A dynamic scheduling method for intelligent charging and swapping networks

Abstract

This invention discloses a dynamic scheduling method for intelligent charging and swapping networks, relating to the field of charging and swapping data processing technology. The method includes: identifying multiple charging and swapping stations in a target area to construct a charging and swapping network; collecting data from the target area to generate a historical operation dataset; determining charging and swapping demand information, predicting the operation of the charging and swapping network, and determining grid load prediction data; performing balance analysis on the charging and swapping network, formulating multiple initial scheduling schemes, and constructing an initial population; simulating scheduling by traversing multiple initial scheduling schemes, performing dynamic optimization, and generating an optimized population; and adaptively and dynamically scheduling the charging and swapping network according to the optimized population. This invention solves the technical problems of poor adaptive adjustment and low operating efficiency and stability of charging and swapping networks in existing technologies, achieving adaptive dynamic scheduling of the charging and swapping network and improving its operating efficiency and stability.

Description

Technical Field

[0001] This invention relates to the field of charging and swapping data processing technology, and specifically to a dynamic scheduling method for intelligent charging and swapping networks. Background Technology

[0002] With the rapid development of the electric vehicle market, the scale of charging and battery swapping networks is constantly expanding, posing greater challenges to their operational efficiency and stability. Currently, numerous problems exist within these networks that urgently need to be addressed. On the one hand, low charging efficiency is a prominent issue. Due to the lack of effective scheduling strategies, charging and battery swapping stations may experience congestion at certain times, leading to excessively long waiting times for users, while resources may remain idle at other times, resulting in resource waste. On the other hand, grid load imbalance also poses a threat to the stable operation of charging and battery swapping networks. Inappropriate charging schedules may cause the grid to bear excessive load at certain times, affecting its safe operation and increasing its operating costs. Furthermore, traditional charging and battery swapping network scheduling methods are typically rigid and unable to dynamically adjust according to real-time demand changes and grid conditions, making it difficult to meet the increasingly diverse needs of users and fully realize the potential of charging and battery swapping networks.

[0003] Existing technologies suffer from poor adaptive adjustment of charging and swapping networks, resulting in low operating efficiency and stability. Summary of the Invention

[0004] This application provides a dynamic scheduling method for intelligent charging and swapping networks, which addresses the technical problems of poor adaptive adjustment, low charging and swapping operation efficiency and stability in existing technologies.

[0005] In view of the above problems, this application provides a dynamic scheduling method for an intelligent charging and swapping network. The method includes: traversing and identifying multiple charging and swapping stations in a target area to construct a charging and swapping network; collecting data from the target area based on the charging and swapping network to generate a historical operation dataset; performing demand analysis based on the historical operation dataset to determine charging and swapping demand information; performing operation prediction on the charging and swapping network according to the charging and swapping demand information to determine grid load prediction data; performing balance analysis on the charging and swapping network based on the grid load prediction data to generate a grid balance coefficient; formulating multiple initial scheduling schemes according to the grid balance coefficient to construct an initial population; traversing the multiple initial scheduling schemes to perform simulated scheduling, evaluating the initial population, dynamically optimizing the initial population based on the evaluation results to generate an optimized population; and adaptively and dynamically scheduling the charging and swapping network according to the optimized population.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] Multiple charging and battery swapping stations in the target area are identified to construct a charging and battery swapping network. Data is collected from the target area based on this network to generate a historical operational dataset. Demand analysis is performed on this historical dataset to determine charging and battery swapping demand information, and operational forecasts are made for the charging and battery swapping network to determine grid load forecast data. Based on the grid load forecast data, a balance analysis is performed on the charging and battery swapping network to generate a grid equilibrium coefficient. Multiple initial scheduling schemes are formulated to construct an initial population. Simulated scheduling is performed on these initial scheduling schemes to evaluate the initial population. The initial population is then dynamically optimized to generate an optimized population. Adaptive dynamic scheduling of the charging and battery swapping network is performed according to the optimized population. This achieves adaptive dynamic scheduling of the charging and battery swapping network, thereby improving its operational efficiency and stability. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A flowchart illustrating a dynamic scheduling method for an intelligent charging and swapping network provided in an embodiment of this application;

[0010] Figure 2 This is a schematic diagram illustrating the process of generating a power grid balance coefficient in a dynamic scheduling method for an intelligent charging and swapping network provided in an embodiment of this application. Detailed Implementation

[0011] This application provides a dynamic scheduling method for intelligent charging and swapping networks, which addresses the technical problems of poor adaptive adjustment, low charging and swapping operation efficiency and stability in existing technologies.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] Examples, such as Figure 1 As shown, this application provides a dynamic scheduling method for a smart charging and swapping network, the method comprising:

[0014] Step S100: Traverse multiple charging and battery swapping stations in the target area and identify them to build a charging and battery swapping network.

[0015] Specifically, the process begins with a comprehensive traversal of multiple charging and battery swapping stations within the target area, visiting each station one by one to ensure no omissions. During this traversal, each station is identified and marked. This marking is crucial for accurate identification and differentiation of each station, and the marking information includes key details such as the station's name, number, and location. Based on this identification, a charging and battery swapping network is then constructed. This network construction considers the relationships between the stations, such as their geographical distribution and power transmission lines. These stations are treated as nodes in the network, and connections are established between them based on their interrelationships. For example, if two charging and battery swapping stations share a power transmission line or have a mutually supportive or collaborative relationship in actual operation, this relationship will be represented by corresponding lines or connections in the network. Through this process of traversing, marking, and building connections, a complete charging and battery swapping network is formed. This network clearly demonstrates the relationships and layout between the charging and battery swapping stations, providing a vital infrastructure for subsequent data analysis and scheduling decisions.

[0016] Step S200: Collect data from the target area based on the charging and swapping network to generate a historical operation dataset.

