Power grid expansion planning method combined with energy storage system

Abstract

The invention relates to the technical field of power grid planning, in particular to a power grid expansion planning method combined with an energy storage system, which adopts an investment strategy of combining the energy storage system and the transformation of related equipment of a power distribution network to realize the collaborative planning and solution of a power distribution network frame line and the energy storage system. According to the method, the power distribution network expansion planning can obtain a more reasonable network frame layout, so that a line which is most likely to be blocked is separated from a high-load operation state, the system power flow is more stable, and the planning result has more reliability and long-term development.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid planning, and in particular to a power grid expansion planning method combined with an energy storage system. Background Art

[0002] As the penetration rate of distributed photovoltaic (PV) in the distribution network continues to increase, the uncertainty of its output is an important issue facing the reliable operation of the distribution network. If the distribution network does not absorb enough PV, its utilization rate will decrease and the equipment operation and maintenance costs will increase. At the same time, with the large-scale integration of PV, the power flow of some lines in the distribution network grid will change frequently, so that the grid line will be blocked, which will not only reduce the absorption rate of distributed photovoltaic, but also affect the rapid response of flexible resources, and indirectly reduce the reliability of the distribution network. During the distribution network expansion planning, preparing sufficient line channel margin for PV absorption can not only respond to distributed energy in a timely manner after planning, but also greatly reduce the investment and construction costs of the expansion planning. Therefore, the distribution network expansion planning needs to not only focus on economy, but also plan and layout the location and capacity of grid lines and distributed power sources, enhance the transmission capacity of grid lines, and provide good protection for the safe operation of distribution network lines.

[0003] A good configuration strategy for the location and capacity of energy storage in the power grid can cope with large loads and the randomness of photovoltaic output. Li Jitong, Wang Zhou and other scholars proposed a distribution network energy storage power station planning scheme under severe weather conditions, taking into account the spatiotemporal distribution characteristics of disasters and distribution network scheduling, and improving the operational flexibility of the distribution network, but the planning object is relatively single. Liu Yuankun, Zhang Weijing and other scholars considered the impact of power market scheduling and constructed planning models with the highest net rate of return and the maximum social welfare as the goals, but the algorithm used had low solution efficiency and accuracy. Liu Zifa, Yu Puyang and others proposed a two-level planning strategy for distributed power sources and generalized energy storage, comprehensively considering the operating characteristics of energy storage and distribution networks, and selecting multiple response capability indicators such as load change rate to improve the efficiency of distribution networks. Cao Xinhui, Che Yong and others proposed the four-quadrant operation characteristics of energy storage, explored the impact of energy storage capacity and location selection on system voltage fluctuations and new energy consumption, and improved the new energy consumption capacity, but this model ignored the impact of energy storage control. Gong Jianfeng, Zhou Zongchuan and other teams built evaluation indicators based on the operating characteristics of flexible resources, effectively improving the flexibility and operating stability of the distribution network, but did not consider the uncertainty of new energy. However, the above scholars only considered energy storage and new energy in the planning model, but did not comprehensively consider the coordinated optimization operation of source-grid-load-storage. The construction of the planning model is not perfect, and there is a lack of comprehensive consideration of the distribution network grid structure.

[0004] The content of distribution network line expansion planning includes finding the optimal distribution network expansion structure and energy storage equipment configuration scheme, with the aim of improving the distribution system's acceptance of distributed renewable energy and power supply reliability. Koutsoukis N, Georgilakis P and others used scenario analysis to deal with the volatility of renewable energy output, and considered substations and renewable energy together, but the genetic algorithm used in model solution would fall into the local optimal solution. Jooshaki M's team constructed a planning model with reliability evaluation indicators and converted the model into a mixed integer linear programming problem for solution, but sacrificed the accuracy of distributed energy output prediction. Xing Haijun, Cheng Haozhong and others took the minimum grid investment and operation cost as the goal, and the planning measures considered the construction of new substations and lines, and considered a variety of management schemes to reduce planning costs. Wu Zhi, Liu Yafei and others adopted a multi-stage collaborative planning scheme for the site selection and sizing of distribution network expansion and distributed energy and reactive compensation devices, and used the improved Benders decomposition method to solve the mixed integer programming model, but lacked consideration of distribution network lines. Summary of the invention

