CN107316097A - Method and system for predicting charging demand of electric vehicle - Google Patents

Abstract

The invention relates to a method for predicting the charging demand of an electric automobile, which comprises the following steps: respectively determining the battery power of each electric automobile in the monitoring area, the estimated parking time at the current position and the probability of charging at the current time according to historical data; and respectively calculating the charging demand probability of each electric automobile, and sequencing the charging demand probability of each electric automobile to form a list of potential charging automobiles. The method can obviously shorten the response time of the user charging request and relieve the peak pressure of the system, thereby providing good user experience.

Description

Electric Vehicle Charging Demand Prediction Method and System

technical field

The present invention relates to the technical field of electric vehicles, and more specifically, to a method and system for predicting charging demand of electric vehicles.

Background technique

Electric vehicles have been gradually popularized, and charging services suitable for electric vehicles are the focus of attention of those skilled in the art.

In the existing charging services, after the user places an order (request for charging), the system passively dispatches service personnel or service vehicles to provide the charging service. There are some problems in this way:

1. The response time of all services cannot be guaranteed. When the service personnel are far away from the user's vehicle, the response time of the service will become longer.

2. There is peak pressure on system scheduling. In some time periods, orders may be relatively concentrated, which will affect the overall operating efficiency of the system;

3. The same user may place orders in different time periods but receive services with different standards (for example, significant differences in response time), thereby affecting user experience.

In addition, for the service mode of mobile charging car, if the driver (service personnel) does not know where the next service order is, he will not know where to drive the charging car and where to park it, but can only passively wait for the system to dispatch , so that the charging resources cannot be fully utilized.

Therefore, in view of the above defects, those skilled in the art need a solution in which the system actively predicts the user's potential charging demand, so as to respond to the user's charging request in advance.

Contents of the invention

A technical purpose of the present invention is to provide a method for predicting the charging demand of an electric vehicle, so as to respond to the user's charging request in advance.

To achieve the above object, the present invention provides a technical solution as follows:

A method for predicting the charging demand of electric vehicles, comprising the following steps: a) For each electric vehicle in the monitoring area, respectively determine its battery power, the estimated parking time at the current location, and determine its charging time at the current time according to historical data probability; b), respectively calculate the charging demand probability of each electric vehicle, and sort the charging demand probability of each electric vehicle to form a list of potential charging vehicles; wherein, the charging demand probability is that the charging demand probability of the electric vehicle is determined when at least the first condition is satisfied. The conditional probability of charging.

Preferably, the method further includes: c), estimating the expected service time required for each potential charging vehicle in the list to be charged at a corresponding service point; d), communicating with each potential charging vehicle on whether to implement the charging service.

Preferably, in step b), the charging demand probability is the conditional probability that the electric vehicle is charged when both the first condition and the second condition are satisfied, wherein the first condition is that the current battery power of the electric vehicle is lower than the first threshold, The second condition is that the estimated parking time of the electric vehicle at the current location exceeds a second threshold.

Preferably, the charging demand probability is calculated according to Bayes' law using the following formula: P(D)=P(C)*P(S|D)*P(T|D), where P(C) represents The probability that an electric vehicle is charging at the corresponding current time, P(S|D) represents the probability that the corresponding current battery power of the electric vehicle is lower than the first threshold when the electric vehicle is charging in the historical data, and P(T|D) represents The probability that the corresponding parking time of the electric vehicle exceeds the second threshold when the electric vehicle is being charged in the historical data.

Preferably, the first threshold is the average value of the corresponding battery power of each electric vehicle when charging, and the second threshold is the average value of the corresponding parking time of each electric vehicle when charging.

Preferably, in step d), if the estimated service time meets the service standard, the user is actively reminded to charge; if the estimated service time cannot meet the service standard, the user is informed of the estimated service time and waits for the user's response within the first time window. instructions, and based on the instructions, dispatching of service personnel or service vehicles to respond to the charging request of the indicated potential charging vehicle.

Preferably, in step d), according to the ranking of the charging demand probabilities of the potential charging vehicles, the potential charging vehicles with high charging demand probabilities are prioritized and actively communicated with.

Preferably, step e) is also included: if the potential charging vehicles do not accept the charging service through communication, or the potential charging vehicles have completed the charging service, adding several charging vehicles with the highest charging demand probability in the monitoring area to the list of potential charging vehicles , and return to step c) to continue execution.

The invention also discloses a system for predicting the charging demand of electric vehicles, including: an area monitoring module, for each electric vehicle in the monitoring area, respectively determine its battery power, the estimated parking time at the current location, and determine its current time according to historical data. The probability of charging; the probability calculation module is used to separately calculate the charging demand probability of each electric vehicle, and sort the charging demand probability of each electric vehicle to form a list of potential charging vehicles.

The electric vehicle charging demand prediction method provided by each embodiment of the present invention actively predicts the charging demand probability of each electric vehicle in the monitoring area, and then communicates with each potential charging vehicle, and dispatches services in advance upon user request or even without waiting for user request. This scheduling method can significantly shorten the response time of user charging requests, relieve the peak pressure of the system, and provide a good user experience.

In addition, the electric vehicle charging demand prediction system can actively predict the potential charging demand of electric vehicles (users), and then respond to user charging requests in advance. The system has higher working efficiency and can effectively avoid peak pressure, and is suitable for promotion in large and medium-sized cities .

Description of drawings

Fig. 1 shows a schematic flowchart of a method for predicting charging demand of an electric vehicle provided by a first embodiment of the present invention.