[0017] Specifically, based on the previously established charging and battery swapping network, a comprehensive data collection effort was initiated in the target area. The charging and battery swapping network provided a clear scope and target for data collection, ensuring accurate acquisition of data related to charging and battery swapping stations. During the data collection process, various information related to the operation of charging and battery swapping stations was gathered, including charging and battery swapping records for each station, such as detailed data on charging time, charging amount, battery swapping time, and number of battery swaps. Simultaneously, information on power supply was collected, such as grid voltage, current, power parameters, and data on the stability and reliability of the power supply. Furthermore, data related to user usage was collected, such as user charging demand patterns, usage frequency, and charging time distribution. This data reflects user behavior and demand characteristics. Through the collection and integration of this multi-faceted data, a historical operation dataset was ultimately generated. This dataset records the operation of the charging and battery swapping network over a period of time, containing rich information and providing a solid data foundation for subsequent demand analysis and operational forecasting. Analysis of this dataset allows for a deeper understanding of the charging and battery swapping network's operational status, identifying potential problems and patterns, and providing crucial information for optimizing network operation and developing reasonable scheduling schemes.

[0018] Step S300: Perform demand analysis based on the historical operation dataset to determine charging and swapping demand information, and perform operation forecasting for the charging and swapping network according to the charging and swapping demand information to determine grid load forecasting data.

[0019] Specifically, time series analysis algorithms are used to perform demand analysis and operational forecasting based on historical operational datasets. First, the historical operational dataset is preprocessed, including data cleaning, filtering, and organization, to ensure data accuracy and completeness. Then, the moving average method in time series analysis is used to analyze the historical data to determine the trends and seasonal variations in charging and swapping demand. Analysis of historical data reveals charging and swapping demand information, such as average charging volume and swapping volume over different time periods, as well as peak and off-peak demand periods. This information provides crucial data for subsequent operational forecasting. Next, based on the charging and swapping demand information, operational forecasting of the charging and swapping network is performed using time series-based forecasting models, such as ARIMA (Autoregressive Moving Average), which considers historical demand trends, seasonality, and cyclical factors, as well as other relevant factors such as the impact of weather and holidays on charging and swapping demand. During the operational forecasting process, the model predicts the operational status of the charging and swapping network over a future period based on the input charging and swapping demand information. By simulating charging and swapping behavior under different scenarios, the load situation of the power grid at various time periods can be determined. Finally, based on the model's prediction results, the power grid load forecast data is determined. This data includes the expected load values, peak loads, and valley loads for different time periods in the future. This forecast data will help us formulate reasonable scheduling schemes to ensure the stable operation of the power grid and meet the charging and swapping needs. For example, when using the ARIMA model for forecasting, the stationarity of historical data is first tested to determine the model order. Then, through model training and parameter tuning, the optimal forecast model is obtained. Finally, the charging and swapping demand information is input into the model to obtain the power grid load forecast data. By employing time series analysis algorithms, accurate demand analysis and operational forecasting are performed, providing a reliable basis for the optimized scheduling of the charging and swapping network.

[0020] Step S400: Perform a balance analysis on the charging and swapping network based on the power grid load forecast data, generate a power grid balance coefficient, formulate multiple initial scheduling schemes according to the power grid balance coefficient, and construct an initial population.

[0021] Specifically, based on previously obtained grid load forecast data, an in-depth balance analysis is conducted on the charging and battery swapping network. The purpose of this balance analysis is to assess the load distribution of the grid in different time periods to determine whether the grid can operate stably. The balance analysis considers all components of the grid, including charging and battery swapping stations and transmission lines. Through detailed analysis of the grid load forecast data, the power balance of the grid in different time periods, as well as the stability of parameters such as voltage and current, are calculated. Based on the results of the balance analysis, a grid equilibrium coefficient is generated. The grid equilibrium coefficient is an indicator reflecting the degree of grid balance; it comprehensively considers factors such as load distribution and power loss. This coefficient provides a direct understanding of the grid's stability and reliability. Next, several initial dispatch schemes are formulated according to the grid equilibrium coefficient. These schemes aim to rationally allocate the power supply of charging and battery swapping stations based on the actual grid conditions to ensure balanced grid operation. When formulating these dispatch schemes, various factors are considered, such as the location and capacity of the charging and battery swapping stations, and user demand. By formulating multiple initial scheduling schemes, an initial population is constructed. This initial population is a set of selectable scheduling schemes that serve as the basis for subsequent optimization. Each initial scheduling scheme has certain characteristics and advantages. Through evaluation and optimization of the initial population, the optimal scheduling scheme can be gradually found. For example, when formulating the initial scheduling scheme, the priority order of power supply to charging and battery swapping stations can be determined based on the grid balance coefficient. For areas with a high grid balance coefficient, the power supply of charging and battery swapping stations can be appropriately increased to improve grid stability; for areas with a low grid balance coefficient, measures can be taken to reduce the power supply to avoid grid overload.

[0022] Step S500: Traverse the multiple initial scheduling schemes to perform simulated scheduling, evaluate the initial population, and dynamically optimize the initial population based on the evaluation results to generate an optimized population.

[0023] Specifically, a genetic algorithm is used to traverse multiple initial scheduling schemes for simulated scheduling, evaluate the initial population, and dynamically optimize it based on the evaluation results to generate an optimized population. Simulated scheduling is performed on multiple initial schemes, treating each scheme as an individual and simulating its operation in the charging and swapping network. Considering practical factors such as the capacity of charging and swapping stations, user demand, and grid load, the fitness value of each individual is calculated during the simulated operation. This fitness value is used to evaluate the individual's performance. Next, the genetic algorithm evaluates the initial population. By simulating the processes of natural selection and genetic mutation, the genetic algorithm performs selection, crossover, and mutation operations on individuals in the population. Based on the individual's fitness value, individuals with higher fitness are selected as parents. Crossover generates offspring individuals, and mutation operations are then performed on the offspring individuals to increase population diversity. During the evaluation process, selection, crossover, and mutation operations are continuously iterated to gradually optimize the population. By comparing the fitness values ​​of different individuals, individuals with higher fitness are retained, while those with lower fitness are eliminated, allowing the population to evolve towards a better direction. Based on the evaluation results, the initial population is dynamically optimized. During each generation of evolution, the probabilities of selection, crossover, and mutation are adjusted according to changes in fitness values ​​to better guide the population's evolutionary direction. For example, if the overall fitness value of a generation is high, it indicates that the population is close to the optimal solution, and the mutation probability can be appropriately reduced to avoid destroying existing superior individuals; if the population's fitness value increases slowly, the mutation probability can be appropriately increased to introduce more diversity. After multiple iterations and optimizations, an optimized population is finally generated. Individuals in the optimized population have higher fitness values, representing better scheduling schemes. These optimized scheduling schemes can better meet the needs of the charging and swapping network, improving the stability and efficiency of the power grid. The application of genetic algorithms can effectively optimize the initial scheduling scheme, generating a better optimized population and providing a reliable scheduling scheme for the efficient operation of the charging and swapping network.

