Charging pricing method, server, system, equipment and medium

Abstract

The present invention provides a charging pricing method, server, system, device and medium, the method comprising: after receiving a charging request from an electric vehicle, obtaining the current state of the electric vehicle, obtaining the current state of the power grid from the power grid provider system, obtaining the current state of the charging service facility from the charging service provider system, and obtaining a trained mathematical model from the central cloud; according to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility, using the distributed reinforcement learning DRL method to use the trained mathematical model to determine the charging pricing strategy of the electric vehicle. The present invention can reduce network burden, reduce transmission energy consumption, effectively reduce user waiting time, respond to dynamic needs of electric vehicle users, optimize the overall efficiency of the power grid, and determine charging pricing strategies for charging service providers for electric vehicles.

Description

Technical Field

[0001] The present invention relates to the field of charging service technology, and in particular to a charging pricing method, an MEC server, a charging pricing system, corresponding computer equipment and a computer-readable storage medium. Background Art

[0002] Existing charging pricing methods mainly process at the granularity of a single charging station and lack global consideration. Charging pricing models are mainly based on static models in specific scenarios and lack the versatility to solve practical problems.

[0003] The charging pricing strategy that performs pricing calculations in the central cloud needs to collect the required raw data from the network edge and transmit it to the central cloud. After the calculation is completed, the processed results are transmitted from the central cloud to the terminal. The data transmission volume is large and the latency is long, which increases the network load pressure and energy consumption. Especially when the supply and demand of charging services change dynamically, the computing requirements and latency generated by the charging pricing strategy will increase significantly. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a charging pricing method, server, system, device and medium in view of the above-mentioned deficiencies in the prior art, so as to solve the problems that the charging pricing method in the prior art lacks globality and universality, has a large data transmission volume, a long time delay, and cannot quickly determine the charging pricing strategy when the supply and demand change dynamically.

[0005] In a first aspect, the present invention provides a charging pricing method, which is applied to an edge computing MEC server, and the method includes:

[0006] After receiving the charging request of the electric vehicle, the current state of the electric vehicle is obtained, the current state of the power grid is obtained from the power grid provider system, the current state of the charging service facility is obtained from the charging service provider system, and the trained mathematical model is obtained from the central cloud;

[0007] According to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility, a distributed reinforcement learning (DRL) method is adopted to determine the charging pricing strategy of the electric vehicle using the trained mathematical model.

[0008] Preferably, after the distributed reinforcement learning (DRL) method is used to determine the charging pricing strategy of the electric vehicle using the trained mathematical model, the method further includes:

[0009] The charging pricing strategy is sent to the charging service provider system, so that the charging service provider system sends the charging price to the electric vehicle according to the charging pricing strategy.

[0010] Preferably, the current state of the electric vehicle includes: the capacity, remaining energy, current location, and charging service type of the electric vehicle;

[0011] The current state of the power grid includes: the current power supply price and the current load state of the power grid;

[0012] The current status of the charging service facilities includes: the number, location, current operation status, current queue length, and current road traffic conditions for the electric vehicle to reach the charging service facilities.

[0013] Preferably, before acquiring the trained mathematical model from the central cloud, the method further comprises:

[0014] The historical status of the electric vehicle, the historical status of the power grid and the historical status of the charging service facility are sent to the central cloud, so that the central cloud can obtain the trained mathematical model based on the training data model of the historical status of the electric vehicle, the historical status of the power grid and the historical status of the charging service facility.

[0015] Preferably, the mathematical model specifically adopts a Markov decision process MDP model.

[0016] Preferably, in the mathematical model:

[0017] The defined states include: electric vehicle state, grid state, and charging service facility state;

[0018] Defined actions include: increase or decrease of charging price;

[0019] Defined incentives include: revenue for charging service providers and / or grid load balancing.

