A system for providing price optimization for electric vehicle charging stations

Abstract

The present invention relates to a system (1) for providing artificial intelligence- based price recommendations specifically designed for electric vehicle charging stations (E).

Description

[0001] DESCRIPTION

[0002] A SYSTEM FOR PROVIDING PRICE OPTIMIZATION FOR ELECTRIC

[0003] VEHICLE CHARGING STATIONS

[0004] Technical Field

[0005] The present invention relates to a system for providing artificial intelligence-based price recommendations specifically designed for electric vehicle charging stations.

[0006] Background of the Invention

[0007] Today, with the growing demand for electric vehicles, the need for efficient electric vehicle charging stations also becomes more important. Electric vehicle charging station pricing models today are largely created by using traditional static pricing techniques. These methods may be easy in terms of implementation; however, they often cannot keep up with dynamic changes in demand and external factors. And this leads to potential revenue losses. The static pricing models used by electric vehicle charging stations lack dynamism. The pricing structure is determined based on average electricity prices, operating costs and a fixed profit margin. Although these static pricing models are simple to implement, they cannot adapt to changing demand and supply conditions. Rule-based dynamic pricing models adjust prices based on predetermined thresholds and triggering factors. When usage rate of a station exceeds a certain value, prices may automatically increase by a predetermined percentage. Although this provides the possibility to respond to demand to a certain extent, it is an outdated approach that lacks predictive capabilities.

[0008] For this reason, considering the studies and deficiencies included in the current technique, it is understood that there is a need for a system for providing artificial intelligence-based price recommendations specifically designed for electric vehicle charging stations.

[0009] The Turkish patent document no. TR2021 / 019983, an application included in the state of the art, discloses an electric charging station. The said invention relates to an electric vehicle charging station which enables users to receive fast service with high or low priority; the charging units in the charging station to be instantly dynamically planned as low priority and high priority; and the service received according to the said planning to be priced.

[0010] The Chinese patent document no. CN113361789, another application included in the state of the art, discloses a charging station service pricing method based on hierarchical game. The said invention describes a charging station service pricing method based on a hierarchical game. According to the method, firstly, a game model of electric vehicle charging and charging station pricing is constructed based on a hierarchical Stackelberg game, wherein the game model comprises game characteristics that pricing of a charging station of a charging operation company affects electric vehicle decisions, decisions of the electric vehicles affect each other, and the decisions of the electric vehicles affect pricing of the charging station; a charging cost optimization problem is defined according to an optimization target of the electric vehicle, and a profit optimization problem is defined according to an optimization target of a charging operation company; and then the convergence of the problem is analyzed based on the optimization theory, the optimization problem of the electric vehicle and the operation company is solved based on the optimization algorithm, and finally the solved pricing result is used as the pricing strategy of the charging operation company for the charging station operated by the charging operation company. According to the method, the rational decision of charging of the electric vehicle is considered, and the income of a charging operation company can be improved to a great extent. The Chinese patent document no. CN115848196, another application included in the state of the art, discloses a charging guiding method based on dynamic demands and new energy consumption. The said invention describes an electric vehicle ordered charging guiding method based on dynamic demands and new energy consumption. The method comprises the steps: directly obtaining battery reserve information by an electric vehicle user, and predicting a next charging time period; the position information and the utilization rate information of the charging stations in the target area along the way and near the parking position and the position information of all the electric vehicles to be charged are directly provided for the user, the user knows the charge of one-time charging, and the power grid load is optimized while it is ensured that the electric quantity of the user is sufficient. According to the dynamic time-of-use electricity price strategy considering the new energy output condition, the electric vehicle is guided to be charged in order to achieve on-site consumption of the new energy, and the minimum total charging cost of the user, the optimal electric quantity reserve during traveling and the minimum power grid load peak-valley difference serve as optimization targets; constraint conditions such as user charging requirements and new energy output are comprehensively considered, and a multi-target optimization model which is based on the dynamic time-of-use electricity price and meets the charging dynamic requirements is established.

[0011] The Spanish patent document no. ES2944384, another application included in the state of the art, discloses a dynamic pricing for electric vehicle charging. The said invention includes defining a region according to different charging zones in the region, each zone including different charging stations for electric vehicles, each located near a road, a parking area or a building area. In the invention, a quota of electricity that is expected to be distributed is assigned to each charging zone in an electrical network; a request is then received at one of the charging stations to deliver electric charge to a corresponding vehicle at a certain time; a base price for charging is charged at the charging station and an electricity distribution run rate for the charging zone associated with the charging station is determined based on an amount of electricity already distributed by all charging stations located in the cargo area; and finally, the base price is adjusted upwards when the execution rate exceeds the quota, but downwards when the execution rate does not reach the quota.