[0005] In view of the problems pointed out in the background technology, the purpose of the present invention is to provide a method for planning the expansion of a power grid combined with an energy storage system, which adopts an investment strategy that combines the energy storage system with the transformation of distribution network related equipment to achieve the coordinated planning and solution of the distribution network grid lines and the energy storage system. The present invention can make the distribution network expansion planning obtain a more reasonable grid layout, so that the lines that are most prone to congestion are out of the high-load operation state, the system flow is more stable, and the planning results are more reliable and long-term.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for planning power grid expansion in combination with an energy storage system comprises the following steps:

[0008] Step S1, obtaining historical photovoltaic output data, load data and original topology information of the target distribution network; obtaining planning parameters, including node data to be connected to the photovoltaic or energy storage system, line data to be planned, photovoltaic data to be planned and energy storage system data to be planned;

[0009] Step S2: Generate a typical photovoltaic load scenario by reducing the dimension of the historical photovoltaic output data and load data of the target distribution network;

[0010] Step S3: modify the target distribution network topology information according to the planning parameters and the original topology information of the target distribution network;

[0011] Step S4, constructing a two-layer planning model of energy storage and power grid according to the modified target distribution network topology information; the two-layer planning model of energy storage and power grid includes an upper model and a lower model, wherein the upper model is to determine the optimal installation location and capacity of the energy storage system and the direction of the expansion of the distribution network grid line, considering the minimum comprehensive cost of energy storage and grid lines as the goal, and the decision variables are the installation location and capacity of the energy storage system and the grid line layout; the lower model considers the operation of the distribution network system, taking the maximum photovoltaic consumption as the goal, and the decision variables include the power flow distribution of the distribution network and the active output of the distributed power source;

[0012] Step S5: A nested improved particle swarm optimization algorithm is used to solve the energy storage and power grid double-layer planning model; the energy storage and power grid double-layer planning model outputs a planning scheme, which includes energy storage investment cost, energy storage operation cost, line expansion investment cost and photovoltaic operation and maintenance cost.

[0013] In step S4, the two-layer planning model of energy storage and power grid is expressed as:

[0014]

[0015] In formula (1), F is the objective function of the upper model, f is the objective function of the lower model; G(·) and H(·) are the constraints of the upper model, g(·) and h(·) are the constraints of the lower model; C I is the construction cost, C O For running costs.

[0016] In step S4, C I Including grid line investment cost Distributed photovoltaic investment cost Energy storage investment cost C O Including the network loss cost for the whole year Cost of purchasing electricity from the large power grid Energy storage equipment operation and maintenance costs Distributed photovoltaic operation and maintenance costs The details are as follows:

[0017] A. Grid line investment cost

[0018]

[0019] In formula (2), r is the discount rate; y is the useful life; K L is the sum of the original route and the route to be planned; x k is a state variable, its value 0 indicates that the kth line is not under construction, and its value 1 indicates that the kth line is under construction; c k is the construction cost of the kth line;

[0020] B. Energy storage investment cost

[0021]

[0022] In formula (3): are the unit power and capacity investment costs of building energy storage devices at node i, respectively; are the energy storage power and capacity constructed at node i respectively;

[0023] C. Network loss cost

[0024]

[0025] In formula (4), T is the total time; R ij is the resistance of line ij; c loss is the unit network loss cost;

[0026] D. Cost of purchasing electricity from the large power grid

[0027]

[0028] In formula (5): c pur is the unit electricity purchase cost; p G,t is the power generated by the generator at time t;

[0029] E. Energy storage equipment operation and maintenance costs

[0030]

[0031] In formula (6): ESS is the set of nodes where ESS is installed; is the operating cost of ESS per unit charge and discharge capacity; are the charging and discharging power of ESS at node i during period t;

[0032] F. Distributed photovoltaic operation costs

[0033]

[0034] In formula (7): λ is the conversion ratio of operation and maintenance cost, which is 0.1; is the power abandonment cost per unit capacity of distributed photovoltaics; is the theoretical value of the photovoltaic grid-connected capacity of node i at time t; is the actual value of the photovoltaic grid-connected capacity of node i at time t.

[0035] In step S4, the constraints of the upper model include topological constraints and reconstruction constraints; wherein the topological constraints are radial constraints during the planning of the distribution network framework, and the total number of branches is the difference between the total number of nodes and the number of root nodes, as shown in the following formula:

[0036]

[0037] In formula (8): D l represents the set of all lines in the distribution network, x ij The circuit is open or closed. 0 means open, 1 means closed, and n s Indicates the number of root nodes in the system;

[0038] The reconstruction constraint is defined as follows: To avoid loops or islands, assume that all lines are connected and form loops, disconnect one branch of the overlapping branches between the loops, and also split one branch of the non-overlapping branches. At a node with n branches, at most n-1 lines are split, which can ensure that the distribution network does not produce islands or loops; as shown in the following formula:

[0039]

[0040] In formula (9): NL k and L k Respectively represent the total number and set of branches in the kth loop, n r Indicates the total number of loops, l k Represents the set of branches connected to the kth node.

