A power distribution network collaborative optimization scheduling method considering flexible load of electric vehicles

Abstract

The application discloses a power distribution network collaborative optimization scheduling method considering flexible load of electric vehicles, and belongs to the technical field of power system optimization scheduling; the application constructs a day-ahead-day-of-week two-stage scheduling framework of an active power distribution network for a new type of power system; in the day-ahead stage, node flexibility price is formed according to node voltage and line congestion sensitivity; master-slave game and Nash negotiation are adopted to optimize distributed power sources, interruptible loads and electric vehicle charging power; and conditional risk value is introduced to control operation risk; in the day-of-week stage, a comprehensive deviation index is formed by using power imbalance, voltage out-of-limit, line overload and charging gap, and a rolling time scale and an optimization window are adaptively selected according to the index grade; and the rolling result is used for online correction of electric vehicle adjustable capacity domain and node flexibility price; the application can effectively solve scheduling uncertainty and multi-resource coordination problems under high proportion of renewable energy and electric vehicle access, and realizes collaborative optimization of economy and voltage stability.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization and dispatching technology, and particularly relates to a distribution network collaborative optimization and dispatching method that takes into account the flexible load of electric vehicles. Background Technology

[0002] The new power system, characterized by a high proportion of renewable energy integration, widespread electrification, and intelligent interaction, serves as a crucial vehicle for achieving a low-carbon energy transition. With the rapid development of electric vehicles and distributed generation technologies, the operational patterns of distribution networks are undergoing profound changes. While the large-scale integration of electric vehicles, as a flexibly adjustable load resource, provides demand-side response potential for the system, the spatiotemporal uncertainty of their charging behavior also presents new challenges to the economic dispatch and voltage stability of the power grid.

[0003] Existing distribution network optimization and dispatching methods often treat electric vehicles as rigid loads or only perform simple peak shaving and valley filling, failing to fully utilize their flexible regulation capabilities. Furthermore, there is insufficient coordination between day-ahead dispatching and intraday rolling dispatching, and a lack of collaborative optimization mechanisms that consider the game-theoretic behavior of multiple stakeholders (distribution network operators, electric vehicle aggregators, etc.) and operational risks. In addition, traditional optimization algorithms suffer from slow convergence and susceptibility to local optima when dealing with complex dispatching models involving mixed integer variables, multiple scenario probabilities, and bidirectional feedback. Therefore, there is an urgent need for a dispatching method that efficiently handles complex constraints and mixed variables to support the reliable operation of active distribution networks under new power systems. Summary of the Invention

[0004] The purpose of this invention is to provide a distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles, aiming to construct a two-stage scheduling framework for active distribution networks oriented towards new power systems, and to achieve collaborative optimization of economy, safety and risk control.

[0005] To achieve the above objectives, the present invention provides a distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles, comprising the following steps: S1. Construct an uncertainty model of the charging behavior of electric vehicles under flexible loads. Based on the probability distribution of the charging start time and daily driving mileage of electric vehicles, use Monte Carlo sampling to generate a typical scenario set of electric vehicle charging power and determine the adjustable power range of electric vehicles in each scenario. S2. Construct the first-stage day-ahead collaborative optimization scheduling model, the model including: The node flexibility price is calculated based on the node voltage sensitivity to active and reactive power injection, the branch power flow sensitivity to node power injection, and the remaining charging demand of electric vehicles. Establish a master-slave game model between distribution network operators and electric vehicle aggregators, with distribution network operators as leaders to set day-ahead purchase and sale prices, and electric vehicle aggregators or microgrid alliances as followers to optimize electric vehicle charging power, controllable distributed power output, and interruptible load reduction. The Conditional Value at Risk (CVaR) is introduced to quantify operational risks under different scenarios, and risk costs are incorporated into the revenue function of distribution network operators to achieve synergistic optimization of revenue and risk. Within electric vehicle aggregators, a cooperative game model is built based on Nash negotiation to achieve energy interaction and fair distribution of benefits among multiple stakeholders; S3. Using the day-ahead economic scheduling results of the first stage in S2 as input, construct the second-stage intraday rolling optimization scheduling model, which includes: During daily operation, actual load, renewable energy output, node voltage, branch load rate, number of electric vehicles online and remaining charging demand are collected to construct a comprehensive deviation index that includes power imbalance, voltage safety, line current carrying capacity and charging compliance. The corresponding rolling scheduling time scale and prediction window are selected according to the level of the comprehensive deviation index. Within the selected time scale and forecast window, with the objectives of adjusting costs, voltage deviation, grid loss, wind and solar curtailment, and the amount of uncharged electric vehicles, the active and reactive power of electric vehicle clusters, energy storage, distributed power sources, and grid-side voltage regulation equipment are jointly adjusted. The actual state of charge, online connection status, node voltage margin, and line load rate of electric vehicles after rolling optimization are written back to correct the node-level adjustable capability domain and node flexibility price of electric vehicles in subsequent time periods. The adaptive alternating direction multiplier method based on voltage and power flow sensitivity partitioning is used to solve the problem and output multi-time scale collaborative scheduling instructions.

