Highway service area power dynamic expansion method considering electric vehicle charging load

Abstract

The application provides a highway service area power dynamic expansion method considering electric vehicle charging load, which is divided into two stages of long-term and short-term scheduling optimization. The long-term scheduling optimization stage is based on the long-term prediction results of wind and light power generation, daily load and electric vehicle charging load power in the service area micro-grid, and adds a target function of minimum grid-connected fluctuation on the basis of the economy and low-carbon performance targets, preliminarily formulates a long-term scheduling optimization plan, and obtains the power change between the service area and the power grid. The short-term scheduling optimization stage is based on the long-term result grid-connected power change and the short-term prediction results of each unit in the system, and carries out a short-term scheduling optimization plan with the optimal operation economy and the minimum grid-connected stability as the targets.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging load prediction, in particular to a method for dynamically expanding the power capacity of a highway service area taking into account the electric vehicle charging load. Background Art

[0002] Under the "dual carbon" goals, energy and transportation networks across the country shoulder the crucial task of achieving a green and low-carbon transition. Currently, the vast majority of highway service areas are located in remote locations with small and dispersed loads, creating significant potential for the development of renewable energy resources such as wind and solar. Furthermore, the integration of a high proportion of high-power-density electric vehicle charging loads can impact the stable operation of the service area's power grid system. Measures such as installing power grids over long distances or expanding transformer capacity increase the cost of supplying power to highway loads, making operation and maintenance difficult and cost-effective. Utilizing unused land resources such as building rooftops and parking sheds to connect to renewable energy sources such as wind and solar, combined with energy storage systems and V2G technology for efficient energy management, can facilitate the local consumption of new energy, reduce carbon emissions, and improve energy efficiency. It also facilitates the dynamic expansion of service area power grids, mitigates the impact of high-power charging stations on the regional grid, and accelerates the electrification of transportation.

[0003] Existing technologies and problems:

[0004] 1. Most highway service areas are located in remote areas. Connecting large numbers of electric vehicles to these areas would exceed the originally designed regional power distribution capacity. This issue of insufficient power distribution capacity in service areas can be addressed by fully utilizing unused land resources, such as building rooftops, parking lots, or charging station sheds, to construct distributed renewable energy facilities, such as wind and photovoltaic power generation. This prioritizes local consumption of renewable energy, storing and utilizing excess energy, and using the grid as a backup energy source.

[0005] 2. Existing methods use historical data and statistical models to infer future charging load demand, ignoring the analysis of charging demand and characteristics in different scenarios. These predictions are only of reference value. In practice, electric vehicle users have varying preferences for distance, time, and cost, creating uncertainty and requiring precise energy scheduling optimization strategies. In highway scenarios, optimizing time-of-use charging prices for electric vehicles can guide charging behavior, simultaneously achieving the goals of reducing carbon emissions from electric vehicle operation and maximizing user energy efficiency. These methods for guiding electric vehicle charging behavior in highway service areas still suffer from the randomness of user preferences.

[0006] 3. Currently, energy scheduling optimization methods for service area microgrids fail to account for the high proportion and high power density of electric vehicle charging loads, and instead optimize scheduling solely based on objective functions such as economic efficiency and low-carbon performance. These scheduling methods are ill-suited to the current scenario of large-scale electric vehicle access to service area loads, which results in increased total load and significantly increased randomness in the service area. Utilizing battery storage and other energy storage devices within the service area to dynamically expand the service area grid would improve the charging capacity and grid-friendliness of highway service areas.

[0007] 4. When the NSGA multi-objective optimization algorithm is used to solve the multi-objective energy scheduling optimization model of a grid-connected microgrid containing solar-storage-hybrid-charging, there is a problem of difficulty in finding a feasible solution due to conflicts between objectives and complex decision variables. Summary of the Invention

[0008] In view of this, the purpose of the present invention is to provide a method for dynamically expanding the power capacity of highway service areas taking into account the charging load of electric vehicles, so as to solve the problem that the original service area exceeds the rated power distribution capacity due to the access of electric vehicle charging load.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: a method for dynamically expanding the power capacity of a highway service area taking into account the charging load of electric vehicles, comprising the following steps:

[0010] Step 1: Identify service areas where dynamic power capacity expansion is required due to planned installation of additional electric vehicle charging facilities;

[0011] Step 2: Design a wind-solar-hybrid storage-charging integrated microgrid based on the existing remaining idle land resources in the service area, load, planned charging piles, and the scale of substations. Determine the scale of wind and solar distributed renewable energy, the capacity allocation of hybrid energy storage with medium and high power density and energy density, and the total energy.