[0041] In step S4, the lower model constraints include system safety constraints, distribution network flow constraints, distributed power output constraints, line power transmission constraints and energy storage constraints of the energy storage system, which are as follows:

[0042] X1. System safety constraints: To maintain the safe operation of the power grid system, the node voltage and the current flowing through the line must be controlled within a safe range. The mathematical expression is:

[0043]

[0044] In formula (10): U i,max and U i,min are the upper and lower limits of the voltage at node i, respectively; U i,t is the voltage of node i at time t;

[0045] X2, power grid flow constraint; select the distribution network DistFlow as the power flow model, as follows:

[0046]

[0047] In formula (11), AL(j,:) is the set of branch end nodes with j as the head node; AL(:,j) is the set of branch head node with j as the end node; P ij,t , P jk,t and Q ij,t , Qjk,t are the active and reactive power on lines ij and jk during period t respectively; X ij is the reactance of line ij; and are respectively the active and reactive power of the original load of node j in period t; U i,t is the voltage of node i;

[0048] X3. Distributed power generation output constraints are as follows:

[0049]

[0050] In formula (12): and They are the actual PV output at time t. The upper and lower limits of

[0051] X4, line power transmission constraints are as follows

[0052]

[0053] Where: P ij,max , Q ij,max They are respectively the upper limits of active and reactive power of line transmission.

[0054] X5, energy storage constraints of the energy storage system, as follows:

[0055]

[0056] In formula (14): They are the state variables of ESS charging and discharging, which are 0-1 variables, representing whether the ESS is charging or discharging; are the charging and discharging power of ESS at node i at time t respectively; is the maximum value of the ESS charging and discharging power of node i; is the power of ESS at node i at time t; is the upper limit of ESS power of node i; η is the charging and discharging efficiency.

[0057] The step S5 comprises the following steps:

[0058] Step S5.1: importing photovoltaic output and load data and modified target distribution network topology information;

[0059] Step S5.2: Initialize the upper particle population, perform binary coding on the new lines and energy storage sites, and generate particle positions and velocities;

[0060] Step S5.3: Calculate the population fitness, obtain the distribution network grid line direction and energy storage site selection location, filter out unreasonable planning results and update the population location;

[0061] Step S5.4: Repeat steps S5.2 to S5.3 until convergence is achieved, i.e., the difference between the planning cost of the upper layer for 10 consecutive iterations does not exceed the preset threshold or reaches the upper limit of the number of iterations of the upper layer;

[0062] Step S5.5: import the planning scheme obtained in the upper layer into the lower layer and initialize the lower layer population, and encode each energy storage capacity integer;

[0063] Step S5.6: Calculate the power grid flow and distributed generation output data, execute the lower-level objective function, update the individual and overall extreme values ​​of the energy storage capacity corresponding to all grid planning results in the lower layer, regard the overall extreme value as the optimal energy storage capacity corresponding to each planning result, and update the speed and position of the lower-level particles.

[0064] Step S5.7: Determine whether the lower-level convergence standard is met, that is, the difference between the distributed photovoltaic absorption rate before and after 10 consecutive iterations is less than 0.001 and 0.05 respectively, or the upper limit of the number of lower-level iterations is reached.

[0065] Step S5.8: If the maximum number of iterations is reached, stop searching and output the results; otherwise return to step S5.2.

[0066] Beneficial effects of the present invention: Compared with traditional distribution network expansion planning, the present invention takes into account the transformation of distribution network-related equipment in the planning model, which can enable the distribution network expansion planning to obtain a more reasonable grid layout, so that the lines most prone to congestion can be separated from the high-load operation state, the system flow is more stable, and the planning results are more reliable and long-term; the present invention can effectively reduce network losses, power purchase costs and carbon emission costs of thermal power units through the adjustment of flexible resources, and put the entire distribution network system in a safer operating state. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 It is a schematic diagram of the energy storage and power grid double-layer planning model of the present invention.

[0068] Figure 2 Flowchart of the improved particle swarm optimization algorithm used in the present invention.

[0069] Figure 3 Schematic diagram of the distribution network structure to be planned in the embodiment.

[0070] Figure 4 This is the distribution network line planning result diagram obtained in Scheme 1 of Example 1.

