Alternating current charging optimization scheduling method and system, and storage medium

Abstract

The invention discloses an AC charging optimization scheduling method and system, and a storage medium, and the method carries out the calculation of the importance degree of each priority parameter to obtain a priority weight, can better match the actual application scene of each electric vehicle compared with a mode of directly distributing a fixed weight, reduces the distribution error of the priority weight, and improves the reliability of the system. A charging influence parameter is generated through the priority parameter and the priority weight, so that the charging influence parameter is utilized to represent the objective influence of the priority index on the charging optimization scheduling of the alternating current charging station; according to the method, fitness calculation is carried out by optimizing the dispatching charging function based on the constraint condition, meanwhile, the dispatching influence of historical data and prediction data of the electric vehicle is considered, the calculated fitness relative error is smaller, the charging dispatching requirements of an alternating current charging station and an electric vehicle user can be met at the same time, optimal charging dispatching of the alternating current charging station is achieved, and the charging efficiency is improved. Under the condition that the charging requirement of an electric vehicle user is met, the profit income of the alternating current charging station is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging, and in particular to an AC charging optimization scheduling method and system, and a storage medium. Background Art

[0002] With the gradual maturity of the new energy vehicle industry and the gradual expansion of market demand, AC charging stations as a power guarantee for new energy vehicles have received more and more attention. AC charging stations are limited by the site, power consumption capacity and other restrictions, and the configured charging resources are usually relatively limited. Considering that the actual charging conditions of different electric vehicles may be different, how to reasonably allocate limited charging resources for each electric vehicle has become an urgent problem to be solved; At present, AC charging stations on the market usually use real-time scheduling charging to distribute charging, that is, real-time distribution is made according to the on-site charging conditions of each electric vehicle user. For example, some electric vehicle users with high urgency will be given priority to charge, or the charging order of electric vehicle users with a long waiting time will be put in front, etc. Although it can solve the actual charging needs of each user in real time to a certain extent, the degree of intelligence is not high, and the actual application is relatively cumbersome. Considering the on-site manpower and material support required for matching, this poses a new challenge to the charging management of AC charging stations. Summary of the invention

[0003] The present invention aims to solve one of the technical problems in the related art at least to a certain extent. To this end, the present invention proposes an AC charging optimization scheduling method and system, and a storage medium.

[0004] In a first aspect, an embodiment of the present invention provides an AC charging optimization scheduling method, comprising:

[0005] When it is determined that the AC charging station has been configured to charge each electric vehicle within a preset historical charging period, historical charging data of all electric vehicles at the AC charging station within the preset historical charging period are obtained, wherein the historical charging data of each electric vehicle includes a historical charging time period, a historical charging duration, a historical charging power, and an average electricity price within the historical charging time period;

[0006] For each electric vehicle, multiple priority indicators are determined based on historical charging data, and each priority indicator is assigned a differentiated value according to its category to obtain multiple priority parameters; the importance of each priority parameter is calculated to obtain multiple priority weights corresponding to each priority indicator; and charging impact parameters are generated based on all priority parameters and all priority weights;

[0007] Performing prediction processing on the historical charging data of each electric vehicle respectively to obtain the predicted charging data of each electric vehicle, wherein the predicted charging data of each electric vehicle includes the predicted charging time period, the predicted charging duration, the predicted charging power and the average electricity price within the predicted charging time period;

[0008] An optimized scheduling charging model corresponding to the AC charging station is established. The optimized scheduling charging model is used as the first-generation chromosome. The historical charging data, predicted charging data, charging influencing parameters of each electric vehicle and the total charging power that the AC charging station can output are used as constraints. The fitness of the first-generation chromosome under the constraints is calculated through the preset optimized scheduling charging function. The optimized scheduling charging function is:

[0009]

[0010] Among them, F is the fitness of the primary chromosome under the constraint conditions, a i is the historical charging time of the i-th electric vehicle, b i is the average electricity price of the ith electric vehicle in the historical charging period, S is the charging influencing parameter, x j is the predicted charging time of the jth electric vehicle, y j is the average electricity price of the j-th electric vehicle in the predicted charging time period, δ is the predicted charging time error of the j-th electric vehicle, ε is the average electricity price error of the j-th electric vehicle in the predicted charging time period, k is the total number of electric vehicles, P is the sum of the predicted charging powers of all electric vehicles, and P0 is the total charging power that the AC charging station can output;

[0011] Based on the calculated fitness of the first-generation chromosomes under the constraint conditions, the first-generation chromosomes are reproduced to obtain a first-generation population, wherein the first-generation population includes a plurality of first-generation chromosomes; the plurality of chromosomes in the first-generation population are replicated, crossed over and mutated to obtain a second-generation population, wherein the second-generation population includes a plurality of second-generation chromosomes; the second-generation population is used as the starting point of genetic iteration, and the iterative operations of reproduction, replication, crossover and mutation are repeated on the plurality of chromosomes in each generation population until the preset iteration conditions are met, thereby obtaining a target population including a plurality of final-generation chromosomes;

[0012] From the target population, the chromosome with the highest fitness under the constraints is selected as the target chromosome, and the charging optimization scheduling strategy of the AC charging station is generated according to the optimization scheduling charging model corresponding to the target chromosome.