[0020] Preferably, the method of determining the charging pricing strategy of the electric vehicle using the trained mathematical model using a distributed reinforcement learning (DRL) method according to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility specifically includes:

[0021] The current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility are taken as the current state S t Inputting the trained mathematical model;

[0022] The DRL method is used according to the current state S t Produce an action a t , according to the action a t Transfer to the next state S (t+1) , and obtain the corresponding reward r(S t , a t ), according to the reward r(S t, a t ) Determine a charging pricing strategy for the electric vehicle.

[0023] In a second aspect, the present invention provides a charging pricing device, which is arranged in an edge computing MEC server, and the device includes:

[0024] An acquisition module is used to acquire the current state of the electric vehicle after receiving a charging request from the electric vehicle, and acquire the current state of the power grid from the power grid provider system, acquire the current state of the charging service facilities from the charging service provider system, and acquire the trained mathematical model from the central cloud;

[0025] A pricing module is connected to the acquisition module and is used to determine the charging pricing strategy of the electric vehicle using the trained mathematical model by adopting a distributed reinforcement learning (DRL) method according to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility.

[0026] In a third aspect, the present invention provides a charging pricing system, comprising:

[0027] Edge computing MEC server, used to execute the charging pricing method as described above;

[0028] An electric vehicle, used to provide the current status of the electric vehicle to the MEC server;

[0029] A power grid provider system, used to provide the current power grid status to the MEC server;

[0030] A charging service provider system, used to provide the current charging service facility status to the MEC server;

[0031] The central cloud is used to provide the trained mathematical model to the MEC server.

[0032] In a fourth aspect, the present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the charging pricing method as described above.

[0033] In a fifth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the charging pricing method as described above is implemented.

[0034] The present invention provides a charging pricing method, server, system, device and medium. The MEC server interacts with electric vehicles, power grid provider systems and charging service provider systems to collect current status information of electric vehicles, power grids and charging service facilities. The MEC server uses a DRL method to determine the charging pricing strategy using a mathematical model trained by a central cloud. The original data no longer needs to be transmitted from the network edge to the central cloud for processing. Only a small amount of data needs to be transmitted between the MEC server and the central cloud, which can reduce the network burden, reduce transmission energy consumption, effectively reduce user waiting time, respond to the dynamic needs of electric vehicle users, optimize the overall efficiency of the power grid, and determine the charging pricing strategy for electric vehicles for charging service providers. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 is a flow chart of a charging pricing method according to Embodiment 1 of the present invention;

[0036] Figure 2 This is a schematic diagram of the structure of a charging pricing device according to Embodiment 2 of the present invention;

[0037] Figure 3 is a structural diagram of a charging pricing system according to Embodiment 3 of the present invention;

[0038] Figure 4 It is a structural diagram of a computer device according to embodiment 4 of the present invention. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0040] It should be understood that the specific embodiments and drawings described herein are only used to explain the present invention rather than to limit the present invention.

[0041] It can be understood that, in the absence of conflict, the various embodiments of the present invention and the various features in the embodiments can be combined with each other.

[0042] It can be understood that, for the convenience of description, the drawings of the present invention only show the parts related to the present invention, while the parts irrelevant to the present invention are not shown in the drawings.

[0043] It can be understood that each unit and module involved in the embodiments of the present invention may correspond to only one physical structure, or may be composed of multiple physical structures, or multiple units and modules may be integrated into one physical structure.

[0044] It can be understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of the present invention may occur in an order different from that marked in the drawings.

[0045] It is understood that the flowcharts and block diagrams of the present invention illustrate the possible architectures, functions, and operations of the systems, devices, equipment, and methods according to the various embodiments of the present invention. Each box in the flowchart or block diagram may represent a unit, module, program segment, or code, which contains executable instructions for implementing the specified functions. Moreover, each box or combination of boxes in the block diagram and flowchart may be implemented by a hardware-based system that implements the specified functions, or may be implemented by a combination of hardware and computer instructions.