[0024] Step S600: Perform adaptive dynamic scheduling of the charging and swapping network according to the optimized population.

[0025] Specifically, the charging and swapping network is adaptively and dynamically scheduled according to the scheduling scheme provided by the optimized population. The network's operational status is monitored in real time, including the power reserves of each charging and swapping station, user charging and swapping demand, and grid load. Power resources are intelligently allocated based on the scheduling scheme in the optimized population to ensure that charging and swapping stations can efficiently provide services to users. For example, when a charging and swapping station has low power reserves while nearby users have high charging demand, power allocation is adjusted to prioritize providing sufficient power to that station to meet user needs. Simultaneously, dynamic adjustments are made based on grid load. If the grid load is too high, the charging power of the charging and swapping stations is appropriately reduced to alleviate grid pressure; conversely, if the grid load is low, the charging power is increased to fully utilize the grid's remaining capacity. Furthermore, the scheduling scheme is optimized and adjusted in real time based on user demand and grid conditions. In case of emergencies, such as a charging and swapping station failure or a sudden increase in user demand, the scheduling scheme is readjusted to ensure the stable operation of the charging and swapping network. By implementing adaptive dynamic scheduling based on optimized populations, the potential of the charging and swapping network can be fully realized, energy utilization efficiency can be improved, and the safe and stable operation of the power grid can be ensured, providing users with more reliable and convenient charging and swapping services.

[0026] In one possible implementation, step S100 further includes:

[0027] Step S110: Traverse the multiple charging and swapping stations in the target area to perform correlation analysis and obtain the correlation relationship between the multiple charging and swapping stations.

[0028] Step S120: The multiple charging and swapping stations are designated as multiple nodes, each of which has a unique coordinate value.

[0029] Step S130: Assign weights to the multiple nodes to generate multiple weight coefficients.

[0030] Step S140: Connect the multiple nodes according to the association relationship and the multiple weight coefficients to construct a network flow model.

[0031] Step S150: Draw the charging and swapping network of the target area based on the network flow model.

[0032] Specifically, multiple charging and battery swapping stations within the target area are examined and analyzed one by one. Various factors are considered to determine the relationships between these stations, such as the distance between them, whether their service areas overlap, the connection status of power transmission lines, and whether they support or cooperate with each other in daily operation. This relevant information is comprehensively collected and organized, and in-depth analysis and calculations are performed to accurately reveal the degree and manner of the relationships between each station. Through this correlation analysis, detailed relationships between multiple charging and battery swapping stations are obtained, providing an important basis for the subsequent construction of the charging and battery swapping network.

[0033] Multiple charging and battery swapping stations in the target area are treated as independent nodes, each assigned a specific node identifier to ensure accurate identification and differentiation during subsequent network construction and analysis. To more precisely describe the location of each node, unique coordinate values ​​are assigned to them. These coordinate values ​​are typically determined based on a geographic coordinate system, such as longitude and latitude. In this way, the location of each charging and battery swapping station can be accurately located in space. For example, charging and battery swapping station A is assigned coordinate values ​​(longitude X1, latitude Y1), charging and battery swapping station B is assigned coordinate values ​​(longitude X2, latitude Y2), and so on. Thus, whether on a map or in subsequent calculations and analyses, the location of each charging and battery swapping station can be accurately found using these unique coordinate values. This method of representing charging and battery swapping stations as nodes with unique coordinate values ​​provides the foundation for constructing the topology of the charging and battery swapping network. These coordinate values ​​facilitate the calculation of distances, directions, and other relationships between nodes, thereby enabling better analysis and optimization of the charging and battery swapping network layout and operation.

[0034] For the previously identified multiple nodes (i.e., charging and battery swapping stations), weights are assigned, encompassing key factors such as distance, time, and cost. First, distance is considered. If distance is included as part of the weight, the actual geographical distance between two charging and battery swapping station nodes is calculated based on geographic information. For example, by measuring the straight-line distance or actual road travel distance between two nodes, nodes that are closer will be assigned a relatively smaller distance weight, because in practical applications, shorter distances usually mean more convenient connections and lower transmission costs. Time is also crucial. For instance, the time required to travel from one charging and battery swapping station to another includes not only travel time but also the time required for charging and battery swapping operations within the station. If the time required for charging, battery swapping operations and transportation from one node to another is long, then the corresponding time weight will be larger. Cost is equally important. Costs include power transmission costs, equipment maintenance costs, and user operating costs. If the power transmission cost between two nodes is high, or if the cost of charging and battery swapping for users differs significantly between different nodes, then these cost differences will be reflected in the weight allocation. By comprehensively considering these factors, the connection relationships between each node are evaluated and quantified, ultimately generating multiple weight coefficients. These weight coefficients will serve as an important basis for subsequent construction of network flow models and network analysis, helping to optimize the layout and scheduling of the charging and swapping network to achieve more efficient and economical charging and swapping services.

[0035] To begin building the network flow model, the process begins by establishing the relationships between multiple nodes corresponding to the previously defined charging and battery swapping stations. If two charging and battery swapping stations are closely related in practice—for example, through frequent power transmission, user movement, or business collaboration—a connection will be established between them. Next, multiple weighting coefficients are incorporated into the connection process. These coefficients represent factors such as distance, time, or cost, and are crucial to the accuracy and practicality of the network flow model. For instance, if two nodes are close together, their connection may be given higher priority or stronger connection strength based on the distance weighting coefficient. During the construction process, for each pair of related nodes, the specific connection method and parameters are determined based on their weighting coefficients. For example, if the time weighting coefficient is large, the model will focus more on time-related factors between the two nodes, such as the response time of power transmission or the time required for users to transfer between the two charging and battery swapping stations. In this way, multiple nodes are connected according to their actual relationships, combined with multiple weighting coefficients reflecting the actual situation, gradually building a complete network flow model. This model can accurately reflect the relationships between nodes in the charging and swapping network, the direction and intensity of power and information flow, and the impact of different factors on network operation, providing strong support and basis for subsequent charging and swapping network analysis, optimization and scheduling.