[0012] Step 3: Establish a multi-objective optimization model for dual-time-scale energy scheduling that takes into account electric vehicle charging load service areas and friendly grid connection;

[0013] Step 4: Determine the multi-objective optimization algorithm to solve the multi-objective and multi-constraint microgrid energy dispatch optimization model of wind-solar-hybrid storage-charging and determine the optimization solution algorithm;

[0014] Step 5: Obtain long-term forecast results for renewable energy output, daily load, and electric vehicle charging load within the service area;

[0015] Step 6: Determine the long-term dispatch optimization plan model based on the operation economy, low carbon and grid-connected volatility, and pass the dispatch plan;

[0016] Step 7: Use multi-objective optimization algorithm to solve the long-term scheduling optimization model;

[0017] Step 8: Select a suitable solution and obtain the long-term scheduling optimization plan;

[0018] Step 9: Obtain short-term forecast results for renewable energy output, daily load, and electric vehicle charging load within the service area;

[0019] Step 10: Determine the short-term dispatch optimization planning model based on the objectives of economic operation, low carbon performance, and grid connection volatility;

[0020] Step 11: Use multi-objective optimization algorithm to solve the long-term scheduling optimization model;

[0021] Step 12: Select a suitable solution to obtain the short-term scheduling optimization plan, which is the final service area microgrid energy scheduling optimization plan to be executed;

[0022] Step 13: Complete the final dispatch strategy and realize dynamic expansion of power capacity in the service area.

[0023] In a preferred embodiment, step 3 includes the following steps:

[0024] Step 31: Develop a long-term scheduling optimization plan;

[0025] Step 32: Develop a short-term scheduling optimization plan;

[0026] Step 33: Establish the objective function;

[0027] Step 34: Constraints.

[0028] In a preferred embodiment, step 31 specifically includes: long-term prediction of the system's wind and solar power output, daily load and electric vehicle charging load in the future, preliminary formulation of the system's long-term scheduling optimization plan, and obtaining the energy interaction plan between the microgrid and the AC distribution network in advance to reduce the impact of the microgrid on the large power grid.

[0029] In a preferred embodiment, the objective function in step 33 is:

[0030]

[0031] Where f OE f CL f G&MG They represent the objective function values ​​of optimal operation economy, minimum carbon emission and minimum grid connection volatility respectively; C MG_load It represents the benefits of directly utilizing distributed renewable energy sources within the microgrid, in CNY; C S C Ab Represent the operating cost of distributed renewable energy and the penalty for wind and solar curtailment respectively; C G&MG CServe They represent the electricity price transaction and service cost from the microgrid to the AC distribution network respectively; C ES represents the operation and maintenance cost of the hybrid energy storage equipment; Represents the operation and maintenance cost of the charging pile.

[0032] In a preferred embodiment, the constraints in step 34 are:

[0033] The electric power balance constraints are as follows:

[0034]

[0035] Where, Indicates the output of the energy supply side of the system at time t, in kW; represents the grid-connected power of the microgrid; Indicates energy storage output; Indicates the output of the charging pile; Indicates the load power change;

[0036] The output power of new energy cannot exceed the upper and lower power limits:

[0037]

[0038] Where, P PV_max 、P PV_min 、P WT_max 、P WT_min Respectively represent the upper and lower output constraints of wind and solar power generation units, in kW;

[0039] The grid-connected interaction constraints are as follows:

[0040]

[0041] Where, P G&MG_max and P G&MG_min Represents the upper and lower limit constraints of the interaction power between the microgrid and the grid side, unit kW;

[0042] The constraints of hybrid energy storage are as follows:

