Multi-scene-based unbalanced active distribution network battery energy storage system configuration method

Abstract

The invention belongs to the technical field of active power distribution network energy storage system configuration, and particularly relates to a multi-scene-based unbalanced active power distribution network battery energy storage system configuration method, which comprises the following steps: modeling a three-phase unbalanced active power distribution network power flow, and building a multi-target BESS optimal configuration model based on an unbalanced active power distribution network; based on the established model, two-stage decomposition is carried out, and solving is carried out by adopting a mixed optimization algorithm combining particle swarm optimization and second-order cone programming; the influence of charging and discharging behaviors of the BESS in multiple operation scenes is considered, a BESS two-stage optimal configuration process based on the network loss sensitivity is formulated, and finally the reasonability and superiority of the submitted BESS optimal configuration method are verified through simulation of a typical example system. According to the method, the respective characteristics of particle swarm optimization and second-order cone programming can be utilized, and the solving efficiency and precision of the BESS optimization configuration model can be remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of active distribution network energy storage system configuration, and specifically relates to a method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios. Background Art

[0002] With the introduction of concepts such as "Energy Internet" and "Carbon Peak and Carbon Neutrality", my country is actively promoting the large-scale development and high-quality development of distributed generation (DG). At the same time, the construction of distributed wind power and distributed photovoltaic projects in load centers and surrounding areas is also continuing to develop steadily. However, the large-scale access of intermittent renewable energy such as wind and solar power has made the frequency characteristics of the distribution system more complex, and problems such as node voltage instability have become increasingly serious. At the same time, the high proportion of new energy grid connection may also cause wind and solar abandonment, resulting in insufficient flexible adjustment capacity of the power system. The deployment of battery energy storage systems (BESS) at key distribution network nodes such as intensive load access or concentrated collection of new energy can effectively maintain the balance of energy supply and demand, alleviate grid congestion and its upgrade needs, and enhance the stability and reliability of the power system. Therefore, it is particularly important to configure BESS reasonably.

[0003] BESS has the following advantages in distribution networks with high DG penetration: 1) improve the reliability and power quality of system power supply; 2) reduce line losses and improve voltage distribution; 3) reduce energy costs and control the consumption of high-cost energy; 4) enhance the supply capacity of peak loads and improve the power factor. To date, the configuration problem of battery energy storage system (BESS) has been widely studied. However, current research still has certain limitations. First, since the operating state of the distribution network is significantly affected by the uncertainty of source and load, the analysis based on a single operating condition (mostly extreme conditions) is difficult to truly reflect the actual operating environment of BESS. Second, the configuration problem of BESS is usually modeled as a mixed-integer nonlinear programming (MINLP) problem, which is mostly solved by heuristic algorithms or mathematical programming methods. Among the heuristic algorithms, particle swarm optimization (PSO) and genetic algorithms (GA) are widely used. They both use similar transformation methods to transform the original optimization problem into a spatial solution problem, and then search for the optimal solution in the space. Mathematical programming methods for solving optimal power flow problems can be roughly divided into three categories: finding local optimal solutions, approximating power flow equality constraints, and convex relaxation of non-convex power flow constraints. The method of finding local optimal solutions is difficult to measure the gap between the local optimal solution and the global optimal solution, and cannot guarantee the optimal solution. Although approximating power flow equality constraints is widely used, it is difficult to apply in distribution networks and is not suitable for the optimization of large-scale systems. In contrast, the third method can obtain the optimal solution to the problem with high relaxation accuracy.

[0004] In summary, the current heuristic algorithms and mathematical programming methods are difficult to strike a balance between solution efficiency and accuracy, especially when dealing with three-phase unbalanced distribution networks. Summary of the invention

[0005] The purpose of the present invention is to provide a configuration method for an unbalanced active distribution network battery energy storage system based on multiple scenarios, which overcomes the shortcomings of the prior art, comprehensively considers the impact of multiple scenario operations, source-load uncertainty, and BESS on network sensitivity characteristics, and takes into account the impact of three-phase imbalance, which is more suitable for a large number of DG grid-connected scenarios, solves the problem of lack of consideration of multiple operating conditions in conventional methods, and realizes the configuration of BESS in a three-phase unbalanced distribution network, which has effectiveness and functional superiority.

