New energy automobile intelligent rapid charging scheduling method based on multi-source data analysis

Abstract

The invention relates to the technical field of big data analysis, and discloses a new energy automobile intelligent fast charging scheduling method based on multi-source data analysis, which comprises the following steps: S1, collecting multi-source charging scheduling data: constructing a charging scheduling multi-source data set; s2, multi-source data preprocessing and fusion: obtaining a charging scheduling fusion decision data set; s3, generating a charging scheduling decision: making the charging scheduling decision; s4, instruction issuing and execution control; s5, real-time monitoring and dynamic adjustment: dynamically adjusting a charging strategy; s6, feedback collection and optimization iteration are carried out; multi-source charging scheduling data are collected in real time through a vehicle-mounted terminal, a charging pile terminal, a power grid monitoring terminal and a traffic environment data terminal, a charging scheduling strategy is formulated through a cloud server, opinions are continuously fed back to a mobile phone user through a man-machine interaction terminal, charging scheduling feedback evaluation is carried out, model parameters are optimized, full-dimensional data driving is achieved, and the charging scheduling efficiency is improved. The charging efficiency is effectively improved, and the user experience is comprehensively optimized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of big data analysis, more specifically, the present application relates to a new energy vehicle intelligent rapid charging scheduling method based on multi-source data analysis. BACKGROUND

[0002] With the global energy transformation and the enhancement of environmental awareness, the new energy vehicle industry has also developed rapidly, and the rapid growth of the number of new energy vehicles has led to a rapid increase in charging demand, and the number, distribution and charging efficiency of charging facilities are also increasingly demanding. A large number of new energy vehicles will lead to a sharp rise in power grid load, affecting the safe and stable operation of the power grid. In addition, with the increasing number of new energy vehicle users, there is a higher demand for the convenience, speed and economy of charging. The existing new energy vehicle charging scheduling method includes obtaining charging demand information, determining priority, determining scheduling decisions, and guiding vehicle charging, which realizes the rational allocation and efficient use of charging resources. It improves the charging efficiency.

[0003] However, it still has some shortcomings in actual use. First, the data is not fully utilized. The existing new energy vehicle charging scheduling method only relies on part of the data, and cannot comprehensively consider the power grid load, renewable energy generation and traffic flow, resulting in inaccurate charging scheduling of new energy vehicles, affecting the safe operation of the power grid. Second, the flexibility and adaptability are insufficient. The existing new energy vehicle charging scheduling method usually uses fixed rules and algorithms to determine the charging priority and scheduling decision, which is difficult to adapt to complex and variable situations. In the face of uncertainty of user charging demand, the charging plan cannot be adjusted in time, resulting in low charging efficiency of new energy vehicles. Third, the user experience optimization is insufficient. The existing new energy vehicle charging scheduling method does not fully consider the user experience, and the charging strategy may affect the flexibility of user charging. Moreover, it lacks analysis of multi-source data, which cannot provide personalized charging services for users, reducing the convenience and economy of charging. SUMMARY

[0004] Therefore, the embodiments of the present application provide a new energy vehicle intelligent rapid charging scheduling method based on multi-source data analysis, which collects multi-source charging scheduling data in real time through a vehicle terminal, a charging pile terminal, a power grid monitoring terminal and a traffic environment data terminal, formulates a charging scheduling strategy through a cloud server, continuously collects user feedback through a human-computer interaction terminal, and evaluates the feedback of the charging scheduling based on the user feedback, effectively solving the problems of insufficient data utilization, insufficient flexibility and adaptability, and insufficient user experience optimization in the background technology.