[0046] It can be understood that the units and modules involved in the embodiments of the present invention can be implemented by software or hardware. For example, the units and modules can be located in a processor.

[0047] Embodiment 1:

[0048] like Figure 1 As shown, Embodiment 1 of the present invention provides a charging pricing method. In this embodiment, the charging pricing method is applied to Figure 2 and 3 The edge computing MEC (Mobile Edge Computing) server 1 shown in FIG. 1 is set in FIG. Figure 3 In the charging pricing system shown in the figure, information interaction with the electric vehicle 2, the grid provider system 3, the charging service provider system 4 and the central cloud 5 is realized, and the charging pricing device set in the MEC server 1 is as follows Figure 2 As shown, it includes an acquisition module 11 and a pricing module 12 , which are used to determine a charging pricing strategy for the electric vehicle 2 for the charging service provider system 4 .

[0049] like Figure 1 As shown, the charging pricing method applied to the MEC server 1 includes:

[0050] S11. After receiving the charging request from the electric car 2, the current state of the electric car 2 is obtained, as well as the current state of the power grid from the power grid provider system 3, the current state of the charging service facilities from the charging service provider system 4, and the trained mathematical model from the central cloud 5.

[0051] Specifically, in this embodiment, if Figure 3As shown, the MEC server 1 is connected to the electric vehicle 2, the power grid provider system 3, the charging service provider system 4, and the central cloud 5. When the electric vehicle 2 sends a charging request according to its own charging demand, the MEC server 1 receives the request and collects the status information related to the charging request at the current moment, including the current state of the electric vehicle 2, the current state of the power grid of the power grid provider system 3, and the current state of the charging service facilities of the charging service provider system 4, so that the MEC server 1 can collect the calculation data required for charging pricing nearby, accurately obtain the dynamic changes of charging supply and demand at the current moment, and at the same time obtain the trained mathematical model for pricing calculation for the charging request from the central cloud 5. The central cloud 5 and the MEC server 1 only need to transmit the model training parameters obtained after the central cloud 5 trains the corresponding charging pricing mathematical model, which reduces the data transmission volume between the MEC server 1 and the central cloud 5, thereby reducing the delay of charging pricing calculation, reducing network load pressure and energy consumption, and the central cloud 5 trains the charging pricing model to obtain the corresponding model training parameters. The trained mathematical model can globally utilize the service capabilities of the charging service facilities to provide charging services for the electric vehicle 2 and optimize the overall efficiency of the power grid.

[0052] In this embodiment, the current state of the electric vehicle 2 includes: the capacity, remaining energy, current location, and charging service type of the electric vehicle 2;

[0053] The current state of the power grid includes: the current power supply price and the current load state of the power grid;

[0054] The current status of the charging service facilities includes: the number, location, current operation status, current queue length, and current road traffic conditions for the electric vehicle 2 to reach the charging service facilities.

[0055] Specifically, in this embodiment, the current state data used for charging pricing calculation is collected from three aspects: the electric vehicle 2, the power grid, and the charging service facilities. The current state of the electric vehicle 2 collects the capacity, remaining energy, current location, and charging service type of the electric vehicle 2. The current state of the power grid collects the current power supply price and current load state of the power grid. The current state of the charging service facilities collects the number, location, current operating state, current queue length, and current road traffic conditions of the electric vehicle 2 reaching the charging service facilities. The current state of the electric vehicle 2 and the current state of the charging service facilities will affect which charging stations the user of the electric vehicle 2 can choose for charging. For example, the charging service type of the electric vehicle 2 depends on the electric vehicle. 2 The charging speed is divided into fast, slow and super fast, which together with the capacity and remaining energy of the electric vehicle 2 determines the charging time required for the electric vehicle 2. The current location of the electric vehicle 2 and the location of the charging service facility determine the distance between the two. Users usually choose the nearest charging service facility for charging, but at the same time they will also consider the current operating status of the charging service facility, the current queue length and the current road traffic conditions to reach the charging service facility. The price will also guide the user's final charging decision, so the current power supply price of the power grid, the number of charging service facilities and other related costs will also be collected. Stable charging service quality is also an important influencing factor, so it is also necessary to collect the current load status of the power grid to ensure the load balance of the power grid during the charging process.