[0006] Preferably, the specific content of S1 is as follows: The daily mileage of electric vehicles follows a log-normal distribution, and its probability density function is expressed as follows: ; In the formula, L Daily mileage; and These are the mean and standard deviation of the log-normal distribution, respectively. The charging start time of electric vehicles follows a piecewise normal distribution, and its probability density function is expressed as follows: ; In the formula, This is the charging start time; and These are the mean and standard deviation of a piecewise normal distribution; Calculate the initial state of charge of the electric vehicle based on daily mileage. The expression is as follows: ; In the formula, Daily travel distance; Maximum driving range; Based on the initial state of charge of electric vehicles Calculate the required charging time The expression is as follows: ; In the formula, This refers to the battery's rated capacity. Rated charging power; For charging efficiency.

[0007] Preferably, the master-slave game model for the day-ahead scheduling phase in S2 is constructed as follows: The node flexibility price is calculated based on the sensitivity of node voltage to active and reactive power injection, the sensitivity of branch power flow to node power injection, and the remaining charging demand of electric vehicles; as shown in the following formula: ; In the formula, T This represents the total number of scheduling periods; W Number of typical scenarios; For the scene w The probability of; , Scenes w Next period t The prices of electricity sold and purchased; , Scenes w Next period t Electricity sales and purchase capacity; λ CVaR is the risk weighting coefficient used to weigh returns against risks; CVaR is the conditional value at risk. for Flexible pricing at specific time points; The objective function of distribution network operators as leaders is to maximize their own profits and control the overall risk level. ; In the formula, T This represents the total number of scheduling periods; W Number of typical scenarios; For the scene w The probability of; , Scenes w Next period t The prices of electricity sold and purchased; , Scenes w Next period t Electricity sales and purchase capacity; λ CVaR is the risk weighting coefficient used to weigh returns against risks; CVaR is the conditional value at risk. The objective function for electric vehicle aggregators or microgrid alliances as followers is to minimize operating costs. ; In the formula, N The number of aggregators or micro-networks; For the first i Internal unit operating costs of each entity; The formula for calculating the Conditional Value at Risk (CVaR) is as follows: ; in, Value at risk (VaR) represents the conditional value of risk and measures the confidence level. β Expected tail returns under the following conditions; Value at Risk (VaR) is the minimum return threshold at a certain confidence level. Confidence level; W This represents the total number of typical scenarios; For the scene W The probability of occurrence; As an auxiliary variable, it is used to linearize the CVaR calculation and represents the deviation of VaR from the scene loss; The constraints are satisfied: ; In the formula, β Confidence level; α Value at Risk (VaR); η w As an auxiliary variable; C t,w For the scene w Next period t The operator's revenue or costs; The cooperative game model based on Nash negotiation is expressed as follows: ; In the formula, The total number of participating entities; For the first i Operating costs when individual entities do not cooperate; For the first i Operating costs of cooperation among individual entities; For the first i The interaction cost of each entity; ϵ i For the first iThe contribution factors of each subject satisfy ∑ ϵ i =1; The problem is transformed into a convex optimization problem by logarithmic transformation: .