[0005] To achieve the above object, the present application provides the following technical solution: a new energy vehicle intelligent rapid charging scheduling method based on multi-source data analysis, comprising a vehicle terminal, a charging pile terminal, a power grid monitoring terminal, a traffic environment data terminal, a cloud server and a man-machine interaction terminal, each terminal realizes data interaction and collaborative control through a communication network architecture, comprising the following steps: S1: Multi-source charging scheduling data acquisition: real-time acquisition of vehicle state data, charging pile state data, power grid operation data and traffic scene data through each acquisition terminal, construction of a charging scheduling multi-source data set, and uploading to the cloud server; S2: Multi-source data preprocessing and fusion: the cloud server performs data preprocessing on the charging scheduling multi-source data set, and integrates the preprocessed data to obtain a charging scheduling fusion decision data set; S3: Charging scheduling decision generation: the cloud server constructs a charging demand evaluation model and a charging pile recommendation degree evaluation model, calculates the charging demand evaluation index and the charging pile recommendation index, and formulates a charging scheduling decision; S4: Instruction issuing and execution control: the cloud server generates scheduling instructions according to the charging scheduling decision, and issues them to the vehicle terminal, the charging pile terminal and the man-machine interaction terminal, and automatically executes the charging operation according to the received scheduling instructions; S5: Real-time monitoring and dynamic adjustment: the cloud server receives information feedback from the vehicle terminal, the charging pile terminal and the man-machine interaction terminal in real time, judges whether the charging is abnormal, and dynamically adjusts the charging strategy according to the charging abnormality; S6: Feedback collection and optimization iteration: the man-machine interaction terminal collects user feedback opinions regularly and uploads them to the cloud server for charging scheduling feedback evaluation, and adjusts the model parameters according to the charging scheduling feedback evaluation results.

[0006] The technical effects and advantages of the present application are as follows: 1. The present application collects multi-dimensional data covering vehicle state, charging pile state, power grid operation and traffic scene through the vehicle terminal, the charging pile terminal, the power grid monitoring terminal and the traffic environment data terminal in real time, constructs a complete charging scheduling multi-source data set, breaks the data island, and integrates the data after data preprocessing to improve the accuracy of scheduling decision, realizes full-dimensional data driving, and is more suitable for actual scene; 2. The present application constructs a charging demand evaluation model and a charging pile recommendation degree evaluation model through the cloud server, dynamically calculates the charging demand index and the charging pile recommendation index through real-time multi-source data, formulates a charging scheduling decision combining the charging demand index and the charging pile recommendation index of the vehicle to be charged, and implements closed-loop control through instruction issuing and feedback optimization, effectively improves the charging efficiency, and is helpful to adapt to complex and variable charging scenes; 3、 The application fuses vehicle data, user demand, traffic data and power grid data, constructs a charging demand evaluation model, provides differentiated and personalized charging strategies for users, continuously collects user feedback through a man-machine interaction terminal, evaluates charging scheduling feedback based on user feedback, reversely optimizes model parameters, improves charging convenience and economy, comprehensively optimizes user experience, and creates personalized precision services. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 It is a schematic diagram of the method steps of the application.

[0008] Figure 2 It is a schematic diagram of the overall structure of the application.

[0009] Figure 3 It is a schematic diagram of the dynamic adjustment steps of the charging strategy of the application. DETAILED DESCRIPTION

[0010] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0011] As shown in the new energy vehicle intelligent rapid charging scheduling method based on multi-source data analysis, it includes a vehicle terminal, a charging pile terminal, a power grid monitoring terminal, a traffic environment data terminal, a cloud server and a man-machine interaction terminal, and each terminal realizes data interaction and collaborative control through a communication network architecture. Figure 1

[0012] In a more specific application of the application, the vehicle terminal is used to obtain the battery SOC, the charging interface state, the vehicle position and the user travel destination, receive the cloud scheduling scheme, control the physical connection between the charging interface and the charging pile, and feedback the execution result, display the charging progress and scheduling suggestion through the vehicle screen, the vehicle terminal includes a vehicle ECU, a GPS module, a battery management system, a communication module and a vehicle display screen, and the vehicle terminal is connected to the cloud server through a 5G network.