[0043]

[0044] In the formula Indicates the amount of energy that can be transferred at time t; SOC ESP_max , SOC ESP_min , SOC ESE_max and SOC ESE_min Represents medium and high power density of hybrid energy storage and SOC upper and lower limits of high energy density; P ESP_max , P ESP_min, P ESE_max and P ESE_min Hybrid energy storage with medium and high power density and high energy density Upper and lower limits, unit kW; u ESP_C and u ESE_C Indicates the number of times the charge and discharge state is switched, u ESP_C_max ,u ESP_C_min ,u ESE_C_max and u ESE_C_min Respectively represent the upper and lower limits of the number of times the charge and discharge states are switched;

[0045] Under the constraints of charging piles:

[0046]

[0047] In the formula represents the SOC of the electric vehicle at time t; SOC Car_max The SOC at the end of charging the electric vehicle; represents the charging power of the electric vehicle at time t; P Car_min P Car_max They represent the upper and lower limits of the charging power of electric vehicles respectively.

[0048] In a preferred embodiment, the step 4 specifically comprises: utilizing the RL TS-NSGA-II multi-objective optimization algorithm, that is, utilizing a method based on reverse learning and two-step solution to optimize and solve;

[0049] The generation of the initial population individuals of the multi-objective optimization algorithm is random within a given range, so the initial individuals are random. Improving the initial population through the reverse learning method can speed up the search for feasible solutions, as shown in formula (7):

[0050]

[0051] Where P 0 P 0 ' are the two initial populations in reverse learning, P U P L are the upper and lower limits of the population within the constraint conditions, and d represents the number of decision variables. After converting the multi-objective algorithm into a single-objective with constraints, reverse learning is performed to compare the fitness function values ​​of the two populations before the optimization begins, and the better population is retained as the initial population in the single-objective optimization process. Then, the differential evolution algorithm is used for rapid global search to find the feasible solution space, as shown in formula (8):

[0052]

[0053] Where F is a scaling factor parameter in the DE variation process, x i represents the i-th individual, xbest Indicates the individual with the smallest transformed objective function value in the current population, x r represents the randomly generated individuals in the mutation process, v ri and v bi They represent the two individual mutation directions of individuals towards randomness and the optimal individual in the current population respectively; finally, when the feasible solution area in the population exceeds k, the population under this single-objective search is retained and used as the initial population in the subsequent NSGA-II multi-objective optimization solution process, and the solution is continued to obtain the final solution.

[0054] Compared with the existing technology, the present invention has the following beneficial effects: in view of the strong randomness and large power fluctuation range of electric vehicle charging load in highway service areas, the problem of the original service area exceeding the rated power distribution capacity due to the access of electric vehicle charging load is solved by high-power, energy-density hybrid energy storage equipment and microgrid energy scheduling optimization plan. A wind-solar-hybrid storage-charging grid-connected microgrid and reasonable objective functions and constraints are designed. The multi-objective optimization solution algorithm of process optimization is used to formulate a grid-friendly dual-time-scale coordinated microgrid energy scheduling plan. The dual-time-scale optimization scheduling adopts the idea of ​​multi-level coordination and step-by-step refinement. First, the overall operation strategy of the microgrid and the AC distribution network is determined by formulating a long-term scheduling optimization plan. Then, the day-ahead plan is revised according to the short-term scheduling optimization to formulate an energy scheduling plan that is optimized for the microgrid operation economy, the lowest carbon emissions, and the lowest grid-connection volatility. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The topology of the highway service area microgrid system of the preferred embodiment of the present invention is as follows;

[0056] Figure 2 This is a flow chart of energy dispatch optimization of a microgrid in a service area according to a preferred embodiment of the present invention;

[0057] Figure 3 This is a flowchart of a multi-objective optimization algorithm solution according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0059] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0060] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application; as used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form, and it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or their combinations.

[0061] refer to Figure 1-3 Aiming at the existing highway service areas with electric vehicle charging needs, the present invention proposes an energy dispatch optimization method for a wind-solar-hybrid storage-charging integrated grid-connected microgrid for expanding the power capacity of the service area and considering friendly grid connection.

