scheme.

[0254] In combination with the minimum comprehensive planning cost, the feasibility level of the scheme is determined, the simulated operation status evaluation value of the energy storage equipment and the minimum comprehensive planning cost are stored as a designated label, and the designated label is compared with the scheme feasibility characteristic value corresponding to each designated label stored in the database to obtain the scheme feasibility characteristic value corresponding to the designated label; the scheme feasibility characteristic value is compared with the scheme feasibility level threshold stored in the database; if the scheme feasibility characteristic value is not lower than the scheme feasibility level threshold, the scheme feasibility level corresponding to the scheme feasibility characteristic value is level two; if the scheme feasibility characteristic value is lower than the scheme feasibility level threshold, the scheme feasibility level corresponding to the scheme feasibility characteristic value is level one.

[0255] This method can provide a reliable data basis for the feasibility analysis of the scheme by storing the operating status evaluation value and comprehensive planning cost data of the energy storage equipment in the form of tags and comparing them with historical data. By comparing with the historical records and thresholds in the database, the feasibility level of the scheme is determined not only based on a single data point, but also based on the powerful conclusions drawn through historical analysis, which improves the accuracy of decision-making. The simulated state evaluation value of the energy storage equipment is automatically compared with the historical characteristic value and the feasibility level threshold, which can quickly and automatically complete the complex evaluation process, greatly reducing human intervention and subjective judgment, improving the efficiency of the evaluation, and being able to meet the evaluation needs of large-scale and complex projects. By combining the operating status evaluation value of the energy storage equipment with the comprehensive planning cost, technical and economic factors can be considered at the same time. Not only the operation of the energy storage equipment is evaluated through technical performance (such as power fluctuation, charging efficiency, energy loss), but also the actual feasibility of the scheme is analyzed through economics (such as comprehensive planning cost). Multi-dimensional comparative analysis ensures that the scheme achieves the optimal balance between technology and economy.

[0256] Example 2: Figure 2 As shown, a storage equipment planning system for a medium and low voltage AC / DC hybrid distribution network is used to implement the method of Example 1, including a scheme decision module, a result optimization module, a distribution network operation cost determination module and a feasibility level determination module.

[0257] The solution decision module is used to decide the solutions of converter capacity, energy storage equipment capacity and line model in the planning layer with an annual cycle and the minimum comprehensive planning cost as the goal, and transmit the solutions to the operation layer.

[0258] The result optimization module is used to cluster the historical data of DG output and load at the operation layer to obtain multiple typical "source-load" scenarios. The operation cost of each typical scenario is minimized, and the results are optimized. The results are the charging and discharging power of the energy storage equipment and the conversion power of the converter at different times, and the results are passed to the planning layer.

[0259] The distribution network operation cost determination module is used at the planning level to multiply the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then add them together to obtain the distribution network operation cost.

[0260] The feasibility level determination module is used to apply the scheme to the power system simulation software, evaluate the simulated operating status of the energy storage equipment in the power system simulation software, and determine the feasibility level of the scheme in combination with the minimum comprehensive planning cost.