[0056] In this embodiment, before obtaining the trained mathematical model from the central cloud 5, the method further includes:

[0057] The historical status of the electric vehicle 2, the historical status of the power grid and the historical status of the charging service facility are sent to the central cloud 5, so that the central cloud 5 can obtain the trained mathematical model based on the training data model of the historical status of the electric vehicle 2, the historical status of the power grid and the historical status of the charging service facility.

[0058] Specifically, in this embodiment, the central cloud 5 is responsible for data training that requires a lot of computing power. The MEC server 1 sends the collected data to the central cloud 5 for training, and the central cloud 5 sends the trained mathematical model to the MEC server 1 for reasoning. The central cloud 5 needs to obtain the trained mathematical model and needs to update the data used for mathematical model training in a timely manner. Therefore, the current status of the electric vehicle 2, the current status of the power grid, and the current status of the charging service facilities previously collected for charging pricing are also sent to the central cloud 5 as historical data, so that the central cloud 5 can collect the historical status of the electric vehicle 2, the historical status of the power grid, and the historical status of the charging service facilities required for training the mathematical model.

[0059] S12. According to the current state of the electric vehicle 2, the current state of the power grid and the current state of the charging service facility, a distributed reinforcement learning (DRL) method is used to determine the charging pricing strategy of the electric vehicle 2 using the trained mathematical model.

[0060] Specifically, in this embodiment, since the charging service facilities are deployed in a distributed manner, the DRL method is adopted to learn the distributed characteristics of the data through multi-layer networks and nonlinear transformations, and to explore the optimal charging pricing strategy through a reward and punishment mechanism. A module in the AI ​​(Antifcial Intelligence) algorithm is deployed on each MEC server 1, which may also be called an intelligent body or agent, and is used to collect and organize data and is responsible for reasoning in the AI ​​algorithm. According to the mathematical model trained by the central cloud 5, the collected data is input into the mathematical model to obtain the charging pricing strategy of the electric vehicle 2. The DRL method avoids the delay caused by the need to transmit a large amount of data in the traditional centralized AI algorithm, can reduce the network burden, reduce transmission energy consumption, and effectively reduce user waiting time.

[0061] In this embodiment, the mathematical model specifically adopts a Markov Decision Process (MDP) model.

[0062] In this embodiment, in the mathematical model:

[0063] The defined states include: electric vehicle 2 state, grid state, and charging service facility state;

[0064] Defined actions include: increase or decrease of charging price;

[0065] Defined incentives include: revenue for charging service providers and / or grid load balancing.

[0066] In this embodiment, according to the current state of the electric vehicle 2, the current state of the power grid and the current state of the charging service facility, a distributed reinforcement learning DRL method is used to use the trained mathematical model to determine the charging pricing strategy of the electric vehicle 2, which specifically includes:

[0067] The current state of the electric vehicle 2, the current state of the power grid and the current state of the charging service facility are taken as the current state S t Inputting the trained mathematical model;

[0068] The DRL method is used according to the current state S t Produce an action a t , according to the action a t Transfer to the next state S(t+1) , and obtain the corresponding reward r(S t , a t ), according to the reward r(S t , a t ) Determine the charging pricing strategy for the electric vehicle 2.