[0013] Optionally, in one embodiment of the present invention, when the priority indicators include electric vehicle type, charging waiting time, charging demand power and target charging cost, the importance of each priority parameter is calculated respectively to obtain multiple priority weights corresponding to each priority indicator, including:

[0014] The importance of each priority parameter is calculated by a preset weight calculation formula group to obtain a plurality of priority weights corresponding to each priority indicator, wherein the weight calculation formula group includes at least the following formulas:

[0015]

[0016] Among them, H1 is the first priority weight corresponding to the electric vehicle type, H2 is the second priority weight corresponding to the charging waiting time, H3 is the third priority weight corresponding to the charging demand power, H4 is the fourth priority weight corresponding to the target charging cost, Q1 is the first priority parameter corresponding to the electric vehicle type, Q2 is the second priority parameter corresponding to the charging waiting time, Q3 is the third priority parameter corresponding to the charging demand power, Q4 is the fourth priority parameter corresponding to the target charging cost, α is the first weight compensation factor adapted to the first priority parameter, β is the second weight compensation factor adapted to the second priority parameter, γ is the third weight compensation factor adapted to the third priority parameter, μ is the fourth weight compensation factor adapted to the fourth priority parameter, H1+H2+H3+H4=1.

[0017] Optionally, in one embodiment of the present invention, generating a charging impact parameter according to all priority parameters and all priority weights includes:

[0018] According to all priority parameters and all priority weights, the charging impact parameter is calculated using a preset weighted impact calculation formula, where the weighted impact calculation formula is as follows:

[0019] S=H1*Q1+H2*Q2+H3*Q3+H4*Q4.

[0020] Optionally, in one embodiment of the present invention, the predictive processing is performed on the historical charging data of each electric vehicle to obtain the predicted charging data of each electric vehicle, including:

[0021] For each electric vehicle's historical charging data, the historical charging data is divided into first charging data and second charging data based on a self-service sampling method;

[0022] The first charging data is used as training data, and the second charging data is used as verification data, which are input into a pre-trained linear regression model to obtain predicted charging data of each electric vehicle.

[0023] Optionally, in one embodiment of the present invention, the first charging data is used as training data and the second charging data is used as verification data to input into a pre-trained linear regression model to obtain predicted charging data of each electric vehicle, including:

[0024] Inputting the first charging data as training data into the pre-trained linear regression model for re-training, and obtaining a feedback training result output by the linear regression model;

[0025] Inputting the second charging data as verification data into the retrained linear regression model for verification, thereby obtaining a verification training result;

[0026] According to the difference between the feedback training results and the verification training results, the retrained linear regression model is updated to obtain the target linear regression model;

[0027] The historical charging data is input into the target linear regression model to obtain the predicted charging data output by the target linear regression model.

[0028] Optionally, in one embodiment of the present invention, the breeding process of the primary chromosomes based on the calculated fitness of the primary chromosomes under the constraint conditions to obtain the first generation population includes:

[0029] Based on the calculated fitness of the primary chromosomes under the constraints, the primary chromosomes are reproduced to obtain the primary population;

[0030] A number of required chromosomes are selected from the primary population to form a first-generation population, wherein the required chromosomes are chromosomes whose fitness under constraints is greater than or equal to a first preset fitness threshold and less than a second preset fitness threshold, and the first preset fitness threshold is less than the second preset fitness threshold.

[0031] Optionally, in one embodiment of the present invention, the preset iteration condition includes at least one of the following: reaching a preset target number of iterations; the fitness of the chromosome under the constraint condition tends to be stable; reaching a preset maximum number of iterations.

[0032] Optionally, in one embodiment of the present invention, the duplication, crossover and mutation processing of multiple chromosomes in the first generation population to obtain the second generation population includes:

[0033] At least one gene on several chromosomes is randomly selected from the first generation population for replication and mutation, and a corresponding proportion of offspring are randomly selected from the first generation population as parents according to a preset crossover rate for pairing. For any two paired chromosomes, at least one gene crossover point at the corresponding position is determined from the two chromosomes, and the genes at all gene crossover points at the corresponding positions of the two chromosomes are exchanged.

[0034] In a second aspect, an embodiment of the present invention provides an AC charging optimization scheduling system, including:

[0035] at least one processor;

[0036] at least one memory for storing at least one program;

[0037] When at least one of the programs is executed by at least one of the processors, the AC charging optimization scheduling method as described in the first aspect is implemented.

[0038] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a program executable by a processor, and the program executable by the processor is used to implement the AC charging optimization scheduling method as described in the first aspect when executed by the processor.

[0039] Compared with the prior art, the AC charging optimization scheduling method, system, and storage medium proposed in the present invention have at least the following beneficial effects:

[0040] 1. By acquiring the historical charging data of all electric vehicles in the preset historical charging period at the AC charging station, and then predicting and processing the historical charging data of each electric vehicle respectively, the predicted charging data of each electric vehicle is obtained, that is, the historical charging situation of the AC charging station is considered, and the historical charging situation of the AC charging station is used as a reference, so as to predict the predicted charging data of the AC charging station in the future. The predicted charging data is used as the key data required for charging optimization scheduling. It can be seen that through the above data processing method, intelligent charging data storage, application and management are realized, which can reduce the charging management burden of the AC charging station;

[0041] 2. When multiple priority indicators are determined, differentiated values ​​are assigned to each priority indicator to obtain multiple priority parameters, so as to realize visualization and transparency of priority indicator evaluation. In particular, the importance of each priority parameter is calculated to calculate the priority weight corresponding to each priority indicator. Compared with the method of directly assigning fixed weights, the weight results obtained by the importance calculation method are independent of each electric vehicle and do not affect each other. Therefore, it can better match the actual application scenario of each electric vehicle and reduce the error of priority weight allocation; and, charging influence parameters are generated through all priority parameters and all priority weights, so as to better characterize the objective influence of the priority indicators of electric vehicles on the charging optimization scheduling of AC charging stations by using charging influence parameters;

[0042] 3. Taking the historical charging data, predicted charging data, charging influence parameters and the total charging power that the AC charging station can output of each electric vehicle as constraints, the optimization scheduling charging function is used for calculation, that is, the expected profit income of the optimized scheduling charging model corresponding to the AC charging station is calculated. The optimized scheduling charging function takes into account both the scheduling influence of the historical data of each electric vehicle and the scheduling influence of the predicted data of each electric vehicle. Finally, the fitness under the constraint conditions is generated by combining the calculated charging influence parameters. The relative error of the fitness is smaller and can meet the charging scheduling needs of AC charging stations and electric vehicle users at the same time.