Claims

1. A method for planning energy storage equipment for a medium and low voltage AC / DC hybrid distribution network, characterized in that: The following steps are involved: In the planning layer, the converter capacity, energy storage device capacity and line model are decided on with an annual cycle and the minimum comprehensive planning cost as the goal, and the plan is transmitted to the operation layer; The operation layer clusters the historical data of DG output and load to obtain multiple typical "source-load" scenarios. The operation layer optimizes the results to minimize the operation cost of each typical scenario. The results are the charging and discharging power of the energy storage device and the conversion power of the converter at different times. The results are passed to the planning layer. The planning layer multiplies the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then adds them together to obtain the distribution network operation cost; Apply the scheme to the power system simulation software, evaluate the simulated operating status of the energy storage equipment in the power system simulation software, and determine the feasibility level of the scheme in combination with the minimum comprehensive planning cost; The operation layer clusters the historical data of DG output and load to obtain multiple typical "source-load" scenarios. The specific analysis process is as follows: Step 2.1: Select the characteristic quantities of DG daily output curve and daily load curve. The selected characteristic quantities are: ； In the formula, is the average daily output of DG, is the DG output during period t, is the variance of DG daily force, is the average daily load, is the load power in period t, is the variance of daily load; T is the total number of time periods; Step 2.2: Set the range of the number of clusters to w belongs to , select w initial cluster centers, traverse the 365 groups of "source-load" sample data throughout the year, calculate the Euclidean distance between the rth sample and each cluster center, and assign the rth sample to the cluster center that minimizes the Euclidean distance. The calculation formula for the Euclidean distance between the rth sample and each cluster center is as follows: ； In the formula, is the Euclidean distance between the rth sample and the wth cluster center, is the hth eigenvector of the rth sample, is the hth eigenvector of the wth cluster center; Step 2.3: Count the number of samples in each class, update w cluster centers according to the mean of the sample data, and repeat steps 2.2 and 2.3 until the cluster center no longer changes; Step 2.4: Calculate clustering effectiveness index based on clustering results , which is used to evaluate the quality of clustering results. The specific calculation formula is: ； In the formula, is the sum of squares of inter-class deviations when the number of clusters is w, reflecting the separation between classes. is the sum of squares of intra-class deviations when the number of clusters is w, reflecting the compactness within the class. N is the total number of clustered samples. As the number of clusters w increases, tends to decline, It tends to rise and reaches a maximum value at a certain w, which is the optimal number of clusters; Step 2.5: Determine whether all cluster analyses within the cluster number range are completed. If completed, proceed to the next step; otherwise, return to step 2.2; Step 2.6: Compare the results for different w values index, The w value corresponding to the maximum index value is the optimal number of clusters, and the "source-load" scenario and its probability corresponding to the number of clusters are output; The goal is to minimize the operating cost of each typical scenario and optimize the results. The specific analysis process is as follows: The operation layer optimizes each typical scenario according to the scheme delivered by the planning layer, with the goal of minimizing the total loss of normal operation of the distribution network. The objective function is: ； In the formula, is the normal operation loss of the distribution network in the kth scenario, is the number of distribution network nodes, The number of VSC connections. is the injected active power of node f in the kth scenario, is the transmission power of the g-th VSC in the k-th scenario, is the weight coefficient of VSC transmission power; The constraints include: Power flow constraints: ； In the formula, The active power injected into node f is The reactive power injected into node f, is the voltage increase at node f, is the real part of the node admittance matrix, is the imaginary part of the node admittance matrix, is the phase angle difference between nodes f and d, is the current between nodes f and d, represents the set of nodes connected to node f, is the voltage increase at node d; Node voltage constraints and line current carrying capacity constraints: ； ； In the formula, is the lower limit of the voltage amplitude at node f, is the upper limit of the voltage amplitude at node f, is the current amplitude of the ith line, is the upper limit of the current amplitude of the ith line; VSC operation constraints: ； In the formula, is the active power flowing into the g-th VSC, is the reactive power flowing into the g-th VSC, is the installed capacity of the g-th VSC; Energy storage device operation constraints: ； ； ； In the formula, is a 0-1 variable for the charging state, is a 0-1 variable for the discharge state, is the maximum charge and discharge power, , are the storage capacity of the energy storage device in time period t and time period t+1 respectively, is the maximum value of the power that can be stored. For charging efficiency, is the discharge efficiency, is the charging power of the energy storage device in time period t, is the discharge power of the energy storage device in time period t.

2. The energy storage equipment planning method for a medium- and low-voltage AC / DC hybrid distribution network according to claim 1 is characterized in that: The minimum comprehensive planning cost is expressed as follows: ； In the formula, is the minimum comprehensive planning cost, is the annual cost of investment, operation and maintenance, and decommissioning of the voltage source converter VSC, The annual cost of investment, operation and maintenance, and decommissioning of energy storage devices. The annual cost of the construction of distribution network lines, The operating cost of the distribution network; ； ； In the formula, is the annual equivalent cost of the converter investment, is the annual equivalent cost of converter operation and maintenance costs, is the equivalent annual cost of the converter scrapping cost, is the equivalent annual cost of energy storage investment, is the annual cost of energy storage operation and maintenance costs, The equivalent annual value of the energy storage scrapping cost; ； Where b is the discount rate, x is the full life cycle of the line, is the number of lines, is the construction cost per unit length of the ith line, is the planned length of the i-th line; ； In the formula, is the number of distribution network operation scenarios obtained by clustering, is the probability of the kth scene, For electricity price, is the normal operation loss of the distribution network in the kth scenario; Constraints: ； ； In the formula, is the installed capacity of VSC, is the installed capacity of the energy storage device, and are all positive integers, For the unit installation capacity of VSC, The unit installed capacity of the energy storage device.

3. The energy storage equipment planning method for a medium- and low-voltage AC / DC hybrid distribution network according to claim 1 is characterized in that: The specific analysis process of the solution for determining the converter capacity, energy storage device capacity and line model is as follows: Step 1.1: Set the maximum number of iterations and the size of the population of the particle swarm algorithm, initialize the particle code and the position and velocity of each particle; Step 1.2: Calculate the fitness of each particle and update the individual optimal value of each particle and the global optimal value of the entire particle group. The fitness is expressed as the minimum comprehensive planning cost. Step 1.3: Introduce adaptive weights and shrinkage factors to update the position and velocity of particles: Adaptive Weights: ； In the formula, is the adaptive weight, the value range is [0.5, 1], rand is a random number between 0 and 1, is the maximum value of the adaptive weight, is the minimum value of the adaptive weight; Shrinkage Factor: ； In the formula, is the shrinkage factor, for , The sum of the two learning factors is greater than 4; During each iteration, the particle updates its velocity and position according to the following formulas; ； ； In the formula, is the velocity of the oth particle at the m+1th iteration, is the velocity of the oth particle at the mth iteration, m is the number of iterations, is the first learning factor, is the second learning factor, , is a random number between [0, 1]. is the position of the oth particle at the mth iteration, is the position of the oth particle at the m+1th iteration, is the optimal position of the oth particle itself at the mth iteration, is the global optimal position of the particle group at the mth iteration; Step 1.4: Determine whether the maximum number of iterations has been reached. If so, output the solution corresponding to the optimal particle. Otherwise, return to step 1.2 to continue iterating and optimizing.