[0069] Specifically, in this embodiment, the DRL method models the pricing decision process as a Markov decision process MDP. MDP is a mathematical model defined as a four-tuple (S, A, R, P): S represents the state space, S t represents the state of the environment where the agent is at time t (i.e., the current state of the electric vehicle 2, the current state of the power grid, and the current state of the charging service facility); A represents the action space, a t represents the action taken by the agent at time t (such as increasing or decreasing the price of charging service); R represents reward / penalty, which represents the feedback of a learning process, where r(S t , a t ) represents S at time t t In this state, the agent makes a t After the action, transfer to S (t+1) The reward obtained by the state (such as the profit obtained by the charging service provider); P represents the strategy space, which represents the state transition probability, that is, the agent is in state S t Execute action a t And transfer to the next state S with a certain probability (t+1) ,In the Markov decision process, the agent interacts with the environment in a trial-and-error manner. ,At each time step, the agent observes the state of the environment and generates an ,action, after which the environment will transfer to the next state, during which the agent will receive a ,reward, which serves as a guiding signal to evaluate the quality of the current ,action. By obtaining more optimized rewards from more actions taken by the ,agent of each MEC server 1, the agent learns the decision-making ,strategy through continuous interaction with the environment to maximize the ,benefit, and achieve optimal control of the charging service price.

[0070] Specifically, in this embodiment, the reward can also be determined based on the multiple objectives to be achieved by the system. If the goal is to maximize profit, the reward can be defined as the profit obtained by the charging service provider, and / or, if the goal is to minimize the peak-to-valley difference, the reward can be defined as grid load balancing.

[0071] Preferably, after the distributed reinforcement learning (DRL) method is used to determine the charging pricing strategy of the electric vehicle 2 using the trained mathematical model, the method further includes:

[0072] The charging pricing strategy is sent to the charging service provider system 4 , so that the charging service provider system 4 sends the charging price to the electric vehicle 2 according to the charging pricing strategy.

[0073] Specifically, in this embodiment, the MEC server 1 does not directly provide information to the electric vehicle 2 user. The MEC server 1 provides charging pricing strategies to the charging service provider based on the user's needs, power company prices and other conditions. The charging service provider then reports it to the user. Each electric vehicle 2 user makes a charging decision based on the service price and waiting time, and each charging service provider modifies the service price to avoid congestion and obtain maximum profit.

[0074] Embodiment 2:

[0075] like Figure 2 As shown, embodiment 2 of the present invention provides a charging pricing device. In this embodiment, the charging pricing device is arranged in an edge computing MEC server 1, and the MEC server 1 is arranged as follows Figure 3 The charging pricing system shown.

[0076] like Figure 2 As shown, the charging pricing device provided in the MEC server includes:

[0077] The acquisition module 11 is used to obtain the current state of the electric vehicle 2 after receiving the charging request of the electric vehicle 2, and obtain the current state of the power grid from the power grid provider system 3, obtain the current state of the charging service facilities from the charging service provider system 4, and obtain the trained mathematical model from the central cloud 5;

[0078] The pricing module 12 is connected to the acquisition module 11, and is used to determine the charging pricing strategy of the electric vehicle 2 using the trained mathematical model by adopting a distributed reinforcement learning DRL method according to the current state of the electric vehicle 2, the current state of the power grid and the current state of the charging service facility.

[0079] Optionally, the current state of the electric vehicle 2 includes: the capacity, remaining energy, current location, and charging service type of the electric vehicle 2;

[0080] The current state of the power grid includes: the current power supply price and the current load state of the power grid;

[0081] The current status of the charging service facilities includes: the number, location, current operation status, current queue length, and current road traffic conditions for the electric vehicle 2 to reach the charging service facilities.

[0082] Optionally, the device further comprises:

[0083] The first sending module is used to send the historical status of the electric vehicle 2, the historical status of the power grid and the historical status of the charging service facility to the central cloud 5, so that the central cloud 5 can obtain the trained mathematical model based on the training data model of the historical status of the electric vehicle 2, the historical status of the power grid and the historical status of the charging service facility.

[0084] Optionally, the mathematical model specifically adopts a Markov decision process MDP model.

[0085] Optionally, in the mathematical model:

[0086] The defined states include: electric vehicle 2 state, grid state, and charging service facility state;

[0087] Defined actions include: increase or decrease of charging price;

[0088] Defined incentives include: revenue for charging service providers and / or grid load balancing.