[0071] Figure 5 This is the distribution network line planning result diagram obtained by Scheme 2 of Example 1.

[0072] Figure 6This is the distribution network line planning result diagram obtained by Scheme 3 of Example 1. DETAILED DESCRIPTION

[0073] The following will be combined with the drawings and embodiments of this specification to clearly and completely describe the technical solution of the present invention. It should be noted that 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.

[0074] As shown in Question 1, the present invention provides a grid expansion planning method combined with an energy storage system, and the specific steps are as described in the invention content section.

[0075] Example 1 is based on the historical data of the distribution network in a certain area throughout the year, and uses Monte Carlo simulation and synchronous back-substitution method to reduce the generated load and photovoltaic output scenarios, and finally generates 5 typical scenarios; the probability of each typical scenario is shown in Table 1.

[0076] Table 1 Typical day scenario probability

[0077]

[0078] This embodiment adds a node based on the IEEE 33-node power distribution system and tests the modified 40-node power distribution network system. The modified 40-node power distribution network structure topology is as follows: Figure 3 The dotted lines are lines to be constructed, and the corresponding new load nodes and lines to be constructed are shown in Table 2.

[0079] Table 2 New node load and branch data

[0080]

[0081]

[0082] 1) Line data: To ensure power supply reliability, lines 1-32 of the original IEEE 33-node system are retained, and lines 33-44 are to be planned. The maximum current of the lines is 0.4kA.

[0083] 2) Photovoltaic data: The number of PV installations is 4, and the installation nodes are 6, 17, 24 and 30. The operation and maintenance cost is 0.35 yuan / (kW·h); the penalty fee for abandonment of light is 0.50 yuan / (kW·h).

[0084] 3) ESS data: The energy storage type is battery energy storage, the number of nodes to be installed is 2-40, that is, all nodes except the first node can be installed, the unit capacity investment cost is 1,000 yuan / (kW·h), the unit power construction cost is 1,500 yuan / (kW·h), the unit capacity operation and maintenance cost is 0.35 yuan / (kW·h), the initial and final state of charge in one cycle is 0.5, the upper and lower limits of the state of charge are 0.9 and 0.2 respectively, and the charge and discharge efficiency is 0.95.

[0085] 4) Other parameters: The node voltage range is 0.95pu-1.05pu, the unit penalty cost of carbon emissions is 80 yuan / t, the carbon emissions per unit power generation of the generator set is set to 0.798t / (MW·h); the discount rate is 5%.

[0086] All tests were modeled and solved using the Matlab R2019b platform with the YALMIP toolkit embedded in it and the Gurobi10.0 solver.

[0087] In order to compare the impact of PV and load uncertainties on distribution network expansion planning, three comparison schemes are set up, namely, considering only PV uncertainty, only load uncertainty, and considering both source and load uncertainty. The comparison between them and the deterministic planning results under historical data is shown in Table 3.

[0088] Table 3 Distribution network expansion planning results under different scenarios

[0089]

[0090]

[0091] As shown in Table 3, compared with the plan under historical data, the plan that takes uncertainty into account has a slight increase in annual comprehensive investment cost, but has improved the new energy consumption rate and can reduce the cost of purchasing electricity from the large power grid. The planning results that take into account both source and load uncertainty have a higher new energy consumption rate than those that only consider photovoltaic and load uncertainty, which are increased by 3.5% and 3.85% respectively.

[0092] In order to compare the advantages and disadvantages of the proposed strategy with other strategies, the following three cases are provided:

[0093] Case 1: Using the energy storage and power grid two-layer planning model proposed in the present invention to obtain the distribution network expansion plan, taking into account both economic cost and grid line current margin;

[0094] Case 2: Distribution network expansion planning based on Case 1 without considering line current margin;

[0095] Case 3: Based on Case 1, the distribution network expansion planning without considering the energy storage device is the traditional distribution network expansion planning strategy.

[0096] The three methods of case setting all use distributed photovoltaic equipment of the same specifications and consider photovoltaic and load uncertainties.

[0097] Through simulation calculation, the grid line planning results of Case 1 and Case 2 are as follows: Figure 4 and Figure 5 The comparison of various planning costs is shown in Table 4.