[0013] The charging pile terminal is used to adjust the charging power according to the cloud instruction, monitor the charging safety in real time, report the idle / occupied state, the charging power and the fault information of the charging pile to the cloud server in real time, support code scanning charging, display the electricity price information and generate the charging bill, the charging pile terminal includes a charging controller, a power module, a communication module, a touch screen, a code scanning gun, a charging pile management system and a billing module, and the charging pile terminal is connected to the cloud server through a 5G network. ​

[0014] The power grid monitoring terminal is used for monitoring power distribution network voltage, current, power factor, real-time electricity price and renewable energy output power, and feeding back power grid bearing capacity to the cloud server, adjusting the power of the charging pile in cooperation with the dispatching instruction issued by the cloud, detecting power grid abnormalities and triggering the protection mechanism, and the power grid monitoring terminal includes a smart grid, a power sensor, an edge computing gateway and a communication module, and the power grid monitoring terminal is connected to the cloud server through a power special network.

[0015] The traffic environment data terminal is used for analyzing real-time traffic, recommending suitable charging routes combined with vehicle positions, and collecting real-time environmental data, and dynamically adjusting charging strategies according to temperature, and the traffic environment data terminal includes a road camera, an RFID traffic monitor, a GPS trajectory acquisition platform, a weather station and a charging pile environment monitoring sensor, and the traffic environment data terminal uploads data to the cloud server through the Internet of Things.

[0016] The cloud server is used for integrating multi-source data transmitted by the vehicle terminal, the charging pile terminal, the power grid monitoring terminal and the traffic environment data terminal, building a real-time data middle platform, generating charging priority ranking, power allocation strategy and charging pile recommendation, and the cloud server includes a cloud computing server cluster, a data storage array and a network security device.

[0017] The human-computer interaction terminal is used for collecting user-submitted charging targets, querying the real-time state of surrounding charging piles, providing functions of remotely reserving charging piles, starting / stopping charging and checking charging bills, receiving charging pile state information pushed by the cloud, and collecting user charging feedback evaluation, and the human-computer interaction terminal includes a smart phone, a smart tablet and a wearable device, and the human-computer interaction terminal is connected to the cloud server through a 5G network and connected to the vehicle terminal and the charging pile terminal through Bluetooth.

[0018] The connection mode of the above vehicle terminal, charging pile terminal, power grid monitoring terminal, traffic environment data terminal, cloud server and human-computer interaction terminal is described in Figure 2 .

[0019] The specific embodiment of the present application includes the following steps: S1: Multi-source charging scheduling data acquisition: real-time acquisition of vehicle state data, charging pile state data, power grid operation data and traffic scene data through each acquisition terminal, construction of a charging scheduling multi-source data set, and uploading to the cloud server; Further, the vehicle-mounted state data specifically includes vehicle-mounted battery state data, vehicle-mounted driving data, and vehicle-mounted charging preference data, the charging pile state data specifically includes charging pile basic data, charging pile real-time state data, and charging pile operation performance data, the power grid operation data specifically includes power grid load data, power grid power quality data, and electricity price policy data, the traffic scene data specifically includes traffic state data, parking lot resource data, and environmental data, and each collection terminal specifically includes a vehicle-mounted terminal, a charging pile terminal, a power grid monitoring terminal, and a traffic environment data terminal.