[0006] To solve the above problems, the technical solutions adopted by the present invention are as follows:

[0007] A method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios includes the following steps:

[0008] S1. Based on the three-phase unbalanced branch power flow model, unbalanced active distribution network power flow modeling is carried out, and a multi-objective BESS optimization configuration model for unbalanced active distribution network is built;

[0009] S2. Based on the multi-objective BESS optimization configuration model, a two-stage decomposition is performed, and a hybrid optimization combining particle swarm optimization and second-order cone programming is used to solve the problem;

[0010] S3. Considering the impact of BESS charging and discharging behavior under multiple operation scenarios, a two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated;

[0011] S4. The rationality and superiority of the two-stage optimization configuration process of BESS are verified through typical example system simulation calculation.

[0012] Furthermore, the unbalanced active distribution network power flow modeling is performed based on the three-phase unbalanced branch power flow model described in S1, and its expression is:

[0013]

[0014] in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; Represents the p-phase voltage at the i and j nodes.

[0015] Furthermore, the multi-objective BESS optimization configuration model for building an unbalanced active distribution network described in S1 is expressed as follows:

[0016] minOF=CI+CM-SLR-SLS

[0017] Among them, CI is the purchase cost of BESS, CM is the operation and maintenance cost, SLR is the benefit of reducing network losses, and SLS is the benefit of load transfer.

[0018] Further, the multi-objective BESS optimization configuration model described in S2 is decomposed into two stages, specifically including the following steps:

[0019] (1) A planning model based on BESS capacity integer variables is constructed, including BESS investment cost and maintenance cost. The expression of the objective function is:

[0020] minOF1=CI+CM

[0021] Among them, CI is the purchase cost of BESS and CM is the operation and maintenance cost.

[0022] (2) Considering the charging and discharging conditions of BESS under different operation scenarios, the optimization results of the planning model are combined for collaborative optimization. The expression of the objective function is:

[0023]

[0024] Where S is the set of operation scenarios considered in the BESS planning problem, which can be obtained through load and DG output forecasting or historical data clustering;

[0025] (3) Based on the two-stage association and their respective mathematical properties, a two-stage hybrid optimization strategy is adopted for effective solution.

[0026] Furthermore, the specific steps of the two-stage hybrid optimization strategy are:

[0027] Step 1: Use the PSO algorithm to solve the first stage;

[0028] Step 1.1, randomly initialize and generate particles with feasible positions and velocities, where each particle represents a possible discrete BESS capacity;

[0029] Step 1.2: Transfer the capacity information of step 1 to the second stage to participate in the economic operation optimization of BESS;

[0030] Step 1.3, calculate the optimal result of each particle and compare them to get the global optimal result;

[0031] Step 1.4: If the PSO termination criteria are met, output the optimal solution, otherwise go to step 1.5;

[0032] Step 1.5, update the particle position and velocity, and repeat steps 1.2 to 1.4 to enter the next round of iteration;

[0033] Step 2: Use SOCP to solve the second stage;

[0034] First, define the following variables:

[0035]

[0036] in, represents the square of the voltage amplitude of node i phase p, represents the square of the current amplitude of the p-phase branch between nodes i and j, represents the active and reactive power transmitted in the branch p phase between nodes i and p, Represents the p-phase voltage at the i and j nodes.

[0037] Substituting the variables into the nonlinear power flow equilibrium constraint,

[0038]

[0039] in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; Qjp is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; represents the square of the voltage amplitude between node i and node j, Represents the square of the current amplitude of the p-phase branch between nodes i and j.