[0098] Table 4 Distribution network expansion planning costs under Cases 1 and 2

[0099]

[0100]

[0101] As can be seen from Table 4, the distribution network expansion planning scheme obtained in Case 1 is not economically superior to the planning scheme in Case 2, and the annual comprehensive cost of the distribution network has increased by 704,800 yuan, an increase of 5.62%. This is because the line selected in Case 1 for line expansion is 40-34-35-36-37-38-39, and the capacity configuration results of the two energy storages are 1.03MW and 0.76MW, while the line selected in Case 2 for grid line expansion is 33-34-35-36-37-38-39, and the capacity configuration results of the two energy storages are 0.63MW and 0.54MW. The distribution network grid line layout tends to build low-cost lines and reduce the configuration and operation and maintenance costs of energy storage capacity to reduce investment costs. However, Case 1 is superior to Case 2 in terms of photovoltaic operation and maintenance costs, electricity purchase costs, line network loss costs and carbon emission costs. The electricity purchase costs of the large power grid were reduced by 549,600 yuan, and the line network loss costs were reduced by 545,600 yuan. This shows that after considering the transformation of relevant equipment, the planning and layout of the grid lines are more reasonable and the flow distribution is more uniform; at the same time, the capacity configuration and site selection of energy storage further optimize the transmission capacity of the grid channel, promote the consumption of distributed photovoltaics, reduce the purchase of electricity from thermal power units, and reduce carbon emission costs.

[0102] At the same time, the economic comparison between Case 1 and Case 3, the various costs are shown in Table 5.

[0103] Table 5 Distribution network expansion planning costs under Cases 1 and 3

[0104]

[0105] Case 3 Overall grid line planning results are as follows Figure 6 shown.

[0106] As shown in Table 5, the planning scheme of Case 1 after considering energy storage is 2.7197 million yuan higher than that of Case 3. This is because Case 3 does not have the cost of energy storage investment, construction and operation and maintenance, but reduces the large power grid power purchase cost, line network loss cost and carbon emission cost by 782,000 yuan, 180,600 yuan and 168,000 yuan respectively, because energy storage can store low power and generate high power, and participate in the optimization of distribution network power flow distribution. It can be seen that the planning scheme considering energy storage can improve the safety of the distribution network, that is, sacrifice part of the economy in exchange for higher stability to ensure the operation of the distribution network.

[0107] Taking the Northeast Grid of Yichuan as an example, when the photovoltaic penetration rate is 80%, the distribution network only considers the transformation of transformer capacity. It is necessary to add 18 transformers to expand the capacity and expand 3 lines, with a total transformation cost of 1.88 million yuan. At the same time, the construction of the 110kV Lujiang transformer will be started in advance in 2025.

[0108] When the photovoltaic penetration rate is 80% and distributed energy storage is taken into consideration, a total of 2 transformers need to be added, with a storage capacity of 1276.6kWh and a storage power of 638.3kW. One line needs to be expanded, and a total renovation cost of 958,300 yuan is required.

[0109] When the photovoltaic penetration rate is 80%, taking into account energy storage, demand-side response and distribution network-related equipment transformation, a total of one transformer needs to be added, with a storage capacity of 1171.2kWh and a storage power of 585.6kW, and one line needs to be expanded, with a total transformation cost of 845,600 yuan.

[0110] Taking the 10kV power grid in the northeast grid of Yichuan County, Luoyang City as an example, the field verification was carried out in combination with MATLAB simulation software. After simulation calculation, the investment strategy considering the transformation of energy storage and distribution network related equipment was obtained, with an energy storage capacity of 1200kWh, an energy storage power of 600kW, and one line expansion, requiring a total transformation fee of 845,600 yuan. According to the investment of new 10kV and below lines, capacity expansion and distribution transformers, energy storage and other devices, the cost classification was calculated. Compared with the traditional plan, the investment in the power grid was delayed by 2.2 million yuan, the network loss was saved by 57,000 yuan, and the investment in the transformation of related equipment was saved by 1.1 million yuan. The power grid's ability to absorb new energy increased by 6.7 percentage points. The proposed plan of the project can realize the coordinated operation of source, grid, load and storage, and improve economic benefits.

[0111] The method proposed in the present invention combines the energy storage system configuration and related data to transform the distributed photovoltaic distribution network expansion planning model to simulate the actual working conditions, and solves it through a nested improved particle swarm optimization algorithm. The particle swarm optimization algorithm makes full use of the location information of excellent particles and searches in parallel in the feasible solution space. At the same time, the algorithm makes full use of the probability transfer rule as the internal working mechanism of the population, and there is no need to add derived information. Since the algorithm was proposed in 1995, it has attracted more and more attention from researchers due to its good optimization performance. To facilitate the understanding of the present invention, a brief introduction is given as follows:

[0112] In the particle swarm optimization algorithm, all particles in the population search for the optimal solution in the search space. Each particle can serve as a potential solution to the target problem, and the information of the particle is affected by its own historical optimal experience and the global historical optimal experience. The quality of each particle is evaluated by calculating the cost function. During the iterative search process of the particle swarm, all particles learn from the optimal particle in the population, and the learning process is accompanied by random perturbation parameters, and finally the position information of the next generation of particles is determined.