[0012] The Chinese patent document no. CN115848196, another application included in the state of the art, discloses an electric vehicle charging station dynamic pricing method for regulating and controlling electric energy quality. The said invention describes an electric vehicle charging station dynamic pricing method for regulating and controlling electric energy quality. The method is mainly used for solving the problem that a plurality of national grid electric vehicle charging stations exist in a region. In the method, based on historical operation data of a charging station, the demand response of a user is simulated according to the comprehensive cost of charging different charging stations by an electric vehicle user, and the effect of regulating and controlling the electric energy quality is achieved by formulating a reasonable service fee price, so that the influence of network loss and voltage deviation caused by access of an electric vehicle load to a power grid is minimized. In the invention, user demand response is simulated through specific factors influencing user decision making, and under the situation that the same operator manages and operates the charging station, on the premise that the total revenue is not changed by formulating the reasonable service fee price, and thus the influence of the electric vehicle load on the electric energy quality of the power grid can be reduced.

[0013] Summary of the Invention

[0014] An object of the present invention is to realize a system developed for providing artificial intelligence-based price recommendations specifically designed for electric vehicle charging stations. Another object of the present invention is to realize a system developed for maximizing the charging station revenue by using a combination of advanced deep learning algorithms and classic statistical methods in order to determine the optimal electric vehicle charging price based on various internal and external factors.

[0015] Detailed Description of the Invention

[0016] “A System for Providing Price Optimization for Electric Vehicle Charging Stations” realized to fulfd the objectives of the present invention is shown in the figure attached, in which:

[0017] Figure 1 is a schematic view of the inventive system.

[0018] The components illustrated in the figures are individually numbered, where the numbers refer to the following:

[0019] 1. System

[0020] 2. Database

[0021] 3. Server

[0022] E. Electric Vehicle Charging Station

[0023] The inventive system (1) developed for providing artificial intelligence-based price optimization for electric vehicle charging stations (E) comprises; at least one database (2) which is configured to keep a record of weather information, traffic information, location information, calendar information, historical data, price flexibility information and competitor pricing information therein; at least one server (3) which is configured to establish communication with the database (2) by using any communication protocol; to predict the busiest and lowest usage times of electric vehicle charging stations (E) and how much energy will be required during which time period by running deep learning algorithms on the data it retrieves from the database (2); to optimize pricing strategies and to determine the most optimal pricing model by taking into account the demand predictions provided by deep learning algorithms through a mixed integer linear programming model.

[0024] The database (2) included in the inventive system (1) is configured to establish connection with the server (3). The database (2) is configured to keep a record of weather information, traffic information, location information, calendar information, historical data, price flexibility information and competitor pricing information therein.

[0025] The server (3) included in the inventive system (1) is configured to establish communication with the database (2) by using any communication protocol included in the state of the art. The server (3) is configured to access data on the database (2) and to store data on the database (2). The server (3) is configured to retrieve weather information, traffic information, location information, calendar information and historical data from the database (2) and to run deep learning algorithms on these data. The server (3) is configured to predict the busiest and lowest usage times of electric vehicle charging stations (E) and how much energy will be required during which time period by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data. The server (3) is configured to combine demand data of multiple stations, to learn similar patterns between these stations and to make more accurate demand predictions for each station by taking into account the simultaneous data patterns of multiple stations and the included external source data by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data. The server (3) included in the inventive system (1) is configured to retrieve price flexibility information and competitor pricing information from the database (2) and to apply a mixed integer linear programming model on these data. The server (3) is configured to analyze peak demand hours and price flexibility; to perform positioning with respect to the pricing strategies of competing stations; to calculate the minimum cost by means of cost and profit margin analysis; and to determine the best pricing under certain constraints by means of mixed integer linear programming optimization techniques by retrieving price flexibility information and competitor pricing information from the database (2). The server (3) is configured to optimize pricing strategies and to determine the most optimal pricing model by taking into account the demand predictions provided by deep learning algorithms through a mixed integer linear programming model.