[0062] In view of the fact that existing electric vehicle charging load forecasting methods ignore the randomness of charging demand in different scenarios, and the service area fails to consider the impact of the high randomness of electric vehicle charging load on the grid stability of the microgrid, the present invention proposes a grid-friendly dual-time scale energy scheduling optimization strategy, which is divided into two stages: long-term and short-term scheduling optimization. The long-term scheduling optimization stage is based on the long-term forecast results of wind-solar power generation, daily load and electric vehicle charging load power in the service area microgrid. On the basis of the goals of economy and low-carbon performance, the objective function of minimizing grid fluctuation is added, and a preliminary long-term scheduling optimization plan is formulated to obtain the power changes between the service area and the power grid. The short-term scheduling optimization stage is based on the long-term results of grid power changes and the short-term forecast results of each unit in the system. The short-term scheduling optimization plan is carried out with the goals of optimal operation economy and minimum grid stability.

[0063] The number of electric vehicle users is increasing year by year, and the demand for charging electric vehicle clusters is growing. In particular, the vast majority of charging in highway service areas is fast charging. Fast charging demand is more random than slow charging, and the power fluctuation range is wider (the average power of highway charging piles is 90kW), and the load in the service area fluctuates widely. In response to the problem of insufficient service area power distribution capacity caused by this situation, the present invention proposes a service area microgrid system with wind-solar-hybrid storage-charging integration containing high power density and high energy density hybrid energy storage. Combined with the above-mentioned grid-connected dual-time scale energy optimization scheduling strategy, hybrid energy storage is used to dynamically expand the service area power capacity, saving the expansion cost of the original power distribution.

[0064] To address the slow or difficult solution problems of the NSGA-type multi-objective optimization algorithm, the multi-objective model is first converted into a constrained single-objective model; the reverse learning method and the differential evolution algorithm with global fast search characteristics are used to quickly find a better initial population and some feasible solution regions; finally, these feasible solution regions are used as the initial population of the multi-objective optimization algorithm to continue solving the problem, shortening the solution time.

[0065] The present invention enables the service area distribution network system to operate at the original power transformation and distribution design capacity, meeting the charging needs of electric vehicles on highways. It also improves the economy, environmental protection, reliability and network connection of the service area microgrid, providing a reference for the transformation of new power systems in highway service areas.

[0066] The specific steps include:

[0067] Step 1: Identify service areas where dynamic power capacity expansion is required due to plans to add electric vehicle charging facilities.

[0068] Step 2: Design a wind-solar-hybrid storage-charging integrated microgrid based on the service area's existing idle land resources, load, planned charging stations, and substation size. Determine the scale of wind and solar distributed renewable energy sources, the capacity allocation for hybrid storage with medium and high power density and energy density, and the total energy capacity.

[0069] Step 3: Establish a multi-objective optimization model for dual-time-scale energy scheduling that takes into account electric vehicle charging load service areas and friendly grid connection.

[0070] Step 31: Long-term scheduling optimization plan:

[0071] Highway service areas are often located in remote areas, where grid transmission and distribution capacity is insufficient to support large-scale electric vehicle loads connected simultaneously, with widely fluctuating loads. Furthermore, given the microgrid's "self-generation for self-use, with surplus power connected to the grid," it's important to plan energy exchange between the microgrid and the AC distribution network in advance. This helps reduce grid-connected volatility during actual operation of the service area microgrid. Long-term forecasts of wind and solar power output, daily loads, and electric vehicle charging loads are developed over the long term to initially formulate a long-term scheduling optimization plan for the system. This allows for a pre-planned energy exchange plan between the microgrid and the AC distribution network, minimizing the impact of the microgrid on the larger grid.

[0072] Step 32: Short-term scheduling optimization plan:

[0073] Long-term scheduling optimization plans are subject to environmental fluctuations, discrepancies between environmental information and actual output, and inaccurate forecasts of wind and photovoltaic power output. Furthermore, the randomness of electric vehicle charging loads can lead to discrepancies between long-term power forecasts and actual conditions. Therefore, the system's long-term scheduling optimization plan is not applicable to the actual conditions on the day of optimization and can only be used as a reference for final decision-making. To further reduce microgrid operating costs, reduce carbon emissions, and achieve more grid-friendly integration, short-term scheduling optimization plans are necessary.