[0113] In the common particle swarm optimization algorithm, each particle is represented by its own position information x i =(x 1 ,x 2 ,…x D ), its best historical position and particle velocity v i =(v 1 ,v 2 ,…,v D ) three D-dimensional vectors. The particle also needs to save the global best historical position information gbest = (gbest 1 ,gbest 2 ,…,gbest D ). During each iterative search of the particle swarm algorithm, the position information of each particle will be evaluated. If the evaluation result is better than its own historical optimal position, the current position information replaces its own historical optimal position. Similarly, if the evaluation result is better than the global historical optimal position, the current position information replaces the global historical optimal position. The algorithm continuously updates the current particle position information, particle velocity information, the particle's own historical optimal information, and the global optimal historical information during the iterative operation. Until the algorithm searches for a feasible and satisfactory solution. The speed and position update formula of the particles in the population at time t+1 is as follows:

[0114]

[0115] In formulas (15) and (16), the dimension of the problem is d = 1, 2, ..., D, c 1 and c 2It is used to represent the acceleration coefficient, which controls the degree to which the particle learns from its own historical best position and the global historical best position. 1 and r 2 It represents a random number between the interval [0,1]. These two numbers are randomly generated during each iteration of the algorithm. ω is the inertia weight, which is used to control the influence coefficient of the previous particle speed on the current particle speed. Among them, the optimization stability of the population is closely related to the inertia weight.

[0116] However, although the conventional PSO algorithm has strong applicability, it is easy to fall into local optimum. The present invention makes improvements in inertia weight and population variation to enhance the global search efficiency of the PSO algorithm.

[0117] 1) Inertia weight update

[0118] The inertia weight can adjust the impact on the current speed. Its increase and decrease are helpful to increase the search range and local precise optimization respectively. In order to balance the conflict between efficient search and fast convergence, the inertia weight is updated using formula (17).

[0119]

[0120] Where: max ,ω min is the maximum and minimum value of inertia weight; k is the number of iterations; k max is the maximum number of iterations.

[0121] 2) Population mutation operation

[0122] In order to improve the diversity of particles, the population is mutated using formula (18).

[0123]

[0124] Where: p m is the mutation probability; ε m is the variation parameter. m The attenuation degree of the mutation probability can be adjusted, that is, the impact of the mutation operation can be attenuated as the number of iterations increases.

[0125] For the two-layer collaborative planning model of energy storage and power grid, a nested improved particle swarm optimization algorithm is used to solve it. The two-layer optimization model will interact when searching for the best result, that is, the optimization of the lower-layer energy storage capacity is based on the results of the upper-layer distribution network grid line direction and energy storage installation location, and the distributed power output and other results are returned to the upper layer as the basis for the next iterative optimization, and the optimal planning solution is obtained through the iteration of the upper and lower layers. Figure 2 This is the improved double-layer nested particle swarm optimization algorithm flow chart. The specific algorithm solution steps are as follows:

[0126] Step S5.1: Importing photovoltaic output and load data and distribution network structure related parameters;

[0127] Step S5.2: Initialize the upper particle population, perform binary coding on the new lines and energy storage sites, and generate particle positions and velocities;

[0128] Step S5.3: Calculate the population fitness, obtain the distribution network grid line direction and energy storage site selection location, filter out unreasonable planning results and update the population location;

[0129] Step S5.4: Repeat steps S5.2 to S5.3 until convergence is achieved, that is, the difference between the upper layer planning cost for 10 consecutive iterations does not exceed 300 yuan or the upper limit of the number of iterations of the upper layer is reached;

[0130] Step S5.5: import the planning scheme obtained in the upper layer into the lower layer and initialize the lower layer population, and encode each energy storage capacity integer;

[0131] Step S5.6: Calculate the power grid flow and distributed generation output data, execute the lower layer objective function, update the individual and overall extreme values ​​of the energy storage capacity corresponding to all grid planning results in the lower layer, regard the overall extreme value as the optimal energy storage capacity corresponding to each planning result, and update the speed and position of the particles in the lower layer;

[0132] Step S5.7: Determine whether the lower-level convergence standard is met, that is, the difference between the distributed photovoltaic absorption rate before and after 10 consecutive iterations is less than 0.001 and 0.05 respectively, or the upper limit of the number of iterations of the lower level is reached;

[0133] Step 8: If the maximum number of iterations is reached, stop searching and output the results; otherwise, return to step S5.2.

