A smart energy supplement task scheduling method and system for a mobile charging vehicle

Abstract

The application discloses a kind of intelligent energy supplement task scheduling methods and systems for mobile charging car, it is related to mobile energy supplement scheduling data processing technical field, obtain user energy supplement request result, mobile charging car state result and candidate charging site result, generate initial collaborative scheduling result;Based on traffic state result and driving stability result generate user trusted time window result, mobile charging car arrives trusted time window result and trusted level result;According to user synchronization overlap degree result, mobile charging car synchronization matching result and synchronization risk result, initial task cluster is reserved, service time is adjusted or synchronization risk user is redistributed, generates task cluster reconstruction result;And according to task execution state result generates feedback correction result, updates the time window of subsequent scheduling period, trusted level, synchronization threshold and reallocation cost parameter, can reduce the waiting risk and task chain delay risk caused by the staggered arrival of multiple users.

Description

Technical Field

[0001] This invention relates to the field of mobile charging dispatch data processing technology, and more specifically, to an intelligent charging task scheduling method and system for mobile charging vehicles. Background Technology

[0002] With the increasing number of new energy vehicles, fixed charging facilities are insufficient to meet users' immediate charging needs in all road sections, parking areas, and temporary charging scenarios. Mobile charging vehicles, as a dispatchable charging resource, can travel to service area nodes, parking area nodes, temporary parking nodes, or other candidate charging locations based on user charging requests to provide charging services for vehicles with low battery levels. Existing mobile charging vehicle scheduling technologies typically acquire data such as user location, user destination, user's current battery level, charging demand, mobile charging vehicle's current location, available service time period, and candidate charging locations. Based on factors such as path cost, waiting cost, vehicle travel cost, candidate location capacity, and mobile charging vehicle service capability, they construct a correspondence between users, candidate charging locations, service time periods, and mobile charging vehicles, thereby forming a charging task allocation result or a mobile charging vehicle task chain result.

[0003] The existing technology has the following shortcomings: In scenarios where multiple users share the same mobile charging vehicle or the same candidate charging location, existing scheduling methods often use deterministic estimated arrival times or static path durations as the basis for generating task clusters. These methods typically do not fully consider the impact of factors such as real-time road speed changes, historical travel time discrepancies, congestion level changes, candidate charging location entry delays, user location update intervals, speed deviations, and route deviations on the actual arrival times. When the actual arrival times of multiple users within the same initial task cluster are staggered, or when the actual arrival time of the mobile charging vehicle does not match the user's arrival time, it can easily lead to increased user waiting time, idle mobile charging vehicles, task cancellation, and delays in subsequent task chains. Existing scheduling methods lack a mechanism for closed-loop correction of reliable time windows, synchronization thresholds, and redistribution costs based on actual execution deviations, which means that the problems of time window prediction deviations and underestimation of synchronization risks may still occur repeatedly in subsequent scheduling cycles.

[0004] To address the above problems, this invention proposes a solution. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent energy replenishment task scheduling method and system for mobile charging vehicles, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A method for intelligent energy replenishment task scheduling for mobile charging vehicles, comprising the following steps; Step S1: Obtain user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generate user reachable candidate location results; construct initial user allocation variables and initial vehicle dispatch variables; generate initial candidate task cluster results; calculate initial cost results; use formulas to generate initial scheduling evaluation results; aiming for a smaller initial scheduling evaluation result; combine user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints to generate initial collaborative scheduling results. Step S2: Based on the initial collaborative scheduling results, obtain the traffic status results and driving stability results, generate the user's baseline travel time results and arrival time fluctuation results, generate the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generate the user's reliable time window results by combining the arrival time fluctuation results, generate the mobile charging vehicle's reliable arrival time window results, and generate the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Step S3: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user's synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user's synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result and the impact result of the subsequent task chain, and divide the synchronization risk result into stable task cluster results, adjustable task cluster results and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results and perform constraint verification. Step S4: Obtain the task execution status result, generate the time window hit result and the waiting risk verification result, generate the feedback correction result based on the time window hit result and the waiting risk verification result, the feedback correction result includes the time window width correction result, the trust level correction result, the synchronization threshold correction result and the reallocation cost correction result, and write the feedback correction result into the scheduling parameter update result.

[0007] In a preferred embodiment, step S1 includes the following: When generating candidate locations reachable by the user, obtain the first travel cost result from the user's current location to the candidate charging location and the second travel cost result from the candidate charging location to the destination; Compare the user's current battery level with the battery consumption result corresponding to the first driving cost result. If the user's current battery level can support reaching the candidate charging location, add the candidate charging location to the user's reachable candidate location result. Based on the user's energy replenishment needs and the second driving cost, determine whether the user can reach their destination after completing the energy replenishment; If the location cannot be reached, the candidate charging location will be removed from the results of the candidate locations reachable by the user. The initial cost results include user detour cost results, user waiting cost results, mobile charging vehicle travel cost results, and mobile charging vehicle waiting cost results; The user's detour cost result is determined based on the first travel cost result, the second travel cost result, and the user's lowest path cost result; The user waiting cost is determined based on the user's initial estimated arrival time and the mobile charging vehicle's initial estimated arrival time. The mobile charging vehicle's operating cost is determined based on the path cost from its current location or previous task location to the candidate charging location. The waiting cost of the mobile charging vehicle is determined based on the initial estimated arrival time of the mobile charging vehicle and the initial estimated arrival time of the user.

[0008] In a preferred embodiment, step S2 includes the following: When generating the user's baseline travel time result, obtain the segment distance result and real-time segment speed result of each segment in the user's recommended route result; Divide the road segment distance result by the real-time road segment speed result to obtain the estimated travel time of the road segment, and then sum and overlay the entry delay result of the candidate charging location to generate the user's baseline travel time result. Arrival time fluctuation results are generated based on historical travel time discrepancy results, congestion level correction results, road segment speed fluctuation correction results, location update correction results, speed deviation correction results, and route deviation correction results; Based on the user's baseline trusted arrival time result, and combined with the arrival time fluctuation result, the user's trusted time window result is generated. The user's trusted time window result includes the earliest trusted arrival time result and the latest trusted arrival time result. When generating the reliable time window result of the mobile charging vehicle arrival, obtain the recommended path result of the mobile charging vehicle from the current location or the location of the previous task cluster to the current candidate charging location, and obtain the remaining service time result of the preceding task. The sum of the travel time of the mobile charging vehicle's recommended route result and the remaining service time of the preceding task is used as the baseline travel time result of the mobile charging vehicle. The arrival reliable time window result of the mobile charging vehicle is generated by combining the mobile charging vehicle's route fluctuation result.

[0009] In a preferred embodiment, step S3 includes the following: When calculating the user synchronization overlap result, the ratio of the length of the common intersection interval of the trusted time window results of all users in the task cluster to the length of the union interval is calculated, and this ratio is used as the user synchronization overlap result. When calculating the synchronization matching result of the mobile charging vehicle, the overlap interval length between the arrival reliable time window result of the mobile charging vehicle and the reliable time window result of each user is calculated, and the synchronization matching result of the mobile charging vehicle is generated based on the overlap interval length and the number of users. For results that need to be split into task clusters, perform synchronization risk user identification processing: For each user within a task cluster, calculate the average overlap of their time window with other users, and calculate the increase in waiting time for the mobile charging vehicle and other users caused by the user remaining in the current task cluster; If the average overlap of the time window is low and the increase in waiting time is high, then a result for users with synchronization risk is generated. Perform reallocation processing on the results of users with synchronous risks. The reallocation processing includes allocation of adjacent service periods, allocation of adjacent candidate charging locations, and allocation of adjacent mobile charging vehicle task chains. Calculate the reallocation cost results. The reallocation cost results include the user's new detour cost, the user's new waiting cost, the mobile charging vehicle's new driving cost, and the subsequent task delay cost. The reallocation method with the lower reallocation cost and the ability to reduce synchronization risk is selected to generate the user reallocation results.