[0074] Step 33: Objective Function:

[0075]

[0076] Where f OE f CL f G&MG They represent the objective function values ​​of optimal operation economy, minimum carbon emission and minimum grid connection volatility respectively; C MG_load It represents the benefits of directly utilizing distributed renewable energy sources within the microgrid, in CNY; C S C Ab Represent the operating cost of distributed renewable energy and the penalty for wind and solar curtailment respectively; C G&MG C Serve They represent the electricity price transaction and service cost from the microgrid to the AC distribution network respectively; C ES represents the operation and maintenance cost of the hybrid energy storage equipment; Represents the operation and maintenance cost of the charging pile.

[0077] Step 34: Constraints:

[0078] The electric power balance constraints are as follows:

[0079]

[0080] Where, Indicates the output of the energy supply side of the system at time t, in kW; represents the grid-connected power of the microgrid; Indicates energy storage output; Indicates the output of the charging pile; Indicates the load power change.

[0081] The output power of renewable energy sources such as wind and solar power cannot exceed the upper and lower power limits:

[0082]

[0083] Where, P PV_max 、P PV_min 、P WT_max 、P WT_min Respectively represent the upper and lower output constraints of wind and solar power generation units, in kW.

[0084] The grid-connected interaction constraints are as follows:

[0085]

[0086] Where, P G&MG_max and P G&MG_min Represents the upper and lower limit constraints of the interactive power between the microgrid and the grid side.

[0087] kW. The hybrid energy storage constraints are as follows:

[0088]

[0089] In the formula Indicates the amount of energy that can be transferred at time t; SOC ESP_max , SOC ESP_min , SOC ESE_max and SOC ESE_min Represents medium and high power density of hybrid energy storage and SOC upper and lower limits of high energy density; P ESP_max , P ESP_min , P ESE_max and P ESE_min Hybrid energy storage with medium and high power density and high energy density Upper and lower limits, unit kW; u ESP_C and u ESE_C Indicates the number of times the charge and discharge state is switched, u ESP_C_max ,u ESP_C_min ,u ESE_C_max and u ESE_C_min They represent the upper and lower limits of the number of times the charge and discharge states can be switched.

[0090] Under the constraints of charging piles:

[0091]

[0092] In the formula represents the SOC of the electric vehicle at time t; SOC Car_max The SOC at the end of charging the electric vehicle; represents the charging power of the electric vehicle at time t; P Car_min P Car_max They represent the upper and lower limits of the charging power of electric vehicles respectively.

[0093] Step 4: Determine the multi-objective optimization algorithm to solve the multi-objective and multi-constraint microgrid energy dispatch optimization model of wind-solar-hybrid storage-charging and determine the optimization solution algorithm.

[0094] A RL multi-objective optimization algorithm TS-NSGA-II (Based On Reverse Learning Non-dominated Sorting Genetic Algorithm) is proposed, which uses a method based on reverse learning and two-step solution to optimize the solution.

[0095] The generation of the initial population individuals of the multi-objective optimization algorithm is random within a given range, so the initial individuals are random. Improving the initial population through the reverse learning method can speed up the search for feasible solutions, as shown in formula (7):

[0096]

[0097] Where P 0 P 0 ' are the two initial populations in reverse learning, P U P L are the upper and lower limits of the population within the constraints, and d represents the number of decision variables. After converting the multi-objective algorithm into a single-objective with constraints, reverse learning is used to compare the fitness function values ​​of the two populations before the optimization begins, and the better population is retained as the initial population in the single-objective optimization process; then, a differential evolution algorithm is used to quickly search globally to find a feasible solution space, as shown in Equation (8):

[0098]

[0099] Where F is a scaling factor parameter in the DE variation process, x i represents the i-th individual, x best Indicates the individual with the smallest transformed objective function value in the current population, x r represents the randomly generated individuals in the mutation process, v ri and v bi They represent the two individual mutation directions of individuals towards randomness and the optimal individual in the current population. Finally, when the feasible solution area in the population exceeds k (0.3 is used as an example in this invention), the population under this single-objective search is retained and used as the initial population in the subsequent NSGA-II multi-objective optimization solution process. The solution is continued to obtain the final solution.