[0043] 4. After determining the fitness of the primary chromosome under the constraints, the primary chromosome is continuously optimized through iterative operations of reproduction, replication, crossover and mutation processing, and finally the chromosome with the highest fitness under the constraints in the target population is determined as the target chromosome, that is, the optimal optimization scheduling charging model is determined to generate the charging optimization scheduling strategy of the AC charging station according to the optimization scheduling charging model, so as to achieve the optimal charging scheduling of the AC charging station and increase the profit income of the AC charging station while meeting the charging needs of electric vehicle users. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of an AC charging optimization scheduling method provided by an embodiment of the present invention;

[0045] Figure 2 yes Figure 1 Flowchart of step S3000 in FIG.

[0046] Figure 3 yes Figure 2 Flowchart of step S3200 in FIG.

[0047] Figure 4 yes Figure 1 The flowchart of the step S5000 in step “based on the calculated fitness of the primary chromosome under the constraint condition, breeding the primary chromosome to obtain the first generation population”;

[0048] Figure 5 It is a structural diagram of an AC charging optimization scheduling system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] like Figure 1 As shown, the AC charging optimization scheduling method provided by an embodiment of the present invention may include but is not limited to steps S1000 to S6000.

[0050] Step S1000: when it is determined that the AC charging station has configured each electric vehicle to charge within the preset historical charging period, the historical charging data of all electric vehicles in the preset historical charging period at the AC charging station are obtained, wherein the historical charging data of each electric vehicle includes the historical charging time period, the historical charging duration, the historical charging power and the average electricity price within the historical charging time period. Obviously, the historical charging duration is less than or equal to the length of the preset historical charging period;

[0051] Step S2000: for each electric vehicle, multiple priority indicators are determined according to historical charging data, and each priority indicator is assigned a differentiated value according to the category of each priority indicator to obtain multiple priority parameters; the importance of each priority parameter is calculated to obtain multiple priority weights corresponding to each priority indicator; and a charging impact parameter is generated according to all priority parameters and all priority weights;

[0052] That is to say, when multiple priority indicators are determined, differentiated values ​​are assigned to each priority indicator to obtain multiple priority parameters, thereby realizing visualization and transparency of priority indicator evaluation. In particular, the importance of each priority parameter is calculated to calculate the priority weight corresponding to each priority indicator. Compared with the method of directly assigning fixed weights, the weight results obtained by the importance calculation method are separate and independent of each other for each electric vehicle, so that the actual application scenario of each electric vehicle can be better matched, and the error of priority weight allocation can be reduced; and the charging influence parameters are generated through all priority parameters and all priority weights, so as to better characterize the objective influence of the priority indicators of electric vehicles on the charging optimization scheduling of AC charging stations by using the charging influence parameters;

[0053] Step S3000: performing prediction processing on the historical charging data of each electric vehicle respectively to obtain the predicted charging data of each electric vehicle, wherein the predicted charging data of each electric vehicle includes the predicted charging time period, the predicted charging duration, the predicted charging power and the average electricity price within the predicted charging time period;

[0054] It can be seen that by acquiring the historical charging data of all electric vehicles in the preset historical charging period of the AC charging station, and then predicting and processing the historical charging data of each electric vehicle respectively, the predicted charging data of each electric vehicle is obtained, that is, the historical charging situation of the AC charging station is considered, and the historical charging situation of the AC charging station is used as a reference, so as to predict the predicted charging data of the AC charging station in the future. The predicted charging data is used as the key data required for charging optimization scheduling. It can be seen that through the above data processing method, intelligent charging data storage, application and management are realized, which can reduce the charging management burden of the AC charging station;

[0055] Step S4000: Establish an optimized scheduling charging model corresponding to the AC charging station, use the optimized scheduling charging model as the first-generation chromosome, take the historical charging data, predicted charging data, charging influencing parameters of each electric vehicle and the total charging power that the AC charging station can output as constraints, and calculate the fitness of the first-generation chromosome under the constraints through a preset optimized scheduling charging function, wherein the optimized scheduling charging function is:

[0056]

[0057] Among them, F is the fitness of the primary chromosome under the constraint conditions, a i is the historical charging time of the i-th electric vehicle, b i is the average electricity price of the ith electric vehicle in the historical charging period, S is the charging influencing parameter, x j is the predicted charging time of the jth electric vehicle, y j is the average electricity price of the j-th electric vehicle in the predicted charging time period, δ is the predicted charging time error of the j-th electric vehicle, ε is the average electricity price error of the j-th electric vehicle in the predicted charging time period, k is the total number of electric vehicles, P is the sum of the predicted charging powers of all electric vehicles, and P0 is the total charging power that the AC charging station can output;

