Intelligent charging pile scheduling method and system based on traffic flow prediction

Abstract

According to the intelligent charging pile scheduling method and system based on traffic flow prediction, the teacher charging pile scheduling network and the auxiliary charging pile scheduling network are constructed and fused, so that the intelligence and accuracy of charging pile scheduling are effectively improved. Two networks are obtained by training a teacher sample traffic flow data sequence and an auxiliary sample traffic flow data sequence respectively, characteristics and differences of different traffic monitoring areas are fully considered, and a student charging pile scheduling network is generated through iterative learning. And the adaptability to traffic flow change and the prediction accuracy are further improved. The finally generated target charging pile scheduling network can quickly decide reasonable charging demand data for the actual traffic flow data of any target traffic monitoring area, and realizes efficient charging pile scheduling according to the reasonable charging demand data, thereby effectively relieving the problem of uneven distribution of charging pile resources, and improving the convenience of the charging service and the user experience.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a smart charging pile scheduling method and system based on traffic flow prediction. Background Art

[0002] With the popularity of electric vehicles and the increasing complexity of the transportation system, the reasonable scheduling of charging piles has become a key issue in improving urban transportation efficiency and user experience. Traditional charging pile scheduling methods often rely on fixed rules or simple prediction models, which are difficult to accurately respond to the dynamic changes and regional differences in traffic flow. Especially in different traffic monitoring areas in the city, due to the diversity of factors such as traffic flow patterns, vehicle types and charging needs, the contradiction between supply and demand of charging piles is more prominent. Summary of the invention

[0003] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a smart charging pile scheduling method based on traffic flow prediction, the method comprising:

[0004] Obtain a teacher charging pile scheduling network, wherein the teacher charging pile scheduling network is generated by parameter learning based on a teacher sample traffic flow data sequence; the teacher sample traffic flow data sequence corresponds to a plurality of first traffic monitoring areas; each first traffic monitoring area includes a plurality of teacher sample traffic flow data; each teacher sample traffic flow data carries a first sample charging demand data;

[0005] Obtain an auxiliary charging pile scheduling network, wherein the auxiliary charging pile scheduling network is generated by parameter learning based on an auxiliary sample traffic flow data sequence; the auxiliary sample traffic flow data sequence corresponds to a plurality of second traffic monitoring areas; each second traffic monitoring area includes a plurality of auxiliary sample traffic flow data; each auxiliary sample traffic flow data carries second sample charging demand data; each second traffic monitoring area does not intersect with each first traffic monitoring area;

[0006] A student charging pile scheduling network is obtained based on the teacher charging pile scheduling network, and a student sample traffic flow data sequence is obtained based on the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence; the student sample traffic flow data sequence includes a plurality of student sample traffic flow data; each student sample traffic flow data carries a third sample charging demand data;

[0007] Iteratively learning network parameters of the student charging pile scheduling network according to the student sample traffic flow data sequence until the network convergence requirements are met, and generating a target charging pile scheduling network;

[0008] Based on the target charging pile scheduling network, a decision is made on the target traffic flow data of any target traffic monitoring area, target charging demand data is generated, and charging pile scheduling is performed on the traffic participants in the target traffic monitoring area according to the target charging demand data.

[0009] On the other hand, an embodiment of the present invention also provides a smart charging pile scheduling system based on traffic flow prediction, including a processor and a machine-readable storage medium, wherein the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0010] Based on the above aspects, the embodiment of the present application effectively improves the intelligence and accuracy of charging pile scheduling by constructing and integrating the teacher charging pile scheduling network and the auxiliary charging pile scheduling network. The two networks trained using the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence not only fully consider the characteristics and differences of different traffic monitoring areas, but also further improve the adaptability and prediction accuracy to traffic flow changes through the student charging pile scheduling network generated by iterative learning. The target charging pile scheduling network finally generated can quickly decide on reasonable charging demand data based on the actual traffic flow data of any target traffic monitoring area, and realize efficient charging pile scheduling accordingly, thereby effectively alleviating the problem of uneven distribution of charging pile resources, improving the convenience and user experience of charging services, and promoting the overall operating efficiency of the transportation system and the rationality of energy utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the execution flow of the smart charging pile scheduling method based on traffic flow prediction provided by an embodiment of the present invention.

[0012] Figure 2 It is a schematic diagram of the hardware architecture of a smart charging pile scheduling system based on traffic flow prediction provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0013] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 It is a flow chart of a smart charging pile scheduling method based on traffic flow prediction provided by an embodiment of the present invention. The smart charging pile scheduling method based on traffic flow prediction is introduced in detail below.

[0014] Step S110, obtaining a teacher charging pile scheduling network, wherein the teacher charging pile scheduling network is generated by parameter learning based on a teacher sample traffic flow data sequence. The teacher sample traffic flow data sequence corresponds to a plurality of first traffic monitoring areas. Each first traffic monitoring area includes a plurality of teacher sample traffic flow data. Each teacher sample traffic flow data carries a first sample charging demand data.

[0015] In this embodiment, it is assumed that in a traffic system of a large city, the server wants to build a teacher charging pile scheduling network. The city is divided into multiple first traffic monitoring areas, such as the city's central business district, several large industrial parks, and major residential areas. Each area has corresponding traffic flow monitoring equipment, which continuously collects traffic flow data.

[0016] In the first traffic monitoring area of ​​the city's central business district, traffic flow data includes traffic information of various types of vehicles (such as private cars, taxis, buses, etc.) in different time periods (such as morning peak, midday trough, evening peak, and weekends). Each teacher's sample traffic flow data carries the first sample charging demand data. Taking taxis as an example, during the evening peak hours on weekdays, the taxi traffic in this area is large. Due to the long-term operation of taxis, the charging demand is more urgent. The first sample charging demand data may show that it is necessary to add temporary charging piles in this area or adjust the charging power of existing charging piles to meet the fast charging needs.

[0017] For large industrial parks, there are many logistics vehicles entering and leaving the park during the daytime on weekdays. These vehicles may be electric trucks or electric forklifts. The first charging demand data in the teacher's sample traffic flow data may indicate that charging piles need to be set up near the cargo loading and unloading area to facilitate logistics vehicles to charge while loading and unloading goods, avoiding long waiting times for charging that affect logistics efficiency.

[0018] The server collects these multiple teacher sample traffic flow data from different first traffic monitoring areas, and then performs parameter learning based on these data. The learning process may involve multiple algorithms, such as neural network algorithms. The server inputs these data into the neural network, and the neurons in the neural network fit the rules in these data by continuously adjusting the weights. After a long learning process, the server finally generates a teacher charging pile scheduling network. This teacher charging pile scheduling network can predict the corresponding charging demand based on the input traffic flow data, providing a basis for subsequent charging pile scheduling.

[0019] Step S120, obtaining an auxiliary charging pile scheduling network, wherein the auxiliary charging pile scheduling network is generated by parameter learning based on the auxiliary sample traffic flow data sequence. The auxiliary sample traffic flow data sequence corresponds to multiple second traffic monitoring areas. Each second traffic monitoring area includes multiple auxiliary sample traffic flow data. Each auxiliary sample traffic flow data carries second sample charging demand data. Each second traffic monitoring area does not intersect with each first traffic monitoring area.

[0020] In this embodiment, it is also necessary to build an auxiliary charging pile scheduling network. In addition to the first traffic monitoring area mentioned above, the city also has other second traffic monitoring areas, such as newly developed areas on the edge of the city, tourist attractions in the suburbs, etc. These areas do not overlap with the first traffic monitoring area.

[0021] Taking the newly developed area on the edge of the city as an example, the traffic flow here is less than that in the first traffic monitoring area, and the traffic composition is also different, which may be more construction vehicles and a small number of commuter vehicles. The auxiliary sample traffic flow data in each second traffic monitoring area carries the second sample charging demand data. Since the working hours of construction vehicles are concentrated in the daytime, and the charging demand is more concentrated near the parking spots of construction vehicles, the second sample charging demand data in the auxiliary sample traffic flow data will reflect these characteristics.

[0022] In suburban tourist attractions, the number of tourist buses and tourists' private cars will increase during the peak tourist season. Tourist buses have special charging needs and may need to be equipped with high-power charging piles in the parking lot to meet their fast charging needs in order to continue their journey. The server collects multiple auxiliary sample traffic flow data from these different second traffic monitoring areas, and uses algorithms similar to those used to build the teacher charging pile scheduling network (such as neural network algorithms) for parameter learning to generate an auxiliary charging pile scheduling network. This network can predict charging needs based on the traffic flow conditions in the second traffic monitoring area, assisting the entire charging pile scheduling decision process.

[0023] Step S130, obtaining a student charging pile scheduling network based on the teacher charging pile scheduling network, and obtaining a student sample traffic flow data sequence based on the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence. The student sample traffic flow data sequence includes a plurality of student sample traffic flow data. Each student sample traffic flow data carries a third sample charging demand data.

[0024] In this embodiment, the student charging pile scheduling network and the student sample traffic flow data sequence are constructed. First, start with the existing teacher charging pile scheduling network. The teacher charging pile scheduling network has learned the relationship pattern between traffic flow and charging demand in the first traffic monitoring area. The server uses this pattern information to initialize the structure and parameters of the student charging pile scheduling network. For example, if the prediction of taxi charging demand during the evening rush hour in the city's central business district in the teacher charging pile scheduling network is achieved through specific neuron connection weights and thresholds, then the student charging pile scheduling network will draw on similar structures and initial parameters.