[0089] Optionally, the pricing module 12 specifically includes:

[0090] An input unit is used to take the current state of the electric vehicle 2, the current state of the power grid and the current state of the charging service facility as the current state S t Inputting the trained mathematical model;

[0091] A pricing unit is used to adopt a DRL method according to the current state S t Produce an action a t , according to the action a t Transfer to the next state S (t+1) , and obtain the corresponding reward r(S t , a t ), according to the reward r(S t , a t ) Determine the charging pricing strategy for the electric vehicle 2.

[0092] Optionally, the device further comprises:

[0093] The second sending module is used to send the charging pricing strategy to the charging service provider system 4, so that the charging service provider system 4 sends the charging price to the electric vehicle 2 according to the charging pricing strategy.

[0094] Embodiment 3:

[0095] like Figure 3 As shown, Embodiment 3 of the present invention provides a charging pricing system, including:

[0096] Edge computing MEC server 1, used to execute the charging pricing method as described in Example 1;

[0097] The electric vehicle 2 is used to send a charging request to the MEC server 1 and provide the current status of the electric vehicle 2;

[0098] A power grid provider system 3, used to provide the current power grid status to the MEC server 1;

[0099] The charging service provider system 4 is used to provide the current charging service facility status to the MEC server 1;

[0100] The central cloud 5 is used to provide the trained mathematical model to the MEC server 1.

[0101] Specifically, in this embodiment, the electric car 2 sends its own data to the MEC server 1, including: the capacity, remaining energy, current location, charging service type, etc. of the electric car 2; the power grid provider system 3 sends the current state of the power grid to the MEC server 1, including: the current power supply price of the power grid, the current load status, etc.; the charging service provider system 4 sends the current status of the charging service facilities to the MEC server 1, including: the number, location, current operating status, current queue length of the charging service facilities, and the current road traffic conditions of the electric car 2 to the charging service facilities, so that the MEC server 1 determines the charging pricing strategy based on the current environmental status and achieves optimal control of the charging service price.

[0102] Embodiment 4:

[0103] like Figure 4 As shown, Embodiment 4 of the present invention provides a computer device, which includes a memory 10 and a processor 20, wherein a computer program is stored in the memory 10, and when the processor 20 runs the computer program stored in the memory 10, the processor 20 executes the charging pricing method as described in Embodiment 1.

[0104] The memory 10 is connected to the processor 20. The memory 10 may be a flash memory, a read-only memory or other memory. The processor 20 may be a central processing unit or a single-chip microcomputer.

[0105] Embodiment 5:

[0106] Embodiment 5 of the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the charging pricing method as described in Embodiment 1 is implemented.

[0107] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, computer program modules or other data). Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable read only memory), flash memory or other memory technology, CD-ROM (Compact Disc Read-Only Memory), digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer.

[0108] Embodiments 1-5 of the present invention provide a charging pricing method, server, system, device and medium, which interact with electric vehicles, power grid provider systems, and charging service provider systems through MEC servers to collect current status information of electric vehicles, power grids and charging service facilities, and use the mathematical model trained by the central cloud to determine the charging pricing strategy by the MEC server using the DRL method. The original data no longer needs to be transmitted from the edge of the network to the central cloud for processing. Only a small amount of data needs to be transmitted between the MEC server and the central cloud, which can reduce the network burden, reduce transmission energy consumption, effectively reduce user waiting time, respond to the dynamic needs of electric vehicle users, optimize the overall efficiency of the power grid, and determine the charging pricing strategy for electric vehicles for charging service providers.