[0058] It can be seen from the optimized scheduling charging function that when P < P0, it indicates that the sum of the predicted charging powers of all electric vehicles is still less than the total charging power that the AC charging station can output. At this time, the AC charging station can provide the corresponding predicted charging power for all electric vehicles. In this case, calculations can be performed based on the predicted charging data. Considering the differential impact of the evolution from historical charging data to predicted charging data, a predicted charging duration error value δ and an average electricity price error value ε within the predicted charging time period are set, so as to calculate the corresponding fitness of the initial generation chromosome under the constraint conditions. When P ≥ P0, it indicates that the sum of the predicted charging powers of all electric vehicles has exceeded the total charging power that the AC charging station can output. At this time, the AC charging station cannot provide the corresponding predicted charging power for all electric vehicles. In this case, it is not applicable to calculate based on the predicted charging data. Considering the authenticity and effectiveness of the historical charging data, in this case, the historical charging data is used for calculation, and the fitness of the initial generation chromosome under the constraint conditions can still be reliably calculated. That is, whether the total charging power that the AC charging station can output can provide the predicted charging power of all electric vehicles, the fitness of the initial generation chromosome under the constraint conditions can be correspondingly calculated. Therefore, the applicable scenario of this embodiment is wider, has good application prospects, and is very convenient and reliable to apply in the actual scenario. That is to say, taking the historical charging data, predicted charging data, charging impact parameters of each electric vehicle, and the total charging power that the AC charging station can output as constraint conditions, calculations are performed through the optimized scheduling charging function, that is, calculating the expected profit income of the optimized scheduling charging model corresponding to the AC charging station. This optimized scheduling charging function not only considers the scheduling impact of the historical data of each electric vehicle, but also considers the scheduling impact of the predicted data of each electric vehicle, and finally generates the fitness under the constraint conditions in combination with the already calculated charging impact parameters. The relative error of this fitness is smaller, and it can simultaneously meet the charging scheduling requirements of the AC charging station and electric vehicle users.

[0059] Step S5000: Based on the calculated fitness of the initial generation chromosome under the constraint conditions, perform reproduction processing on the initial generation chromosome to obtain the first-generation population, where the first-generation population includes multiple first-generation chromosomes; perform replication, crossover, and mutation processing on the multiple chromosomes in the first-generation population to obtain the second-generation population, where the second-generation population includes multiple second-generation chromosomes; taking the second-generation population as the genetic iteration starting point, repeat the iterative operations of reproduction, replication, crossover, and mutation processing on the multiple chromosomes in each generation of population until the preset iteration conditions are met, to obtain the target population including multiple final-generation chromosomes, where the preset iteration conditions can be set accordingly according to the specific iteration scenario, which is not limited here. For example, it can include but is not limited to at least one of the following: reaching the preset target iteration number; the fitness of the chromosome under the constraint conditions tends to be stable; reaching the preset maximum iteration number.

[0060] It should be noted that the target number of iterations is less than or equal to the maximum number of iterations, that is, the target number of iterations is the number of iterations to be achieved. It can be set to the maximum number of iterations or to be less than the maximum number of iterations. The maximum number of iterations is used as the upper limit of the iteration and must not be exceeded to avoid adverse situations such as the crash of the iteration system. To determine whether the fitness of the chromosome under the constraints is stable, the difference variance between the fitness variance values ​​of all chromosomes in the population of this iteration and the fitness variance values ​​of all chromosomes in the population of the previous iteration can be calculated to determine whether the difference variance is within the preset variance error range. If so, it can be considered that the fitness of the chromosome under the constraints is stable. Otherwise, it is considered that the fitness of the chromosome under the constraints is unstable.

[0061] Step S6000: Select the chromosome with the highest fitness under the constraint conditions from the target population as the target chromosome, and generate the charging optimization scheduling strategy of the AC charging station according to the optimization scheduling charging model corresponding to the target chromosome;

[0062] It can be seen that after determining the fitness of the first-generation chromosomes under the constraints, the first-generation chromosomes are continuously optimized through iterative operations of reproduction, replication, crossover and mutation processing, and finally the chromosome with the highest fitness under the constraints in the target population is determined as the target chromosome, that is, the optimal optimization scheduling charging model is determined, and the charging optimization scheduling strategy of the AC charging station is generated according to the optimization scheduling charging model, so as to achieve the optimal charging scheduling of the AC charging station and increase the profit income of the AC charging station while meeting the charging needs of electric vehicle users.

[0063] In one embodiment, there may be multiple ways to determine in step S1000 that the AC charging station has been configured to charge each electric vehicle within a preset historical charging period, which needs to be determined specifically according to the working mode, content, etc. of the AC charging station. For example, the AC charging station is configured with multiple AC charging piles, that is, the power distribution of all AC charging piles is issued by the AC charging station. At this time, the total charging power that the AC charging station can output is the sum of the charging powers that the AC charging station can issue and configure for each AC charging pile. As long as an electric vehicle is charged at any AC charging pile, the electric vehicle can be considered as a user of the AC charging station. In this way, it is only necessary to count the charging conditions of each AC charging pile within the preset historical charging period to determine all the electric vehicles charged within this period, thereby obtaining the historical charging data of all the electric vehicles at the AC charging station within the preset historical charging period. Of course, it is not ruled out that the electric vehicle goes directly to the AC charging station for charging, and the charging data of this part of the electric vehicle should be included in the historical charging data of all electric vehicles.

[0064] In one embodiment, the categories of priority indicators can be multiple, which are determined by technicians in this field according to actual application scenarios, and are not limited here. For example, they can be, but are not limited to, electric vehicle types, charging waiting time, charging demand power, and target charging cost, where electric vehicle types can be divided into cars, medium-sized cars, and large cars according to common body types, and each type of vehicle can be further divided according to its purpose, which will not be repeated here; charging waiting time, that is, the time the electric vehicle waits for charging after entering the AC charging station or arriving at the AC charging pile under the AC charging station; charging demand power, that is, the total power required for the electric vehicle to be charged this time; target charging cost, that is, the charging amount expected to be consumed by the electric vehicle for this time. In addition, the method of determining multiple priority indicators according to historical charging data in step S2000 can be multiple, for example, the actual charging cost of the electric vehicle in the historical charging time period can be determined by the product of the historical charging time and the average electricity price in the historical charging time period, and the corresponding value or range of the target charging cost can be estimated based on this data; for example, the corresponding value or range of the charging demand power of the electric vehicle can be estimated by the historical charging power.