[0010] In a preferred embodiment, step S4 includes the following: When generating time window hit results, for each user, the result of the user's actual arrival time is compared with the result of the user's trusted time window; If the user's actual arrival time is within the user's trusted time window, then a user time window hit result is generated. If the result is earlier than the earliest reliable arrival time, a user early arrival bias result is generated; if the result is later than the latest reliable arrival time, a user late arrival bias result is generated. For mobile charging vehicles, the actual arrival time of the mobile charging vehicle is compared with the reliable arrival time window result of the mobile charging vehicle to generate the mobile charging vehicle time window hit result, the mobile charging vehicle early deviation result, or the mobile charging vehicle late deviation result. When generating the waiting risk verification result, the actual waiting risk result is generated based on the user's actual waiting time and the mobile charging vehicle's actual waiting time. Based on the waiting risk results after reconstruction, predict the synchronization risk results are generated, and the actual waiting risk results are compared with the predicted synchronization risk results. When generating feedback correction results, increase the arrival time fluctuation result in the corresponding scenario when the user's lateness deviation result occurs frequently in a certain scenario; When a certain data source continuously causes a large prediction deviation, the confidence level contribution of the corresponding data source is reduced, and the synchronization threshold result of the next scheduling cycle is adjusted according to the difference between the actual waiting risk result and the predicted synchronization risk result. When a certain type of redistribution method frequently causes new user waiting or task delays in actual implementation, the redistribution cost corresponding to that type of redistribution method should be increased.

[0011] An intelligent energy replenishment task scheduling system for mobile charging vehicles includes: an initial scheduling module, a time window generation module, a task cluster reconstruction module, and a closed-loop correction module, with signal connections between the modules; Initial scheduling module: Obtains user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generates user reachable candidate location results; constructs initial user allocation variables and initial vehicle dispatch variables; generates initial candidate task cluster results; calculates initial cost results; generates initial scheduling evaluation results using formulas; with the goal of minimizing the initial scheduling evaluation results, it generates initial collaborative scheduling results by combining user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints. Time window generation module: Based on the initial collaborative scheduling results, it obtains the traffic status results and driving stability results, generates the user's baseline travel time results and arrival time fluctuation results, generates the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generates the user's reliable time window results by combining the arrival time fluctuation results, generates the mobile charging vehicle's reliable arrival time window results, and generates the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Task Cluster Reconstruction Module: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result, and the impact result of subsequent task chains, and divide the task cluster results into stable task cluster results, adjustable task cluster results, and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results, and perform constraint verification. Closed-loop correction module: Obtains task execution status results, generates time window hit results and waiting risk verification results, generates feedback correction results based on time window hit results and waiting risk verification results, including time window width correction results, trust level correction results, synchronization threshold correction results and reallocation cost correction results, and writes the feedback correction results into the scheduling parameter update results.

[0012] The technical effects and advantages of the intelligent energy replenishment task scheduling method for mobile charging vehicles of the present invention are as follows: This invention employs a continuous processing approach—initial scheduling, reliable time window generation, task cluster reconstruction, and closed-loop correction—to prevent mobile charging tasks from being allocated solely based on user location, vehicle location, and static path cost. Instead, after generating the initial collaborative scheduling result, it further incorporates real-time road speed results, historical travel time results, congestion level results, candidate charging location entry delay results, and user driving stability results to perform interval-based predictions of the arrival times of users and mobile charging vehicles at candidate charging locations. This yields reliable time window results that reflect road traffic fluctuations and user driving uncertainties. Synchronization risk is then generated using user synchronization overlap results, mobile charging vehicle synchronization matching results, reliability level results, and the impact of subsequent task chains. The system can identify risks such as asynchronous user arrivals, mobile charging vehicles waiting in vacancies, or disruptions to subsequent task chains within the same task cluster. It then performs different processing on task clusters with different risk states, including retention, service time adjustment, or reallocation of users with synchronization risks. This improves the synchronization reliability during task cluster execution. By reverse-engineering the task execution status results, it generates time window width correction results, reliability level correction results, synchronization threshold correction results, and reallocation cost correction results. This allows subsequent scheduling cycles to update parameters based on actual arrival deviations and actual waiting risks, avoiding long-term reliance on fixed thresholds and static empirical parameters during scheduling. This enhances the scheduling adaptability of mobile charging vehicles under different road conditions, different candidate charging locations, and different user driving behaviors. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the structure of an intelligent energy replenishment task scheduling method for mobile charging vehicles according to the present invention.

[0014] Figure 2 This is a schematic diagram for generating a reliable time window. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] Example: Please refer to Figures 1-2 As shown, this invention discloses an intelligent energy replenishment task scheduling method for mobile charging vehicles, including the following steps: Step S1: Obtain user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generate user reachable candidate location results; construct initial user allocation variables and initial vehicle dispatch variables; generate initial candidate task cluster results; calculate initial cost results; use formulas to generate initial scheduling evaluation results; aiming for a smaller initial scheduling evaluation result; combine user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints to generate initial collaborative scheduling results. Step S2: Based on the initial collaborative scheduling results, obtain the traffic status results and driving stability results, generate the user's baseline travel time results and arrival time fluctuation results, generate the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generate the user's reliable time window results by combining the arrival time fluctuation results, generate the mobile charging vehicle's reliable arrival time window results, and generate the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Step S3: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user's synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user's synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result and the impact result of the subsequent task chain, and divide the synchronization risk result into stable task cluster results, adjustable task cluster results and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results and perform constraint verification. Step S4: Obtain the task execution status result, generate the time window hit result and the waiting risk verification result, generate the feedback correction result based on the time window hit result and the waiting risk verification result, the feedback correction result includes the time window width correction result, the trust level correction result, the synchronization threshold correction result and the reallocation cost correction result, and write the feedback correction result into the scheduling parameter update result.