[0025] Next, the server obtains the student sample traffic flow data sequence based on the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence. Assume that the server selects the teacher sample traffic flow data from the teacher sample traffic flow data sequence. For example, the data of the morning rush hour on weekdays is selected from the teacher sample traffic flow data in the central business district of the city. The server performs multiple rounds of data optimization on the teacher sample traffic flow data sequence.

[0026] In the first round of data optimization, a teacher sample traffic flow data is randomly selected from the teacher sample traffic flow data sequence, such as the logistics vehicle flow data of a large industrial park during the daytime on weekdays as the first candidate sample traffic flow data. Then, the first characteristic distance between each second candidate sample traffic flow data (except the selected first candidate sample traffic flow data) and the first candidate sample traffic flow data in the teacher sample traffic flow data sequence is calculated. Taking the calculation of the first characteristic distance between the private car flow data on weekday evenings in a residential area (as the second candidate sample traffic flow data) and the first candidate sample traffic flow data as an example, the server will obtain the first characteristic weighted value of the first candidate sample traffic flow data (for example, the weighted value determined according to factors such as the type and load of the logistics vehicle), the first characteristic discrete degree value (for example, the degree of fluctuation of the logistics vehicle flow in different time periods), and obtain the second characteristic weighted value of the second candidate sample traffic flow data (for example, the weighted value determined by factors such as different models of private cars and the commuting habits of car owners), the second characteristic discrete degree value (for example, the fluctuation of private car flow in different time periods at night), and determine the covariance degree value between the two (for example, whether the logistics vehicle and private car flows have related changes due to some common factors). Then, the similarity of traffic scale is calculated based on the weighted value of the first feature and the weighted value of the second feature (for example, comparing the similarity of the overall traffic scale of logistics vehicles and private cars), the similarity of traffic change is calculated based on the first feature discrete degree value and the second feature discrete degree value (for example, comparing the similarity of the traffic changes of the two in different time periods), the similarity of traffic distribution is calculated based on the first feature discrete degree value, the second feature discrete degree value and the covariance degree value (for example, comparing the similarity of traffic distribution of the two in different areas), and finally, the first feature distance is calculated based on the similarity of traffic scale, similarity of traffic change and similarity of traffic distribution.

[0027] Arrange each second candidate sample traffic flow data in descending order according to the first characteristic distance, and generate a candidate sample traffic flow data sequence based on the first second candidate sample traffic flow data of a second set size (for example, the first 10) in the descending order and the first candidate sample traffic flow data.

[0028] If it is not the first round of data optimization, for each second candidate sample traffic flow data, obtain its second characteristic distance with each obtained candidate sample traffic flow data, and calculate the weighted characteristic distance of each second characteristic distance. For example, for a private car traffic data of a residential area that has not been selected (as the second candidate sample traffic flow data), calculate its second characteristic distance with the candidate sample traffic flow data obtained in the previous rounds of optimization (such as the previously selected industrial park logistics vehicle traffic data, tourist bus traffic data at tourist attractions, etc.), and obtain the weighted characteristic distance after comprehensively considering the weights of different rounds, and output the second candidate sample traffic flow data with the largest weighted characteristic distance as the first candidate sample traffic flow data.

[0029] The server loads the candidate sample traffic flow data sequence into the auxiliary sample traffic flow data sequence until the training scale in the auxiliary sample traffic flow data sequence meets the first set scale (such as reaching 1000 data), and then generates a student sample traffic flow data sequence. Each student sample traffic flow data in this sequence carries the third sample charging demand data, which reflects the charging demand information contained in the teacher sample traffic flow data and the auxiliary sample traffic flow data.

[0030] Step S140, iteratively learning network parameters of the student charging pile scheduling network according to the student sample traffic flow data sequence until the network convergence requirements are met, and then generating a target charging pile scheduling network.

[0031] In this embodiment, for each student sample traffic flow data in the student sample traffic flow data sequence, for example, the student sample traffic flow data is about the traffic flow data of a certain period of time in the newly developed area on the edge of the city, the student sample traffic flow data is loaded into the teacher charging pile scheduling network to generate the first charging demand prediction data. Assume that the teacher charging pile scheduling network predicts that a certain number of charging piles need to be added in a certain parking lot to meet the charging demand based on the traffic conditions of construction vehicles and commuting vehicles in the newly developed area. This is the first charging demand prediction data.

[0032] At the same time, the student sample traffic flow data is loaded into the auxiliary charging pile scheduling network to generate the second charging demand prediction data. The auxiliary charging pile scheduling network may predict that the charging demand is relatively dispersed based on the sparse overall traffic flow characteristics of the newly developed area, and there is no need to set up charging piles on a large scale. This is the second charging demand prediction data.

[0033] The student sample traffic flow data is then loaded into the student charging pile scheduling network to generate the third charging demand prediction data. It is assumed that the charging demand initially predicted by the student charging pile scheduling network is different from the predictions of the teacher charging pile scheduling network and the auxiliary charging pile scheduling network.

[0034] Then, the server determines the global network error parameter based on the first loss function value between the first charging demand prediction data and the third charging demand prediction data, the second loss function value between the second charging demand prediction data and the third charging demand prediction data, and the third loss function value between the third charging demand prediction data and the third sample charging demand data.

[0035] When calculating the first loss function value and the second loss function value, take the construction engineering vehicles in the newly developed area as the first charging pile scheduling subject and the commuting vehicles as the non-charging pile scheduling subject as an example. The server obtains the first dynamic flow characteristics (such as the vehicle's driving speed, parking frequency, etc.) of the first charging pile scheduling subject (construction engineering vehicles) in the student sample traffic flow data and the second dynamic flow characteristics (such as the round-trip route and parking time of the commuting vehicle) of the non-charging pile scheduling subject (commuting vehicle), and obtains the third dynamic flow characteristics (such as the working mode of special vehicles, etc.) of the second charging pile scheduling subject (assuming it is other special vehicle types) in the student sample traffic flow data. Determine the third characteristic distance between the third charging demand forecast data and the first charging demand forecast data (for example, by comparing the differences in the predicted number of charging piles, location, etc.), and determine the fourth characteristic distance between the third charging demand forecast data and the second charging demand forecast data. According to the first charging demand prediction data, the first subject quantity (such as the number of construction engineering vehicles) and the second subject quantity (such as the number of commuting vehicles) corresponding to the first charging pile scheduling subject in the student sample traffic flow data are obtained, and the third subject quantity (such as the number of special vehicles) corresponding to the second charging pile scheduling subject in the student sample traffic flow data is obtained according to the second charging demand prediction data. The fourth loss function value for the first charging pile scheduling subject between the first charging demand prediction data and the third charging demand prediction data is calculated according to the third characteristic distance, the first dynamic flow feature, and the first subject quantity (for example, calculated according to factors such as the degree to which the charging demand of construction engineering vehicles is not met), and the fifth loss function value for the non-charging pile scheduling subject between the first charging demand prediction data and the third charging demand prediction data is calculated according to the third characteristic distance, the third dynamic flow feature, and the second subject quantity (for example, calculated according to factors such as the impact of commuting vehicles on charging facilities). The first loss function value is calculated according to the fourth loss function value and the fifth loss function value, and the second loss function value is calculated according to the fourth characteristic distance, the second dynamic flow feature, and the third subject quantity.

[0036] Obtain a first weight value corresponding to the first loss function value (for example, determined according to the importance of the first charging pile scheduling subject), obtain a second weight value corresponding to the second loss function value (for example, determined according to the importance of the second charging pile scheduling subject), and obtain a third weight value corresponding to the third loss function value (for example, determined according to the importance of the accuracy of the third sample charging demand data). Determine the first error parameter of the student charging pile scheduling network based on the first weight value, the first loss function value, the second weight value, and the second loss function value, and determine the second error parameter of the student charging pile scheduling network based on the third weight value and the third loss function value. Determine the global network error parameter based on the first error parameter and the second error parameter.

[0037] The neuron weight information of the student charging pile scheduling network is updated according to the global network error parameters, and the updated student charging pile scheduling network is used as the student charging pile scheduling network corresponding to the next round of network parameter learning process. The server repeats this process until the network convergence requirements are met (for example, the loss function value is lower than a certain set threshold, or the change of the network parameter is less than a certain set value), at which time the target charging pile scheduling network is generated.

[0038] Step S150, based on the target charging pile scheduling network, a decision is made on the target traffic flow data of any target traffic monitoring area, target charging demand data is generated, and charging pile scheduling is performed on the traffic participants in the target traffic monitoring area according to the target charging demand data.

[0039] In this embodiment, the server selects a target traffic monitoring area, such as a city central business district, as the target traffic monitoring area. The target traffic flow data includes the flow information of various traffic participants (such as private cars, taxis, buses, etc.). The server inputs the target traffic flow data into the target charging pile scheduling network to generate target charging demand data.