[0065] In one embodiment, in step S2000, each priority indicator is assigned a value according to its category, and there may be multiple ways to obtain multiple priority parameters. The specific value assignment size may be set accordingly according to the actual scenario, and there is no limitation here. For example, considering the charging difference caused by different types of cars, medium-sized cars and large cars, the priority parameter of the car may be assigned a value of 1, the priority parameter of the medium-sized car may be assigned a value of 1.5, and the priority parameter of the large car may be assigned a value of 2. Alternatively, a charging waiting time threshold range is set, and the threshold range includes an upper limit value and a lower limit value of the charging waiting time. If the charging waiting time is less than or equal to the lower limit value of the charging waiting time, it means that the waiting time is not very long at this time, and the priority parameter of the charging waiting time is assigned a value of 1; if the charging waiting time is between the lower limit value and the upper limit value of the charging waiting time, it means that the waiting time is relatively long at this time, and the priority parameter of the charging waiting time is assigned a value of 1.5; if the charging waiting time is greater than or equal to the lower limit value of the charging waiting time, it means that the waiting time is already very long at this time, and the electric vehicle user is likely to be impatient, so the priority needs to be advanced, and the priority parameter of the charging waiting time is assigned a value of 2. It can be understood that the above-mentioned assignment method can also be optimized according to different scenarios. For example, for large vehicles, the uses of large vehicles can be further divided, such as into transport vehicles, passenger vehicles and tow trucks, etc., and then differentiated assignments can be made for large vehicles of different uses, thereby obtaining multiple priority parameters corresponding to large vehicles of different uses.

[0066] In one embodiment, the optimized scheduling charging model in step S4000 is used to achieve optimized scheduling through the determined relevant data of the electric vehicle. For example, the optimized scheduling charging model corresponding to the first-generation chromosome realizes optimized scheduling of the electric vehicle in the future time period based on the historical charging data, predicted charging data and charging influencing parameters of each electric vehicle. As the first-generation chromosome is continuously iterated and updated, the optimized scheduling charging model is also continuously updated, but it still realizes optimized scheduling of the electric vehicle in the corresponding time period based on the charging data and charging influencing parameters of each electric vehicle. In other words, the architecture of the optimized scheduling charging model is not limited. As long as the above-mentioned optimized scheduling requirements are met, it can be nested in any relevant existing model, for example, it can be applied to but not limited to convolutional neural networks, clustering models or random forest models, etc. Since the architecture and working principles of this part of the existing models are well known to those skilled in the art, they are not elaborated here.

[0067] In one embodiment of the present invention, when the priority indicators include electric vehicle type, charging waiting time, charging demand power and target charging cost, the steps in step S2000 respectively calculate the importance of each priority parameter to obtain multiple priority weights corresponding to each priority indicator, which may include but is not limited to step S2100.

[0068] Step S2100: Calculate the importance of each priority parameter using a preset weight calculation formula group to obtain a plurality of priority weights corresponding to each priority indicator, wherein the weight calculation formula group includes at least the following formulas:

[0069]

[0070] Among them, H1 is the first priority weight corresponding to the electric vehicle type, H2 is the second priority weight corresponding to the charging waiting time, H3 is the third priority weight corresponding to the charging demand power, H4 is the fourth priority weight corresponding to the target charging cost, Q1 is the first priority parameter corresponding to the electric vehicle type, Q2 is the second priority parameter corresponding to the charging waiting time, Q3 is the third priority parameter corresponding to the charging demand power, Q4 is the fourth priority parameter corresponding to the target charging cost, α is the first weight compensation factor adapted to the first priority parameter, β is the second weight compensation factor adapted to the second priority parameter, γ is the third weight compensation factor adapted to the third priority parameter, μ is the fourth weight compensation factor adapted to the fourth priority parameter, H1+H2+H3+H4=1.

[0071] In this step, the importance of each priority parameter is calculated using a preset weight calculation formula group. Since the weight calculation formula group includes multiple different formulas, the importance of each priority parameter can be calculated separately using different formulas, thereby obtaining the priority weight corresponding to each priority parameter. This is separate for each electric vehicle and does not affect each other, so it can better match the actual application scenarios of each electric vehicle and reduce the error in priority weight allocation.

[0072] It should be noted that the specific values ​​of α, β, γ and μ can be configured according to the specific circumstances of each priority parameter, and there is no restriction here. For example, if the priority indicator corresponding to the first priority parameter has a higher priority, then for the sake of balance, α can be set to be relatively small. Conversely, if the corresponding priority indicator has a lower priority, α can be set to be relatively large. Other parameters such as β, γ and μ can be set similarly and will not be elaborated here.

[0073] In one embodiment of the present invention, when the priority indicators include electric vehicle type, charging waiting time, charging demand power and target charging cost, the step in step S2000 generates charging impact parameters based on all priority parameters and all priority weights, which may include but is not limited to step S2200.

[0074] Step S2200: According to all priority parameters and all priority weights, a preset weighted influence calculation formula is used to calculate a charging influence parameter, wherein the weighted influence calculation formula is as follows:

[0075] S=H1*Q1+H2*Q2+H3*Q3+H4*Q4.

[0076] In this step, all priority parameters and all priority weights are calculated using the preset weighted impact calculation formula. Since both the impact of all priority parameters and the impact of all priority weights are taken into account, the charging impact parameters finally calculated can better characterize the actual impact of the priority indicators on charging, thereby better matching the actual application scenarios of each electric vehicle and reflecting the actual demand for electric vehicle charging.