[0134] From the above embodiments and field simulation results, the following conclusions can be drawn:

[0135] 1) Compared with the traditional distribution network expansion planning, the grid expansion planning method of the present invention takes into account the energy storage system. Although it is not economically advantageous, it can effectively reduce network losses, power purchase costs of thermal power units and carbon emissions through the adjustment of flexible resources, and put the entire distribution network system in a safer operating state.

[0136] 2) The present invention considers the transformation of distribution network related equipment in the planning model, which can make the distribution network expansion planning obtain a more reasonable grid layout, so that the lines most prone to congestion can be separated from the high-load operation state, the system flow is more stable, and the planning results are more reliable and long-term.

[0137] 3) The strategy of the present invention, which takes into account the uncertainty of photovoltaic output and load fluctuation, is slightly higher than the deterministic result in planning cost, but has achieved better optimization and improvement in photovoltaic absorption rate. Compared with the results that only consider the uncertainty of the source side or the load side, it is also more advantageous in photovoltaic absorption rate, while reducing the cost of purchasing electricity from the large power grid.

[0138] The parts not described in detail in this invention are prior art.

Claims

1. A method for planning power grid expansion in combination with an energy storage system, characterized by: The following steps are involved: Step S1, obtaining historical photovoltaic output data, load data and original topology information of the target distribution network; Obtain planning parameters, including node data to be connected to the photovoltaic or energy storage system, line data to be planned, photovoltaic data to be planned, and energy storage system data to be planned; Step S2: Generate a typical photovoltaic load scenario by reducing the dimension of the historical photovoltaic output data and load data of the target distribution network; Step S3: modify the target distribution network topology information according to the planning parameters and the original topology information of the target distribution network; Step S4, constructing a two-layer planning model of energy storage and power grid according to the modified target distribution network topology information; the two-layer planning model of energy storage and power grid includes an upper model and a lower model, wherein the upper model is to determine the optimal installation location and capacity of the energy storage system and the direction of the expansion of the distribution network grid line, considering the minimum comprehensive cost of energy storage and grid lines as the goal, and the decision variables are the installation location and capacity of the energy storage system and the grid line layout; the lower model considers the operation of the distribution network system, taking the maximum photovoltaic consumption as the goal, and the decision variables include the power flow distribution of the distribution network and the active output of the distributed power source; Step S5: A nested improved particle swarm optimization algorithm is used to solve the energy storage and power grid double-layer planning model; the energy storage and power grid double-layer planning model outputs a planning scheme, which includes energy storage investment cost, energy storage operation cost, line expansion investment cost and photovoltaic operation and maintenance cost.

2. A method for planning power grid expansion in combination with an energy storage system according to claim 1, characterized in that: In step S4, the two-layer planning model of energy storage and power grid is expressed as: In formula (1), F is the objective function of the upper model, f is the objective function of the lower model; G(·) and H(·) are the constraints of the upper model, g(·) and h(·) are the constraints of the lower model; C I is the construction cost, C O For running costs.

3. A method for planning power grid expansion in combination with an energy storage system according to claim 2, characterized in that: In step S4, C I Including grid line investment cost Distributed photovoltaic investment cost Energy storage investment cost C O Including the network loss cost for the whole year Cost of purchasing electricity from the large power grid Energy storage equipment operation and maintenance costs Distributed photovoltaic operation and maintenance costs The details are as follows: A. Grid line investment cost In formula (2), r is the discount rate; y is the useful life; K L is the sum of the original route and the route to be planned; x k is a state variable, its value 0 indicates that the kth line is not under construction, and its value 1 indicates that the kth line is under construction; c k is the construction cost of the kth line; B. Energy storage investment cost In formula (3): are the unit power and capacity investment costs of building energy storage devices at node i, respectively; are the energy storage power and capacity constructed at node i respectively; C. Network loss cost In formula (4), T is the total time; R ij is the resistance of line ij; c loss is the unit network loss cost; D. Cost of purchasing electricity from the large power grid In formula (5): c pur is the unit electricity purchase cost; p G,t is the power generated by the generator at time t; E. Energy storage equipment operation and maintenance costs In formula (6): ESS is the set of nodes where ESS is installed; is the operating cost of ESS per unit charge and discharge capacity; are the charging and discharging power of ESS at node i during period t; F. Distributed photovoltaic operation costs In formula (7): λ is the conversion ratio of operation and maintenance cost, which is 0.1; is the power abandonment cost per unit capacity of distributed photovoltaics; is the theoretical value of the photovoltaic grid-connected capacity of node i at time t; is the actual value of the photovoltaic grid-connected capacity of node i at time t.