[0040] Assume that the target traffic flow data shows that during the evening rush hour on weekdays, taxi traffic increases significantly and buses are also at their peak. The target charging pile dispatching network generates target charging demand data based on this data, which may include the need to add temporary charging piles at locations where taxis frequently pick up and drop off passengers (such as the entrance of large shopping malls and near office buildings). For buses, the charging power of charging piles needs to be adjusted at bus hubs to meet the demand for fast charging. At the same time, it is predicted that the charging demand of the entire commercial area will increase significantly during the evening rush hour.

[0041] The server then preprocesses the target charging demand data of the target traffic monitoring area to extract key charging demand data. For example, the charging demand period is determined to be the evening peak on weekdays (17:00 - 19:00), the charging demand hotspots are the entrances of large shopping malls, near office buildings and bus hubs, the charging demand type ratio is 50% for taxi charging demand, 30% for buses, and 20% for private cars, and the predicted charging demand is that a total of 100 charging positions need to be provided (calculated based on the charging demand of different types of vehicles).

[0042] Based on these key charging demand data, charging resources are evaluated and a charging resource evaluation report is generated. The server queries the distribution of charging piles in the target traffic monitoring area and finds that there are currently 20 charging piles near shopping malls, 10 charging piles near office buildings, and 30 charging piles at bus hubs. The available status is that some charging piles are in use and some are idle. The charging power varies, and there are fast charging piles and slow charging piles. In terms of expansion potential, the expansion potential is small near shopping malls and office buildings due to limited space, while there is a certain amount of expansion space at bus hubs.

[0043] According to the charging resource assessment report and key charging demand data, the charging pile scheduling strategy is constructed. The server determines the charging piles that need to be scheduled, such as deploying 5 charging piles from other areas with lower charging demand to the entrance of the shopping mall, and adjusting the charging power of 10 charging piles to the fast charging mode at the bus hub; the scheduling time is to complete the scheduling before the evening peak; the scheduling location is to accurately place the deployed charging piles at the designated location at the entrance of the shopping mall; the status of the charging piles after scheduling is turned on and can be used normally.

[0044] According to the charging pile scheduling strategy, the server sends a scheduling instruction to the charging pile management system. The scheduling instruction includes the identification of each charging pile, the scheduling time (such as scheduling starts at 16:30), the scheduling location (such as the specific coordinates of the shopping mall entrance) and the charging status after scheduling (such as fast charging status). After receiving the scheduling instruction, the charging pile management system controls the charging pile to perform corresponding operations, such as moving the designated charging pile to the designated location at the shopping mall entrance, adjusting the 10 charging piles at the bus hub to fast charging power, or changing the charging status to a usable state.

[0045] After the charging pile scheduling operation is executed, the server monitors the scheduling effect of the charging pile in real time. The server collects real-time charging data of the charging pile (such as the charging current, voltage, charging time, etc. of each charging pile), traffic flow data (such as the flow changes of taxis, buses, and private cars during the evening peak hours), and user feedback data (such as whether taxi drivers find charging more convenient, whether buses can be charged in time, etc.), and evaluates the actual scheduling effect of the charging pile scheduling. If it is found that although charging piles have been added at the entrance of the shopping mall, taxis cannot reach the charging piles in time due to traffic congestion, and the scheduling effect does not meet the expected conditions, the charging pile scheduling strategy is adjusted, such as re-planning the placement of the charging piles to avoid traffic congestion sections, and the scheduling operation is performed again.

[0046] At the same time, the server optimizes and learns the charging pile scheduling strategy based on historical charging pile scheduling data (previous scheduling experience data in this area or other areas), real-time charging data, traffic flow data, and user feedback data. The correlation and regularity in these data are analyzed through machine learning algorithms. For example, it is found that the charging demand of buses in a specific time period is closely related to the operating schedule of bus routes, and the charging demand of private cars is related to the business hours and promotional activities of shopping malls. The key factors affecting the scheduling effect of charging piles, such as traffic congestion and the rationality of the layout of charging facilities, are extracted, and the charging pile scheduling strategy is intelligently adjusted according to these key factors to generate an optimized charging pile scheduling strategy. The optimized charging pile scheduling strategy is input into the charging pile scheduling system, and the scheduling parameters and scheduling strategy library of the charging pile scheduling system are updated so that the charging demand can be met more accurately and efficiently in future charging pile scheduling decisions.

[0047] Based on the above steps, the embodiment of the present application effectively improves the intelligence and accuracy of charging pile scheduling by constructing and integrating the teacher charging pile scheduling network and the auxiliary charging pile scheduling network. The two networks trained using the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence not only fully consider the characteristics and differences of different traffic monitoring areas, but also further improve the adaptability and prediction accuracy to traffic flow changes through the student charging pile scheduling network generated by iterative learning. The target charging pile scheduling network finally generated can quickly decide on reasonable charging demand data based on the actual traffic flow data of any target traffic monitoring area, and realize efficient charging pile scheduling accordingly, thereby effectively alleviating the problem of uneven distribution of charging pile resources, improving the convenience and user experience of charging services, and promoting the overall operating efficiency of the transportation system and the rationality of energy utilization.

[0048] In a possible implementation, each round of network parameter learning process specifically includes:

[0049] Step S141: for each student sample traffic flow data, the student sample traffic flow data is loaded into the teacher charging pile scheduling network to generate first charging demand forecast data. The student sample traffic flow data is loaded into the auxiliary charging pile scheduling network to generate second charging demand forecast data. The student sample traffic flow data is loaded into the student charging pile scheduling network to generate third charging demand forecast data.

[0050] Step S142, determining the global network error parameter based on the first loss function value between the first charging demand forecast data and the third charging demand forecast data, the second loss function value between the second charging demand forecast data and the third charging demand forecast data, and the third loss function value between the third charging demand forecast data and the third sample charging demand data.

[0051] Step S143, updating the neuron weight information of the student charging pile scheduling network according to the global network error parameter, and using the updated student charging pile scheduling network as the student charging pile scheduling network corresponding to the next round of network parameter learning process.

[0052] In this embodiment, when each round of network parameter learning is performed, the operation is carried out using data from different traffic monitoring areas in the city as an example. For each student sample traffic flow data in the student sample traffic flow data sequence, such as the traffic flow data of a certain working day in a newly developed area on the edge of the city, the server first loads it into the teacher charging pile scheduling network. In this area, the traffic conditions of construction vehicles and commuting vehicles are important factors. The teacher charging pile scheduling network generates the first charging demand prediction data based on the relationship pattern between traffic flow and charging demand in various areas of the city that it has learned before. For example, based on factors such as the working hours of construction vehicles, parking rules, and round-trip routes of commuting vehicles, it is predicted that a certain number of charging piles need to be set up near several construction sites in the newly developed area, and an appropriate number of charging piles are also needed in areas where commuting vehicles are parked to meet the charging needs.

[0053] Next, the server loads the same student sample traffic flow data into the auxiliary charging pile scheduling network. The auxiliary charging pile scheduling network generates the second charging demand prediction data based on the relationship between different traffic characteristics and charging demand that it has learned. Since the auxiliary charging pile scheduling network takes into account the relatively sparse overall traffic flow in the newly developed area, it may predict that charging facilities do not need to be too concentrated, but can be dispersed in different small parking areas, and the number of charging piles may be less than the prediction of the teacher charging pile scheduling network.

[0054] Then, the server loads the student sample traffic flow data into the student charging pile scheduling network to generate the third charging demand prediction data. The prediction data initially generated by the student charging pile scheduling network may be different from the previous two because it is still in the process of learning and adjusting.

[0055] Afterwards, the server begins to determine the global network error parameter based on the first loss function value between the first charging demand forecast data and the third charging demand forecast data, the second loss function value between the second charging demand forecast data and the third charging demand forecast data, and the third loss function value between the third charging demand forecast data and the third sample charging demand data. When calculating the first loss function value, the construction vehicle is used as the first charging pile scheduling subject, and the commuting vehicle is used as the non-charging pile scheduling subject for analysis. The server obtains the first dynamic flow characteristics of the construction vehicle in the student sample traffic flow data, such as the average driving speed of the vehicle, the parking interval time, etc., and the second dynamic flow characteristics of the commuting vehicle, such as the concentrated parking area during the peak period of the commuting vehicle. At the same time, the third dynamic flow characteristics of other special vehicles (assuming that they are test vehicles for model verification) that may exist are obtained, such as special charging time requirements. Determine the third characteristic distance between the third charging demand forecast data and the first charging demand forecast data, for example, by comparing the differences between the two in charging pile location prediction, quantity prediction, etc. Determine the fourth characteristic distance between the third charging demand forecast data and the second charging demand forecast data, which may be the difference in the degree of dispersion of the charging facility layout, etc. According to the first charging demand prediction data, the first subject number corresponding to the construction engineering vehicles and the second subject number corresponding to the commuting vehicles in the student sample traffic flow data are obtained, and the third subject number corresponding to the special vehicles is obtained according to the second charging demand prediction data. The fourth loss function value for the construction engineering vehicles between the first charging demand prediction data and the third charging demand prediction data is calculated according to the third characteristic distance, the first dynamic flow feature, and the first subject number, for example, according to the degree of influence caused by the construction engineering vehicles if they cannot be charged at the predicted charging pile location. The fifth loss function value for commuting vehicles between the first charging demand prediction data and the third charging demand prediction data is calculated according to the third characteristic distance, the third dynamic flow feature, and the second subject number. The first loss function value is calculated based on the fourth loss function value and the fifth loss function value, and the second loss function value is calculated based on the fourth characteristic distance, the second dynamic flow feature, and the third subject number.