[0100] Step 5: Obtain long-term forecast results for renewable energy output, daily load, and electric vehicle charging load within the service area. The data has a 1-hour resolution.

[0101] Step 6: Determine the long-term scheduling optimization plan model with the goals of operating economy, low carbon, and grid-connected volatility. Through the scheduling plan, the wind-solar-hybrid storage-charging integrated grid-connected microgrid can achieve the best operating economy while maintaining low carbon emissions and minimizing the volatility of the grid-connected net power curve, thereby reducing the impact of the microgrid on the large power grid.

[0102] Step 7: Use multi-objective optimization algorithm to solve the long-term scheduling optimization model.

[0103] Step 8: Select a suitable solution and obtain the long-term scheduling optimization plan.

[0104] The solution obtained by the multi-objective optimization algorithm is a Pareto solution, which is an optimal solution set that meets the three objectives of economic optimization, low carbon emission, and minimum volatility. The dispatching center needs to select an optimal solution (this invention takes the solution with the highest economic benefit as an example) as the final execution scheduling plan.

[0105] Step 9: Obtain short-term forecasts for renewable energy output, daily load, and electric vehicle charging load within the service area. The data is at a 15-minute resolution. The day-ahead forecast is generated every 4 hours, and the dispatch strategy is updated every 15 minutes.

[0106] Step 10: Determine the short-term dispatch optimization plan model with the goals of economic operation, low carbon emission and grid-connected volatility. On the basis of the long-term dispatch optimization plan, the results of grid-side dispatch optimization are used as constraints to reduce the volatility of grid connection. Similarly, the short-term dispatch optimization plan model is used with the goals of optimal microgrid operation, low carbon emission and minimum volatility.

[0107] Step 11: Use multi-objective optimization algorithm to solve the long-term scheduling optimization model.

[0108] Step 12: Select a suitable solution to obtain the short-term scheduling optimization plan, which is the final service area microgrid energy scheduling optimization plan to be executed.

[0109] Step 13: Complete the final dispatch strategy to achieve dynamic power expansion in the service area. Upload it to the service network's integrated wind-solar-hybrid storage-charging microgrid operation and management platform, and update the dispatch plan every 15 minutes.