[0077] like Figure 2 As shown, in one embodiment of the present invention, step S3000 may include but is not limited to steps S3100 to S3200.

[0078] Step S3100: for each electric vehicle's historical charging data, divide the historical charging data into first charging data and second charging data based on a self-service sampling method;

[0079] Step S3200: input the first charging data as training data and the second charging data as verification data into a pre-trained linear regression model to obtain predicted charging data of each electric vehicle.

[0080] In this step, considering that the historical charging data is the real charging data of the electric vehicle, it can be considered that the historical charging data is reliable. In order to better utilize the reliability of the historical charging data, the historical charging data is divided into the first charging data and the second charging data based on the self-service sampling method, that is, the historical charging data is divided into different parts of data at the beginning, and then the first charging data is used as training data and the second charging data is used as verification data for training respectively. This not only ensures the stability of the training effect, but also can further verify and update the training results, so as to obtain a linear regression model that meets the requirements better, and the output predicted charging data of each electric vehicle is more accurate and has a smaller error.

[0081] It should be noted that the pre-training method of the linear regression model can be determined by technical personnel in this field according to the actual scenario, which is not elaborated here; the historical charging data, the first charging data and the second charging data in the above embodiments are all general descriptions, and the same is applicable to their specific parameters. For example, when the historical charging data includes the historical charging time period and the historical charging power, the historical charging time period is divided into the first historical charging time period and the second historical charging time period, and the first charging data and the second charging data can be the charging power corresponding to the first historical charging time period and the charging power corresponding to the second historical charging time period, respectively. The summary of the charging powers corresponding to these two sections is the historical charging power. In this way, the charging powers corresponding to different charging time periods can be distinguished, and the data can be well distinguished for training and verification. In other cases, such as when the historical charging data includes the historical charging time, the average electricity price in the historical charging time period, etc., the first charging data and the second charging data can be set similarly, which is not elaborated here.

[0082] like Figure 3 As shown, in one embodiment of the present invention, step S3200 may include but is not limited to steps S3210 to S3240.

[0083] Step S3210: input the first charging data as training data into the pre-trained linear regression model for re-training, and obtain a feedback training result output by the linear regression model;

[0084] Step S3220: input the second charging data as verification data into the retrained linear regression model for verification, to obtain a verification training result;

[0085] Step S3230: updating the retrained linear regression model according to the difference between the feedback training result and the verification training result to obtain a target linear regression model;

[0086] Step S3240: input the historical charging data into the target linear regression model to obtain the predicted charging data output by the target linear regression model.

[0087] In this step, after the pre-training of the linear regression model, that is, the training is completed, the first charging data is input as training data into the linear regression model again for training to strengthen the training characteristics of the linear regression model and obtain the feedback training result output by the linear regression model. On this basis, the second charging data is input as verification data into the re-trained linear regression model for verification to detect the training effect of the linear regression model and obtain the verification training result. Finally, by comparing the difference between the feedback training results and the verification training results obtained from the two trainings, the linear regression model is updated so that the linear regression model is further strengthened on the original basis, thereby obtaining the target linear regression model, which is used to stably output predicted charging data with smaller error.

[0088] In one embodiment, there may be multiple ways to update the retrained linear regression model based on the difference between the feedback training results and the verification training results in step S3230, which are not limited here. For example, based on the difference between the feedback training results and the verification training results, the specific difference result data between the two can be determined, and the difference result data can be input into the linear regression model again as feedback for feedback training, thereby realizing the update of the linear regression model.

[0089] like Figure 4 As shown, in one embodiment of the present invention, the steps in step S5000, based on the calculated fitness of the primary chromosomes under constraints, reproduce the primary chromosomes to obtain the first generation population, may include but is not limited to steps S5100 to S5200.

[0090] Step S5100: Based on the calculated fitness of the primary chromosome under the constraint condition, the primary chromosome is reproduced to obtain a primary population;

[0091] Step S5200: Select a number of required chromosomes from the primary population to form a first-generation population, wherein the required chromosomes are chromosomes whose fitness under constraints is greater than or equal to a first preset fitness threshold and less than a second preset fitness threshold, and the first preset fitness threshold is less than the second preset fitness threshold.

[0092] In this step, the fitness range of the selected chromosome is limited by setting the first preset fitness threshold and the second preset fitness threshold. If the fitness of a chromosome under the constraint condition is less than the first preset fitness threshold, it means that the chromosome cannot meet the application requirements of the AC charging station, so it should be removed; if the fitness of a chromosome under the constraint condition is greater than the second preset fitness threshold, it means that the chromosome may have abnormal conditions (such as abnormal mutation, replacement, etc.). In other words, the fitness of the chromosome exceeds the idealized second preset fitness threshold, so it should also be removed. Finally, the chromosomes whose fitness under the constraint condition is between the first preset fitness threshold and the second preset fitness threshold are selected from the first generation population to form a first generation population that meets the requirements of the charging scenario.

[0093] It should be noted that the roulette wheel selection strategy can be used but is not limited to be used to carry out primary reproduction of primary chromosomes, so that a primary population with better stability can be obtained. The principle of the roulette wheel selection strategy is to calculate the probability of the chromosome appearing in the offspring according to its fitness, and randomly select chromosomes to form the offspring population according to this probability. That is to say, the higher the chromosome fitness, the greater the probability of it being selected.

[0094] In one embodiment, the specific numerical values ​​of the first preset fitness threshold and the second preset fitness threshold can be multiple, and can be set accordingly in different application scenarios. For example, the first preset fitness threshold can be but not limited to a constant between 0 and 1, and the second preset fitness threshold can be but not limited to a constant between 1 and 3, etc., which is not limited here.