[0017] In step S1, the user's energy replenishment request result, the mobile charging vehicle status result, and the candidate charging location result are obtained. The user-reachable candidate location result is generated. Initial user allocation variables and initial vehicle dispatch variables are constructed. Initial candidate task cluster results are generated. Initial cost results are calculated. Initial scheduling evaluation results are generated using a formula. With the goal of minimizing the initial scheduling evaluation result, and considering user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service count constraints, and candidate charging location capacity constraints, initial collaborative scheduling results are generated. Specific content includes: This step establishes the initial scheduling foundation required for subsequent trusted time window generation and task cluster reconstruction, and obtains the results of user recharging requests, mobile charging vehicle status, and candidate charging locations within the current scheduling period. The results of a user's power replenishment request include the user set, the user's current location, the user's departure time, the user's destination, the user's expected arrival time, the user's current battery level, the user's power replenishment needs, and the user's lowest path cost. The user set result is denoted as the user set, and any user in the user set is denoted as a user. The user's current location result is used to indicate the road network node and corresponding time when the user initiates the charging request; the user's departure time result is used to indicate the time when the user starts executing this trip from the current location; the user's destination result is used to indicate the target node that the user needs to reach after completing the charging; the user's expected arrival time result is used to indicate the latest allowed time for the user to arrive at the destination; the user's current battery level result is used to determine whether the user can reach the candidate charging location; the user's charging demand result is used to indicate the amount of energy or charging time that the user needs to obtain at the temporary charging meeting point; the user's minimum path cost result is used to indicate the minimum travel cost when the user directly reaches the destination from the current location, and this minimum travel cost serves as the benchmark for subsequently calculating the user's detour cost result; The mobile charging vehicle status results include the mobile charging vehicle set result, the current location result of the mobile charging vehicle, the service time period result of the mobile charging vehicle, the parallel service capability result of the mobile charging vehicle, the maximum number of service times of the mobile charging vehicle, and the current task chain result of the mobile charging vehicle. The mobile charging vehicle set result is denoted as the mobile charging vehicle set, and any mobile charging vehicle in the mobile charging vehicle set is denoted as a mobile charging vehicle. The current location result of the mobile charging vehicle indicates the road network node where the mobile charging vehicle is located at the beginning of the current scheduling cycle; the service time period result of the mobile charging vehicle indicates the time period during which the mobile charging vehicle can perform charging tasks; the parallel service capability result of the mobile charging vehicle indicates the maximum number of users that the same mobile charging vehicle can serve at the same candidate charging location and within the same service time period; the maximum number of service times of the mobile charging vehicle indicates the maximum number of service time periods that the same mobile charging vehicle can execute within the current scheduling cycle; the current task chain result of the mobile charging vehicle indicates the order of tasks that the mobile charging vehicle has confirmed but not yet completed. When the current scheduling cycle is the initial scheduling cycle, the current task chain result of the mobile charging vehicle can be an empty task chain result; when the current scheduling cycle is the rolling scheduling cycle, the current task chain result of the mobile charging vehicle is used as the constraint basis for subsequent insertion or adjustment of tasks. The candidate charging location results include the candidate charging location set results and the candidate charging location capacity results. The candidate charging location set results are denoted as the candidate charging location set, and any candidate charging location in the candidate charging location set is denoted as a candidate charging location. The candidate charging location results can be formed by service area nodes, parking area nodes, temporary parking nodes, or other road network nodes that allow user vehicles and mobile charging vehicles to meet. The candidate charging location capacity results are used to represent the maximum number of users that can be allocated to the candidate charging location during the same service period. The candidate charging location capacity results are used in this step to constrain the user scale of the initial task cluster results. After obtaining the user's energy replenishment request result, the mobile charging vehicle status result, and the candidate charging location result, a user reachable candidate location result is generated for each user. First, the first driving cost result from the user's current location to each candidate charging location is obtained. The first driving cost result is used to represent the driving distance, driving time, or energy consumption conversion cost required for the user to travel from the current location to the candidate charging location. At the same time, the second driving cost result of the user from the candidate charging location to the destination is obtained. The second driving cost result is used to represent the driving distance, driving time or energy consumption cost of the user from the candidate charging location to the destination after completing the charging. The first and second driving cost results are both calculated based on the road network path. If there are multiple optional paths, the path with the lower path cost in the current scheduling cycle is selected as the basis for calculation. Compare the user's current battery level with the battery consumption result corresponding to the first travel cost result. If the user's current battery level can support the user to reach the candidate charging location, add the candidate charging location to the user's reachable candidate location result. If the user's current battery level does not allow the user to reach the candidate charging location, then the candidate charging location will be excluded. Based on the user's energy replenishment needs and the second driving cost, determine whether the user can reach their destination after completing energy replenishment at the candidate charging location. If the destination cannot be reached, the candidate charging location will be removed from the results of the candidate locations reachable by the user. After generating the candidate locations reachable by users, we construct the initial user allocation variable results and the initial vehicle dispatch variable results. The initial user allocation variable results are used to indicate whether a user is assigned to a candidate charging location and a certain service period. When a user is assigned to the candidate charging location and receives charging during the service period, the corresponding initial user allocation variable result is a valid allocation result; Otherwise, the allocation result is invalid. The initial vehicle dispatch variable result is used to indicate whether a mobile charging vehicle is dispatched to a candidate charging location and performs energy replenishment service during a certain service period. When a mobile charging vehicle is dispatched to the candidate charging location and performs service during the service period, the corresponding initial vehicle dispatch variable result is a valid dispatch result; otherwise, it is an invalid dispatch result. Initial candidate task cluster results are generated based on the initial user allocation variable results and the initial vehicle dispatch variable results. Each initial candidate task cluster result includes candidate charging location results, service period results, assigned user set results, corresponding mobile charging vehicle results, and initial estimated arrival time results. It should be noted that the assigned user set consists of users assigned to the same candidate charging location and within the same service period; the corresponding mobile charging vehicle result is the mobile charging vehicle dispatched to the candidate charging location during the service period; the initial estimated arrival time result includes the user's initial estimated arrival time result and the mobile charging vehicle's initial estimated arrival time result. The user's initial estimated arrival time result is determined based on the travel time corresponding to the user's departure time result and the first travel cost result; the mobile charging vehicle's initial estimated arrival time result is determined based on the travel time from the mobile charging vehicle's current location result or the previous task location result to the candidate charging location. Calculate the initial cost results, which include the user detour cost results, the user waiting cost results, the mobile charging vehicle travel cost results, and the mobile charging vehicle waiting cost results. The user detour cost results are determined based on the first travel cost results, the second travel cost results, and the user's lowest path cost results. When the path cost of a user reaching their destination via a candidate charging location is higher than the user's lowest path cost, the difference is taken as the user's detour cost. The user's waiting cost is determined based on the user's initial estimated arrival time and the mobile charging vehicle's initial estimated arrival time. When a user arrives at a candidate charging location earlier than the mobile charging vehicle, the time difference between the two is taken as the user's waiting time and converted into the user's waiting cost result according to the user's waiting cost coefficient. The mobile charging vehicle's travel cost result is determined based on the path cost of the mobile charging vehicle from its current location or the previous task location to the candidate charging location. The waiting cost of the mobile charging vehicle is determined based on the initial estimated arrival time of the mobile charging vehicle and the initial estimated arrival time of the user. When the mobile charging vehicle arrives earlier than the user in the task cluster, the waiting time of the mobile charging vehicle is determined based on the arrival time of the user in the task cluster and converted into the waiting cost of the mobile charging vehicle. Before calculating the initial cost results, the user detour cost results, user waiting cost results, mobile charging vehicle travel cost results, and mobile charging vehicle waiting cost results have been obtained or calculated. The initial scheduling evaluation results are generated using the following formula: ;in, This is the initial scheduling evaluation result; The cost of detours for users; Waiting for the user to pay for the results; The results regarding the operating costs of mobile charging vehicles; Waiting for the cost results of mobile charging vehicles; Weighted results for user detour costs; Waiting for the cost weighting results for the user; Weighting the operating costs of mobile charging vehicles; Waiting for the cost weighting result for the mobile charging vehicle, the weighting result is set before the start of the current scheduling cycle and remains unchanged within the same scheduling cycle; With the goal of minimizing the initial scheduling evaluation result, the initial collaborative scheduling result is generated by combining the user unique allocation constraint result, the user power reach constraint result, the user destination reach constraint result, the mobile charging vehicle single-period service constraint result, the mobile charging vehicle maximum service times constraint result, and the candidate charging location capacity constraint result. The user unique allocation constraint result is used to ensure that each user is only allocated to one candidate charging location and one service period in the current scheduling cycle. The user's battery reachability constraint results are used to ensure that the user's current battery level can support their arrival at the assigned candidate charging location. The user destination arrival constraint result is used to ensure that the user can reach the destination before the expected arrival time result after completing the refueling; The mobile charging vehicle's single-time service constraint results are used to ensure that the same mobile charging vehicle only goes to one candidate charging location within the same service time period. The constraint result of the maximum number of times a mobile charging vehicle can serve is used to ensure that the number of service periods executed by the same mobile charging vehicle in the current scheduling cycle does not exceed the maximum number of times a mobile charging vehicle can serve. The capacity constraint results of candidate charging locations are used to ensure that the number of users allocated to the same candidate charging location and the same service period does not exceed the smaller of the candidate charging location capacity result and the mobile charging vehicle parallel service capacity result. This step outputs the initial collaborative scheduling results, which include user candidate allocation results, initial task cluster results, mobile charging vehicle candidate dispatch results, initial task chain results, user initial estimated arrival time results, and mobile charging vehicle initial estimated arrival time results. User candidate allocation results are used to determine the candidate charging location and service time period assigned to each user; initial task cluster results are used to determine which users are served by the same mobile charging vehicle at the same candidate charging location; mobile charging vehicle candidate dispatch results are used to determine the mobile charging vehicle corresponding to each task cluster; initial task chain results are used to determine the order in which each mobile charging vehicle executes each task cluster. This step unifies the user charging request results, mobile charging vehicle status results, and candidate charging location results into initial collaborative scheduling results within the current scheduling cycle. This establishes a calculable and constrained initial correspondence between users, candidate charging locations, mobile charging vehicles, and service periods. Under constraints such as user battery availability, destination arrival, mobile charging vehicle service per time period, maximum number of mobile charging vehicle services, and candidate charging location capacity, initial task cluster results and initial task chain results are pre-formed. This avoids indiscriminate searching across all road network nodes and all vehicle combinations in subsequent reliable time window calculations. It provides scheduling-significant candidate objects for subsequent arrival time fluctuation analysis, enabling step S2 to expand the reliable time window around the determined user candidate allocation results and initial task cluster results, thereby improving the targeting and computational efficiency of subsequent synchronization risk assessment.