[0040] Finally, SOCP is used for relaxation, and its expression is:

[0041]

[0042] The equivalent of this is:

[0043]

[0044] Furthermore, the two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated as described in S3, and the specific steps include:

[0045] (1) Determine the BESS installation location using CLSI that considers the impact of BESS charging and discharging in multiple operating scenarios;

[0046] (2) For the selected nodes, the PSO-SOCP hybrid optimization strategy is used to solve the two-stage optimization problem of BESS capacity and operation planning, and the configuration capacity and charging and discharging plan of the BESS at the corresponding node are obtained;

[0047] (3) Taking the previous configuration result as a known quantity, repeat the above operation to perform the next BESS configuration until the overall configuration efficiency deteriorates or the planned BESS configuration quantity is reached.

[0048] Furthermore, the specific steps of the BESS sequential configuration strategy based on network loss sensitivity are:

[0049] (1) Calculate the network loss sensitivity CLSI of each phase of all nodes;

[0050] (2) Calculate the average of the three-phase CLSI as the equivalent CLSI of the corresponding node;

[0051] (3) Find the three-phase node with the highest equivalent network loss sensitivity as the BESS configuration location;

[0052] (4) Solve the two-stage optimization configuration problem to determine the optimal BESS capacity of each phase of the configuration node selected in step (3) and its charging and discharging plan under various scenarios;

[0053] (5) Taking the configuration result of this time as a known quantity, repeat steps (1) to (4) until the number of BESS configurations reaches the requirement or the configuration benefits deteriorate.

[0054] Furthermore, the calculation of the network loss sensitivity CLSI of each phase of all nodes comprises the following specific steps:

[0055] Define a grid loss sensitivity CLSI that considers the impact of BESS charging and discharging under multiple operation scenarios:

[0056]

[0057] Among them, λd and λc are the network loss sensitivity weights corresponding to charging and discharging, are the network loss sensitivities corresponding to the r-th configuration charge and discharge, and the specific calculation formula is as follows:

[0058]

[0059] in, is the network loss differential function of the r-th configuration distribution network; The DES charge and discharge differential functions are configured for the rth time respectively; are the charging network losses of the rth and r-1th configurations respectively; are the discharge network losses of the rth and r-1th configurations respectively; are the DES charging powers of the rth and r-1th configurations respectively; are the DES discharge powers of the rth and r-1th configurations respectively.

[0060] Here, it is assumed that the BESS is charged during the load valley and discharged during the load peak. In order to make CLSI fit multiple operation scenarios as much as possible, the calculated Take the average value and get the final CLSIr used:

[0061]

[0062] Compared with the prior art, the present invention has the following beneficial effects:

[0063] The present invention discloses a method for configuring a battery energy storage system for an unbalanced active distribution network based on multiple scenarios, which comprehensively considers the effects of multiple scenario operations, source-load uncertainty, and BESS on network sensitivity characteristics, and also takes into account the effects of three-phase imbalance. It is more suitable for a large number of DG grid-connected scenarios, solves the problem of lack of consideration of multiple operating conditions in conventional methods, and realizes the configuration of BESS in a three-phase unbalanced distribution network, which has effectiveness and functional superiority. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 The relationship between the two-stage optimization of the present invention;

[0065] Figure 2 It is the algorithm flow chart of the present invention;

[0066] Figure 3 A wiring diagram of a real 101-node example in Australia used for the simulation of the present invention;

[0067] Figure 4 It is a network loss sensitivity curve diagram after sequential configuration of the BESS of the present invention;

[0068] Figure 5 It is a distribution network net load curve diagram before and after BESS configuration in multiple operation scenarios of the present invention;

[0069] Figure 6 This is a comparison diagram of the net load of the distribution network before and after the BESS configuration based on a single spring scenario of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] like Figure 1-Figure 2 As shown, the present invention provides a method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios, comprising the following steps:

[0072] S1. Based on the three-phase unbalanced branch power flow model, unbalanced active distribution network power flow modeling is carried out, and a multi-objective BESS optimization configuration model for unbalanced active distribution network is built;

[0073] S2, based on the multi-objective BESS optimization configuration model, a two-stage decomposition is performed, and a hybrid optimization combining particle swarm optimization and second-order cone programming is used to solve it;

[0074] S3. Considering the impact of BESS charging and discharging behavior under multiple operation scenarios, a two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated;

[0075] S4. The rationality and superiority of the two-stage optimization configuration process of BESS are verified through typical example system simulation calculation.