[0020] In the embodiment, it is specifically pointed out that the vehicle-mounted battery state data specifically includes battery remaining capacity, state of charge, battery temperature, battery health status, and charging interface type, the vehicle-mounted driving data specifically includes vehicle GPS coordinates, driving speed, and destination coordinates, the vehicle-mounted charging preference data specifically includes target charging amount, expected charging completion time, and expected charging price interval, the charging pile basic data specifically includes charging pile ID, charging pile geographic location coordinates, charging type, maximum output power, and charging interface compatibility, the charging pile real-time state data specifically includes current use state, remaining charging position quantity, and real-time charging price, the power grid load data specifically includes regional power grid real-time load rate and transformer capacity utilization rate, the power grid power quality data specifically includes voltage fluctuation value, frequency fluctuation value, and harmonic content, the electricity price policy data specifically includes real-time electricity price and peak-valley electricity price difference, the traffic state data specifically includes real-time road condition and average vehicle speed, the parking lot resource data specifically includes parking lot location, remaining parking spaces, parking fee standard, and charging pile quantity, and the environmental data specifically includes real-time temperature and precipitation probability.

[0021] S2: Multi-source data preprocessing and fusion: the cloud server preprocesses the charging scheduling multi-source data set, and integrates the preprocessed data to obtain a charging scheduling fusion decision data set; Further, the data preprocessing of the charging scheduling multi-source data set needs to standardize the formats of different collection terminals, process missing values, remove duplicate values, and detect abnormal values, and the acquisition of the charging scheduling fusion decision data set needs to associate the preprocessed charging scheduling multi-source data set in time and space dimensions, associate business logic, and reconstruct data structure.

[0022] In this embodiment, it is necessary to specifically explain that: format standardization of the charging scheduling multi-source dataset refers to unifying the format of data collected from different terminals and converting it into a unified time format; missing value handling refers to interpolating missing data; duplicate value removal refers to identifying and deleting duplicate records by establishing unique data identifiers; outlier detection is used to eliminate outliers in the charging scheduling multi-source dataset caused by sensor failures or communication interference; spatiotemporal correlation of the charging scheduling multi-source dataset specifically refers to matching related data based on time and geographical location; business logic correlation refers to logically integrating data according to the actual business process of new energy vehicle charging scheduling; and data structure reconstruction refers to structuring the correlated data and storing it in a format suitable for subsequent charging demand assessment and charging pile recommendation assessment.

[0023] S3: Charging scheduling decision generation: The cloud server builds a charging demand assessment model and a charging pile recommendation assessment model, calculates the charging demand assessment index and the charging pile recommendation index, and formulates charging scheduling decisions. Furthermore, the calculation of the charging demand assessment index requires obtaining the battery capacity, remaining battery capacity, scheduled charging time, current charging time, and estimated charging duration from the charging scheduling fusion decision dataset. This yields standardized battery capacity E, remaining battery capacity Q, scheduled charging time t, current charging time t0, and estimated charging duration Δt. These are then imported into the charging demand assessment model using the formula: , The charging demand assessment index I is calculated, where Q i Q represents the current battery charge level. i-1 This represents the remaining battery power 1 hour ago at the current time. a1, a2, and a3 represent the weighting coefficients of the remaining battery power, power fluctuation rate, and time urgency change rate at the current time, respectively, and their sum is 1.

[0024] In this embodiment, it is necessary to specifically explain the use of The formula standardizes the data, where X takes the values ​​of battery capacity, remaining battery power, scheduled charging time, current charging time, and estimated charging time. It is used to normalize data of different dimensions so that the value of the charging demand index is mapped between 0 and 1.

[0025] Furthermore, the calculation of the charging pile recommendation index requires obtaining the charging power P and hourly idle time t of the charging piles from the charging scheduling fusion decision dataset. f Actual distance d between vehicle and charging station, and estimated navigation time t. n The real-time electricity price p and the current grid load L are imported into the charging pile recommendation evaluation model, using the formula: , The charging pile recommendation index R is calculated, wherein P e represents the charging pile rated power, t ft is 60 minutes, and λ is a distance attenuation factor, t nmin represents the shortest time of the vehicle to be charged to the charging pile, p min represents the minimum grid price, L max represents the maximum load of the grid, b1, b2 and b3 represent the weight coefficients of the charging pile parameter, the traffic parameter and the grid parameter respectively, and the sum is 1.