[0018] In step S2, based on the initial coordinated scheduling results, traffic status results and driving stability results are obtained, and user baseline travel time results and arrival time fluctuation results are generated. Based on the current scheduling time results and user baseline travel time results, user baseline reliable arrival time results are generated. Combined with arrival time fluctuation results, user reliable time window results are generated, and mobile charging vehicle arrival reliable time window results are generated. Finally, based on data integrity results, data freshness results, and prediction consistency results, a reliability level result is generated. Specific content includes: After generating the initial collaborative scheduling results, the arrival times of users and mobile charging vehicles at candidate charging locations are predicted in intervals to generate reliable time window results. The inputs of this step are the user candidate allocation results, initial task cluster results, initial task chain results, initial estimated arrival times of users and initial estimated arrival times of mobile charging vehicles output from step S1. This step obtains traffic status results and driving stability results, and expands the deterministic predicted arrival time based on these results to obtain a reliable time window result that can represent time fluctuations; Based on the user candidate allocation results, the candidate charging location corresponding to each user is determined, and the user recommended path result is determined based on the user's current location result and the candidate charging location result. The user recommended path result is the path result of the user from the current location to the assigned candidate charging location, which includes several road segment results. For each road segment, obtain the road segment distance result, real-time road segment speed result, historical travel time result, congestion level result, and road segment speed fluctuation result. The road segment distance result is used to represent the length of the road segment. Real-time road segment speed results are used to represent the current traffic speed of the road segment; historical travel time results are used to represent the historical travel time statistics of the road segment during the same time period; congestion level results are used to represent the current traffic status of the road segment; road segment speed fluctuation results are used to represent the speed change of the road segment within the most recent observation window; Obtain the entry delay result of the candidate charging location. The entry delay result of the candidate charging location is used to indicate the queuing, turning, deceleration or stopping delay that may occur when a user enters the assigned candidate charging location. The entry delay result of the candidate charging location can be generated based on the historical entry time result of the candidate charging location in the current time period, the current vehicle aggregation status result and the current road entrance traffic status result. If the candidate charging location is a service area or parking area, the entry delay result of the candidate charging location is also used to reflect the increase in time caused by entrance congestion or parking waiting. If the real-time entry status cannot be obtained at present, the median entry time of the candidate charging location in the same time period will be used as the entry delay result of the candidate charging location. It also obtains user driving stability results, which include user location update interval results, user speed deviation results, and user route deviation results. The user location update interval results are obtained from the time difference between two adjacent user location receptions. The user speed deviation result is obtained from the difference between the user's actual driving speed and the predicted speed corresponding to the user's recommended route result; The user route deviation result is determined by the deviation distance or the number of deviation segments between the user's current location result and the user's recommended path result. When the user location update interval is large, it indicates that the user status observation is discontinuous. When the user speed deviation is large, it means that the user's actual travel status is inconsistent with the recommended path prediction. When the user's route deviates significantly from the result, it indicates that the user has strayed from the initially recommended path; Generate the user's baseline travel time result, which is the baseline time required for the user to travel from the current location to the assigned candidate charging location under the current travel conditions. Calculate the estimated travel time result for each segment of the user's recommended path result. For any road segment, if the real-time road segment speed result is valid, the estimated travel time for that road segment is obtained by dividing the road segment distance result by the real-time road segment speed result. If the real-time road segment speed result is invalid, the historical travel time result will be used as the estimated travel time result for that road segment. The estimated travel time results for each road segment are summed up, and the entry delay results for candidate charging locations are overlaid to obtain the user's baseline travel time result; The user's baseline travel time is calculated using the following formula. ;in, The baseline travel time for user P to reach candidate charging location k; This represents the set of road segments that user P will travel from their current location to the candidate charging location k; r represents any road segment in the set of road segments. This represents the segment distance result for segment r. This is the real-time segment speed result for segment r. When the real-time segment speed result is invalid, it is replaced by the historical traffic speed. The candidate charging location k is included in the delayed results; Furthermore, arrival time fluctuation results are generated. These results describe the possible fluctuation range of the user's actual arrival time at the candidate charging location relative to the baseline reliable arrival time. The arrival time fluctuation results are generated based on historical travel time discrepancies, congestion level correction results, road segment speed fluctuation correction results, location update correction results, speed deviation correction results, and route deviation correction results. The historical travel time discrepancies are obtained by summing the dispersion of the historical travel times of each road segment in the user's recommended path results. The congestion level correction result is determined based on the congestion level result; the higher the congestion level, the greater the congestion level correction result. The correction result for road segment speed fluctuation is determined based on the road segment speed fluctuation result; the greater the speed fluctuation, the greater the correction result for road segment speed fluctuation. The location update correction result is determined based on the user's location update interval; the larger the location update interval, the larger the location update correction result. The speed deviation correction result is determined based on the user's speed deviation result; the larger the speed deviation, the larger the speed deviation correction result. The route deviation correction result is determined based on the user's route deviation result; the more obvious the route deviation, the greater the route deviation correction result. Generate a reliable arrival time result for the user baseline based on the current scheduling time result and the user baseline travel time result. The current scheduling time result is the time result when this step is executed. The user's baseline reliable arrival time result is equal to the sum of the current scheduling time result and the user's baseline travel time result; Based on the user's baseline trusted arrival time result, and combined with the arrival time fluctuation result, the user's trusted time window result is generated. The user's trusted time window result includes the earliest trusted arrival time result and the latest trusted arrival time result. The earliest reliable arrival time result is obtained by subtracting the arrival time fluctuation result from the user's baseline reliable arrival time result, and the latest reliable arrival time result is obtained by adding the arrival time fluctuation result to the user's baseline reliable arrival time result. Generate the user's reliable time window results using the following formula: ;in, The reliable time window result for user P to reach candidate charging location k; The baseline reliable arrival time result for user P to reach candidate charging location k; The arrival time fluctuation results for user P to candidate charging location k are given. The left endpoint of the user's reliable time window result is the earliest reliable arrival time result, and the right endpoint is the latest reliable arrival time result. It also generates a reliable time window result for the arrival of the mobile charging vehicle based on the initial task chain result, and determines the position result of the corresponding mobile charging vehicle before the execution of the task cluster for any initial task cluster. If the task cluster is the first task cluster in the current task chain result of the mobile charging vehicle, then the position result is the current position result of the mobile charging vehicle; If the task cluster is not the first task cluster, then the location result is the candidate charging location result of the previous task cluster, and the recommended path result of the mobile charging vehicle from the location result to the current candidate charging location result is obtained. Obtain the road segment distance, real-time road segment speed, historical travel time, congestion level, and road segment speed fluctuation results along the route; At the same time, the remaining service time result of the preceding task is obtained. The remaining service time result of the preceding task is used to represent the recharging service time that the mobile charging vehicle still needs to complete the previous task cluster. The sum of the travel time of the mobile charging vehicle's recommended path result and the remaining service time result of the preceding task is used as the benchmark travel time result of the mobile charging vehicle. The reliable arrival time window results of the mobile charging vehicle are generated by combining the path fluctuation results of the mobile charging vehicle. The reliable arrival time window results of the mobile charging vehicle include the earliest reliable arrival time result and the latest reliable arrival time result of the mobile charging vehicle. Furthermore, a trust level result is generated, which is used to represent the reliability of the trust time window result for each user and the trust time window result for each mobile charging vehicle. The trust level result is generated based on the data integrity result, data freshness result, and prediction consistency result. The data integrity result indicates whether the real-time road segment speed result, historical travel time result, congestion level result, location update result, and candidate charging location entry delay result are complete; The data freshness result represents the time interval between the above data and the result at the current scheduling time; Prediction consistency results indicate the degree of difference between real-time traffic predictions, historical traffic predictions, and actual user driving trends. When data integrity results are high, data freshness results are high, and prediction consistency results are high, the credibility level is high. When any one of these factors is lower, the confidence level decreases. This step ultimately outputs the user's trusted time window result, the mobile charging vehicle's trusted time window result, and the trust level result. The user's trusted time window result represents the time interval between the user's arrival at the assigned candidate charging location; the mobile charging vehicle's trusted time window result represents the time interval between the mobile charging vehicle's arrival at the assigned candidate charging location; and the trust level result represents the reliability of the above time window results. This step transforms the initial estimated arrival time of users and the initial estimated arrival time of mobile charging vehicles obtained from the deterministic path time in step S1 into reliable time window results for users and reliable time window results for mobile charging vehicles that can characterize road traffic fluctuations and driving status uncertainties. By using real-time road segment speed results, historical travel time results, congestion level results, candidate charging location entry delay results, and user driving stability results, the single arrival time is corrected in intervals. This avoids underestimating the risks of users arriving late, arriving early, or mobile charging vehicles being empty by using only a fixed estimated arrival time. The generation of reliability level results enables subsequent steps to distinguish the reliability of different time window results, thus providing a quantitative basis for step S3 to determine whether multiple users within the same initial task cluster have the conditions for synchronous arrival.