[0008] Preferably, the intraday rolling scheduling model constructed in S3 is as follows: The comprehensive deviation index Γ(t) is: ; In the formula, This is the normalized deviation between the actual net load and the planned net load. This is a normalized index for node voltage exceeding limits or minimum voltage margin. This is a normalized indicator for branch load rates exceeding the warning value. This addresses the charging fulfillment gap for the remaining connectivity time of electric vehicles. , , and These are the weight coefficients of the corresponding sub-items, and their sum is 1; when When the value is below the first threshold, the base time scale is used. When the value is between the first and second thresholds, a shortened time scale is used. An emergency time scale is adopted when the value is not lower than the second threshold.

[0009] The basic time scale, shortened time scale, and emergency time scale are 15 minutes, 5 minutes, and 1 minute, respectively, or multiples thereof. Different levels correspond to different prediction window lengths. When the comprehensive deviation index increases, the control step size is shortened and the prediction window is reduced to improve the control response speed.

[0010] Preferably, the intraday optimization scheduling divides the distribution network into multiple regions based on the sensitivity of node voltage to power injection and the coupling strength of branch power flow. Each region independently optimizes its distributed power sources, energy storage, electric vehicle clusters, and reactive power regulation equipment. The active power, reactive power, and boundary node voltage at the region boundary are used as consistency variables for alternating direction multiplier iteration. When the ratio of the original residual to the dual residual exceeds the preset range, the penalty factor is adaptively adjusted, and the optimal solution of the previous rolling window is used for hot start.

[0011] Preferably, the new power system used in the distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles includes an active distribution network; the active distribution network includes generation-grid-load-energy storage resources, and the generation-grid-load-energy storage resources include controllable distributed power sources, uncontrollable renewable energy generator sets, main grid interconnection lines, interruptible loads, electric vehicle clusters, and reactive power compensation devices.

[0012] Preferably, the multi-timescale coordinated scheduling instructions include: charging power, discharging power and reactive power instructions for electric vehicle clusters, energy storage charging and discharging power instructions, active and reactive power output instructions for controllable distributed power sources, and operation instructions for on-load tap-changing transformers, capacitor banks or static var compensators.

[0013] Therefore, the present invention employs the above-mentioned distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles, and has the following beneficial effects: This invention constructs a two-stage closed-loop feedback framework that coordinates day-ahead economic dispatch and intraday rolling optimization. It integrates flexible electric vehicle loads, distributed power sources, interruptible loads, and reactive power compensation devices into a unified optimization system. Furthermore, it introduces master-slave game theory and Nash negotiation to achieve multi-stakeholder interest coordination, balancing system operational economy, voltage stability, and risk control. This addresses the problems of separation between economic dispatch and voltage control, and insufficient multi-stakeholder coordination in traditional methods. A probabilistic model of electric vehicle charging behavior is introduced. Based on the distribution characteristics of daily mileage and charging start time, Monte Carlo sampling is used to generate typical scenario sets, accurately characterizing the uncertainty of charging load and improving the dispatching efficiency. The plan is adaptable to actual operating scenarios; it quantifies operational risks through Conditional Value at Risk (CVaR) and designs a dynamic time scale switching mechanism based on the real-time deviation index, realizing flexible connection between day-ahead planning and intraday adjustments, effectively reducing the impact of wind and solar power output and load forecast deviations; the comprehensive deviation index simultaneously reflects power, voltage, line and charging service quality, and adaptively selects the control step size and forecast window according to the operating status, reducing redundant calculations during stable operation and improving response speed when disturbances increase, accelerating convergence speed while ensuring global optimization capabilities, and is suitable for day-ahead and intraday dispatching of large-scale new power system active distribution networks.

[0014] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0015] Figure 1 This is an overall flowchart of a power distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles according to the present invention; Figure 2 This is a day-ahead predicted power output curve of wind power and photovoltaic units in an embodiment of the present invention; Figure 3 This is a diagram showing the Monte Carlo simulation results of electric vehicle charging load in an embodiment of the present invention; Figure 4 This is a graph showing the power purchased from the main grid during the day-ahead dispatch phase in an embodiment of the present invention. Figure 5 This is a diagram showing the output results of the controllable distributed power source in an embodiment of the present invention; Figure 6 This is a diagram showing the interruptible load reduction results in an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages disclosed in the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout.

[0017] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such as a process, method, system, product, or server that includes a series of steps or units, not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or device.