[0036] Based on the previously constructed network flow model, the charging and swapping network for the target area is drawn. First, the geographical information of the target area, such as a map or geographic coordinate system, is used as a foundation. Then, each node in the network flow model, i.e., the charging and swapping stations, is accurately placed on the map according to its corresponding unique coordinate value. Each node is appropriately identified and labeled according to its importance and attributes in the network; for example, different icons or colors can be used to distinguish charging and swapping stations of different sizes or types. Next, based on the connection relationships between nodes determined in the model, related nodes are connected with lines. The thickness, color, or style of these lines can be set according to the connection weight coefficient. For example, lines between nodes with larger weight coefficients, indicating closer connections or greater importance, can be drawn thicker or use more prominent colors. During the drawing process, auxiliary information and annotations can be added, such as marking relevant weight information like distance, time, or cost on the lines, or marking key data such as the capacity and service range of the charging and swapping stations near the nodes. Through this drawing process, a visually intuitive and clear charging and swapping network diagram is finally presented. This diagram allows for a quick and accurate understanding of the layout and structure of the charging and swapping network in the target area, including the location of each charging and swapping station, the degree of interconnection between them, and the key features of the entire network.

[0037] In one possible implementation, such as Figure 2 As shown, step S400 further includes:

[0038] Step S410: Perform time sequence analysis based on the charging and swapping network to obtain a time series.

[0039] Step S420: Retrieve the charging and swapping demand information and perform demand analysis on the charging and swapping network according to the time series to obtain the total power demand data of the swapping network.

[0040] Step S430: Align the power grid load forecast data with the total power demand data according to the time series to generate a data alignment result.

[0041] Step S440: Based on the data alignment results, compare the power grid load forecast data with the total power demand data to calculate and obtain multiple load imbalance quantities.

[0042] Step S450: Based on the multiple load imbalances, perform an impact analysis, and calculate the grid balance coefficient of the charging and swapping network according to the impact factors and the multiple load imbalances.

[0043] Specifically, an operational time-series analysis is conducted based on the charging and battery swapping network. First, operational data of the network at different time periods is collected, including the working status of charging and battery swapping stations and the timing of power inflows and outflows. Then, this data is organized and categorized chronologically to obtain a time series. This time series provides a foundational time framework for subsequent time-based analysis.

[0044] Existing charging and battery swapping demand information is retrieved, which includes the characteristics of different users' charging and battery swapping needs under different scenarios. Then, based on the previously obtained time series, the charging and battery swapping network is divided into different time points or time periods. Within each specific time interval, various factors are comprehensively considered, such as changes in the number of users, differences in charging or battery swapping needs of different types of vehicles, and the service capacity of charging and battery swapping stations, to conduct an in-depth analysis of charging and battery swapping demand. By calculating the sum of the electricity demand corresponding to all charging and battery swapping needs within each time interval, the total electricity demand data of the battery swapping network at different times is finally obtained. This data accurately reflects the changing pattern of electricity demand of the charging and battery swapping network over time.

[0045] Since both grid load forecast data and total electricity demand data are closely related to time, the data points of both are first matched one by one based on the common time dimension of time series. For each specific time point or time period, the corresponding grid load forecast value and total electricity demand value are found and matched to ensure that the data of the two can be correlated under the same time identifier. After this precise time alignment operation, a clear data alignment result is generated. This result enables the accurate comparison of the relationship and difference between grid load forecast and actual total electricity demand at the same time in subsequent analysis.

[0046] Based on the data alignment results, key calculations are performed. First, the grid load forecast data and total electricity demand data within the same time series are analyzed. For each specific time period, such as an hourly, daily, or weekly interval, the corresponding grid load forecast value and total electricity demand value are compared one by one. If the grid load forecast value and total electricity demand value are not equal within a certain time period, the load imbalance is calculated, usually by the difference between the two values. For example, if the grid load forecast value is 100 units of electricity within an hourly time period, while the actual total electricity demand is 80 units, then the load imbalance for that hour is 20 units. This comparison and calculation is performed for each time period over time, thereby obtaining the load imbalance values ​​for multiple different time periods. These load imbalance values ​​reflect the differences in grid supply and demand at different points in time. For example, during the morning rush hour, the simultaneous charging of numerous electric vehicles causes a sharp increase in total electricity demand. This results in a significant difference between the forecast and the predicted grid load, leading to a relatively large load imbalance. Conversely, during off-peak hours at night, electricity demand is relatively low, and the load imbalance is smaller or even negative (when the predicted grid load exceeds the total electricity demand). Accurate calculation and analysis of these load imbalances across different time periods provide a deeper understanding of the grid's supply and demand balance at different times, offering crucial data support for subsequent grid balancing adjustments and optimizations.

[0047] Impact analysis is conducted based on multiple load imbalances. These load imbalances reflect the differences between power grid supply and demand at different time periods. A series of influencing factors are then identified, including changes in the operating status of charging and battery swapping stations (such as equipment failures or maintenance leading to service interruptions at some stations), changes in user behavior (such as sudden increases or decreases in user charging demand within a specific time period, influenced by seasonality, promotional activities, or policy adjustments), external environmental factors (such as the impact of weather changes on electric vehicle range and charging demand, and changes in battery performance under high or low temperatures leading to changes in charging frequency), and the stability and reliability of the power grid itself (such as power grid upgrades and line faults). These influencing factors are then combined with multiple load imbalances for calculation. For example, for a specific time period's load imbalance, if the influencing factor of charging and battery swapping station equipment failure exists, it will be assigned a certain weight based on its impact on charging and battery swapping capacity. This weighted average method is then used to integrate and process the weights of each influencing factor and the load imbalance values ​​for that time period. Finally, through a complex calculation and analysis process, the grid balance coefficient of the charging and swapping network is obtained. This coefficient is a key indicator that comprehensively reflects the grid balance state in the charging and swapping network. If the grid balance coefficient is close to 1, it indicates that the grid is in a relatively balanced state; if the coefficient deviates significantly from 1, it indicates that there is a significant imbalance in the grid, requiring corresponding adjustment and optimization measures, such as adjusting the power distribution of charging and swapping stations, guiding users to rationally arrange charging and swapping times, or upgrading and transforming the grid, to improve the stability and reliability of the grid and ensure the efficient operation of the charging and swapping network.

[0048] In one possible implementation, step S440 further includes:

[0049] Step S441: Analyze the total power demand data to obtain charging demand information and discharging demand information for multiple time periods.