[0019] In step S3, for each initial task cluster result, the user's trusted time window result and the mobile charging vehicle's trusted arrival time window result are extracted. The user synchronization overlap result and the mobile charging vehicle synchronization matching result are calculated. Based on the user synchronization overlap result, the mobile charging vehicle synchronization matching result, the trust level result, and the impact result of subsequent task chains, a synchronization risk result is generated. According to the comparison between the synchronization risk result and the preset synchronization threshold result, the results are divided into stable task cluster results, adjustable task cluster results, and task cluster results that need to be split. Service time adjustment processing is performed on the adjustable task cluster results, and synchronization risk user identification processing and reallocation processing are performed on the task cluster results that need to be split. The task cluster reconstruction result is generated and constraint verification is performed. The specific contents include: This step performs synchronization overlap analysis and synchronization risk assessment on the results of each initial task cluster, generates task cluster reconstruction results, determines whether users within the same task cluster can arrive synchronously in time based on the initial task cluster results, and performs adjustments or splits on task clusters with synchronization risks. Extract the results of any initial task cluster, which include candidate charging location results, service period results, assigned user set results, corresponding mobile charging vehicle results, user initial estimated arrival time results, mobile charging vehicle initial estimated arrival time results, and initial task chain results. Extract the user trust time window result for each user in the assigned user set result, and extract the mobile charging vehicle arrival trust time window result, the trust level result for each user, and the trust level result for the mobile charging vehicle. Calculate the user synchronization overlap result. The user synchronization overlap result is used to represent the degree of overlap between the user reliable time window results of multiple users within the same task cluster. When the user reliable time window results of multiple users have a long common overlap interval, it means that these users are more likely to arrive at the candidate charging location at a similar time. When the trusted time window results of multiple users have no common overlapping interval, or the length of the common overlapping interval is short, it indicates that there is a risk of asynchronous arrival of users in the task cluster. First, calculate the common intersection interval of the trusted time window results of all users in the task cluster, then calculate the union interval of the trusted time window results of all users, and use the ratio between the length of the common intersection interval and the length of the union interval as the result of user synchronization overlap. Before calculating the user synchronization overlap result, the results of the user set within the task cluster, the user reliable time window, and the time interval length are all known. The user synchronization overlap result is calculated using the following formula: ;in, The user synchronization overlap result for task cluster a; is the set of assigned users in task cluster a; p is any user in the set of assigned users. The reliable time window result for user p arriving at candidate charging location k; This is the result for the length of the time interval; When there is no common intersection interval among the trusted time window results of all users, the length of the common intersection interval is set to zero, and the user synchronization overlap is set to zero. Subsequently, the mobile charging vehicle synchronization matching result is calculated. The mobile charging vehicle synchronization matching result is used to represent the degree of overlap between the mobile charging vehicle arrival reliable time window result and the user reliable time window result within the task cluster. Calculate the overlap interval between the arrival time window result of the mobile charging vehicle and the arrival time window result of each user, and generate the mobile charging vehicle synchronization matching result based on the length of the overlap interval and the number of users; When the result of the mobile charging vehicle arriving at the trusted time window effectively overlaps with the result of the trusted time window of most users, the synchronization matching result of the mobile charging vehicle is relatively high. When the mobile charging vehicle arrives at the trusted time window result significantly earlier or later than the trusted time window result of most users, the synchronization matching result of the mobile charging vehicle is low. A synchronization risk result is generated. The synchronization risk result is determined based on the user synchronization overlap result, the mobile charging vehicle synchronization matching result, the trust level result, and the impact result of subsequent task chains. The impact result of subsequent task chains is generated based on the initial task chain result and is used to indicate whether the delay of the current task cluster will affect the subsequent tasks of the same mobile charging vehicle. If there are multiple task clusters after the current task cluster, and the service time of the subsequent task clusters is tight, then the impact of the subsequent task chain on the result will be high. If the current task cluster is the last task cluster in the task chain, or if the subsequent task clusters have a large time redundancy, the impact of the subsequent task chain on the result will be lower. The lower the user synchronization overlap, the lower the mobile charging vehicle synchronization matching result, the lower the trust level result, and the higher the impact of the subsequent task chain on the result, the higher the synchronization risk result. Based on the synchronization risk results and the preset synchronization threshold results, the initial task cluster results are divided into stable task cluster results, adjustable task cluster results, and task cluster results that need to be split. The preset synchronization threshold results are the thresholds used to judge the synchronization reliability of task clusters in the current scheduling cycle. If the synchronization risk result is less than the first synchronization threshold result, then the task cluster is determined as a stable task cluster result; If the synchronization risk result is not less than the first synchronization threshold result and less than the second synchronization threshold result, then the task cluster is determined as an adjustable task cluster result. If the synchronization risk result is not less than the second synchronization threshold result, then the task cluster is determined as a task cluster that needs to be split, and the first synchronization threshold result is less than the second synchronization threshold result. For stable task cluster results, retain the candidate charging location results, assigned user set results, corresponding mobile charging vehicle results and service time period results generated in step S1, and append the user reliable time window results and the mobile charging vehicle arrival reliable time window results to the task execution information. For adjustable task cluster results, perform service time adjustment processing. Specifically, the service time adjustment result is determined based on the common overlapping interval of the trusted time window results of each user in the assigned user set result; If the result of the mobile charging vehicle arriving at the reliable time window is earlier than the result of the common overlapping interval, then the result of the mobile charging vehicle's target arrival time will be delayed. If the mobile charging vehicle arrives at the reliable time window later than the result of the common overlap interval, the priority of this task cluster in the initial task chain result is increased or the waiting redundancy between its predecessor tasks is shortened, so that the result of the mobile charging vehicle's target arrival time is closer to the result of the user's common overlap interval. For the results of task clusters that need to be split, perform synchronous risk user identification processing, calculate the average overlap of the time window of each user with other users in the task cluster, and calculate the increase in waiting time of mobile charging vehicle and other users caused by the user staying in the current task cluster. If a user's average time window overlap is low and the increase in waiting time it causes is high, then a synchronization risk user result is generated for that user. The synchronization risk user result indicates that the user is not suitable to continue to be retained in the current task cluster. The results of users with synchronous risks are reassigned. The reassignment process includes adjacent service time allocation, adjacent candidate charging location allocation, and adjacent mobile charging vehicle task chain allocation. The adjacent service time allocation is used to keep the candidate charging location unchanged while changing the service time. Adjacent candidate charging location assignment is used to select new candidate charging locations from the user's reachable candidate location results; Adjacent mobile charging vehicle task chain assignment is used to assign the user to the task chain result of another mobile charging vehicle; When performing the reallocation process, the reallocation cost results are calculated. The reallocation cost results include the user's new detour cost results, the user's new waiting cost results, the mobile charging vehicle's new driving cost results, and the subsequent task delay cost results. The user's new detour cost results are determined by the difference between the candidate charging location path cost after reallocation and the original allocated path cost. The waiting cost for new users is determined by the waiting difference between the reallocated user's reliable time window result and the mobile charging vehicle's reliable arrival time window result. The additional driving cost of the mobile charging vehicle is determined by the difference in path costs before and after the change in the mobile charging vehicle task chain result; The cost of subsequent task delays is determined by the impact of the current redistribution scheme on the service time of subsequent tasks of the mobile charging vehicle. A redistribution method with lower redistribution cost and reduced synchronization risk is selected to generate user redistribution results. Based on the stable task cluster results, service time adjustment results, and user reallocation results, task cluster reconstruction results are generated. The task cluster reconstruction results include reconstructed user grouping results, reconstructed charging assembly point results, reconstructed mobile charging vehicle dispatch results, reconstructed service time results, reconstructed task chain results, and reconstructed waiting risk results. The reconstructed user grouping results are used to represent the user combinations after adjustment or splitting. The reconstructed charging assembly point results are used to represent the candidate charging locations corresponding to each reconstructed task cluster. The reconstructed mobile charging vehicle dispatch results are used to represent the mobile charging vehicle corresponding to each reconstructed task cluster. The reconstructed service time result is used to represent the target time when the mobile charging vehicle arrives at the candidate charging location and begins service. The reconstructed task chain result is used to represent the order in which each mobile charging vehicle executes the reconstructed task cluster; The post-reconstruction waiting risk result is used to represent the waiting risk that may still exist after reconstruction. The task cluster reconstruction results are constrained and verified. The constraint verification includes user power availability verification, user destination arrival time verification, candidate charging location capacity verification, mobile charging vehicle single-period service verification, and mobile charging vehicle maximum service count verification. If the reconstructed task cluster results meet all constraints, they are used as the execution scheduling results. If any constraint is not met, return to the reallocation process and reselect the service period, candidate charging location, or mobile charging vehicle task chain until a task cluster reconstruction result that satisfies the constraints is generated. Based on the user reliable time window results and mobile charging vehicle arrival reliable time window results generated in step S2, this step performs synchronization reliability verification and reconstruction processing on the initial task cluster results formed in step S1. This ensures that the initial task cluster results not only meet the requirements of optimal static cost and basic constraint feasibility, but also meet the execution requirements of coordinated arrival times for multiple users. By identifying stable task cluster results, adjustable task cluster results, and task cluster results that need to be split through user synchronization overlap results, mobile charging vehicle synchronization matching results, and synchronization risk results, this step adopts retention, service time adjustment, or synchronization risk user reallocation processing for different task cluster states. This reduces the waiting risk caused by staggered arrival of users within the same charging assembly point and the risk of subsequent task chain delays. The task cluster reconstruction results formed in this step can serve as the scheduling basis for actual execution and provide a clear evaluation object for execution deviation verification and parameter iteration correction in step S4.