[0095] In one embodiment of the present invention, the steps in step S5000, which replicate, cross over and mutate multiple chromosomes in the first generation population to obtain the second generation population, may include but is not limited to step S5300.

[0096] Step S5300: randomly select at least one gene on several chromosomes from the first generation population for replication and mutation, and randomly select a corresponding proportion of offspring from the first generation population as parents for pairing according to a preset crossover rate. For any two paired chromosomes, determine at least one gene crossover point at the corresponding position from the two chromosomes, and exchange the genes at all gene crossover points at the corresponding positions in the two chromosomes, wherein each gene in the chromosome represents the charging income corresponding to the charging power output by the AC charging station per unit time.

[0097] In this step, by randomly selecting a corresponding proportion of offspring as parents for pairing in pairs, the mutual differences in offspring pairing are reflected, that is, there is no restriction on the pairing selection of offspring chromosomes. On this basis, at least one gene crossover point at the corresponding position is determined from the two chromosomes respectively, so that the genes on the corresponding gene crossover points are crossed. When the overall gene crossover points and the number of crossovers involved in the crossover reach a certain level, a good crossover effect can be achieved, thereby ensuring reliable and effective crossover processing of multiple chromosomes in the population.

[0098] In one embodiment, the preset crossover rate and the corresponding ratio of paired chromosomes can be randomly set by a computer or controlled by a computer to be set to corresponding values. For example, a preliminary test can be performed with a crossover rate of 0.2. If it does not meet the requirements of the iterative scenario in the later stage, the crossover rate can be updated. Generally speaking, the crossover rate needs to be set according to the actual scenario. If the crossover rate is higher, the population update speed may be too fast, which will affect the population update performance. However, if it is too low, the population search will not achieve the expected effect due to the small sample size.

[0099] Figure 5 FIG. 1 is a schematic diagram of the structure of an AC charging optimization scheduling system 1000 provided in one embodiment of the present invention. Figure 5 As shown, the AC charging optimization scheduling system 1000 includes a memory 1100 and a processor 1200. The number of the memory 1100 and the processor 1200 can be one or more. Figure 5 In the embodiment, a memory 1100 and a processor 1200 are taken as an example; the memory 1100 and the processor 1200 in the device may be connected via a bus or other means. Figure 5 The example of connecting through bus is taken in the following.

[0100] The memory 1100 is a computer-readable storage medium that can be used to store software programs, computer executable programs, and modules, such as program instructions / modules corresponding to the AC charging optimization scheduling method provided in any embodiment of the present invention. The processor 1200 implements the above-mentioned AC charging optimization scheduling method by running the software programs, instructions, and modules stored in the memory 1100.

[0101] The memory 1100 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function. In addition, the memory 1100 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 1100 may further include a memory remotely arranged relative to the processor 1200, and these remote memories may be connected to the device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0102] An embodiment of the present invention further provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the AC charging optimization scheduling method provided in any embodiment of the present invention.

[0103] An embodiment of the present invention further provides a computer program product, including a computer program or computer instructions, wherein the computer program or computer instructions are stored in a computer-readable storage medium, a processor of a computer device reads the computer program or computer instructions from the computer-readable storage medium, and the processor executes the computer program or computer instructions, so that the computer device executes the AC charging optimization scheduling method provided in any embodiment of the present invention.

[0104] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.

[0105] In hardware implementation, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some physical components or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0106] The terms "component", "module", "system", etc. used in this specification are used to represent computer-related entities, hardware, firmware, a combination of hardware and software, software, or software in execution. For example, a component can be, but is not limited to, a process running on a processor, a processor, an object, an executable file, an execution thread, a program, or a computer. By way of illustration, both applications running on a computing device and a computing device can be components. One or more components may reside in a process or an execution thread, and a component may be located on a computer or distributed between two or more computers. In addition, these components may be executed from various computer-readable media having various data structures stored thereon. Components may communicate, for example, through local or remote processes based on signals having one or more data packets (e.g., data from two components interacting with another component between a local system, a distributed system, or a network, such as the Internet interacting with other systems through signals).

Claims

1. An AC charging optimization scheduling method, characterized in that: include: When it is determined that the AC charging station has been configured to charge each electric vehicle within a preset historical charging period, historical charging data of all electric vehicles at the AC charging station within the preset historical charging period are obtained, wherein the historical charging data of each electric vehicle includes a historical charging time period, a historical charging duration, a historical charging power, and an average electricity price within the historical charging time period; For each electric vehicle, multiple priority indicators are determined based on historical charging data, and each priority indicator is assigned a differentiated value according to its category to obtain multiple priority parameters; the importance of each priority parameter is calculated to obtain multiple priority weights corresponding to each priority indicator; and charging impact parameters are generated based on all priority parameters and all priority weights; The historical charging data of each electric vehicle is respectively predicted and processed to obtain the predicted charging data of each electric vehicle, wherein the predicted charging data of each electric vehicle includes the predicted charging time period, the predicted charging duration, the predicted charging power and the average electricity price within the predicted charging time period; An optimized scheduling charging model corresponding to the AC charging station is established. The optimized scheduling charging model is used as the first-generation chromosome. The historical charging data, predicted charging data, charging influencing parameters of each electric vehicle and the total charging power that the AC charging station can output are used as constraints. The fitness of the first-generation chromosome under the constraints is calculated through the preset optimized scheduling charging function. The optimized scheduling charging function is: Among them, F is the fitness of the primary chromosome under the constraint conditions, a i is the historical charging time of the i-th electric vehicle, b i is the average electricity price of the i-th electric vehicle in the historical charging period, S is the charging influencing parameter, x j is the predicted charging time of the jth electric vehicle, y j is the average electricity price of the j-th electric vehicle in the predicted charging time period, δ is the predicted charging time error of the j-th electric vehicle, ε is the average electricity price error of the j-th electric vehicle in the predicted charging time period, k is the total number of electric vehicles, P is the sum of the predicted charging powers of all electric vehicles, and P0 is the total charging power that the AC charging station can output; Based on the calculated fitness of the first-generation chromosomes under the constraint conditions, the first-generation chromosomes are reproduced to obtain a first-generation population, wherein the first-generation population includes a plurality of first-generation chromosomes; the plurality of chromosomes in the first-generation population are replicated, crossed over and mutated to obtain a second-generation population, wherein the second-generation population includes a plurality of second-generation chromosomes; the second-generation population is used as the starting point of genetic iteration, and the iterative operations of reproduction, replication, crossover and mutation are repeated on the plurality of chromosomes in each generation population until the preset iteration conditions are met, thereby obtaining a target population including a plurality of final-generation chromosomes; From the target population, the chromosome with the highest fitness under the constraints is selected as the target chromosome, and the charging optimization scheduling strategy of the AC charging station is generated according to the optimization scheduling charging model corresponding to the target chromosome.