[0018] The following is combined with Figures 1-6 The embodiments of the present invention will be described in detail below.

[0019] Example 1: This embodiment uses an improved IEEE 33-node active distribution network test system as an example to verify the effectiveness of the proposed two-stage scheduling method. The system includes three controllable distributed power sources (including fuel cells and micro gas turbines), three interruptible loads, three capacitor banks, two wind turbine generators, one photovoltaic generator, and two electric vehicle charging stations. The system's base capacity is set at 10 MW, and the base voltage is 12.66 kV. The scheduling cycle is 24 hours, with a time step of 1 hour.

[0020] like Figure 1 The diagram shows the overall flowchart of a distribution network collaborative optimization scheduling method considering the flexible load of electric vehicles according to the present invention. This embodiment provides a distribution network collaborative optimization scheduling method considering the flexible load of electric vehicles, and the specific implementation details are as follows: Based on data from the U.S. National Household Mobility Survey, and assuming that the daily mileage of electric vehicles follows a log-normal distribution, the parameters are taken as follows: , The charging start time follows a piecewise normal distribution, and the parameters are taken as follows: , The electric vehicle has a battery capacity of 30 kWh, a charging efficiency of 0.95, a slow charging power of 3.3 kW, and a fast charging power of 28 kW. The charging load for each time period was generated using Monte Carlo simulation, and the results are as follows: Figure 3 As shown.

[0021] S1. Construct a charging behavior model for electric vehicle flexible loads, and calculate the charging power demand of electric vehicles based on the probability distribution of electric vehicle charging start time and daily driving mileage. The daily mileage of electric vehicles follows a log-normal distribution, and its probability density function is expressed as follows: ; In the formula, L Daily mileage; and These are the mean and standard deviation of the log-normal distribution, respectively. The charging start time of electric vehicles follows a piecewise normal distribution, and its probability density function is expressed as follows: ; In the formula, This is the charging start time; and These are the mean and standard deviation of a piecewise normal distribution; Calculate the initial state of charge of the electric vehicle based on daily mileage. The expression is as follows: ; In the formula, Daily travel distance; Maximum driving range; Based on the initial state of charge of electric vehicles Calculate the required charging time The expression is as follows: ; In the formula, This refers to the battery's rated capacity. Rated charging power; For charging efficiency.

[0022] S2. Construct the first-stage day-ahead collaborative optimization scheduling model, the model including: Establish a master-slave game model between distribution network operators and electric vehicle aggregators, with distribution network operators as leaders to set day-ahead purchase and sale prices, and electric vehicle aggregators or microgrid alliances as followers to optimize electric vehicle charging power, controllable distributed power output, and interruptible load reduction. The Conditional Value at Risk (CVaR) is introduced to quantify operational risks under different scenarios, and risk costs are incorporated into the revenue function of distribution network operators to achieve synergistic optimization of revenue and risk. Within electric vehicle aggregators, a cooperative game model is built based on Nash negotiation to achieve energy interaction and fair distribution of benefits among multiple stakeholders; S3. Using the day-ahead economic scheduling results of the first stage in S2 as input, construct the second-stage intraday rolling optimization scheduling model, which includes: The comprehensive deviation index Γ(t) is: ; in, This is the normalized deviation between the actual net load and the planned net load. This is a normalized index for node voltage exceeding limits or minimum voltage margin. This is a normalized indicator for branch load rates exceeding the warning value. This addresses the charging fulfillment gap for the remaining connectivity time of electric vehicles. , , and These are the weight coefficients of the corresponding sub-items, and their sum is 1; when When the value is below the first threshold, the base time scale is used. When the value is between the first and second thresholds, a shortened time scale is used. An emergency time scale is adopted when the value is not lower than the second threshold.

[0023] The basic time scale, shortened time scale, and emergency time scale are 15 minutes, 5 minutes, and 1 minute, respectively, or multiples thereof. Different levels correspond to different prediction window lengths. When the comprehensive deviation index increases, the control step size is shortened and the prediction window is reduced to improve the control response speed.