[0020] In step S4, the task execution status result is obtained, and the time window hit result and the waiting risk verification result are generated. Based on the time window hit result and the waiting risk verification result, a feedback correction result is generated. The feedback correction result includes the time window width correction result, the trust level correction result, the synchronization threshold correction result, and the reallocation cost correction result. The feedback correction result is written into the scheduling parameter update result, specifically including: After generating the task cluster reconstruction result and issuing it for execution, the task execution status result is continuously obtained. Based on the task execution status result, the calculation rules of the reliable time window result and the synchronization risk result in the subsequent scheduling cycle are corrected. This step forms a closed loop iteration that can update parameters based on the actual execution deviation. The task execution status results include the user's actual arrival time, the mobile charging vehicle's actual arrival time, the user's actual waiting time, the mobile charging vehicle's actual waiting time, the task's actual start time, the task's actual completion time, the user's route deviation, the task cancellation, and the subsequent task delay. The user's actual arrival time indicates the time when the user actually arrives at the reconstructed charging assembly point. The actual arrival time of the mobile charging vehicle is used to indicate the actual time when the mobile charging vehicle arrives at the reconstructed charging assembly point. The actual waiting time result is used to indicate the time a user waits for the mobile charging vehicle or for the charging to begin after arriving at the location. The actual waiting time of the mobile charging vehicle is used to represent the time it takes for the mobile charging vehicle to arrive and for the user to arrive or for the service to start after it arrives. The actual start time and actual completion time of the task are used to determine whether the task execution exceeds the service time after reconstruction. The user route deviation result is used to determine whether the user has deviated from the reconstructed charging assembly point path determined in step S3; The task cancellation result indicates whether the user has cancelled the charging service; the subsequent task delay result indicates whether the execution of the current task has caused a delay in the start time of the subsequent tasks of the mobile charging vehicle. Generate time window hit results. For each user, compare the user's actual arrival time result with the user's trusted time window result generated for that user. If the user's actual arrival time falls within the user's trusted time window, then a user time window hit result is generated. If the user's actual arrival time is earlier than the earliest reliable arrival time of the user's reliable time window, then a user advance bias result is generated. If the user's actual arrival time is later than the latest reliable arrival time of the user's reliable time window, then a user late arrival deviation result is generated. For mobile charging vehicles, the actual arrival time of the mobile charging vehicle is compared with the reliable arrival time window result of the mobile charging vehicle to generate the mobile charging vehicle time window hit result, the mobile charging vehicle early deviation result, or the mobile charging vehicle late deviation result. Generate waiting risk verification results. These results are used to compare the differences between the reconstructed waiting risk results and the actual waiting situation. Based on the actual waiting time of the user and the actual waiting time of the mobile charging vehicle, generate actual waiting risk results. Based on the waiting risk results after reconstruction, a predicted synchronization risk result is generated. If the actual waiting risk result is higher than the predicted synchronization risk result, it indicates that step S3 underestimated the synchronization risk of the task cluster. If the actual waiting risk is lower than the predicted synchronization risk, it means that the synchronization risk of the task cluster has been overestimated. If the task is canceled and the reason for the cancellation is related to excessive waiting or asynchronous arrival, the actual waiting risk is increased. If subsequent task delays occur, these delays will be included in the actual waiting risk outcome. After generating the actual waiting risk result and the predicted synchronization risk result, a feedback correction result is generated. The feedback correction result includes the time window width correction result, the confidence level correction result, the synchronization threshold correction result, and the reallocation cost correction result. The time window width correction result is used to correct the arrival time fluctuation result in step S2. When user late arrival deviation results frequently occur in a certain type of road segment, a certain level of congestion, or a certain type of candidate charging location, increase the arrival time fluctuation result in the corresponding scenario. When the actual arrival time of the user falls in the middle of the trusted time window for a long time, and the trusted time window is too wide, causing the task cluster to be over-split in step S3, the arrival time fluctuation results in the corresponding scenario are reduced. The trusted level correction results are used to correct the impact of data integrity results, data freshness results and prediction consistency results on the trusted level results in step S2. When a certain data source consistently causes large prediction deviations, the confidence level contribution of that data source is reduced. When a data source can reliably explain the actual arrival deviation, the credibility level contribution of that data source is increased. The synchronization threshold correction result is used to correct the judgment boundary between the stable task cluster result, the adjustable task cluster result, and the task cluster that needs to be split in step S3. In this step, the synchronization threshold result, the actual waiting risk result, and the predicted synchronization risk result used in the current scheduling cycle have been obtained. The synchronization threshold result is updated according to the following formula: ;in, The synchronization threshold result to be used in the next scheduling cycle; The synchronization threshold result used in the current scheduling cycle; Update the step size result for the threshold; To actually wait for the risk outcome; To predict the outcome of synchronous risk; When the actual waiting risk is higher than the predicted synchronization risk, the synchronization threshold for the next scheduling cycle is increased. When the actual waiting risk result is lower than the predicted synchronization risk result, the synchronization threshold result for the next scheduling cycle is reduced. The reallocation cost correction result is used to correct the calculation of the reallocation cost result in step S3. If a certain type of reallocation method frequently causes users to wait for new tasks, cancel tasks, or delay subsequent tasks in actual execution, the reallocation cost corresponding to that type of reallocation method will be increased. If a certain type of redistribution method can reliably reduce the actual waiting risk, then reduce the redistribution cost corresponding to that type of redistribution method. In the next scheduling cycle, step S3 can more accurately select adjacent service time allocation, adjacent candidate charging location allocation, or adjacent mobile charging vehicle task chain allocation. The feedback correction results are written into the scheduling parameter update results. The scheduling parameter update results include path time fluctuation parameter results, candidate charging location entry delay parameter results, user driving stability parameter results, trust level parameter results, synchronization threshold parameter results, and redistribution cost parameter results. The path time fluctuation parameter results are used in the subsequent step S2 to generate arrival time fluctuation results. The candidate charging location entry delay parameter result is used in subsequent step S2 to correct the candidate charging location entry delay result; The user driving stability parameter results are used in subsequent step S2 to adjust the impact of position update interval, speed deviation and route deviation on the reliable time window results; The trust level parameter results are used in the subsequent step S2 to generate trust level results; the synchronization threshold parameter results are used in the subsequent step S3 to divide the task cluster status. The results of the reallocation cost parameters are used in the subsequent step S3 to select the reallocation method; This step reverses the actual execution status after the task cluster reconstruction results are issued, applying it to the trusted time window generation rules and synchronization risk judgment rules. This creates a closed-loop self-calibration mechanism in the scheduling process. By analyzing the time window hit results, waiting risk verification results, and feedback correction results, it identifies the deviations between the user's trusted time window results, the mobile charging vehicle's trusted time window results, the reconstructed waiting risk results, and the actual execution status. Based on these deviations, it generates time window width correction results, trusted level correction results, synchronization threshold correction results, and redistribution cost correction results. Subsequent scheduling cycles can update the calculation processes of steps S2 and S3 based on path time fluctuation parameters, candidate charging location entry delay parameters, user driving stability parameters, trusted level parameters, synchronization threshold parameters, and redistribution cost parameters. This avoids the scheduling model relying on fixed thresholds and static empirical parameters in the long term, improving the adaptability of multi-user mobile charging tasks under different road conditions, different candidate charging locations, and different user driving behaviors.