[0050] Step S442: Calculate based on the multi-time period charging demand information to determine the expected charging demand threshold range.

[0051] Step S443: Calculate based on the multi-time period discharge demand information to determine the expected discharge demand threshold range.

[0052] Step S444: Based on the data alignment result, compare the power grid load forecast data with the expected charging demand threshold range by traversing the time series, extract the power grid load forecast data that is not in the expected charging demand threshold range, and calculate to obtain multiple charging load imbalances. The multiple charging load imbalances have a corresponding relationship with the time series.

[0053] Step S445: Based on the data alignment result, compare the power grid load forecast data with the expected discharge demand threshold range by traversing the time series, extract the power grid load forecast data that is not in the expected discharge demand threshold range, calculate and obtain multiple discharge load imbalances, and the multiple discharge load imbalances have a corresponding relationship with the time series.

[0054] Step S446: Integrate the multiple charging load imbalances and the multiple discharging load imbalances to obtain the multiple load imbalances.

[0055] Specifically, a deep analysis of total electricity demand data is conducted. Since total electricity demand data includes electricity demand information of charging and swapping networks at different time periods, it is broken down into multi-time period charging demand information and multi-time period discharging demand information through data processing and analysis technology. The charging demand information reflects the electricity required by electric vehicles and other devices when charging at different times, while the discharging demand information involves energy storage devices or power feedback under certain specific conditions.

[0056] To determine the expected charging demand threshold range using a moving average method combined with standard deviation, the process begins by collecting and organizing charging demand data across multiple time periods. This data reflects changes in charging demand over different time periods. Next, a moving average is used to smooth the data. A suitable time window is selected, such as three time periods (this window can be adjusted based on actual conditions and data characteristics). For each time period, the average charging demand is calculated for that period and the two preceding periods. This yields a new set of data processed by the moving average. Then, the standard deviation of this new set of data is calculated. The standard deviation reflects the dispersion of the data. When determining the expected charging demand threshold range, the average value is added to or subtracted from the standard deviation to construct the range. Typically, adding or subtracting two times the standard deviation is chosen as the boundary of the range.

[0057] Similarly, the moving average method combined with standard deviation is used to determine the expected discharge demand threshold range. Discharge demand data for different time periods are collected and analyzed, and through corresponding calculations and evaluations, a suitable expected discharge demand threshold range is determined.

[0058] Based on the data alignment results, this ensures that the grid load forecast data and the expected charging demand threshold range are compared within the same time series, where each time point corresponds to specific data. Starting from the earliest time point, the comparison is performed chronologically, comparing the grid load forecast data at each time point with the corresponding expected charging demand threshold range. If the grid load forecast data falls within the expected charging demand threshold range, it indicates that the charging demand and grid load are relatively balanced at that time point, requiring no further processing. However, when the grid load forecast data at a certain time point is found to be outside the expected charging demand threshold range, this data is extracted. For example, if the grid load forecast data is higher than the upper limit or lower than the lower limit of the expected charging demand threshold range, this indicates an imbalance. These extracted grid load forecast data points outside the range are then calculated, typically by subtracting the data from the boundary values ​​of the expected charging demand threshold range. For instance, if the grid load forecast data is higher than the upper limit, the upper limit value is subtracted; if it is lower than the lower limit, the lower limit value is subtracted, thus obtaining the charging load imbalance at that time point. Because the operation is performed sequentially according to the time series, there is a one-to-one correspondence between the multiple charging load imbalances obtained and the time series. It is possible to clearly know at which specific time point each charging load imbalance occurred. This is of great significance for subsequent analysis of the changing trend of charging load over time and for identifying the key time points where imbalance problems occur. It also helps to optimize and adjust the charging and swapping network and the power grid more accurately.

[0059] Similar to step S444, based on the data alignment results, the grid load forecast data and the expected discharge demand threshold range are compared according to the time series. For the grid load forecast data at each time point, it is determined whether it is within the expected discharge demand threshold range. If not, these data are extracted and calculated to obtain multiple discharge load imbalance quantities. These imbalance quantities also have a corresponding relationship with the time series, which can reflect the discharge imbalance status at different time points.

[0060] The obtained charging load imbalance reflects the difference between grid load and charging demand during the charging process, while the discharging load imbalance reflects the corresponding imbalance during the discharging process. For the integration process, these two types of imbalances are considered comprehensively using a simple addition operation, summing the charging and discharging load imbalances at corresponding time points. This is because both charging and discharging ultimately affect the load balance of the entire charging and swapping network. Addition provides a comprehensive imbalance value reflecting the overall imbalance situation. For example, at a specific time point, a positive charging load imbalance indicates that charging demand is excessive relative to grid load; simultaneously, a negative discharging load imbalance indicates that discharging capacity is surplus relative to grid capacity. Adding them together provides a more comprehensive understanding of the overall impact of the charging and swapping network on grid balance at that time point. During the integration process, it is also necessary to pay attention to the consistency of the time series, because both charging and discharging load imbalances have a corresponding relationship with the time series. Therefore, it is essential to ensure that the operations are performed at the same time points during integration to guarantee that the multiple load imbalances obtained after integration also accurately correspond to the time series. Through such integration, the final multiple load imbalances can comprehensively reflect the overall balance status between the charging and swapping network and the power grid during charging and discharging processes. This provides comprehensive and accurate data support for further analysis of the power grid's operating status, load adjustment and optimization, and the formulation of reasonable charging and swapping strategies.

[0061] In one possible implementation, step S500 further includes:

[0062] Step S510: Build a simulated environment information for the charging and swapping network based on the historical operation dataset.

[0063] Step S520: Randomly execute the multiple initial scheduling schemes according to the simulated environment information to perform scheduling simulation on the charging and swapping network, and obtain multiple initial scheduling results.

[0064] Step S530: Use the fitness function to evaluate the multiple initial scheduling results and generate multiple scheduling fitnesss.

[0065] Step S540: Add the plurality of scheduling fitness values ​​to the evaluation result.

[0066] Specifically, to build a simulation environment for the charging and swapping network, historical operational datasets are first acquired. This dataset contains various information about the network's past operation, such as charging and swapping demand at different times, grid load changes, and the operational status of charging and swapping stations. Then, this data is used to construct a simulation environment that closely approximates the actual operating environment of the charging and swapping network. This simulation environment strives to reproduce as many factors and conditions as possible from the real-world scenario, including the location of charging and swapping stations, equipment performance, power transmission capacity, and user behavior patterns. In this way, a reliable foundational environment is provided for subsequent scheduling simulations, ensuring that the simulation results more closely resemble reality.