[0026] In this embodiment, it needs to be specifically explained that the closer the ratio of the actual power to the rated power is to 1, the more fully the power performance of the charging pile can be utilized, the longer the idle time is, the more idle the charging pile is, the actual distance of the vehicle to the pile refers to the actual distance of the vehicle to be charged to the target charging pile, the farther the distance of the charging pile is, the worse the experience is, the higher the ratio of the shortest navigation time to the actual navigation time of the target charging pile is, the easier the target charging pile is to reach, the larger the ratio of the minimum price to the current price is, the cheaper the charging of the target charging pile is, and the higher the grid redundancy is, indicating that the grid is more free.

[0027] Further, the formulation of the charging scheduling decision needs to set a target area, obtain n vehicles to be charged and m charging piles in the target area, and obtain the expected waiting time W j of each charging pile, calculate the charging demand evaluation index corresponding to the n vehicles to be charged through the charging demand evaluation model, arrange the vehicles to be charged in descending order to obtain a vehicle priority queue and the charging demand evaluation index I i corresponding to each vehicle to be charged, and obtain the charging pile recommendation index R ij of the m charging piles corresponding to the n vehicles to be charged respectively, and select the optimal charging pile c ij corresponding to the i-th vehicle to be charged by using the formula , wherein λ1 is a time sensitivity coefficient, and the optimal charging pile of the vehicle to be charged is combined with the resource allocation constraint condition to perform charging scheduling.

[0028] In this embodiment, it needs to be specifically explained that the time sensitivity coefficient can be dynamically adjusted according to the historical charging scheduling effect, and the resource allocation constraint condition specifically includes capacity constraint, time window constraint and grid load constraint, wherein the capacity constraint specifically means that if the number of the vehicles to be charged allocated to the charging pile reaches the upper limit, the allocation of the vehicles to be charged to the charging pile is stopped; the time window constraint specifically means that if the expected charging time of the vehicle to be charged exceeds the available time window of the charging pile, the charging pile should be selected again; and the grid load principle specifically means that if the total load of the grid in the target area after allocation exceeds the threshold, the allocation of the low-priority vehicle to be charged should be delayed.

[0029] S4: Instruction issuing and execution control: the cloud server generates scheduling instructions according to the charging scheduling decision and issues them to the vehicle terminal, the charging pile terminal and the human-computer interaction terminal, and automatically executes the charging operation according to the received scheduling instructions; Further, the generation of the scheduling instructions needs the cloud server to make decision decomposition and protocol adaptation based on the charging scheduling decision. The decision decomposition refers to decomposing the charging scheduling decision into vehicle-side scheduling instructions, pile-side scheduling instructions and human-computer interaction scheduling instructions, and delivering the vehicle-side scheduling instructions to the vehicle terminal, the pile-side scheduling instructions to the charging pile terminal and the human-computer interaction scheduling instructions to the human-computer interaction terminal, and each terminal executes the corresponding operation according to the received scheduling instructions.

[0030] In this embodiment, it needs to be specifically explained that the vehicle-side scheduling instructions specifically include the ID and position of the target charging pile, the expected departure time, the recommended navigation path and the reserved charging period, the pile-side scheduling instructions specifically include the locking of the charging pile port and the expected charging power, the human-computer interaction scheduling instructions specifically include the scheduling result display, the operation confirmation button and the pop-up window warning, wherein the scheduling result display specifically includes the charging pile position, the expected cost and the waiting time, the pop-up window warning refers to the pop-up window reminding the user in advance if a special reservation occurs, and the human-computer interaction terminal refers to the owner APP or the car central control screen.