[0024] like Figure 2 The figure shows the day-ahead forecast output curves of wind and solar power units, which demonstrate the output fluctuation trends of wind and solar power over 24 hours, providing basic data for renewable energy consumption and day-ahead dispatch. like Figure 4 The figure shown is a graph of the main grid power purchase results during the day-ahead dispatch phase, reflecting the optimized adjustment scheme of the main grid power purchase at different time periods, matching the changes in load and distributed power output; like Figure 5 and Figure 6 The figure shows the output of controllable distributed power sources and the reduction of interruptible loads. It presents the output plan of each controllable distributed power source and the power reduction of interruptible loads during different time periods, verifying the effect of economic dispatch on the coordinated optimization of multiple resources.

[0025] The new power system includes an active distribution network, and the generation-grid-load-energy storage resources include controllable distributed power sources, uncontrollable renewable energy generator sets, main grid interconnection lines, interruptible loads, electric vehicle clusters, and reactive power compensation devices.

[0026] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for collaborative optimization scheduling of distribution networks considering the flexible load of electric vehicles, characterized in that, Includes the following steps: S1. Construct an uncertainty model of the charging behavior of electric vehicles under flexible loads. Based on the probability distribution of the charging start time and daily driving mileage of electric vehicles, use Monte Carlo sampling to generate a typical scenario set of electric vehicle charging power and determine the adjustable power range of electric vehicles in each scenario. S2. Construct the first-stage day-ahead collaborative optimization scheduling model, the model including: The node flexibility price is calculated based on the node voltage sensitivity to active and reactive power injection, the branch power flow sensitivity to node power injection, and the remaining charging demand of electric vehicles. Establish a master-slave game model between distribution network operators and electric vehicle aggregators, with distribution network operators as leaders to set day-ahead purchase and sale prices, and electric vehicle aggregators or microgrid alliances as followers to optimize electric vehicle charging power, controllable distributed power output, and interruptible load reduction. The Conditional Value at Risk (CVaR) is introduced to quantify operational risks under different scenarios, and risk costs are incorporated into the revenue function of distribution network operators to achieve synergistic optimization of revenue and risk. Within electric vehicle aggregators, a cooperative game model is built based on Nash negotiation to achieve energy interaction and fair distribution of benefits among multiple stakeholders; S3. Using the day-ahead economic scheduling results of the first stage in S2 as input, construct the second-stage intraday rolling optimization scheduling model, which includes: During daily operation, actual load, renewable energy output, node voltage, branch load rate, number of electric vehicles online and remaining charging demand are collected to construct a comprehensive deviation index that includes power imbalance, voltage safety, line current carrying capacity and charging compliance. The corresponding rolling scheduling time scale and prediction window are selected according to the level of the comprehensive deviation index. Within the selected time scale and forecast window, with the objectives of adjusting costs, voltage deviation, grid loss, wind and solar curtailment, and the amount of uncharged electric vehicles, the active and reactive power of electric vehicle clusters, energy storage, distributed power sources, and grid-side voltage regulation equipment are jointly adjusted. The actual state of charge, online connection status, node voltage margin, and line load rate of electric vehicles after rolling optimization are written back to correct the node-level adjustable capability domain and node flexibility price of electric vehicles in subsequent time periods. The adaptive alternating direction multiplier method based on voltage and power flow sensitivity partitioning is used to solve the problem and output multi-time scale collaborative scheduling instructions.

2. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The specific content of S1 is as follows: The daily mileage of electric vehicles follows a log-normal distribution, and its probability density function is expressed as follows: ； In the formula, L Daily mileage; and These are the mean and standard deviation of the log-normal distribution, respectively. The charging start time of electric vehicles follows a piecewise normal distribution, and its probability density function is expressed as follows: ； In the formula, This is the charging start time; and These are the mean and standard deviation of a piecewise normal distribution; Calculate the initial state of charge of the electric vehicle based on daily mileage. The expression is as follows: ； In the formula, Daily driving distance; Maximum driving range; Based on the initial state of charge of electric vehicles Calculate the required charging time The expression is as follows: ； In the formula, This refers to the battery's rated capacity. Rated charging power; For charging efficiency.

3. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The node flexibility price in step S2 is determined according to the following formula: ； In the formula, Based on the price of electricity, The voltage scarcity coefficient is formed by the node voltage margin and voltage sensitivity. The congestion coefficient is determined by the remaining current-carrying capacity and power flow sensitivity of the branch. The fulfillment coefficient formed for the remaining charging demand of electric vehicles. , and These are the weighting coefficients. for Flexible pricing at specific time points.

4. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The master-slave game model for the day-ahead scheduling phase in S2 is constructed as follows: The objective function of distribution network operators as leaders is to maximize their own profits and control the overall risk level. ； In the formula, T This represents the total number of scheduling periods; W Number of typical scenarios; For the scene w The probability of; , Scenes w Next period t The prices of electricity sold and purchased; , Scenes w Next period t Electricity sales and purchase capacity; λ CVaR is the risk weighting coefficient used to weigh returns against risks; CVaR is the conditional value at risk. The objective function for electric vehicle aggregators or microgrid alliances as followers is to minimize operating costs. ； In the formula, N The number of aggregators or micro-networks; For the first i Internal unit operating costs of each entity; The formula for calculating the Conditional Value at Risk (CVaR) is as follows: ； in, Value at risk (VaR) represents the conditional value of risk and measures the confidence level. β Expected tail returns under the following conditions; Value at Risk (VaR) is the minimum return threshold at a certain confidence level. Confidence level; W This represents the total number of typical scenarios; For the scene W The probability of occurrence; As an auxiliary variable, it is used to linearize the CVaR calculation and represents the deviation of VaR from the scene loss; The constraints are satisfied: ； In the formula, β Confidence level; α Value at Risk (VaR); η w As an auxiliary variable; C t,w For the scene w Next period t The operator's revenue or costs; The cooperative game model based on Nash negotiation is expressed as follows: ； In the formula, The total number of participating entities; For the first i Operating costs when individual entities do not cooperate; For the first i Operating costs of cooperation among individual entities; For the first i The interaction cost of each entity; ϵ i For the first i The contribution factors of each subject satisfy ∑ ϵ i =1; The problem is transformed into a convex optimization problem by logarithmic transformation: 。 5. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The intraday rolling scheduling model in S3 is as follows: The comprehensive deviation index Γ(t) is: ； In the formula, This is the normalized deviation between the actual net load and the planned net load. This is a normalized index for node voltage exceeding limits or minimum voltage margin. This is a normalized indicator for branch load rates exceeding the warning value. This addresses the charging fulfillment gap for the remaining connectivity time of electric vehicles. , , and These are the weight coefficients of the corresponding sub-items, and their sum is 1; when When the value is below the first threshold, the base time scale is used. When the value is between the first and second thresholds, a shortened time scale is used. An emergency time scale is adopted when the value is not lower than the second threshold.

6. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The basic time scale, shortened time scale, and emergency time scale are one of 15 minutes, 5 minutes, and 1 minute, or integer multiples thereof. Different levels correspond to different prediction window lengths. When the comprehensive deviation index increases, the control step size is shortened and the prediction window is reduced to improve the control response speed.

7. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, In step S3, the distribution network is divided into multiple regions according to the sensitivity of node voltage to power injection and the coupling strength of branch power flow. Each region independently optimizes its distributed power sources, energy storage, electric vehicle clusters and reactive power regulation equipment, and performs alternating direction multiplier iteration with active power, reactive power and boundary node voltage at the region boundary as consistency variables. When the ratio of the original residual to the dual residual exceeds the preset range, the penalty factor is adaptively adjusted and the optimal solution of the previous rolling window is used for hot start.

8. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The new power system applied by the distribution network collaborative optimization scheduling method that takes into account the flexible load of electric vehicles includes an active distribution network; the active distribution network includes generation-grid-load-energy storage resources, and the generation-grid-load-energy storage resources include controllable distributed power sources, uncontrollable renewable energy generator sets, main grid interconnection lines, interruptible loads, electric vehicle clusters, and reactive power compensation devices.

9. The method for coordinated optimization scheduling of distribution networks considering the flexible load of electric vehicles according to claim 1, characterized in that, The multi-timescale coordinated scheduling instructions include: charging power, discharging power, and reactive power instructions for electric vehicle clusters; energy storage charging and discharging power instructions; active and reactive power output instructions for controllable distributed power sources; and operation instructions for on-load tap-changing transformers, capacitor banks, or static var compensators.

Images