4. A method for planning power grid expansion in combination with an energy storage system according to claim 2, characterized in that: In step S4, the constraints of the upper model include topological constraints and reconstruction constraints; wherein the topological constraints are radial constraints during the planning of the distribution network framework, and the total number of branches is the difference between the total number of nodes and the number of root nodes, as shown in the following formula: In formula (8): D l represents the set of all lines in the distribution network, x ij The circuit is open or closed. 0 means open, 1 means closed, and n s Indicates the number of root nodes in the system; The reconstruction constraint is defined as follows: To avoid loops or islands, assume that all lines are connected and form loops, disconnect one branch of the overlapping branches between the loops, and also split one branch of the non-overlapping branches. At a node with n branches, at most n-1 lines are split, which can ensure that the distribution network does not produce islands or loops; as shown in the following formula: In formula (9): NL k and L k Respectively represent the total number and set of branches in the kth loop, n r Indicates the total number of loops, l k Represents the set of branches connected to the kth node.

5. The method for planning power grid expansion in combination with an energy storage system according to claim 1, characterized in that: In step S4, the lower model constraints include system safety constraints, distribution network flow constraints, distributed power output constraints, line power transmission constraints and energy storage constraints of the energy storage system, which are as follows: X1. System safety constraints: To maintain the safe operation of the power grid system, the node voltage and the current flowing through the line must be controlled within a safe range. The mathematical expression is: In formula (10): U i,max and U i,min are the upper and lower limits of the voltage at node i; U i,t is the voltage of node i at time t; X2, power grid flow constraint; select the distribution network DistFlow as the power flow model, as follows: In formula (11), AL(j,:) is the set of branch end nodes with j as the head node; AL(:,j) is the set of branch head node with j as the end node; P ij,t , P jk,t and Q ij,t , Q jk,t are the active and reactive power on lines ij and jk during period t respectively; X ij is the reactance of line ij; and are respectively the active and reactive power of the original load of node j in period t; U i,t is the voltage of node i; X3. Distributed power generation output constraints are as follows: In formula (12): and They are the actual PV output at time t. The upper and lower limits of X4, line power transmission constraints are as follows Where: P ij,max , Q ij,max They are respectively the upper limits of active and reactive power of line transmission. X5, energy storage constraints of the energy storage system, as follows: In formula (14): They are the state variables of ESS charging and discharging, which are 0-1 variables, representing whether the ESS is charging or discharging; are the charging and discharging power of ESS at node i at time t; P i ESS,max is the maximum value of the ESS charging and discharging power of node i; is the power of ESS at node i at time t; is the upper limit of ESS power of node i; η is the charge and discharge efficiency.

6. A method for planning power grid expansion in combination with an energy storage system according to claim 1, characterized in that: The step S5 comprises the following steps: Step S5.1: importing photovoltaic output and load data and modified target distribution network topology information; Step S5.2: Initialize the upper particle population, perform binary coding on the new lines and energy storage sites, and generate particle positions and velocities; Step S5.3: Calculate the population fitness, obtain the distribution network grid line direction and energy storage site selection location, filter out unreasonable planning results and update the population location; Step S5.4: Repeat steps S5.2 to S5.3 until convergence is achieved, i.e., the difference between the planning cost of the upper layer for 10 consecutive iterations does not exceed the preset threshold or reaches the upper limit of the number of iterations of the upper layer; Step S5.5: import the planning scheme obtained in the upper layer into the lower layer and initialize the lower layer population, and encode each energy storage capacity integer; Step S5.6: Calculate the power grid flow and distributed generation output data, execute the lower-level objective function, update the individual and overall extreme values ​​of the energy storage capacity corresponding to all grid planning results in the lower layer, regard the overall extreme value as the optimal energy storage capacity corresponding to each planning result, and update the speed and position of the lower-level particles. Step S5.7: Determine whether the lower-level convergence standard is met, that is, the difference between the distributed photovoltaic absorption rate before and after 10 consecutive iterations is less than 0.001 and 0.05 respectively, or the upper limit of the number of lower-level iterations is reached. Step S5.8: If the maximum number of iterations is reached, stop searching and output the results; otherwise return to step S5.2.

Images