[0031] After receiving the instructions delivered by the cloud server, the vehicle terminal, the charging pile terminal and the human-computer interaction terminal need to execute the corresponding operation instructions. The vehicle terminal should automatically wake up the navigation according to the instructions, plan the recommended path according to the cloud server, and report the current vehicle state to the cloud server in real time. When reaching the vicinity of the charging pile, the charging process is automatically triggered; the charging pile terminal should immediately lock the charging pile port when receiving the reservation instruction, mark that it has been reserved, and automatically check whether the charging pile has failed 10 minutes before the reservation period, and report the charging pile state in real time; after receiving the instructions, the APP of the human-computer interaction terminal displays the reservation content and provides confirmation and cancellation buttons. If the owner manually cancels the reservation, it is immediately uploaded to the cloud server, and the charging scheduling plan is re-executed. The vehicle charging progress is synchronized in real time, and after the charging is completed, a charging completion reminder is automatically pushed to the owner APP.

[0032] S5: Real-time monitoring and dynamic adjustment: the cloud server receives the information fed back by the vehicle terminal, the charging pile terminal and the human-computer interaction terminal in real time, judges whether the charging is abnormal, and dynamically adjusts the charging strategy according to the charging abnormality; Further, as shown in Figure 3 The dynamic adjustment steps of the charging strategy are as follows: A1: The vehicle terminal, the charging pile terminal and the human-computer interaction terminal monitor and collect the charging state data in real time and upload them to the cloud server; In this embodiment, it needs to be specifically pointed out that the vehicle-mounted terminal acquires the SOC and the battery temperature in real time during the vehicle charging process, the charging pile terminal acquires the charging pile output power and the gun head connection state in real time during the vehicle charging process, and the man-machine interaction terminal acquires the reservation cancellation and the manual adjustment of the charging capacity in real time during the vehicle charging process.

[0033] A2: The cloud server performs threshold judgment and trend prediction based on the charging state data, and identifies the charging process exception; In this embodiment, it needs to be specifically pointed out that the cloud server establishes a charging state digital twin model based on the received charging state data, identifies the abnormal condition in the vehicle charging process, and the charging process exception includes vehicle-side exception, pile-side exception and user-side exception. The vehicle-side exception specifically refers to battery overheating, SOC jump and communication interruption. The SOC jump refers to that the SOC decreases by more than 3% within 10 seconds. The pile-side exception specifically refers to power sudden drop, gun head exception and power grid overload. The power sudden drop refers to that the charging pile output power decreases significantly within 10 seconds. The gun head exception refers to that the vehicle-pile connection state is abnormal. The power grid overload refers to that the regional power grid load is greater than the rated load for a long time. The user-side exception specifically refers to reservation breach and user manual intervention. The reservation breach refers to that the vehicle does not arrive at the specified charging position during the reservation period. The user manual intervention refers to that the vehicle owner manually adjusts the charging parameters for two times in succession.

[0034] A3: When the charging process exception is detected, the cloud server performs hierarchical processing according to the preset hierarchical error correction strategy.

[0035] In this embodiment, it needs to be specifically pointed out that the hierarchical error correction strategy specifically includes light fault automatic repair, heavy fault manual intervention and global strategy re-optimization. The light fault automatic repair refers to that when the battery overheats, the charging pile should reduce the output power, and trigger the active cooling of the vehicle battery. When the pile power suddenly drops, the charging pile performs fault self-checking. The heavy fault manual intervention refers to that when the communication interruption occurs, the cloud server should trigger the offline plan, and push the information of manually connecting the charging pile to the man-machine interaction terminal. When the power grid is overloaded, the cloud server should start the plan, and push the message of power grid tension and charging time extension to the man-machine interaction terminal. The global strategy re-optimization refers to that all available resources are re-collected, the scheduling decision is recalculated in combination with the new constraint strategy, and the scheduling change notification is pushed to the allocated vehicle.

[0036] S6: Feedback collection and optimization iteration: the user feedback opinions are collected regularly through the man-machine interaction terminal, and uploaded to the cloud server for charging scheduling feedback evaluation. The model parameters are adjusted according to the charging scheduling feedback evaluation result.