[0076] In S1, based on the three-phase unbalanced branch power flow model, the unbalanced active distribution network power flow modeling is carried out, and its expression is:

[0077]

[0078] in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; Represents the p-phase voltage at the i and j nodes.

[0079] The multi-objective BESS optimization configuration model for unbalanced active distribution network is constructed in S1, and its expression is:

[0080] minOF=CI+CM-SLR-SLS

[0081] Among them, CI is the purchase cost of BESS, CM is the operation and maintenance cost, SLR is the benefit of reducing network losses, and SLS is the benefit of load transfer.

[0082] In S2, a two-stage decomposition is performed based on the multi-objective BESS optimization configuration model, which specifically includes the following steps:

[0083] (3) Construct a planning model based on BESS capacity integer variables, including BESS investment cost and maintenance cost. The expression of the objective function is:

[0084] minOF1=CI+CM

[0085] Among them, CI is the purchase cost of BESS and CM is the operation and maintenance cost.

[0086] (4) Considering the charging and discharging conditions of BESS under different operation scenarios, the optimization results of the planning model are combined for collaborative optimization. The expression of the objective function is:

[0087]

[0088] Where S is the set of operation scenarios considered in the BESS planning problem, which can be obtained through load and DG output forecasting or historical data clustering;

[0089] (3) Based on the two-stage association and their respective mathematical properties, a two-stage hybrid optimization strategy is adopted for effective solution.

[0090] The specific steps of the two-stage hybrid optimization strategy are:

[0091] Step 1: Use the PSO algorithm to solve the first stage;

[0092] Step 1.1, randomly initialize and generate particles with feasible positions and velocities, where each particle represents a possible discrete BESS capacity;

[0093] Step 1.2: Transfer the capacity information of step 1 to the second stage to participate in the economic operation optimization of BESS;

[0094] Step 1.3, calculate the optimal result of each particle and compare them to get the global optimal result;

[0095] Step 1.4: If the PSO termination criteria are met, output the optimal solution, otherwise go to step 1.5;

[0096] Step 1.5, update the particle position and velocity, and repeat steps 1.2 to 1.4 to enter the next round of iteration;

[0097] Step 2: Use SOCP to solve the second stage;

[0098] First, define the following variables:

[0099]

[0100] in, represents the square of the voltage amplitude of node i phase p, represents the square of the current amplitude of the p-phase branch between nodes i and j, represents the active and reactive power transmitted in the branch p phase between nodes i and p, Represents the p-phase voltage at the i and j nodes.

[0101] Bring the variables into the nonlinear power flow equilibrium constraint,

[0102]

[0103] in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; represents the square of the voltage amplitude between node i and node j, Represents the square of the current amplitude of the p-phase branch between nodes i and j.

[0104] Finally, SOCP is used for relaxation, and its expression is:

[0105]

[0106] The equivalent of this is:

[0107]

[0108] In S3, a two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated. The specific steps include:

[0109] (1) Determine the BESS installation location using CLSI that considers the impact of BESS charging and discharging in multiple operating scenarios;

[0110] (2) For the selected nodes, the PSO-SOCP hybrid optimization strategy is used to solve the two-stage optimization problem of BESS capacity and operation planning, and the configuration capacity and charging and discharging plan of the BESS at the corresponding node are obtained;

[0111] (3) Taking the previous configuration result as a known quantity, repeat the above operation to perform the next BESS configuration until the overall configuration efficiency deteriorates or the planned BESS configuration quantity is reached.