4. The energy storage equipment planning method for a medium- and low-voltage AC / DC hybrid distribution network according to claim 1 is characterized in that: The power flow constraint is transformed into a mixed integer second-order cone rule model for solution. The specific analysis process is as follows: The nonlinear power flow constraint is converted into a linear power flow constraint by variable substitution. The substitution method is as follows: ； The following conversion formula is obtained: ； Introduce the rotating cone constraint: ； in, The node f is in phase The normalized representation of the square of the voltage under is the phase between node f and node d The cosine component of the voltage product under Indicates that node f is in phase The voltage amplitude under Indicates that node d is in phase The voltage amplitude under The phase between node f and node d is The phase angle difference is Indicates the phase relationship between node f and node d The sinusoidal component of the voltage product under The node f is in phase Active power injection under The node f is in phase Reactive power injection under The phase between node f and node d is The square of the current under Node f is in phase The real part of the self-admittance matrix under The phase between node f and node d is The real part of the admittance matrix under The phase between node f and node d is The imaginary part of the admittance matrix under The node f is in phase The imaginary part of the self-admittance matrix under For node d in phase The normalized representation of the square of the voltage under.

5. The method for planning energy storage equipment for a medium- and low-voltage AC / DC hybrid distribution network according to claim 1 is characterized in that: The simulated operation status of the energy storage equipment in the power system simulation software is evaluated, and the specific analysis process is as follows: Obtaining a simulated operating state data set of an energy storage device in a power system simulation software, wherein the simulated operating state data set of an energy storage device in the power system simulation software specifically includes the instantaneous power fluctuation frequency of the energy storage device, the charging efficiency of the energy storage device, and the total energy loss of the energy storage device; Based on the data set of simulated operation status of energy storage equipment obtained in the power system simulation software, a comprehensive analysis is performed to obtain the simulated operation status evaluation value of the energy storage equipment. The simulated operation status evaluation value of the energy storage equipment is used as the analysis basis for determining the feasibility level of the scheme.

6. The method for planning energy storage equipment for a medium- and low-voltage AC / DC hybrid distribution network according to claim 5 is characterized in that: The simulated operation status evaluation value of the energy storage device, the specific analysis process is as follows: ； In the formula, is the evaluation value of the simulated operating status of the energy storage device, bd is the instantaneous power fluctuation frequency of the energy storage device, xl is the charging efficiency of the energy storage device, sh is the total energy loss of the energy storage device, and e is a natural constant.

7. The energy storage equipment planning method for a medium- and low-voltage AC / DC hybrid distribution network according to claim 1 is characterized in that: The feasibility level of the scheme is determined by combining the minimum comprehensive planning cost. The specific analysis process is as follows: The simulated operation status evaluation value of the energy storage device and the minimum comprehensive planning cost are stored as a designated tag, and the designated tag is compared with the feasibility characteristic value of the scheme corresponding to each designated tag stored in the database to obtain the feasibility characteristic value of the scheme corresponding to the designated tag; Compare the feasibility characteristic value of the scheme with the feasibility level threshold of the scheme stored in the database; If the feasibility characteristic value of the scheme is not lower than the feasibility level threshold of the scheme, the feasibility level of the scheme corresponding to the feasibility characteristic value of the scheme is level 2; If the feasibility characteristic value of the scheme is lower than the feasibility level threshold of the scheme, the feasibility level of the scheme corresponding to the feasibility characteristic value of the scheme is level one.

8. A system for planning energy storage equipment for a medium and low voltage AC / DC hybrid distribution network, used to implement the method described in any one of claims 1 to 7, characterized in that: It includes a scheme decision module, a result optimization module, a distribution network operation cost determination module and a feasibility level determination module, among which: The solution decision module is used to determine the solution of converter capacity, energy storage equipment capacity and line model in the planning layer with an annual cycle and the minimum comprehensive planning cost as the goal, and transmit the solution to the operation layer; The result optimization module is used to cluster the historical data of DG output and load in the operation layer to obtain multiple typical "source-load" scenarios. The goal is to minimize the operating cost of each typical scenario, and the results are the charging and discharging power of the energy storage device and the conversion power of the converter at different times. The results are passed to the planning layer; The distribution network operation cost determination module is used to multiply the normal operation loss of the distribution network in each typical scenario by the corresponding number of days and then add them together to obtain the distribution network operation cost; The feasibility level determination module is used to apply the scheme to the power system simulation software, evaluate the simulated operating status of the energy storage equipment in the power system simulation software, and determine the feasibility level of the scheme in combination with the minimum comprehensive planning cost.

Images