[0021] This invention discloses an intelligent energy replenishment task scheduling system for mobile charging vehicles, comprising: an initial scheduling module, a time window generation module, a task cluster reconstruction module, and a closed-loop correction module, with signal connections between the modules; Initial scheduling module: Obtains user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generates user reachable candidate location results; constructs initial user allocation variables and initial vehicle dispatch variables; generates initial candidate task cluster results; calculates initial cost results; generates initial scheduling evaluation results using formulas; with the goal of minimizing the initial scheduling evaluation results, it generates initial collaborative scheduling results by combining user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints. Time window generation module: Based on the initial collaborative scheduling results, it obtains the traffic status results and driving stability results, generates the user's baseline travel time results and arrival time fluctuation results, generates the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generates the user's reliable time window results by combining the arrival time fluctuation results, generates the mobile charging vehicle's reliable arrival time window results, and generates the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Task Cluster Reconstruction Module: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result, and the impact result of subsequent task chains, and divide the task cluster results into stable task cluster results, adjustable task cluster results, and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results, and perform constraint verification. Closed-loop correction module: Obtains task execution status results, generates time window hit results and waiting risk verification results, generates feedback correction results based on time window hit results and waiting risk verification results, including time window width correction results, trust level correction results, synchronization threshold correction results and reallocation cost correction results, and writes the feedback correction results into the scheduling parameter update results.

[0022] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0023] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0024] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0025] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0026] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0027] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent energy replenishment task scheduling for mobile charging vehicles, characterized in that, Includes steps; Step S1: Obtain user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generate user reachable candidate location results; construct initial user allocation variables and initial vehicle dispatch variables; generate initial candidate task cluster results; calculate initial cost results; use formulas to generate initial scheduling evaluation results; aiming for a smaller initial scheduling evaluation result; combine user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints to generate initial collaborative scheduling results. Step S2: Based on the initial collaborative scheduling results, obtain the traffic status results and driving stability results, generate the user's baseline travel time results and arrival time fluctuation results, generate the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generate the user's reliable time window results by combining the arrival time fluctuation results, generate the mobile charging vehicle's reliable arrival time window results, and generate the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Step S3: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user's synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user's synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result and the impact result of the subsequent task chain, and divide the synchronization risk result into stable task cluster results, adjustable task cluster results and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results and perform constraint verification. Step S4: Obtain the task execution status result, generate the time window hit result and the waiting risk verification result, generate the feedback correction result based on the time window hit result and the waiting risk verification result, the feedback correction result includes the time window width correction result, the trust level correction result, the synchronization threshold correction result and the reallocation cost correction result, and write the feedback correction result into the scheduling parameter update result.

2. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 1, characterized in that, In step S1, when generating the candidate locations reachable by the user, the first driving cost result from the user's current location to the candidate charging location and the second driving cost result from the candidate charging location to the destination are obtained. Compare the user's current battery level with the battery consumption result corresponding to the first driving cost result. If the user's current battery level can support reaching the candidate charging location, add the candidate charging location to the user's reachable candidate location result. Based on the user's energy replenishment needs and the second driving cost, determine whether the user can reach their destination after completing the energy replenishment; If the location cannot be reached, the candidate charging location will be removed from the list of reachable locations.

3. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 2, characterized in that, The initial cost results include user detour cost results, user waiting cost results, mobile charging vehicle travel cost results, and mobile charging vehicle waiting cost results; The user's detour cost result is determined based on the first travel cost result, the second travel cost result, and the user's lowest path cost result; The user waiting cost is determined based on the user's initial estimated arrival time and the mobile charging vehicle's initial estimated arrival time. The mobile charging vehicle's operating cost is determined based on the path cost from its current location or previous task location to the candidate charging location. The waiting cost of the mobile charging vehicle is determined based on the initial estimated arrival time of the mobile charging vehicle and the initial estimated arrival time of the user.

4. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 1, characterized in that, In step S2, when generating the user's baseline travel time result, the segment distance result and real-time segment speed result of each segment in the user's recommended route result are obtained; Divide the road segment distance result by the real-time road segment speed result to obtain the estimated travel time of the road segment, and then sum and overlay the entry delay result of the candidate charging location to generate the user's baseline travel time result. The arrival time fluctuation results are generated based on historical travel time dispersion results, congestion level correction results, road segment speed fluctuation correction results, location update correction results, speed deviation correction results, and route deviation correction results.

5. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 4, characterized in that, Based on the user's baseline trusted arrival time result, and combined with the arrival time fluctuation result, the user's trusted time window result is generated. The user's trusted time window result includes the earliest trusted arrival time result and the latest trusted arrival time result. When generating the reliable time window result of the mobile charging vehicle arrival, obtain the recommended path result of the mobile charging vehicle from the current location or the location of the previous task cluster to the current candidate charging location, and obtain the remaining service time result of the preceding task. The sum of the travel time of the mobile charging vehicle's recommended route result and the remaining service time of the preceding task is used as the baseline travel time result of the mobile charging vehicle. The arrival reliable time window result of the mobile charging vehicle is generated by combining the mobile charging vehicle's route fluctuation result.

6. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 1, characterized in that, In step S3, when calculating the user synchronization overlap result, the ratio of the common intersection interval length of the trusted time window results of all users in the task cluster to the union interval length is calculated, and this ratio is used as the user synchronization overlap result. When calculating the synchronization matching result of the mobile charging vehicle, the overlap interval length between the arrival reliable time window result of the mobile charging vehicle and the reliable time window result of each user is calculated, and the synchronization matching result of the mobile charging vehicle is generated based on the overlap interval length and the number of users.

7. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 6, characterized in that, For results that need to be split into task clusters, perform synchronization risk user identification processing: For each user within a task cluster, calculate the average overlap of their time window with other users, and calculate the increase in waiting time for the mobile charging vehicle and other users caused by the user remaining in the current task cluster; If the average overlap of the time window is low and the increase in waiting time is high, then a result for users with synchronization risk is generated. Perform reallocation processing on the results of users with synchronous risks. The reallocation processing includes allocation of adjacent service periods, adjacent candidate charging locations, and adjacent mobile charging vehicle task chains, and calculate the reallocation cost results. The reallocation cost results include the user's new detour cost results, the user's new waiting cost results, the mobile charging vehicle's new driving cost results, and the subsequent task delay cost results. The reallocation method with the lower reallocation cost results and the ability to reduce synchronization risk results is selected to generate user reallocation results.

8. The intelligent energy replenishment task scheduling method for mobile charging vehicles according to claim 1, characterized in that, In step S4, when generating the time window hit result, for each user, the result of the user's actual arrival time is compared with the result of the user's trusted time window; If the user's actual arrival time is within the user's trusted time window, then a user time window hit result is generated. If the result is earlier than the earliest reliable arrival time, a user early arrival bias result is generated; if the result is later than the latest reliable arrival time, a user late arrival bias result is generated. For mobile charging vehicles, the actual arrival time of the mobile charging vehicle is compared with the reliable arrival time window result of the mobile charging vehicle to generate the mobile charging vehicle time window hit result, the mobile charging vehicle early deviation result, or the mobile charging vehicle late deviation result.

9. A method for intelligent energy replenishment task scheduling for mobile charging vehicles according to claim 8, characterized in that, When generating the waiting risk verification result, the actual waiting risk result is generated based on the user's actual waiting time and the mobile charging vehicle's actual waiting time. Based on the waiting risk results after reconstruction, predict the synchronization risk results are generated, and the actual waiting risk results are compared with the predicted synchronization risk results. When generating feedback correction results, increase the arrival time fluctuation result in the corresponding scenario when the user's lateness deviation result occurs frequently in a certain scenario; When a certain data source continuously causes a large prediction deviation, the confidence level contribution of the corresponding data source is reduced, and the synchronization threshold result of the next scheduling cycle is adjusted according to the difference between the actual waiting risk result and the predicted synchronization risk result. When a certain type of redistribution method frequently causes new user waiting or task delays in actual implementation, the redistribution cost corresponding to that type of redistribution method should be increased.

10. An intelligent energy replenishment task scheduling system for mobile charging vehicles, used to implement the intelligent energy replenishment task scheduling method for mobile charging vehicles as described in any one of claims 1-9, characterized in that... ； Initial scheduling module: Obtains user energy replenishment request results, mobile charging vehicle status results, and candidate charging location results; generates user reachable candidate location results; constructs initial user allocation variables and initial vehicle dispatch variables; generates initial candidate task cluster results; calculates initial cost results; generates initial scheduling evaluation results using formulas; with the goal of minimizing the initial scheduling evaluation results, it generates initial collaborative scheduling results by combining user unique allocation constraints, user energy reachability constraints, user destination reach constraints, mobile charging vehicle single-period service constraints, mobile charging vehicle maximum service times constraints, and candidate charging location capacity constraints. Time window generation module: Based on the initial collaborative scheduling results, it obtains the traffic status results and driving stability results, generates the user's baseline travel time results and arrival time fluctuation results, generates the user's baseline reliable arrival time results based on the current scheduling time results and the user's baseline travel time results, generates the user's reliable time window results by combining the arrival time fluctuation results, generates the mobile charging vehicle's reliable arrival time window results, and generates the reliability level results based on the data integrity results, data freshness results, and prediction consistency results. Task Cluster Reconstruction Module: For each initial task cluster result, extract the user's trusted time window result and the mobile charging vehicle's trusted time window result, calculate the user synchronization overlap result and the mobile charging vehicle's synchronization matching result, generate a synchronization risk result based on the user synchronization overlap result, the mobile charging vehicle's synchronization matching result, the trust level result, and the impact result of subsequent task chains, and divide the task cluster results into stable task cluster results, adjustable task cluster results, and task cluster results that need to be split according to the comparison between the synchronization risk result and the preset synchronization threshold result. Perform service time adjustment processing on the adjustable task cluster results, and perform synchronization risk user identification processing and reallocation processing on the task cluster results that need to be split, generate task cluster reconstruction results, and perform constraint verification. Closed-loop correction module: Obtains task execution status results, generates time window hit results and waiting risk verification results, generates feedback correction results based on time window hit results and waiting risk verification results, including time window width correction results, trust level correction results, synchronization threshold correction results and reallocation cost correction results, and writes the feedback correction results into the scheduling parameter update results.

Images