[0112] The specific steps of the BESS sequential configuration strategy based on network loss sensitivity are as follows:

[0113] (1) Calculate the network loss sensitivity CLSI of each phase of all nodes;

[0114] (2) Calculate the average of the three-phase CLSI as the equivalent CLSI of the corresponding node;

[0115] (3) Find the three-phase node with the highest equivalent network loss sensitivity as the BESS configuration location;

[0116] (4) Solve the two-stage optimization configuration problem to determine the optimal BESS capacity of each phase of the configuration node selected in step (3) and its charging and discharging plan under various scenarios;

[0117] (5) Taking the configuration result of this time as a known quantity, repeat steps (1) to (4) until the number of BESS configurations reaches the requirement or the configuration benefits deteriorate.

[0118] Calculate the network loss sensitivity CLSI of each phase of all nodes. The specific steps are:

[0119] Define a grid loss sensitivity CLSI that considers the impact of BESS charging and discharging under multiple operation scenarios:

[0120]

[0121] Among them, λd and λc are the network loss sensitivity weights corresponding to charging and discharging, are the network loss sensitivities corresponding to the r-th configuration charge and discharge, and the specific calculation formula is as follows:

[0122]

[0123] in, is the network loss differential function of the r-th configuration distribution network; The DES charge and discharge differential functions are configured for the rth time respectively; are the charging network losses of the rth and r-1th configurations respectively; are the discharge network losses of the rth and r-1th configurations respectively; are the DES charging powers of the rth and r-1th configurations respectively; are the DES discharge powers of the rth and r-1th configurations respectively.

[0124] Here, it is assumed that the BESS is charged during the load valley and discharged during the load peak. In order to make CLSI fit multiple operation scenarios as much as possible, the calculated Take the average value and get the final CLSIr used:

[0125]

[0126] like Figure 3 As shown, the present invention selects a real distribution network system in Australia as the test environment. The 415V three-phase distribution network is powered by a distribution transformer with a capacity of 200kVA and a voltage of 22kV / 415V, and includes 101 nodes and 77 users. Among them, 51 users are connected in a single phase, and 34 users have installed a single-phase rooftop photovoltaic system. The average capacity of the rooftop photovoltaic system is 1.82kWp, the total installed capacity is 63.81kWp, and the penetration rate is 31.9%.

[0127] Figure 4 The network loss sensitivity curve of the BESS after sequential configuration of the present invention is shown in the figure. The four curves of different colors represent the network loss sensitivity of the four configuration distribution networks. The following table shows the results of the sequential optimization configuration of BESS. Figure 4 It can be seen that as the configuration continues, the single BESS configuration capacity generally shows a downward trend, because the configuration cost gradually dominates the objective function. It is worth noting that the BESS capacity of the fifth configuration is 0, which shows that as the cost increases, continuing to configure BESS will lead to a loss of net benefits, that is, continuing to configure is not economically feasible. Therefore, configuring 4 BESS is the optimal configuration solution for this real network.

[0128]

[0129] Figure 5 The net load curves of the four seasons operation scenarios before and after BESS configuration are shown. As can be seen from the figure, BESS plays a significant role in peak shaving and valley filling. At the same time, the different operation modes of BESS four seasons allow it to play an effective role in every season of the year. It should be noted that the load levels and peak-to-valley differences of the four seasons are quite different. For example, the peak shaving and valley filling effect of BESS in winter is not as significant as that in spring. This is because the load in winter is heavier and the load in spring is relatively lighter.