[0056] After determining the first loss function value, the second loss function value, and the third loss function value, the server obtains the first weight value corresponding to the first loss function value, which is determined based on the importance of the first charging pile scheduling subject (construction engineering vehicle) in the entire transportation system and the degree of influence on the charging demand. Obtain the second weight value corresponding to the second loss function value, which is also determined based on the relevant importance of the special vehicle. Obtain the third weight value corresponding to the third loss function value, which is determined considering the importance of the accuracy of the third sample charging demand data to the entire learning process. Determine the first error parameter of the student charging pile scheduling network based on the first weight value, the first loss function value, the second weight value, and the second loss function value, and determine the second error parameter of the student charging pile scheduling network based on the third weight value and the third loss function value. Finally, determine the global network error parameter based on the first error parameter and the second error parameter.

[0057] Finally, the server updates the neuron weight information of the student charging pile scheduling network based on the global network error parameters. For example, if the global network error parameters indicate that there is a large deviation between the predicted charging demand data and the actual demand data (comprehensively reflected by the teacher and auxiliary networks and sample data), the server will adjust the connection weights between the neurons so that the student charging pile scheduling network can more accurately predict the charging demand the next time it processes similar student sample traffic flow data. The updated student charging pile scheduling network is used as the student charging pile scheduling network corresponding to the next round of network parameter learning process, so that it can be continuously optimized in subsequent learning rounds and gradually approach an accurate charging demand prediction model.

[0058] In a possible implementation, step S130 includes:

[0059] Acquire multiple teacher sample traffic flow data from the teacher sample traffic flow data sequence, and load the multiple teacher sample traffic flow data into the auxiliary sample traffic flow data sequence to generate a student sample traffic flow data sequence.

[0060] In a possible implementation, the step of obtaining a plurality of teacher sample traffic flow data from the teacher sample traffic flow data sequence, and loading the plurality of teacher sample traffic flow data into the auxiliary sample traffic flow data sequence to generate a student sample traffic flow data sequence includes:

[0061] Step S131, performing multiple rounds of data optimization on the teacher sample traffic flow data sequence to generate a candidate sample traffic flow data sequence corresponding to each data optimization.

[0062] Step S132, loading the candidate sample traffic flow data sequence into the auxiliary sample traffic flow data sequence until the training scale in the auxiliary sample traffic flow data sequence meets the first set scale, thereby generating a student sample traffic flow data sequence.

[0063] Wherein, step S131 specifically includes:

[0064] Step S1311, obtain the first candidate sample traffic flow data from the teacher sample traffic flow data sequence, and calculate the first characteristic distance between each second candidate sample traffic flow data in the teacher sample traffic flow data sequence and the first candidate sample traffic flow data based on the first candidate sample traffic flow data, wherein the second candidate sample traffic flow data is other sample traffic flow data in the teacher sample traffic flow data sequence except the first candidate sample traffic flow data.

[0065] Step S1312, arrange each second candidate sample traffic flow data in descending order according to the first characteristic distance, and generate the candidate sample traffic flow data sequence according to the first candidate sample traffic flow data and the first candidate sample traffic flow data of the first set scale in the descending arrangement result.

[0066] In this embodiment, first, the server obtains a plurality of teacher sample traffic flow data from the teacher sample traffic flow data sequence, and then loads these data into the auxiliary sample traffic flow data sequence to generate a student sample traffic flow data sequence.

[0067] In this process, multiple rounds of data optimization of the teacher sample traffic flow data sequence are involved to generate candidate sample traffic flow data sequences corresponding to each round of data optimization. Taking the urban traffic system as an example, the teacher sample traffic flow data sequence covers the traffic flow data of multiple first traffic monitoring areas in the city, including the city's central business district, large industrial parks, and residential areas.

[0068] When multiple rounds of data optimization are performed, the first round of data optimization begins, and the server obtains the first candidate sample traffic flow data from the teacher sample traffic flow data sequence. For example, the server selects the traffic flow data during the morning rush hour of a certain weekday from the teacher sample traffic flow data in the city's central business district as the first candidate sample traffic flow data. This data contains various traffic characteristics such as flow information, driving speed, and parking frequency of various types of vehicles (such as private cars, taxis, buses, etc.) in the business district during this period. It also carries the first sample charging demand data, which reflects the charging demand of different types of vehicles during this period, such as the power consumption of taxis after the morning rush hour operation and the location of the charging piles required.

[0069] Next, the server calculates the first characteristic distance between each second candidate sample traffic flow data and the first candidate sample traffic flow data in the teacher sample traffic flow data sequence based on the first candidate sample traffic flow data. The second candidate sample traffic flow data here refers to other sample traffic flow data in the teacher sample traffic flow data sequence except the first candidate sample traffic flow data. Taking the traffic flow data of a certain working day in a large industrial park as the second candidate sample traffic flow data as an example, the server will obtain the first characteristic weighted value of the first candidate sample traffic flow data, which is determined based on multiple characteristic factors of the traffic flow data during the morning rush hour in the central business district of the city. For example, the respective weights are determined based on factors such as the travel ratio of private cars, the passenger load rate of taxis, and the full load rate of buses, and then the first characteristic weighted value is obtained comprehensively. At the same time, the first characteristic discrete degree value of the first candidate sample traffic flow data is obtained, for example, it is determined based on the fluctuation of the traffic flow during the morning rush hour in the commercial district in different time periods (such as 8:00-8:30, 8:30-9:00, etc.). For the second candidate sample traffic flow data, its second characteristic weighted value is also obtained, for example, determined according to the different types of logistics vehicles in the industrial park, the load conditions, and the proportion of commuter vehicles, and the second characteristic discrete degree value is determined according to the fluctuation of traffic flow in different periods of the day in the industrial park. Then, the covariance degree value between the first candidate sample traffic flow data and the second candidate sample traffic flow data is determined, which may involve analyzing whether there are some common influencing factors between the morning peak traffic in the commercial area and the daytime traffic in the industrial park, such as the overall weather conditions of the city, large-scale activities, etc., and the common influence of the traffic flow of the two.

[0070] For the traffic flow data of the second candidate sample, the similarity of the traffic scale is calculated based on the weighted value of the first feature and the weighted value of the second feature. For example, the overall vehicle flow scale during the morning rush hour in the commercial district is compared with the overall vehicle flow scale during the day in the industrial park. Taking into account the different weights of the vehicle types of the two, the similarity of the traffic scale is calculated. The similarity of the traffic change is calculated based on the first feature discrete degree value and the second feature discrete degree value. For example, the change range of the traffic flow during the morning rush hour in the commercial district in a short period of time is compared with the change range of the traffic flow during the day in the industrial park in the corresponding period. The similarity of the traffic distribution is calculated based on the first feature discrete degree value, the second feature discrete degree value and the covariance degree value. For example, the similarity of the traffic flow distribution in different areas of the commercial district (such as the surrounding areas of shopping malls and office buildings) and the traffic flow distribution in different areas of the industrial park (such as production areas, office areas, etc.) is analyzed. After comprehensively considering the above factors, the first feature distance is calculated.

[0071] The server performs such calculations on all second candidate sample traffic flow data, and then arranges each second candidate sample traffic flow data in descending order according to the first feature distance. Assuming that the second set scale is 10, the server generates a candidate sample traffic flow data sequence based on the first 10 second candidate sample traffic flow data and the first candidate sample traffic flow data in the descending order. This sequence includes traffic flow data that has a high correlation with the first candidate sample traffic flow data in multiple features (determined according to the first feature distance).

[0072] For data optimization other than the first round, the server also needs to obtain the first candidate sample traffic flow data from the teacher sample traffic flow data sequence. At this time, for each second candidate sample traffic flow data, the server obtains its second characteristic distance from each obtained candidate sample traffic flow data. For example, for the traffic flow data of a certain weekday night in a residential area (as the second candidate sample traffic flow data), the server will calculate its second characteristic distance from the candidate sample traffic flow data obtained in the previous rounds of optimization (such as the previously selected commercial area morning peak, industrial park daytime and other traffic flow data). This calculation process also needs to consider a variety of factors, such as the characteristic weighted value of traffic flow in different areas, the discrete degree value, and the covariance degree value. Then calculate the weighted characteristic distance of each second characteristic distance, which is obtained by comprehensively considering the weights of the optimization results of the previous rounds. Finally, the server outputs the second candidate sample traffic flow data with the largest weighted characteristic distance of each second characteristic distance as the first candidate sample traffic flow data.

[0073] After each round of data optimization generates a candidate sample traffic flow data sequence, the server loads the candidate sample traffic flow data sequence into the auxiliary sample traffic flow data sequence. The auxiliary sample traffic flow data sequence corresponds to the second traffic monitoring area in the city, such as the newly developed area on the edge of the city, tourist attractions in the suburbs, etc. The server continues this loading process until the training scale in the auxiliary sample traffic flow data sequence meets the first set scale. For example, the first set scale is 1,000 data. When the amount of data in the auxiliary sample traffic flow data sequence reaches 1,000, a student sample traffic flow data sequence is generated. Each data in this student sample traffic flow data sequence carries the third sample charging demand data, which integrates the charging demand information in the teacher sample traffic flow data and the auxiliary sample traffic flow data, providing a rich and representative data basis for the subsequent learning of the student charging pile scheduling network.