[0109] It is to be understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of the present invention, but the present invention is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A charging pricing method, characterized in that: Applied to edge computing MEC server, data is exchanged between MEC server and electric vehicle, power grid provider system, and charging service provider system, and current status information of electric vehicle, power grid and charging service facilities is collected. The mathematical model trained by the central cloud is used to determine the charging pricing strategy by MEC server using DRL method. The method includes: After receiving the charging request of the electric vehicle, the current state of the electric vehicle is obtained, the current state of the power grid is obtained from the power grid provider system, the current state of the charging service facility is obtained from the charging service provider system, and the trained mathematical model is obtained from the central cloud; According to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility, a distributed reinforcement learning (DRL) method is adopted to determine the charging pricing strategy of the electric vehicle using the trained mathematical model.

2. The method according to claim 1, characterized in that After the distributed reinforcement learning (DRL) method is used to determine the charging pricing strategy of the electric vehicle using the trained mathematical model, the method further includes: The charging pricing strategy is sent to the charging service provider system, so that the charging service provider system sends the charging price to the electric vehicle according to the charging pricing strategy.

3. The method according to claim 1, characterized in that The current state of the electric vehicle includes: the capacity, remaining energy, current location, and charging service type of the electric vehicle; The current state of the power grid includes: the current power supply price and the current load state of the power grid; The current status of the charging service facilities includes: the number, location, current operation status, current queue length, and current road traffic conditions for the electric vehicle to reach the charging service facilities.

4. The method according to claim 1, characterized in that: Before acquiring the trained mathematical model from the central cloud, the method further includes: The historical status of the electric vehicle, the historical status of the power grid and the historical status of the charging service facility are sent to the central cloud, so that the central cloud can obtain the trained mathematical model based on the training data model of the historical status of the electric vehicle, the historical status of the power grid and the historical status of the charging service facility.

5. The method according to any one of claims 1 to 4, characterized in that: The mathematical model specifically adopts a Markov decision process MDP model.

6. The method according to claim 5, characterized in that In the mathematical model: The defined states include: electric vehicle state, grid state, and charging service facility state; Defined actions include: increase or decrease of charging price; Defined incentives include: revenue for charging service providers and / or grid load balancing.

7. The method according to claim 6, characterized in that The method of using a distributed reinforcement learning (DRL) method and the trained mathematical model to determine the charging pricing strategy of the electric vehicle according to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility specifically includes: The current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility are taken as the current state S t Inputting the trained mathematical model; The DRL method is used according to the current state S t Produce an action a t , according to the action a t Transfer to the next state S (t+1) , and obtain the corresponding reward r(S t , a t ), according to the reward r(S t , a t ) Determine a charging pricing strategy for the electric vehicle.

8. A charging pricing device, characterized in that: The device is set up on the edge computing MEC server, and the data is exchanged with the electric vehicle, the power grid provider system, and the charging service provider system through the MEC server to collect the current status information of the electric vehicle, the power grid, and the charging service facilities. The mathematical model trained by the central cloud is used to determine the charging pricing strategy by the MEC server using the DRL method. The device includes: An acquisition module is used to acquire the current state of the electric vehicle after receiving a charging request from the electric vehicle, and acquire the current state of the power grid from the power grid provider system, acquire the current state of the charging service facilities from the charging service provider system, and acquire the trained mathematical model from the central cloud; A pricing module is connected to the acquisition module and is used to determine the charging pricing strategy of the electric vehicle using the trained mathematical model by adopting a distributed reinforcement learning (DRL) method according to the current state of the electric vehicle, the current state of the power grid and the current state of the charging service facility.

9. A charging pricing system, characterized in that: include: An edge computing MEC server, configured to execute the charging pricing method as described in any one of claims 1 to 7; An electric vehicle, used to provide the current status of the electric vehicle to the MEC server; A power grid provider system, used to provide the current state of the power grid to the MEC server; A charging service provider system, used to provide the current status of charging service facilities to the MEC server; The central cloud is used to provide the trained mathematical model to the MEC server.

10. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the charging pricing method as described in any one of claims 1 to 7.

11. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the charging pricing method as described in any one of claims 1 to 7 is implemented.

Images