[0067] Based on the prepared simulation environment information, the scheduling simulation begins. Multiple initial scheduling schemes are pre-set different scheduling strategies. These strategies differ in aspects such as power allocation to charging and battery swapping stations, charging / swapping sequence arrangement, and equipment usage priority. The charging and battery swapping network is scheduled sequentially in the simulation environment according to these schemes, and the operational performance of the charging and battery swapping network under different schemes is observed and recorded. For example, under one initial scheduling scheme, the charging and battery swapping needs of high-demand areas are prioritized, while another scheme focuses more on balancing the load among various charging and battery swapping stations. By simulating the operation of the charging and battery swapping network under different schemes, multiple initial scheduling results are obtained. These results reflect the specific impact of different scheduling schemes on the operation of the charging and battery swapping network.

[0068] The fitness function is meticulously designed based on the specific needs and optimization objectives of the charging and swapping network. For the initial scheduling evaluation of the charging and swapping network, the fitness function comprehensively considers multiple key factors. For example, it considers grid stability; if a scheduling scheme causes excessive grid load fluctuations, a penalty will be imposed in the fitness function calculation, reducing its fitness. It also considers charging and swapping efficiency, including time costs and energy losses. If a scheduling scheme can make the charging and swapping process more efficient, reducing user waiting time and energy waste, it will obtain a higher fitness. In practice, for each initial scheduling result, its relevant parameters and data are input into the fitness function, which calculates according to preset algorithms and rules. For example, grid stability is measured by calculating the grid load variance at different time points; charging and swapping efficiency can be quantified based on indicators such as average charging and swapping time and energy conversion efficiency. Then, after the fitness function is calculated, a corresponding scheduling fitness value is generated for each initial scheduling result. This value is a quantitative evaluation result, which reflects the performance of the scheduling result in meeting the optimization goals of the charging and swapping network. If the fitness value of a scheduling result is high, it means that it performs better after comprehensively considering various factors and is more in line with the needs and expectations of the charging and swapping network.

[0069] Once multiple scheduling fitness values ​​are obtained, they need to be added to the evaluation results. First, the evaluation results are a dataset or document used to record and comprehensively analyze the performance of various scheduling schemes. Adding scheduling fitness values ​​is an orderly and precise operation. For each initial scheduling scheme, its corresponding fitness value is associated with the relevant scheduling scheme and recorded in the evaluation results according to a specific format and rules. This is done to maintain data integrity and traceability, ensuring that it is clear which scheduling scheme generated each fitness value. During the addition process, records are kept in tabular form, with one column recording the key characteristics or number of the scheduling scheme and the other column recording the corresponding scheduling fitness value. By adding multiple scheduling fitness values ​​to the evaluation results, a comprehensive and detailed evaluation system is built, providing a solid data foundation and strong decision support for further selecting the optimal scheduling scheme or improving and optimizing existing schemes.

[0070] In one possible implementation, step S500 further includes:

[0071] Step S550: Sort the multiple initial scheduling schemes in descending order based on the multiple scheduling fitnesss to determine the scheduling scheme sequence; extract the first initial scheduling scheme in the first order and the second initial scheduling scheme in the second order based on the scheduling scheme sequence; perform cross-validation on the first initial scheduling scheme and the second initial scheduling scheme as parents to generate the first scheduling scheme; perform random mutation on the first scheduling scheme to obtain the first scheduling mutation scheme; iterate through the scheduling scheme sequence to obtain N scheduling schemes and N scheduling mutation schemes, where N is an integer greater than 1; dynamically optimize the initial population based on the N scheduling schemes and the N scheduling mutation schemes to generate the optimized population.

[0072] Specifically, for multiple initial scheduling schemes, each scheme has a corresponding scheduling fitness. These scheduling fitness values ​​are key indicators for evaluating the performance of each scheme. Based on these fitness values, they are sorted in descending order of performance. During the sorting process, schemes with higher fitness values ​​are ranked higher, meaning they perform better under the current evaluation criteria. Through this sorting operation, all initial scheduling schemes form a sequence of scheduling schemes ranked from best to worst performance.

[0073] In the established sequence of scheduling schemes, the order reflects their relative merits, as the sequence is sorted in descending order. First, the scheme at the very beginning of the sequence is identified and defined as the first initial scheduling scheme. This is typically the best-performing scheme among all initial schemes based on evaluation. Next, the scheme at the second position is identified as the second initial scheduling scheme. These two schemes are relatively early in the sequence, indicating their superior performance and characteristics among the many initial schemes. By extracting these two schemes, they can be used as key targets for subsequent operations (such as cross-validation), leveraging their advantages and characteristics to generate new, better scheduling schemes or providing a foundation and reference for further optimization.

[0074] Using a genetic algorithm, the first and second initial scheduling schemes are considered as parents, each carrying specific scheduling strategies and parameter information. Cross-validation is a genetic operation method. By selecting certain key parts from these two parent schemes, such as specific power allocation strategies for charging and battery swapping stations and equipment usage sequences, they are then cross-combined. At the same or specific locations in the two schemes, a portion of information is randomly selected and exchanged. For example, the time allocation strategy (first half) is selected from the first initial scheduling scheme, and the equipment allocation strategy (second half) is selected from the second initial scheduling scheme, and then they are combined. After this cross-validation operation, a new scheduling scheme, the first scheduling scheme, is generated. This new scheme integrates the different characteristics and advantages of the two parent schemes, resulting in better performance than the parent schemes, or bringing new scheduling ideas and possibilities, providing new approaches and directions for finding more suitable and efficient scheduling schemes.

[0075] After obtaining the first scheduling scheme, a random mutation operation was introduced to further increase the diversity of scheduling schemes and explore a wider solution space. Random mutation is a process of making small random changes to the original scheme. The first scheduling scheme includes a series of specific scheduling parameters and strategies. During random mutation, one or more elements are randomly selected from these parameters and strategies for modification. For example, the charging power setting of a charging / swapping station might be changed, or the charging / swapping sequence within a certain time period might be adjusted. This modification is completely random, occurring with a certain probability, and the magnitude of the modification is also randomly determined within a certain range. Through this random mutation operation, the first mutated scheduling scheme is obtained. This mutated scheme has a certain inheritance relationship with the first scheduling scheme, but also possesses new characteristics due to random changes. It may bring better scheduling results, or it may simply serve as an exploratory attempt, providing more possibilities and choices for the entire scheduling scheme optimization process.