[0074] In a possible implementation manner, if the first candidate sample traffic flow data is the sample traffic flow data corresponding to the first round of data optimization, step S1311 includes:

[0075] Randomly select a teacher sample traffic flow data from the teacher sample traffic flow data sequence and output it as the first candidate sample traffic flow data.

[0076] If the first candidate sample traffic flow data is not sample traffic flow data corresponding to the first round of data optimization, step S1311 includes:

[0077] For each second candidate sample traffic flow data, the second characteristic distance between the second candidate sample traffic flow data and each acquired candidate sample traffic flow data is obtained, and the weighted characteristic distance of each second characteristic distance is calculated.

[0078] The second candidate sample traffic flow data having the largest weighted characteristic distance among the second characteristic distances is output as the first candidate sample traffic flow data.

[0079] In a possible implementation manner, the calculating the first characteristic distance between each second candidate sample traffic flow data and the first candidate sample traffic flow data in the teacher sample traffic flow data sequence includes:

[0080] Step S1311-1, obtain the first feature weighted value and the first feature discrete degree value of the first candidate sample traffic flow data, and obtain the second feature weighted value and the second feature discrete degree value of each second candidate sample traffic flow data, and determine the covariance degree value between the first candidate sample traffic flow data and each second candidate sample traffic flow data.

[0081] Step S1311 - 2 : for each second candidate sample traffic flow data, calculate the similarity of the flow scale between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature weighted value and the second feature weighted value.

[0082] Step S1311 - 3 , calculating the flow change similarity between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature discrete degree value and the second feature discrete degree value.

[0083] Step S1311 - 4 , calculating the flow distribution similarity between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature discrete degree value, the second feature discrete degree value and the covariance degree value.

[0084] Step S1311-5, calculating the first characteristic distance between the second candidate sample traffic flow data and the first candidate sample traffic flow data based on the traffic scale similarity, the traffic change similarity and the traffic distribution similarity.

[0085] In this embodiment, when the first candidate sample traffic flow data is the sample traffic flow data corresponding to the first round of data optimization, the server arbitrarily selects a teacher sample traffic flow data from the teacher sample traffic flow data sequence and outputs it as the first candidate sample traffic flow data. For example, in a teacher sample traffic flow data sequence containing traffic flow data of multiple areas in a city, this sequence covers traffic flow information of different areas such as the city's central business district, large industrial parks, and residential areas, as well as the corresponding first sample charging demand data. The server may directly select the traffic flow data of a certain weekday during the morning rush hour as the first candidate sample traffic flow data from the traffic flow data of the city's central business district. The data during this morning rush hour contains various traffic information, such as the number of traffic flows of private cars, taxis, buses and other types of vehicles, the speed of different sections of the road, the number of stops and other traffic characteristics, as well as the first sample charging demand data generated based on these traffic conditions, such as the power consumption of taxis after operating during this period, and information on which areas may need to be charged.

[0086] When the first candidate sample traffic flow data is not the sample traffic flow data corresponding to the first round of data optimization, the operation is relatively complicated. For each second candidate sample traffic flow data, the server obtains the second characteristic distance between the second candidate sample traffic flow data and each obtained candidate sample traffic flow data, and calculates the weighted characteristic distance of each second characteristic distance. Taking urban traffic data as an example, it is assumed that several rounds of data optimization have been carried out before, and there are some candidate sample traffic flow data, such as the morning rush hour traffic flow data of the city center business district selected in the first round, and the daytime traffic flow data of the large industrial park selected in the second round. Now the new second candidate sample traffic flow data is to be processed, such as the traffic flow data of a residential area on a weekday night. The server will calculate the second characteristic distance between the residential area night traffic flow data and each candidate sample traffic flow data obtained before (the morning rush hour of the commercial area, the daytime of the industrial park, etc.). When calculating this distance, many factors should be considered, such as the proportion of different vehicle types in the traffic flow data, the speed distribution, the parking frequency and other characteristics. For each previous candidate sample traffic flow data, the server assigns different weights according to its own feature importance. For example, the taxi traffic ratio in the commercial area during the morning rush hour is high, so its weight is high, and the logistics vehicles in the industrial park during the daytime have a high weight. Then, the second feature distance with the residential area's night traffic flow data is comprehensively calculated. After that, the weighted feature distance of each second feature distance is calculated. This weight is obtained by comprehensively considering the importance of each candidate sample traffic flow data in the previous rounds of optimization. Finally, the second candidate sample traffic flow data with the largest weighted feature distance of each second feature distance is output as the first candidate sample traffic flow data. For example, after comparing the traffic flow data of multiple residential areas at different time periods or other areas, the traffic flow data of a specific residential area on weekdays at night is determined as the new first candidate sample traffic flow data.

[0087] When calculating the first characteristic distance between each second candidate sample traffic flow data and the first candidate sample traffic flow data in the teacher sample traffic flow data sequence, first obtain the first characteristic weighted value and the first characteristic discrete degree value of the first candidate sample traffic flow data, and obtain the second characteristic weighted value and the second characteristic discrete degree value of each second candidate sample traffic flow data, and determine the covariance degree value between the first candidate sample traffic flow data and each second candidate sample traffic flow data. For example, the first candidate sample traffic flow data is the traffic flow data of a large industrial park during the daytime on weekdays, in which logistics vehicles account for a large proportion. Its first characteristic weighted value will be determined according to factors such as the load type and the proportion of the number of logistics vehicles, and the first characteristic discrete degree value will be determined according to the flow fluctuation of logistics vehicles entering and leaving the park at different times during the day. For a second candidate sample traffic flow data, such as the traffic flow data of a certain weekday at the evening peak in the central business district of the city, its second characteristic weighted value will be determined according to the proportion of private cars, taxis, and buses in the commercial district during the evening peak, and the second characteristic discrete degree value will be determined according to the traffic flow fluctuation at different times during the evening peak. The covariance degree value considers whether there are common influencing factors between the two, such as the degree of common impact on the logistics transportation of industrial parks and the traffic flow of commercial areas when a city holds a large-scale event.

[0088] Next, for each second candidate sample traffic flow data, the similarity of the flow scale between the second candidate sample traffic flow data and the first candidate sample traffic flow data is calculated based on the first feature weighted value and the second feature weighted value. For example, the overall flow scale of logistics vehicles in the industrial park during the day is compared with the overall vehicle flow scale of the evening peak in the commercial area, and the weighted values ​​of the respective vehicle types are taken into account to calculate the similarity of the flow scale. The similarity of the flow change between the second candidate sample traffic flow data and the first candidate sample traffic flow data is calculated based on the first feature discrete degree value and the second feature discrete degree value, for example, the change range of the logistics vehicle flow in the industrial park during the day in different time periods is compared with the change range of the evening peak traffic flow in the commercial area during the corresponding time period. The similarity of the flow distribution between the second candidate sample traffic flow data and the first candidate sample traffic flow data is calculated based on the first feature discrete degree value, the second feature discrete degree value and the covariance degree value, for example, the similarity of the logistics vehicle flow distribution in different areas of the industrial park (production area, office area, etc.) and the evening peak traffic flow distribution in different areas of the commercial area (around the shopping mall, around the office building, etc.) is analyzed. Finally, the first characteristic distance between the second candidate sample traffic flow data and the first candidate sample traffic flow data is calculated based on the similarity of traffic scale, traffic change and traffic distribution. This first characteristic distance comprehensively reflects the degree of difference between the two traffic flow data in multiple dimensions, providing an important basis for subsequent data screening and optimization, and helping the server to more accurately construct the student sample traffic flow data sequence, thereby providing a high-quality data foundation for the learning and optimization of the charging pile scheduling network.

[0089] In a possible implementation, step S142 includes:

[0090] Step S1421, obtain a first weight value corresponding to the first loss function value, obtain a second weight value corresponding to the second loss function value, and obtain a third weight value corresponding to the third loss function value.

[0091] Step S1422, determine the first error parameter of the student charging pile scheduling network based on the first weight value, the first loss function value, the second weight value, and the second loss function value, and determine the second error parameter of the student charging pile scheduling network based on the third weight value and the third loss function value.

[0092] Step S1423: determining the global network error parameter according to the first error parameter and the second error parameter.

[0093] In a possible implementation, the step of determining the first loss function value and the second loss function value includes:

[0094] The first dynamic flow feature of the first charging pile scheduling subject and the second dynamic flow feature of the non-charging pile scheduling subject in the student sample traffic flow data are obtained, and the third dynamic flow feature of the second charging pile scheduling subject in the student sample traffic flow data is obtained.

[0095] A third characteristic distance between the third charging demand forecast data and the first charging demand forecast data is determined, and a fourth characteristic distance between the third charging demand forecast data and the second charging demand forecast data is determined.

[0096] The first number of entities corresponding to the first charging pile scheduling entity and the second number of entities corresponding to the non-charging pile scheduling entity in the student sample traffic flow data are obtained based on the first charging demand prediction data, and the third number of entities corresponding to the second charging pile scheduling entity in the student sample traffic flow data is obtained based on the second charging demand prediction data.