Claims

1. A method for dynamically expanding power capacity in highway service areas taking into account electric vehicle charging loads, characterized in that: The following steps are involved: Step 1: Identify service areas where dynamic power capacity expansion is required due to planned installation of additional electric vehicle charging facilities; Step 2: Design a wind-solar-hybrid storage-charging integrated microgrid based on the existing remaining idle land resources in the service area, load, planned charging piles, and the scale of substations. Determine the scale of wind and solar distributed renewable energy, the capacity allocation of hybrid energy storage with medium and high power density and energy density, and the total energy. Step 3: Establish a multi-objective optimization model for dual-time-scale energy scheduling that takes into account electric vehicle charging load service areas and friendly grid connection; Step 4: Determine a multi-objective optimization algorithm to solve the multi-objective and multi-constraint microgrid energy dispatch optimization model of wind-solar-hybrid storage-charging; Step 5: Obtain long-term forecast results for renewable energy output, daily load, and electric vehicle charging load within the service area; Step 6: Determine the long-term dispatch optimization planning model with the objectives of economic operation, low carbon performance, and grid connection volatility; Step 7: Use multi-objective optimization algorithm to solve the long-term scheduling optimization plan model; Step 8: Select a suitable solution and obtain the long-term scheduling optimization plan; Step 9: Obtain short-term forecast results for renewable energy output, daily load, and electric vehicle charging load within the service area; Step 10: Determine the short-term dispatch optimization planning model based on the objectives of economic operation, low carbon performance, and grid connection volatility; Step 11: Use multi-objective optimization algorithm to solve the short-term scheduling optimization plan model; Step 12: Select a suitable solution to obtain the short-term scheduling optimization plan, which is the final service area microgrid energy scheduling optimization plan to be executed; Step 13: Complete the final dispatch strategy and realize dynamic power expansion in the service area; The step 3 comprises the following steps: Step 31: Develop a long-term scheduling optimization plan; Step 32: Develop a short-term scheduling optimization plan; Step 33: Establish the objective function; Step 34: Establish constraints; Step 31 specifically includes: long-term forecasting of the system's wind and solar output, daily load, and electric vehicle charging load over a period of time in the future, preliminary formulation of a long-term scheduling optimization plan for the system, and obtaining an energy interaction plan between the microgrid and the AC distribution network in advance to reduce the impact of the microgrid on the main grid; The objective function in step 33 is: (1) In the formula 、 、 They represent the objective function values ​​of optimal operation economy, minimum carbon emission, and minimum grid connection volatility respectively; It represents the benefits generated by the direct use of distributed renewable energy by the loads in the microgrid; 、 denote the operating costs of distributed renewable energy and the wind and solar curtailment penalties, respectively; They represent the electricity price transaction and service cost from the microgrid to the AC distribution network respectively; represents the operation and maintenance cost of the hybrid energy storage equipment; Indicates the operation and maintenance cost of the charging pile; represents the grid-connected power of the microgrid; t represents the time; The step 4 specifically comprises: using the RL TS-NSGA-II multi-objective optimization algorithm, that is, optimizing and solving the problem using a method based on reverse learning and two-step solution; The generation of the initial population individuals of the multi-objective optimization algorithm is random within a given range, so the initial individuals are random. The search speed of feasible solutions is accelerated by improving the initialization population through the reverse learning method, as shown in formula (7): (7) In the formula 、 are the two initial populations in reverse learning, 、 are the upper and lower limits of the population within the constraint conditions, and d represents the number of decision variables. After converting the multi-objective optimization algorithm into a single-objective with constraints, reverse learning is performed to compare the fitness function values ​​of the two populations before the optimization begins, and the better population is retained as the initial population in the single-objective optimization process. Then, the differential evolution algorithm is used to quickly search globally to find the feasible solution space, as shown in formula (8): (8) Where, is a scaling factor parameter in the DE mutation process, represents the i-th individual, represents the individual with the smallest transformed objective function value in the current population, represents individuals randomly generated during the mutation process, and They represent the two individual mutation directions of individuals towards randomness and the optimal individual in the current population respectively; finally, when the feasible solution area in the population exceeds k, the population under this single-objective search is retained and used as the initial population in the subsequent NSGA-II multi-objective optimization solution process, and the solution is continued to obtain the final solution.

2. The method for dynamically expanding power capacity in highway service areas taking into account electric vehicle charging loads according to claim 1, characterized in that: The constraints in step 34 are: The electric power balance constraints are as follows: (2) Where, Indicates the output of the energy supply side in the system at time t; represents the grid-connected power of the microgrid; Indicates energy storage output; Indicates the output of the charging pile; Indicates the load power change; The output power of new energy cannot exceed the upper and lower power limits: (3) Where, 、 、 、 Respectively represent the upper and lower limits of the output of the solar and wind power generation units; The grid-connected interaction constraints are as follows: (4) Where, and Indicates the upper and lower limits of the interactive power between the microgrid and the grid; The constraints of hybrid energy storage are as follows: (5) In the formula and They represent the transferable energy of high energy density and high power density energy storage at time t respectively; , , and Represents medium and high power density of hybrid energy storage and high energy density SOC upper and lower limits; , , and Hybrid energy storage with medium and high power density and high energy density Upper and lower limits; and Indicates the number of times the charge and discharge states are switched. , , and Respectively represent the upper and lower limits of the number of times the charge and discharge states are switched; Under the constraints of charging piles: (6) In the formula represents the SOC of the electric vehicle at time t; The SOC at the end of charging the electric vehicle; represents the charging power of the electric vehicle at time t; 、 They represent the upper and lower limits of the charging power of electric vehicles respectively.

Images