[0076] During the iterative process of traversing the sequence of scheduling schemes, each scheme is operated on sequentially, starting from the beginning of the sequence. In each iteration, corresponding processing steps are performed based on the current scheduling scheme. For each scheduling scheme, new scheduling schemes and scheduling mutation schemes are generated through operations such as cross-validation and random mutation. As the iteration continues, these newly generated schemes are continuously accumulated, eventually resulting in N scheduling schemes and N scheduling mutation schemes, where N is an integer greater than 1, representing the number of schemes generated after multiple iterations.

[0077] The initial population is dynamically optimized based on these N scheduling schemes and N scheduling mutation schemes. The initial population is the set of scheduling schemes initially given, representing the initial solution space of the problem. During the optimization process, the newly generated scheduling schemes and scheduling mutation schemes are compared and evaluated with the schemes in the initial population. If the newly generated scheme performs better in certain performance indicators (such as scheduling fitness), it can replace the corresponding poor scheme in the initial population. By continuously performing such replacement operations, the overall quality of the population is gradually improved.

[0078] In one possible implementation, step S550 further includes:

[0079] Step S551: Iterate through multiple scheduling control parameters of the first scheduling scheme and randomly select to determine the mutation point.

[0080] Step S552: Construct a mutation distribution expression based on the multiple scheduling fitnesss, and calculate and determine the mutation probability through the mutation distribution expression.

[0081] Step S553: ​​Calculate the scheduling response coefficients of the multiple scheduling control parameters to dynamically adjust the mutation probability and determine the dynamic mutation probability.

[0082] Step S554: Perform a mutation operation based on the mutation point and the dynamic mutation probability to obtain the first scheduling mutation scheme.

[0083] Specifically, the process begins by iterating through multiple scheduling control parameters of the first scheduling scheme. These parameters include the power allocation ratio of charging and battery swapping stations, the charging and battery swapping priority settings for different time periods, and the equipment's on / off policies. During the iteration, a mutation point is determined through random selection. This mutation point is the key location for subsequent mutation operations; it can be a specific parameter or a specific value of a parameter. For example, in a charging and battery swapping network scheduling scheme, the charging power setting of a specific charging and battery swapping station during a specific time period is randomly selected as the mutation point.

[0084] Based on the previously obtained scheduling fitness values, a mutation distribution expression is constructed. Scheduling fitness reflects the relative merits of different scheduling schemes. Using these fitness values ​​to construct the expression allows the mutation operation to be correlated with the performance of the scheme. This mutation distribution expression is then used to calculate and determine the mutation probability, which dictates the likelihood of performing a mutation operation at the mutation point. For example, if a scheduling scheme has low fitness, its mutation probability will be increased accordingly, in the hope of finding a better scheme through mutation.

[0085] Calculate the dispatch response coefficients for multiple dispatch control parameters. The dispatch response coefficient measures the degree of influence of each dispatch control parameter on the entire dispatch system. For example, changes in certain parameters have a significant impact on grid load balance, resulting in relatively high dispatch response coefficients. These dispatch response coefficients are applied to dynamically adjust the mutation probability, thereby determining the dynamic mutation probability. The purpose of this is to more rationally adjust the probability of mutation operations based on the importance and degree of influence of the parameters, making mutation operations more targeted and adaptive.

[0086] Once the mutation point and dynamic mutation probability are determined, the mutation operation is executed to obtain the first scheduling mutation scheme. The mutation point is a specific location or parameter selected in the first scheduling scheme. This mutation point could be a specific charging time setting at a charging / swapping station, the operating power parameter of a specific device, or other key factors related to scheduling. The dynamic mutation probability provides a quantifiable standard for the occurrence of the mutation operation. If a value between 0 and 1 is randomly generated, a mutation operation is triggered when this value is less than or equal to the dynamic mutation probability. This mutation operation based on the mutation point and dynamic mutation probability provides a way to diversify scheduling schemes and explore better solutions. It allows the algorithm to try more extensively in the search space, avoiding getting trapped in local optima, thereby finding a scheduling scheme with higher fitness and better meeting actual needs.

[0087] In one possible implementation, step S552 further includes:

[0088] Step S5521: Construct the mutation distribution expression based on the multiple scheduling fitnesss, wherein the mutation distribution expression is:

[0089] ;

[0090] in, The mutation probability, It is a constant. This is the value with the highest fitness in the initial population. The fitness value for the first scheduling scheme, This represents the average fitness value of the initial population.

[0091] Step S5522: When the fitness value of the first scheduling scheme is less than the initial population average fitness value, the expression for the variation distribution is modified. Replace with ,and Less than The constant.

[0092] Step S5523: Calculate the mutation probability by combining the mutation distribution expression with multiple scheduling fitnesss.

[0093] Specifically, the purpose of constructing the mutation distribution expression is to determine the mutation probability based on the fitness of the scheduling scheme. The expression is as follows: , This represents the probability of mutation, which is the final value to be determined. It is a constant used to calculate the mutation probability in the initial case. This represents the value with the highest fitness in the initial population, reflecting the fitness level of the optimal scheduling scheme in the population. It is the fitness value of the first scheduling scheme, used to measure the merits of that specific scheduling scheme. The initial average fitness value reflects the average fitness level of the entire population. When calculating the mutation probability, adjustments are made based on the relationship between the fitness value of the first scheduling scheme and the initial average fitness value. If the fitness value of the first scheduling scheme is less than the initial average fitness value, it indicates that the scheduling scheme is ineffective, and the mutation distribution expression will be adjusted accordingly. Replace with , And less than The constant is used to increase the likelihood of mutating poorly performing scheduling schemes, with the aim of finding better solutions through mutation operations.

[0094] This detailed calculation process, based on the specific fitness of the scheduling schemes, allows for the reasonable and accurate determination of mutation probabilities. This is crucial for subsequent random mutation operations, guiding the algorithm to more effectively optimize and improve the scheduling schemes, thereby increasing the efficiency and likelihood of finding the optimal scheme. For example, if the fitness of multiple scheduling schemes differs significantly, the mutation probabilities determined through this calculation method can better encourage the algorithm to attempt more mutations on schemes with lower fitness, potentially leading to the discovery of better scheduling schemes and improving the overall scheduling performance of the charging and swapping network.