[0097] The fourth loss function value between the first charging demand forecast data and the third charging demand forecast data for the first charging pile scheduling subject is calculated based on the third characteristic distance, the first dynamic flow feature, and the first subject quantity, and the fifth loss function value between the first charging demand forecast data and the third charging demand forecast data for the non-charging pile scheduling subject is calculated based on the third characteristic distance, the third dynamic flow feature, and the second subject quantity.

[0098] The first loss function value is calculated based on the fourth loss function value and the fifth loss function value, and the second loss function value is calculated based on the fourth feature distance, the second dynamic flow feature, and the third subject quantity.

[0099] In this embodiment, taking the charging pile scheduling in urban traffic as an example, the determination of these weight values ​​is based on the importance of different factors. Assume that the first charging pile scheduling subject is a taxi in the central business district of the city, the non-charging pile scheduling subject is a private car in the area, and the second charging pile scheduling subject is a bus. If the accuracy of the charging demand prediction of the taxi has a greater impact on the entire network, then the first weight value corresponding to the first loss function value will be relatively high. The determination of this weight value may be based on factors such as the proportion of taxis in urban traffic, their long operating time and frequent charging demand. For buses, if their charging demand pattern in the entire transportation system is relatively special and has a certain influence on the overall charging facility layout, then the second weight value corresponding to the second loss function value will also be set according to its degree of importance. The third weight value corresponding to the third loss function value considers more the importance of the accuracy of the third sample charging demand data itself to the entire learning process. For example, if the third sample charging demand data is based on accurate traffic flow statistics and actual charging records, its weight value may be higher.

[0100] Next, the server determines the first error parameter of the student charging pile scheduling network based on the first weight value, the first loss function value, the second weight value, and the second loss function value, and determines the second error parameter of the student charging pile scheduling network based on the third weight value and the third loss function value. For example, when calculating the first error parameter, if the first weight value is 0.4, the first loss function value is 0.3 (this value is obtained through subsequent detailed calculations), the second weight value is 0.3, and the second loss function value is 0.2, then the first error parameter may obtain a specific value through a certain specific calculation method (such as weighted summation). For the second error parameter, assuming that the third weight value is 0.3 and the third loss function value is 0.4, the corresponding value is also obtained through specific calculations.

[0101] Then, the global network error parameter is determined based on the first error parameter and the second error parameter. This global network error parameter comprehensively reflects the difference between the entire student charging pile scheduling network and the teacher charging pile scheduling network, the auxiliary charging pile scheduling network and the actual third sample charging demand data in predicting charging demand, and is an important basis for adjusting the neuron weight information of the student charging pile scheduling network.

[0102] There are also detailed steps when determining the first loss function value and the second loss function value. The server first obtains the first dynamic flow characteristics of the first charging pile scheduling subject (such as a taxi) in the student sample traffic flow data and the second dynamic flow characteristics of the non-charging pile scheduling subject (such as a private car), and also obtains the third dynamic flow characteristics of the second charging pile scheduling subject (such as a bus) in the student sample traffic flow data. For the first dynamic flow characteristics of taxis, factors such as the average speed of taxis in different time periods, empty driving rate, and changes in driving routes after picking up passengers may be included. The second dynamic flow characteristics of private cars may include information such as the peak travel hours, average parking time, and parking location distribution of private cars. The third dynamic flow characteristics of buses may involve the busyness of bus routes, station stop times, and full load rates.

[0103] Then, the server determines the third characteristic distance between the third charging demand forecast data and the first charging demand forecast data, and determines the fourth characteristic distance between the third charging demand forecast data and the second charging demand forecast data. Taking the city's central business district as an example, if the third charging demand forecast data predicts that 5 charging piles need to be set up in a certain area, and the first charging demand forecast data predicts that 8 charging piles need to be set up, then there is a difference in the number of charging piles. At the same time, there may also be differences in other aspects such as the prediction of the location of the charging piles. These factors are combined to calculate the third characteristic distance. The calculation of the fourth characteristic distance is similar, comparing the differences in the number and location of charging piles between the third charging demand forecast data and the second charging demand forecast data.

[0104] Afterwards, the first subject quantity corresponding to the first charging pile dispatching subject (taxi) and the second subject quantity corresponding to the non-charging pile dispatching subject (private car) in the student sample traffic flow data are obtained based on the first charging demand forecast data, and the third subject quantity corresponding to the second charging pile dispatching subject (bus) in the student sample traffic flow data is obtained based on the second charging demand forecast data. For example, in a specific time period, according to the first charging demand forecast data, the number of taxis in the area is 100 (the first subject quantity) and the number of private cars is 500 (the second subject quantity); according to the second charging demand forecast data, the number of buses is 30 (the third subject quantity).

[0105] Then, the fourth loss function value between the first charging demand forecast data and the third charging demand forecast data for the first charging pile dispatching subject (taxi) is calculated based on the third characteristic distance, the first dynamic flow feature, and the first subject number, and the fifth loss function value between the first charging demand forecast data and the third charging demand forecast data for the non-charging pile dispatching subject (private car) is calculated based on the third characteristic distance, the third dynamic flow feature, and the second subject number. For example, if the third characteristic distance is large, the first dynamic flow feature of the taxi shows that its empty driving rate is low (meaning that the operating time is long and the charging demand is more urgent), and the number of taxis (the number of the first subject) is large, then the fourth loss function value may be large, reflecting the large difference between the first charging demand forecast data and the third charging demand forecast data in the prediction of taxi charging demand. The calculation of the fifth loss function value for private cars is similar, considering factors such as the third characteristic distance, the second dynamic flow feature of private cars, and the number of private cars (the number of the second subject).

[0106] Finally, the first loss function value is calculated based on the fourth loss function value and the fifth loss function value, and the second loss function value is calculated based on the fourth characteristic distance, the second dynamic flow feature, and the third number of subjects. For example, if the fourth loss function value is 0.2 and the fifth loss function value is 0.1, then the first loss function value may be calculated as 0.3 by some calculation method (such as summation or weighted summation, etc.). For the second loss function value, a specific value is obtained by similar calculation based on the fourth characteristic distance, the third dynamic flow feature of the bus, and the number of buses (the third number of subjects). The accurate calculation of these loss function values ​​is crucial to determine the global network error parameters, which in turn affects the learning and optimization process of the student charging pile scheduling network, enabling it to more accurately predict charging demand and perform effective charging pile scheduling.

[0107] In a possible implementation, step S150 specifically includes:

[0108] Step S151, preprocessing the target charging demand data of the target traffic monitoring area to extract key charging demand data, wherein the key charging demand data includes charging demand time period, charging demand hotspot area, charging demand type ratio and predicted charging demand amount.

[0109] Step S152, performing charging resource assessment based on the key charging demand data, and generating a charging resource assessment report, wherein the charging resource assessment report includes the distribution, available status, charging power, and expansion potential of the charging piles in the target traffic monitoring area.

[0110] Step S153, constructing a charging pile scheduling strategy according to the charging resource evaluation report and the key charging demand data, wherein the charging pile scheduling strategy includes the charging piles to be scheduled, the scheduling time, the scheduling location, and the status of the charging piles after scheduling.

[0111] Step S154, according to the charging pile scheduling strategy, a scheduling instruction is sent to the charging pile management system to control the charging pile to move or adjust the charging status. The scheduling instruction includes the identification of the charging pile, the scheduling time, the scheduling location and the charging status after scheduling. After receiving the scheduling instruction, the charging pile management system controls the charging pile to perform corresponding operations, and the operations include moving to a designated location, adjusting the charging power or changing the charging status.

[0112] Step S155, after the charging pile scheduling operation is executed, the scheduling effect of the charging pile is monitored in real time, specifically by collecting the real-time charging data, traffic flow data and user feedback data of the charging pile, and evaluating the actual scheduling effect of the charging pile scheduling. If the scheduling effect does not meet the expected conditions, the charging pile scheduling strategy is adjusted and the scheduling operation is executed again.

[0113] And, step S156, based on the historical charging pile scheduling data, real-time charging data, traffic flow data and user feedback data, the charging pile scheduling strategy is optimized and learned, so as to analyze the correlation and regularity in the historical charging pile scheduling data, real-time charging data, traffic flow data and user feedback data through a machine learning algorithm, extract the key factors affecting the charging pile scheduling effect, and intelligently adjust the charging pile scheduling strategy according to the key factors to generate an optimized charging pile scheduling strategy.

[0114] Step S157, inputting the optimized charging pile scheduling strategy into the charging pile scheduling system, and updating the scheduling parameters and scheduling strategy library of the charging pile scheduling system.