[0130] The following table is a quantitative comparative analysis of the proposed hybrid optimization algorithm and the single optimization of PSO and SOCP. It is worth noting that the configuration of the three optimization methods here is based on simplification, which is only for the purpose of comparing and verifying their computing performance, and the configuration results are not completely referenceable. As can be seen from the table, in terms of accuracy, the PSO-SOCP hybrid optimization is 51.93% higher than the single SOCP optimization; although PSO has higher accuracy for MINLP problems in theory, it may fall into local optimal solutions due to the need to search the complete solution set space, and incur huge computational costs. In terms of speed, the PSO-SOCP hybrid optimization has been significantly improved compared to PSO, and the gap with SOCP is acceptable. Therefore, the above results on solution speed and accuracy show that the proposed PSO-SOCP hybrid optimization can take into account both solution speed and accuracy, and effectively support the application of the proposed BESS optimization configuration in unbalanced distribution networks.

[0131]

[0132] Figure 6 The net load curve of the distribution network before and after the BESS configuration based on a single spring scenario is shown in the following table. The BESS sequential optimization configuration results considering a single operation scenario are shown in the following table. Figure 5 and Figure 6By comparison, it can be seen that the BESS configuration results after considering multiple operating scenarios have higher adaptability to complex operating scenarios and can obtain better overall benefits. Specifically, the BESS configuration based on a single typical spring day scenario does have a better peak shaving and valley filling effect in spring, but it is far inferior to the BESS configuration based on multiple operating scenarios in the other three seasons. At the same time, the estimated profit of the BESS configured with multiple operating sequences is $188,598.4, while the estimated profit of the BESS configured based on a single operating scenario is only $134,013.98, a decrease of 28.94%.

[0133]

[0134] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios, characterized in that: The following steps are involved: S1. Based on the three-phase unbalanced branch power flow model, unbalanced active distribution network power flow modeling is carried out, and a multi-objective BESS optimization configuration model for unbalanced active distribution network is built; S2. Based on the multi-objective BESS optimization configuration model, a two-stage decomposition is performed, and a hybrid optimization combining particle swarm optimization and second-order cone programming is used to solve the problem; S3. Considering the impact of BESS charging and discharging behavior under multiple operation scenarios, a two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated; S4. The rationality and superiority of the two-stage optimization configuration process of BESS are verified through typical example system simulation calculation.

2. According to a multi-scenario unbalanced active distribution network battery energy storage system configuration method according to claim 1, it is characterized by: The unbalanced active distribution network power flow modeling is performed based on the three-phase unbalanced branch power flow model described in S1, and its expression is: in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; Represents the p-phase voltage at the i and j nodes.

3. According to a multi-scenario unbalanced active distribution network battery energy storage system configuration method according to claim 2, it is characterized in that: The multi-objective BESS optimization configuration model for building an unbalanced active distribution network described in S1 is expressed as: minOF=CI+CM-SLR-SLS Among them, CI is the purchase cost of BESS, CM is the operation and maintenance cost, SLR is the benefit of reducing network losses, and SLS is the benefit of load transfer.

4. According to a multi-scenario unbalanced active distribution network battery energy storage system configuration method according to claim 1, it is characterized in that: The two-stage decomposition described in S2 based on the multi-objective BESS optimization configuration model specifically includes the following steps: (1) A planning model based on BESS capacity integer variables is constructed, including BESS investment cost and maintenance cost. The expression of the objective function is: Among them, CI is the purchase cost of BESS and CM is the operation and maintenance cost. (2) Considering the charging and discharging conditions of BESS under different operation scenarios, the optimization results of the planning model are combined for collaborative optimization. The expression of the objective function is: Where S is the set of operation scenarios considered in the BESS planning problem, which can be obtained through load and DG output forecasting or historical data clustering; (3) Based on the two-stage association and their respective mathematical properties, a two-stage hybrid optimization strategy is adopted for effective solution.