[0026] The server (3) included in the inventive system (1) is configured to provide model-independent explanations through the ANCHOR method and to highlight influential points in the input data caused by a particular prediction. The server (3) is configured to attribute the output of the SHAP model to each input feature and to interpret feature effects. The server (3) is configured to enable complex analytics and calculations to be converted into actionable information for the management.

[0027] Industrial Application of the Invention

[0028] In the inventive system, the server (3) retrieves weather information, traffic information, location information, calendar information and historical data from the database (2) and runs deep learning algorithms on these data. The server (3) predicts the busiest and lowest usage times of electric vehicle charging stations (E) and how much energy will be required during which time period by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data. The server (3) combines demand data of multiple stations, learns similar patterns between these stations and makes more accurate demand predictions for each station by taking into account the simultaneous data patterns of multiple stations and the included external source data by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data. The server (3) retrieves price flexibility information and competitor pricing information from the database (2) and applies a mixed integer linear programming model on these data. The server (3) analyzes peak demand hours and price flexibility; performs positioning with respect to the pricing strategies of competing stations; calculates the minimum cost by means of cost and profit margin analysis; and determines the best pricing under certain constraints by means of mixed integer linear programming optimization techniques by retrieving price flexibility information and competitor pricing information from the database (2). The server (3) optimizes pricing strategies and determines the most optimal pricing model by taking into account the demand predictions provided by deep learning algorithms through a mixed integer linear programming model. In this way, it is enabled to find the price range that will maximize the station revenue on a daily basis and meet customer demand by predicting the charging demands and energy consumption of electric vehicles and generating decision variables and criteria.

[0029] Within these basic concepts; it is possible to develop various embodiments of the inventive “A System (1) for Providing Price Optimization for Electric Vehicle Charging Stations”; the invention cannot be limited to examples disclosed herein and it is essentially according to claims.

Claims

CLAIMS1. A system (1) developed for providing artificial intelligence-based price optimization for electric vehicle charging stations (E) comprising at least one database (2) which is configured to keep a record of weather information, traffic information, location information, calendar information, historical data, price flexibility information and competitor pricing information therein; and characterized by at least one server (3) which is configured to establish communication with the database (2) by using any communication protocol; to predict the busiest and lowest usage times of electric vehicle charging stations (E) and how much energy will be required during which time period by running deep learning algorithms on the data it retrieves from the database (2); to optimize pricing strategies and to determine the most optimal pricing model by taking into account the demand predictions provided by deep learning algorithms through a mixed integer linear programming model.

2. A system (1) according to Claim 1; characterized by the database (2) which is configured to establish connection with the server (3).

3. A system (1) according to Claim 1 or 2; characterized by the database (2) which is configured to keep a record of weather information, traffic information, location information, calendar information, historical data, price flexibility information and competitor pricing information therein.

4. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to establish communication with the database (2) by using any communication protocol.

5. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to access data on the database (2) and to store data on the database (2).

6. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to retrieve weather information, traffic information, location information, calendar information and historical data from the database (2) and to run deep learning algorithms on these data.

7. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to predict the busiest and lowest usage times of electric vehicle charging stations (E) and how much energy will be required during which time period by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data.

8. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to combine demand data of multiple stations, to learn similar patterns between these stations and to make more accurate demand predictions for each station by taking into account the simultaneous data patterns of multiple stations and the included external source data by retrieving weather information, traffic information, location information, calendar information and historical data from the database (2) and running deep learning algorithms on these data.

9. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to retrieve price flexibility information and competitor pricing information from the database (2) and to apply a mixed integer linear programming model on these data.

10. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to analyze peak demand hours and price flexibility; to perform positioning with respect to the pricing strategies of competing stations; to calculate the minimum cost by means of cost and profit margin analysis; and to determine the best pricing under certain constraints by means of mixed integer linear programming optimization techniques by retrieving price flexibility information and competitor pricing information from the database (2).

11. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to optimize pricing strategies and to determine the most optimal pricing model by taking into account the demand predictions provided by deep learning algorithms through a mixed integer linear programming model.

12. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to provide model-independent explanations through the ANCHOR method and to highlight influential points in the input data caused by a particular prediction.

13. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to attribute the output of the SHAP model to each input feature and to interpret feature effects.

14. A system (1) according to any one of the preceding claims; characterized by the server (3) which is configured to enable complex analytics and calculations to be converted into actionable information for the management.

Images