[0115] In this embodiment, taking the city's central business district as the target traffic monitoring area as an example, the target charging demand data includes the charging demand information of various traffic participants (such as taxis, private cars, buses, etc.) in the area at different times. During the preprocessing process, the server will accurately extract the key charging demand data from these data. In terms of charging demand time period, it may be analyzed that the charging demand increases significantly during the evening peak (17:00-19:00) on weekdays. This is because the traffic flow is large, the vehicle operation time is long, and the power consumption is fast. Charging demand hotspots may be concentrated at the entrance of large shopping malls, around office buildings, and near bus hubs. This is because shopping malls and office buildings are crowded places, taxis and private cars frequently get on and off, and the vehicle stay time is relatively short but the charging demand is concentrated; bus hubs are places where buses are concentrated and turned around. Buses need to be charged in time after a period of operation in order to continue to run. In terms of the proportion of charging demand types, after analysis, it may be concluded that taxi charging demand accounts for 30%, because there are more taxis operating in commercial areas and the operating hours are long; private car charging demand accounts for 40%, because there are a large number of private cars and some car owners will charge when shopping or working in commercial areas; bus charging demand accounts for 30%, which depends on the bus operating routes and the number of vehicles. In terms of the predicted charging demand, it is calculated based on factors such as traffic flow, vehicle type and operating hours. For example, during the evening peak period, a total of 100 charging positions are required, of which 30 are required for taxis, 40 for private cars and 30 for buses.

[0116] Next, the server conducts a charging resource assessment based on these key charging demand data and generates a charging resource assessment report. The server queries the distribution of existing charging piles in the target traffic monitoring area and finds that there are currently 20 charging piles around shopping malls, 15 charging piles near office buildings, and 30 charging piles at bus hubs. For the available status, through communication with the charging pile management system, it is obtained that some charging piles are in use and some are idle. For example, there are 10 in use and 10 idle around shopping malls; 8 in use and 7 idle near office buildings; 20 in use and 10 idle at bus hubs. In terms of charging power, each charging pile is different. There are fast charging piles and slow charging piles. Fast charging piles are mainly distributed in bus hubs to meet the fast charging needs of buses, while the proportion of fast and slow charging piles around shopping malls and office buildings is relatively balanced. In terms of expansion potential, shopping malls and office buildings have small expansion potential due to the limitations of building layout and land resources, while there is a certain amount of space around bus hubs to increase the number of charging piles.

[0117] Then, the server builds a charging pile scheduling strategy based on the charging resource assessment report and key charging demand data. The charging piles that need to be scheduled are determined based on the predicted charging demand and the existing distribution and availability of charging piles. For example, 5 charging piles are deployed from areas with low charging demand to the entrance of the shopping mall to meet the increased charging demand of private cars and taxis; the charging power of 10 charging piles is adjusted to fast charging mode at the bus hub to improve the charging efficiency of buses. The scheduling time is set to complete the scheduling before the evening peak (16:00 - 16:30) to ensure that the charging piles can operate normally when the peak of charging demand arrives. The scheduling location is clearly defined as placing the deployed charging piles accurately at the designated vacant position at the entrance of the shopping mall, and the adjustment of the charging piles in the bus hub is clear to the specific pile position. The status of the charging piles after scheduling is set to be turned on and can be used normally. For the charging piles with adjusted charging power, it is ensured that they operate according to the set fast charging power.

[0118] After that, the server sends a dispatch instruction to the charging pile management system according to the constructed charging pile dispatch strategy. The dispatch instruction contains detailed information, such as the identification of each charging pile that needs to be dispatched, which is the unique identifier of the charging pile in the management system, used to accurately identify each charging pile; the dispatch time is clearly set to start at 16:00; the dispatch location is accurate to the specific coordinates of the shopping mall entrance or the specific pile position in the bus hub; the charging status after dispatch is clearly set to open and fast charging (for charging piles with adjusted power) or normal charging (for deployed charging piles). After receiving the dispatch instruction, the charging pile management system controls the charging pile to perform corresponding operations according to the instruction. For charging piles that need to be moved, move them to the designated location; for charging piles that need to adjust the charging power, adjust their charging power to the specified power; for charging piles that need to change the charging status, change their status to the set status.

[0119] After the charging pile scheduling operation is executed, the server monitors the scheduling effect of the charging pile in real time. The server collects real-time charging data of the charging pile, including the charging current, voltage, charging time and other information of each charging pile, to understand the actual charging situation of the charging pile; collects traffic flow data, such as the flow changes of taxis, private cars and buses in the target traffic monitoring area during the evening peak period, to determine whether the charging demand is effectively met; collects user feedback data, such as whether taxi drivers feel that charging is more convenient and faster, whether private car owners can successfully find idle charging piles, and whether buses can charge in time to ensure normal operation. The actual scheduling effect of the charging pile scheduling is evaluated based on these data. If it is found that although the charging piles at the entrance of the shopping mall are added, the taxi cannot reach the charging pile in time due to traffic congestion, or the charging time of the bus is still too long to affect the operation, that is, the scheduling effect does not meet the expected conditions, the server adjusts the charging pile scheduling strategy. For example, re-plan the placement of the charging pile to avoid traffic congestion, or increase the number of fast charging piles at the bus hub, and perform the scheduling operation again.

[0120] At the same time, the server optimizes and learns the charging pile scheduling strategy based on historical charging pile scheduling data, real-time charging data, traffic flow data, and user feedback data. The correlation and regularity of these data are deeply analyzed through machine learning algorithms. For example, from the historical charging pile scheduling data, it is found that on certain specific dates (such as holidays or promotion days), the traffic flow and charging demand in the commercial area will change specially; from the real-time charging data, it is found that there is a certain fluctuation pattern in the charging time and charging power demand of different types of vehicles; from the traffic flow data, the impact of traffic congestion on the arrival of vehicles at the charging pile is analyzed; from the user feedback data, the user's expectations and dissatisfaction with the location of the charging pile, charging speed, etc. are understood. The server extracts the key factors that affect the scheduling effect of the charging pile, such as traffic congestion, the rationality of the layout of charging facilities, and the charging characteristics of different types of vehicles, and intelligently adjusts the charging pile scheduling strategy based on these key factors. For example, the layout of the charging pile is dynamically adjusted according to the traffic congestion situation, and the charging power distribution is optimized according to the charging characteristics of different types of vehicles, etc., to generate an optimized charging pile scheduling strategy. Finally, the optimized charging pile scheduling strategy is input into the charging pile scheduling system, and the scheduling parameters and scheduling strategy library of the charging pile scheduling system are updated, so that the charging demand can be met more accurately and efficiently in future charging pile scheduling decisions, thereby improving the operating efficiency and user satisfaction of the entire transportation system.

[0121] Figure 2 The hardware structure of the intelligent charging pile scheduling system 100 based on traffic flow prediction for implementing the intelligent charging pile scheduling method based on traffic flow prediction provided by an embodiment of the present invention is shown as follows: Figure 2As shown, the smart charging pile scheduling system 100 based on traffic flow prediction may include a processor 110, a machine-readable storage medium 120, a bus 130 and a communication unit 140.

[0122] The machine-readable storage medium 120 may store data and / or instructions. In some embodiments, the machine-readable storage medium 120 may store data obtained from an external terminal. In some embodiments, the machine-readable storage medium 120 may store data and / or instructions that the intelligent charging pile scheduling system based on traffic flow prediction 100 uses to execute or use to complete the exemplary method described in the present invention.

[0123] During the specific implementation process, one or more processors 110 execute computer executable instructions stored in the machine-readable storage medium 120, so that the processor 110 can execute the smart charging pile scheduling method based on traffic flow prediction in the above method embodiment. The processor 110, the machine-readable storage medium 120 and the communication unit 140 are connected through the bus 130, and the processor 110 can be used to control the sending and receiving actions of the communication unit 140.

[0124] The specific implementation process of the processor 110 can refer to the various method embodiments executed by the above-mentioned smart charging pile scheduling system 100 based on traffic flow prediction. The implementation principles and technical effects are similar, and this embodiment will not be repeated here.

[0125] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer executable instructions are preset. When a processor executes the computer executable instructions, the above-mentioned smart charging pile scheduling method based on traffic flow prediction is implemented.

[0126] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, various features are sometimes combined into one embodiment, drawing or description thereof.

Claims

1. A smart charging pile scheduling method based on traffic flow prediction, characterized in that: The method comprises: Obtain a teacher charging pile scheduling network, wherein the teacher charging pile scheduling network is generated by parameter learning based on a teacher sample traffic flow data sequence; the teacher sample traffic flow data sequence corresponds to a plurality of first traffic monitoring areas; each first traffic monitoring area includes a plurality of teacher sample traffic flow data; each teacher sample traffic flow data carries a first sample charging demand data; Obtain an auxiliary charging pile scheduling network, wherein the auxiliary charging pile scheduling network is generated by parameter learning based on an auxiliary sample traffic flow data sequence; the auxiliary sample traffic flow data sequence corresponds to a plurality of second traffic monitoring areas; each second traffic monitoring area includes a plurality of auxiliary sample traffic flow data; each auxiliary sample traffic flow data carries second sample charging demand data; each second traffic monitoring area does not intersect with each first traffic monitoring area; A student charging pile scheduling network is obtained based on the teacher charging pile scheduling network, and a student sample traffic flow data sequence is obtained based on the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence; the student sample traffic flow data sequence includes a plurality of student sample traffic flow data; each student sample traffic flow data carries a third sample charging demand data; Iterative network parameter learning is performed on the student charging pile scheduling network according to the student sample traffic flow data sequence until the network convergence requirement is met, and a target charging pile scheduling network is generated; Based on the target charging pile scheduling network, a decision is made on the target traffic flow data of any target traffic monitoring area, target charging demand data is generated, and charging pile scheduling is performed on the traffic participants in the target traffic monitoring area according to the target charging demand data.

2. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 1 is characterized in that: The process of each round of network parameter learning includes: For each student sample traffic flow data, the student sample traffic flow data is loaded into the teacher charging pile scheduling network to generate first charging demand prediction data; the student sample traffic flow data is loaded into the auxiliary charging pile scheduling network to generate second charging demand prediction data; the student sample traffic flow data is loaded into the student charging pile scheduling network to generate third charging demand prediction data; Determine a global network error parameter according to a first loss function value between the first charging demand prediction data and the third charging demand prediction data, a second loss function value between the second charging demand prediction data and the third charging demand prediction data, and a third loss function value between the third charging demand prediction data and the third sample charging demand data; The neuron weight information of the student charging pile scheduling network is updated according to the global network error parameter, and the updated student charging pile scheduling network is used as the student charging pile scheduling network corresponding to the next round of network parameter learning process.

3. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 1 is characterized in that: The step of obtaining the student sample traffic flow data sequence based on the teacher sample traffic flow data sequence and the auxiliary sample traffic flow data sequence comprises: Acquire multiple teacher sample traffic flow data from the teacher sample traffic flow data sequence, and load the multiple teacher sample traffic flow data into the auxiliary sample traffic flow data sequence to generate a student sample traffic flow data sequence.

4. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 3 is characterized in that: The method of obtaining a plurality of teacher sample traffic flow data from the teacher sample traffic flow data sequence and loading the plurality of teacher sample traffic flow data into the auxiliary sample traffic flow data sequence to generate a student sample traffic flow data sequence includes: Perform multiple rounds of data optimization on the teacher sample traffic flow data sequence to generate a candidate sample traffic flow data sequence corresponding to each data optimization; Loading the candidate sample traffic flow data sequence into the auxiliary sample traffic flow data sequence until the training scale in the auxiliary sample traffic flow data sequence meets the first set scale, thereby generating a student sample traffic flow data sequence; The method of performing multiple rounds of data optimization on the teacher sample traffic flow data sequence to generate a candidate sample traffic flow data sequence corresponding to each data optimization specifically includes: Acquire first candidate sample traffic flow data from the teacher sample traffic flow data sequence, and calculate the first characteristic distance between each second candidate sample traffic flow data in the teacher sample traffic flow data sequence and the first candidate sample traffic flow data based on the first candidate sample traffic flow data, wherein the second candidate sample traffic flow data is other sample traffic flow data in the teacher sample traffic flow data sequence except the first candidate sample traffic flow data; The second candidate sample traffic flow data are arranged in descending order according to the first characteristic distance, and the candidate sample traffic flow data sequence is generated according to the first second candidate sample traffic flow data of the second set size in the descending arrangement result and the first candidate sample traffic flow data.

5. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 4 is characterized in that: If the first candidate sample traffic flow data is the sample traffic flow data corresponding to the first round of data optimization; then obtaining the first candidate sample traffic flow data from the teacher sample traffic flow data sequence includes: Randomly select a teacher sample traffic flow data from the teacher sample traffic flow data sequence and output it as the first candidate sample traffic flow data; If the first candidate sample traffic flow data is sample traffic flow data not corresponding to the first round of data optimization; then obtaining the first candidate sample traffic flow data from the teacher sample traffic flow data sequence includes: For each second candidate sample traffic flow data, obtaining a second characteristic distance between the second candidate sample traffic flow data and each obtained candidate sample traffic flow data, and calculating a weighted characteristic distance of each second characteristic distance; The second candidate sample traffic flow data having the largest weighted characteristic distance among the second characteristic distances is output as the first candidate sample traffic flow data.

6. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 4 is characterized in that: The calculating of the first characteristic distance between each second candidate sample traffic flow data and the first candidate sample traffic flow data in the teacher sample traffic flow data sequence comprises: Obtaining a first feature weighted value and a first feature discrete degree value of the first candidate sample traffic flow data, and obtaining a second feature weighted value and a second feature discrete degree value of each second candidate sample traffic flow data, and determining a covariance degree value between the first candidate sample traffic flow data and each second candidate sample traffic flow data; For each second candidate sample traffic flow data, calculating the traffic scale similarity between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature weighted value and the second feature weighted value; Calculating the flow change similarity between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature discrete degree value and the second feature discrete degree value; Calculating the flow distribution similarity between the second candidate sample traffic flow data and the first candidate sample traffic flow data according to the first feature discrete degree value, the second feature discrete degree value and the covariation degree value; The first characteristic distance between the second candidate sample traffic flow data and the first candidate sample traffic flow data is calculated based on the traffic scale similarity, the traffic change similarity and the traffic distribution similarity.

7. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 2 is characterized in that: The determining of the global network error parameter according to a first loss function value between the first charging demand prediction data and the third charging demand prediction data, a second loss function value between the second charging demand prediction data and the third charging demand prediction data, and a third loss function value between the third charging demand prediction data and the third sample charging demand data includes: Obtain a first weight value corresponding to the first loss function value, obtain a second weight value corresponding to the second loss function value, and obtain a third weight value corresponding to the third loss function value; Determine a first error parameter of the student charging pile scheduling network according to the first weight value, the first loss function value, the second weight value, and the second loss function value, and determine a second error parameter of the student charging pile scheduling network according to the third weight value and the third loss function value; The global network error parameter is determined according to the first error parameter and the second error parameter.

8. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 7 is characterized in that: The step of determining the first loss function value and the second loss function value comprises: Obtaining a first dynamic flow feature of a first charging pile scheduling subject and a second dynamic flow feature of a non-charging pile scheduling subject in the student sample traffic flow data, and obtaining a third dynamic flow feature of a second charging pile scheduling subject in the student sample traffic flow data; Determine a third characteristic distance between the third charging demand forecast data and the first charging demand forecast data, and determine a fourth characteristic distance between the third charging demand forecast data and the second charging demand forecast data; According to the first charging demand prediction data, the number of first entities corresponding to the first charging pile scheduling entity and the number of second entities corresponding to the non-charging pile scheduling entity in the student sample traffic flow data are obtained; according to the second charging demand prediction data, the number of third entities corresponding to the second charging pile scheduling entity in the student sample traffic flow data is obtained; Calculating a fourth loss function value between the first charging demand forecast data and the third charging demand forecast data for the first charging pile scheduling subject according to the third characteristic distance, the first dynamic flow feature, and the first subject quantity, and calculating a fifth loss function value between the first charging demand forecast data and the third charging demand forecast data for the non-charging pile scheduling subject according to the third characteristic distance, the third dynamic flow feature, and the second subject quantity; The first loss function value is calculated based on the fourth loss function value and the fifth loss function value, and the second loss function value is calculated based on the fourth feature distance, the second dynamic flow feature, and the third subject quantity.

9. The intelligent charging pile scheduling method based on traffic flow prediction according to claim 1 is characterized in that: The step of dispatching charging piles for traffic participants in the target traffic monitoring area according to the target charging demand data specifically includes: Preprocessing the target charging demand data of the target traffic monitoring area to extract key charging demand data, wherein the key charging demand data includes charging demand time period, charging demand hotspot area, charging demand type ratio and predicted charging demand amount; Perform charging resource evaluation based on the key charging demand data and generate a charging resource evaluation report, wherein the charging resource evaluation report includes the distribution, available status, charging power and expansion potential of charging piles in the target traffic monitoring area; Constructing a charging pile scheduling strategy based on the charging resource assessment report and the key charging demand data, wherein the charging pile scheduling strategy includes the charging piles to be scheduled, the scheduling time, the scheduling location, and the status of the charging piles after scheduling; According to the charging pile scheduling strategy, a scheduling instruction is sent to the charging pile management system to control the charging pile to move or adjust the charging state, wherein the scheduling instruction includes the identification of the charging pile, the scheduling time, the scheduling location, and the charging state after scheduling. After receiving the scheduling instruction, the charging pile management system controls the charging pile to perform corresponding operations, wherein the operations include moving to a designated location, adjusting the charging power, or changing the charging state; After the charging pile scheduling operation is executed, the scheduling effect of the charging pile is monitored in real time, specifically by collecting the real-time charging data, traffic flow data and user feedback data of the charging pile, and evaluating the actual scheduling effect of the charging pile scheduling. If the scheduling effect does not meet the expected conditions, the charging pile scheduling strategy is adjusted and the scheduling operation is executed again; And, based on historical charging pile scheduling data, real-time charging data, traffic flow data and user feedback data, the charging pile scheduling strategy is optimized and learned, so as to analyze the correlation and regularity in the historical charging pile scheduling data, real-time charging data, traffic flow data and user feedback data through a machine learning algorithm, extract the key factors affecting the charging pile scheduling effect, and intelligently adjust the charging pile scheduling strategy according to the key factors to generate an optimized charging pile scheduling strategy; The optimized charging pile scheduling strategy is input into the charging pile scheduling system, and the scheduling parameters and scheduling strategy library of the charging pile scheduling system are updated.

10. A smart charging pile dispatching system based on traffic flow prediction, characterized in that: The smart charging pile scheduling system based on traffic flow prediction includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the smart charging pile scheduling method based on traffic flow prediction as described in any one of claims 1 to 9 above.

Images