5. According to a multi-scenario unbalanced active distribution network battery energy storage system configuration method according to claim 4, it is characterized in that: The specific steps of the two-stage hybrid optimization strategy are: Step 1: Use the PSO algorithm to solve the first stage; Step 1.1, randomly initialize and generate particles with feasible positions and velocities, where each particle represents a possible discrete BESS capacity; Step 1.2: Transfer the capacity information of step 1 to the second stage to participate in the economic operation optimization of BESS; Step 1.3, calculate the optimal result of each particle and compare them to get the global optimal result; Step 1.4: If the PSO termination criteria are met, output the optimal solution, otherwise go to step 1.5; Step 1.5, update the particle position and velocity, and repeat steps 1.2 to 1.4 to enter the next round of iteration; Step 2: Use SOCP to solve the second stage; First, define the following variables: in, represents the square of the voltage amplitude of node i phase p, represents the square of the current amplitude of the p-phase branch between nodes i and j, represents the active and reactive power transmitted in the branch p phase between nodes i and p, represents the voltage of p-phase between nodes i and j; Substituting the variables into the nonlinear power flow equilibrium constraint, in, is the active and reactive output of DG at phase p of node i; represents the active and reactive power transmitted in the branch p phase between nodes i and p; It represents the sum of active and reactive power transmitted in phase p by the other k nodes connected to node j except node i; Qjp is the active and reactive power of phase p at node j; represents the resistance and reactance of the p-phase branch between nodes i and j; represents the square of the voltage amplitude between node i and node j, represents the square of the current amplitude of the p-phase branch between nodes i and j; Finally, SOCP is used for relaxation, and its expression is: The equivalent of this is:

6. According to a multi-scenario unbalanced active distribution network battery energy storage system configuration method according to claim 1, it is characterized by: The two-stage optimization configuration process of BESS based on grid loss sensitivity is formulated as described in S3. The specific steps include: (1) Determine the BESS installation location using CLSI that considers the impact of BESS charging and discharging in multiple operating scenarios; (2) For the selected nodes, the PSO-SOCP hybrid optimization strategy is used to solve the two-stage optimization problem of BESS capacity and operation planning, and the configuration capacity and charging and discharging plan of the BESS at the corresponding node are obtained; (3) Taking the previous configuration result as a known quantity, repeat the above operation to perform the next BESS configuration until the overall configuration efficiency deteriorates or the planned BESS configuration quantity is reached.

7. The method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios according to claim 6 is characterized in that: The specific steps of the BESS sequential configuration strategy based on network loss sensitivity are: (1) Calculate the network loss sensitivity CLSI of each phase of all nodes; (2) Calculate the average of the three-phase CLSI as the equivalent CLSI of the corresponding node; (3) Find the three-phase node with the highest equivalent network loss sensitivity as the BESS configuration location; (4) Solve the two-stage optimization configuration problem to determine the optimal BESS capacity of each phase of the configuration node selected in step (3) and its charging and discharging plan under various scenarios; (5) Taking the configuration result of this time as a known quantity, repeat steps (1) to (4) until the number of BESS configurations reaches the requirement or the configuration benefits deteriorate.

8. The method for configuring an unbalanced active distribution network battery energy storage system based on multiple scenarios according to claim 7 is characterized in that: The specific steps of calculating the network loss sensitivity CLSI of each phase of all nodes are as follows: Define a grid loss sensitivity CLSI that considers the impact of BESS charging and discharging under multiple operation scenarios: Among them, λd and λc are the network loss sensitivity weights corresponding to charging and discharging, LSI c r are the network loss sensitivities corresponding to the r-th configuration charge and discharge, and the specific calculation formula is as follows: in, is the network loss differential function of the r-th configuration distribution network; The DES charge and discharge differential functions are configured for the rth time respectively; are the charging network losses of the rth and r-1th configurations respectively; are the discharge network losses of the rth and r-1th configurations respectively; are the DES charging powers of the rth and r-1th configurations respectively; are the DES discharge powers of the rth and r-1th configurations respectively. Assume that BESS is charged at low load and discharged at peak load. In order to make CLSI fit multiple operation scenarios as much as possible, the calculated Take the average value and get the final CLSIr used:

Images