[0095] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0096] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0097] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A dynamic scheduling method for an intelligent charging and swapping network, characterized in that, The method includes: Identify and traverse multiple charging and battery swapping stations in the target area to construct a charging and battery swapping network; Data is collected from the target area based on the charging and swapping network to generate a historical operation dataset; Demand analysis is performed based on the historical operation dataset to determine charging and battery swapping demand information. Operation forecasts are then made for the charging and battery swapping network based on the charging and battery swapping demand information to determine grid load forecast data. Based on the power grid load forecast data, a balance analysis is performed on the charging and swapping network to generate a power grid balance coefficient. Multiple initial scheduling schemes are formulated according to the power grid balance coefficient, and an initial population is constructed. The multiple initial scheduling schemes are traversed for simulated scheduling, the initial population is evaluated, and the initial population is dynamically optimized based on the evaluation results to generate an optimized population. The charging and swapping network is adaptively and dynamically scheduled according to the optimized population described above. Based on the evaluation results, the initial population is dynamically optimized to generate an optimized population. The method includes: Based on the multiple scheduling fitness values, the multiple initial scheduling schemes are sorted in descending order to determine the scheduling scheme sequence; Based on the aforementioned scheduling scheme sequence, extract the first initial scheduling scheme with the first position and the second initial scheduling scheme with the second position. The first initial scheduling scheme and the second initial scheduling scheme are used as parents for cross-validation to generate the first scheduling scheme. Based on the first scheduling scheme, a random mutation is performed to obtain the first scheduling mutation scheme; Iterate through the sequence of scheduling schemes to obtain N scheduling schemes and N scheduling mutation schemes, where N is an integer greater than 1; The initial population is dynamically optimized based on the N scheduling schemes and the N scheduling mutation schemes to generate the optimized population.

2. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 1, characterized in that, The method involves iterating through and identifying multiple charging and battery swapping stations in the target area to construct a charging and battery swapping network, including: The multiple charging and battery swapping stations in the target area are traversed for correlation analysis to obtain the correlation relationships between the multiple charging and battery swapping stations; The multiple charging and battery swapping stations are considered as multiple nodes, and each node has a unique coordinate value. Weights are assigned to the multiple nodes to generate multiple weight coefficients; The multiple nodes are connected according to the association relationship and the multiple weight coefficients to construct a network flow model; The charging and swapping network of the target region is drawn based on the network flow model.

3. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 1, characterized in that, Based on the power grid load forecast data, a balance analysis is performed on the charging and swapping network to generate a power grid equilibrium coefficient. The method includes: Based on the charging and swapping network, an operational time sequence analysis is performed to obtain a time series. The charging and swapping demand information is retrieved, and the demand analysis of the charging and swapping network is performed according to the time series to obtain the total power demand data of the swapping network. The power grid load forecast data and the total electricity demand data are aligned according to the time series to generate a data alignment result. Based on the data alignment results, the power grid load forecast data is compared with the total power demand data to calculate multiple load imbalances. An impact analysis is performed based on the multiple load imbalances, and the grid balance coefficient of the charging and swapping network is obtained by calculating the impact factors in conjunction with the multiple load imbalances.

4. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 3, characterized in that, Based on the data alignment results, the power grid load forecast data and the total power demand data are compared and calculated to obtain multiple load imbalance quantities. The method includes: The total electricity demand data is analyzed to obtain charging demand information and discharging demand information for multiple time periods. Based on the multi-time period charging demand information, the expected charging demand threshold range is determined. Based on the multi-time period discharge demand information, the expected discharge demand threshold range is determined. Based on the data alignment results, the power grid load forecast data is compared with the expected charging demand threshold range by traversing the time series. Power grid load forecast data that is not in the expected charging demand threshold range is extracted and calculated to obtain multiple charging load imbalances. The multiple charging load imbalances have a corresponding relationship with the time series. Based on the data alignment results, the power grid load forecast data is compared with the expected discharge demand threshold range by traversing the time series. Power grid load forecast data that is not in the expected discharge demand threshold range is extracted and calculated to obtain multiple discharge load imbalances. The multiple discharge load imbalances have a corresponding relationship with the time series. The multiple charging load imbalances and the multiple discharging load imbalances are integrated to obtain the multiple load imbalances.

5. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 1, characterized in that, The method involves traversing multiple initial scheduling schemes to simulate scheduling and evaluating the initial population, including: Based on the aforementioned historical operational dataset, a simulated environment for the charging and swapping network is constructed. Based on the simulated environment information, the multiple initial scheduling schemes are randomly executed to perform scheduling simulation on the charging and swapping network, and multiple initial scheduling results are obtained. The fitness function is used to evaluate the multiple initial scheduling results to generate multiple scheduling fitnesss; The multiple scheduling fitness values ​​are added to the evaluation results.

6. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 1, characterized in that, The method for obtaining a first scheduling mutation scheme by randomly mutating the first scheduling scheme includes: The multiple scheduling control parameters of the first scheduling scheme are randomly selected to determine the mutation point; A mutation distribution expression is constructed based on the multiple scheduling fitness values, and the mutation probability is calculated and determined using the mutation distribution expression. The scheduling response coefficients of the multiple scheduling control parameters are calculated to dynamically adjust the mutation probability, thereby determining the dynamic mutation probability; Based on the mutation points and the dynamic mutation probabilities, a mutation operation is performed to obtain the first scheduling mutation scheme.

7. The dynamic scheduling method for an intelligent charging and swapping network as described in claim 6, characterized in that, A mutation distribution expression is constructed based on the multiple scheduling fitness values, and the mutation probability is calculated and determined using the mutation distribution expression. The method includes: The mutation distribution expression is constructed based on the multiple scheduling fitness values, and the mutation distribution expression is as follows: ； in, The mutation probability, It is a constant. This is the value with the highest fitness in the initial population. The fitness value for the first scheduling scheme, This represents the average fitness value of the initial population. When the fitness value of the first scheduling scheme is less than the initial population average fitness value, the expression for the variation distribution will be changed. Replace with ,and Less than The constant; The mutation probability is determined by combining the mutation distribution expression with multiple scheduling fitness values.

Images