2. The AC charging optimization scheduling method according to claim 1, characterized in that: When the priority indicators include electric vehicle type, charging waiting time, charging demand power and target charging cost, the importance of each priority parameter is calculated respectively to obtain multiple priority weights corresponding to each priority indicator, including: The importance of each priority parameter is calculated by a preset weight calculation formula group to obtain a plurality of priority weights corresponding to each priority indicator, wherein the weight calculation formula group includes at least the following formulas: Among them, H1 is the first priority weight corresponding to the electric vehicle type, H2 is the second priority weight corresponding to the charging waiting time, H3 is the third priority weight corresponding to the charging demand power, H4 is the fourth priority weight corresponding to the target charging cost, Q1 is the first priority parameter corresponding to the electric vehicle type, Q2 is the second priority parameter corresponding to the charging waiting time, Q3 is the third priority parameter corresponding to the charging demand power, Q4 is the fourth priority parameter corresponding to the target charging cost, α is the first weight compensation factor adapted to the first priority parameter, β is the second weight compensation factor adapted to the second priority parameter, γ is the third weight compensation factor adapted to the third priority parameter, μ is the fourth weight compensation factor adapted to the fourth priority parameter, H1+H2+H3+H4=1.

3. The AC charging optimization scheduling method according to claim 2, characterized in that: The generating of charging impact parameters according to all priority parameters and all priority weights includes: According to all priority parameters and all priority weights, the charging impact parameter is calculated using a preset weighted impact calculation formula, where the weighted impact calculation formula is as follows: S=H1*Q1+H2*Q2+H3*Q3+H4*Q4.

4. The AC charging optimization scheduling method according to claim 1, characterized in that: The predictive processing is performed on the historical charging data of each electric vehicle to obtain the predicted charging data of each electric vehicle, including: For each electric vehicle's historical charging data, the historical charging data is divided into first charging data and second charging data based on a self-service sampling method; The first charging data is used as training data, and the second charging data is used as verification data, which are input into a pre-trained linear regression model to obtain predicted charging data of each electric vehicle.

5. The AC charging optimization scheduling method according to claim 4, characterized in that: The first charging data is used as training data and the second charging data is used as verification data to input into the pre-trained linear regression model to obtain the predicted charging data of each electric vehicle, including: Inputting the first charging data as training data into the pre-trained linear regression model for re-training, and obtaining a feedback training result output by the linear regression model; Inputting the second charging data as verification data into the retrained linear regression model for verification, thereby obtaining a verification training result; According to the difference between the feedback training results and the verification training results, the retrained linear regression model is updated to obtain the target linear regression model; The historical charging data is input into the target linear regression model to obtain the predicted charging data output by the target linear regression model.

6. The AC charging optimization scheduling method according to claim 1, characterized in that: The method of performing reproduction processing on the primary chromosomes based on the calculated fitness of the primary chromosomes under the constraint conditions to obtain the first generation population includes: Based on the calculated fitness of the primary chromosomes under the constraints, the primary chromosomes are reproduced to obtain the primary population; A number of required chromosomes are selected from the primary population to form a first-generation population, wherein the required chromosomes are chromosomes whose fitness under constraints is greater than or equal to a first preset fitness threshold and less than a second preset fitness threshold, and the first preset fitness threshold is less than the second preset fitness threshold.

7. The AC charging optimization scheduling method according to claim 1, characterized in that: The preset iteration condition includes at least one of the following: reaching a preset target number of iterations; the fitness of the chromosome under the constraint condition tends to be stable; reaching a preset maximum number of iterations.

8. The AC charging optimization scheduling method according to claim 1, characterized in that: The duplication, crossover and mutation processing of multiple chromosomes in the first generation population to obtain the second generation population includes: At least one gene on several chromosomes is randomly selected from the first generation population for replication and mutation, and a corresponding proportion of offspring are randomly selected from the first generation population as parents according to a preset crossover rate for pairing. For any two paired chromosomes, at least one gene crossover point at the corresponding position is determined from the two chromosomes, and the genes at all gene crossover points at the corresponding positions of the two chromosomes are exchanged.

9. An AC charging optimization scheduling system, characterized in that: include: at least one processor; at least one memory for storing at least one program; When at least one of the programs is executed by at least one of the processors, the AC charging optimization scheduling method as described in any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: A processor-executable program is stored therein, and when the processor-executable program is executed by the processor, it is used to implement the AC charging optimization scheduling method as described in any one of claims 1 to 8.

Images