[0037] Further, the user feedback includes pre-charging user feedback, charging user feedback and post-charging user feedback, the cloud server carries out data structural processing on the received user feedback, carries out feedback evaluation, experience loss analysis and attribution analysis, and the cloud server optimizes model parameters based on the results of feedback evaluation.

[0038] In the embodiment, it is necessary to be specifically explained that the experience loss analysis refers to the calculation of the experience loss function combined with the user feedback data after data structural processing, the weighted analysis of the user satisfaction, charging efficiency and charging cost to obtain the experience loss function, the attribution analysis specifically refers to using decision tree and correlation analysis to find the association between feedback and charging scheduling, and the hierarchical optimization model parameters specifically include real-time parameter micro-adjustment, periodic model reconstruction and new strategy gray test, wherein the real-time parameter micro-adjustment refers to automatic adjustment for high-frequency feedback problems, the periodic model reconstruction refers to retraining the model and updating the model parameter weight using full feedback data every week, and the new strategy gray test refers to small-scale gray release of the optimized parameters or model, selecting part of charging piles for testing, comparing the experience loss functions of the gray group and the control group, if obvious improvement occurs, the parameters are pushed, if the improvement effect is poor, the parameters are rolled back, and feedback analysis optimization is re-performed.

[0039] Secondly, the drawings in the disclosed embodiment of the application only involve the structures involved in the disclosed embodiment of the application, other structures can refer to the usual design, and in the case of no conflict, the same embodiment and different embodiments of the application can be combined with each other. Finally, the above only describes the preferred embodiments of the application and is not used to limit the application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application should be included in the protection scope of the application.

Claims

1. A method for intelligent fast charging scheduling of new energy vehicles based on multi-source data analysis, characterized in that, This includes vehicle-mounted terminals, charging pile terminals, power grid monitoring terminals, traffic environment data terminals, cloud servers, and human-machine interaction terminals. These terminals achieve data interaction and collaborative control through a communication network architecture, including the following steps: S1: Multi-source charging scheduling data acquisition: Real-time acquisition of vehicle status data, charging pile status data, power grid operation data and traffic scenario data through various acquisition terminals to construct a multi-source charging scheduling dataset and upload it to the cloud server; S2: Multi-source data preprocessing and fusion: The cloud server preprocesses the multi-source dataset for charging scheduling and integrates the preprocessed data to obtain the charging scheduling fusion decision dataset. S3: Charging scheduling decision generation: The cloud server builds a charging demand assessment model and a charging pile recommendation assessment model, calculates the charging demand assessment index and the charging pile recommendation index, and formulates charging scheduling decisions. S4: Command Issuance and Execution Control: The cloud server generates scheduling commands based on charging scheduling decisions and issues them to the vehicle terminal, charging pile terminal, and human-machine interaction terminal. The charging operation is automatically executed according to the received scheduling commands. S5: Real-time monitoring and dynamic adjustment: The cloud server receives information from the vehicle terminal, charging pile terminal and human-machine interaction terminal in real time, determines whether there is an abnormality in charging, and dynamically adjusts the charging strategy according to the charging abnormality. S6: Feedback Collection and Optimization Iteration: User feedback is collected regularly through the human-computer interaction terminal and uploaded to the cloud server for charging scheduling feedback evaluation. Model parameters are adjusted based on the charging scheduling feedback evaluation results.

2. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The vehicle status data specifically includes vehicle battery status data, vehicle driving data, and vehicle charging preference data. The charging pile status data specifically includes charging pile basic data, charging pile real-time status data, and charging pile operating performance data. The power grid operation data specifically includes power grid load data, power grid power quality data, and electricity pricing policy data. The traffic scenario data specifically includes traffic status data, parking lot resource data, and environmental data. The data collection terminals specifically include vehicle terminals, charging pile terminals, power grid monitoring terminals, and traffic environment data terminals.

3. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The data preprocessing of the charging scheduling multi-source dataset requires format standardization, missing value handling, duplicate value removal, and outlier detection for different acquisition terminals. The acquisition of the charging scheduling fusion decision dataset requires spatiotemporal dimension association, business logic association, and data structure reconstruction of the preprocessed charging scheduling multi-source dataset.

4. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The calculation of the charging demand assessment index requires obtaining the battery capacity, remaining battery power, scheduled charging time, current charging time, and estimated charging duration from the charging scheduling fusion decision dataset. This yields standardized battery capacity E, remaining battery power Q, scheduled charging time t, current charging time t0, and estimated charging duration Δt. These are then imported into the charging demand assessment model using the formula: ， The charging demand assessment index I is calculated, where Q i Q represents the current battery charge level. i-1 This represents the remaining battery power 1 hour ago at the current time. a1, a2, and a3 represent the weighting coefficients of the remaining battery power, power fluctuation rate, and time urgency change rate at the current time, respectively, and their sum is 1.

5. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The calculation of the charging pile recommendation index requires obtaining the charging power P and hourly idle time t of the charging piles from the charging scheduling fusion decision dataset. f Actual distance d between vehicle and charging station, and estimated navigation time t. n The real-time electricity price p and the current grid load L are imported into the charging pile recommendation evaluation model, using the formula: ， The charging pile recommendation index R is calculated, where P e This indicates the rated power of the charging pile, t ft The time interval is 60 minutes, λ is the distance decay factor, and t is... nmin p represents the shortest time it takes for a vehicle to reach a charging station. min L represents the minimum grid electricity price. max The maximum load of the power grid is represented by b1, b2 and b3, which represent the weighting coefficients of the charging pile parameters, traffic parameters and power grid parameters, respectively, and the sum is 1.

6. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The charging scheduling decision requires setting a target area, obtaining n vehicles waiting to be charged and m charging piles within that target area, and obtaining the estimated waiting time W for each charging pile. j The charging demand assessment index for n vehicles to be charged is calculated using a charging demand assessment model. These vehicles are then sorted in descending order to obtain a vehicle priority queue and the corresponding charging demand assessment index I for each vehicle. i And obtain the charging pile recommendation index R for each of the m charging piles corresponding to n vehicles to be charged. ij Using the formula Select the optimal charging station c corresponding to the i-th vehicle to be charged. ij λ1 is the time sensitivity coefficient, which combines the optimal charging pile of the vehicle to be charged with resource allocation constraints to perform charging scheduling.

7. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The generation of the scheduling instructions requires the cloud server to perform decision decomposition and protocol adaptation based on the charging scheduling decision. Decision decomposition refers to breaking down the charging scheduling decision into vehicle-side scheduling instructions, charging pile-side scheduling instructions, and human-machine interaction scheduling instructions. The vehicle-side scheduling instructions are transmitted to the vehicle terminal, the charging pile-side scheduling instructions are transmitted to the charging pile terminal, and the human-machine interaction scheduling instructions are transmitted to the human-machine interaction terminal. Each terminal performs the corresponding operation according to the received scheduling instructions.

8. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The dynamic adjustment steps of the charging strategy are as follows: A1: The vehicle-mounted terminal, charging pile terminal, and human-machine interaction terminal monitor and collect charging status data in real time and upload it to the cloud server; A2: The cloud server uses charging status data to make threshold judgments and trend predictions to identify abnormalities in the charging process; A3: When an abnormality is detected in the charging process, the cloud server performs tiered processing according to the preset tiered error correction strategy.

9. The intelligent fast charging scheduling method for new energy vehicles based on multi-source data analysis according to claim 1, characterized in that: The user feedback includes user feedback before charging, user feedback during charging, and user feedback after charging. The cloud server performs data structuring processing on the received user feedback, conducts feedback evaluation, performs experience loss analysis and attribution analysis, and optimizes model parameters hierarchically based on the feedback evaluation results.

Images