Fig. 2 shows a schematic diagram of the module structure of the electric vehicle charging demand prediction system provided by the second embodiment of the present invention.

detailed description

As shown in FIG. 1 , a first embodiment of the present invention provides a method for predicting charging demand of an electric vehicle, which includes the following steps.

Step S10, respectively determine the battery power, estimated parking time and probability of charging at the current time of each electric vehicle in the monitoring area.

Specifically, firstly, all registered electric vehicles driving or parked in the monitoring area are monitored, and the battery power and estimated parking time of each electric vehicle are obtained; then, according to historical data, it is determined that each electric vehicle is charged at the current time The probability.

Among them, the battery power can be characterized by SOC parameters, and the estimated parking time can be obtained through statistical analysis based on historical data, or can be provided by the user. For any electric vehicle, historical data records, for example, information such as the parking time of the electric vehicle at each service point in the monitoring area on a certain day in history, whether it is charged or not. Based on historical data, it is possible to calculate the probability that the electric vehicle will be charged at the current time.

Step S12, calculating the charging demand probability of each electric vehicle, and sorting to form a list of potential charging vehicles.

Specifically, in this step S12, the charging demand probabilities of the electric vehicles in the monitoring area are respectively calculated, and the charging demand probabilities of the electric vehicles are sorted in descending order to form a list of potential charging vehicles.

Wherein, the charging demand probability is a conditional probability, for example, it is the conditional probability that the electric vehicle is charged when at least the first condition is satisfied. The first condition may be expressed as that the current battery power of the electric vehicle is lower than the first threshold.

Preferably, the charging demand probability is a conditional probability that the electric vehicle is charged when both the first condition and the second condition are satisfied. Wherein, the first condition may be expressed that the current battery power of the electric vehicle is lower than the first threshold, and the second condition may be expressed that the estimated parking time of the electric vehicle at the current location exceeds the second threshold.

As a conditional probability, the charging demand probability can be calculated according to Bayesian theorem. The calculation formula is P(D)=P(C)*P(S|D)*P(T|D), where P( C) indicates the probability that an electric vehicle in the historical data is charging at the corresponding current time, and P(S|D) indicates that the corresponding current battery power of the electric vehicle in the historical data is lower than the first threshold when charging (that is, satisfying The probability of the first condition), P(T|D) represents the probability that the corresponding parking time of the electric vehicle exceeds the second threshold (that is, meets the second condition) when the electric vehicle is charging in the historical data.

The following is just an example: assume that the current SoC of the electric vehicle is S _t , and the estimated parking time is T _e , and S _a is the average value of the SoC of the corresponding battery power of each electric vehicle during charging obtained from the historical charging service records, T _a is the average value of the corresponding parking time of each electric vehicle during charging obtained from the historical charging service records, then the charging demand probability can be calculated as follows:

1) If S _t <= S _a and T _e <= T _a , then P(D)=P(C)*P(S|D)*(1-P(T|D));

2) If S _t <= S _a and T _e >T _a , then P(D)=P(C)*P(S|D)*P(T|D);

3) If S _t >S _a and T _e <=T _a , then P(D)=P(C)*(1-P(S|D))*(1-P(T|D));

4) If S _t >S _a and T _e >T _a , then P(D)=P(C)*(1-P(S|D))*P(T|D).

After obtaining the charging demand probability of each electric vehicle, or sort them in descending order, and use the N electric vehicles with the highest charging demand probability to form a list of potential charging vehicles, or, use the charging demand probability greater than a certain threshold to form a list of potential charging vehicles. N electric vehicles form a list of potential charging vehicles.

As an improvement to the above embodiment, the method for predicting the charging demand of electric vehicles may further include the following steps:

Step S14, estimating the expected service time required for each potential charging vehicle in the list to be charged at the corresponding service point.

Specifically, based on the list of potential charging vehicles generated in step S12, in step S14, the estimated service time required for each potential charging vehicle to be charged at the corresponding service point is estimated.

Wherein, the corresponding service point may be the current location of each potential charging vehicle, or the nearest service point location, or the service point location assigned by the system for its scheduling, or other suitable locations suitable for implementing charging services. The estimated service time may be the time required for the potential charging vehicle to be fully charged, or the time required for charging to the user's expected power, or the charging time specified by the user.

As an example, the estimated service time may represent a time period for a service person or service vehicle to respond to a charging request of a potential charging vehicle until the potential charging vehicle completes charging and returns to the user.

Step S16 , communicating with the potential charging vehicle on whether to implement the charging service.

Any electric vehicle in the list of potential charging vehicles may choose to require charging or not to charge. In view of this, based on the estimated service time obtained in step S14, in step S16, the system and the potential charging vehicle will implement the charging service. Communication, if the user requests charging service, the system will dispatch service personnel or service vehicles to a service point suitable for charging service, and can also instruct the potential charging car to go to the service point.

In the process of communication, according to the current power of the charging car and the estimated parking time, the system can also propose different levels of recommendations to the user. Under appropriate circumstances (for example, the battery of the charging car is in an emergency), the system will dispatch service personnel or service vehicles to a service point suitable for charging service while communicating, so as to respond to the user's future charging request in advance.

As another improved implementation, in the communication process, if the estimated service time from step S14 meets a certain service standard (for example, the first time required for the service personnel to respond to the charging request of the potential charging car is less than 15 minutes), then Inform the user to wait for charging (the service point can also be assigned to the electric vehicle to be charged at the same time); if the estimated service time cannot meet the service standard (that is, the above-mentioned first time exceeds 15 minutes), the estimated service time (or the above-mentioned first time ) to inform the user, and wait for the user's instruction within the first time window (for example, 5 minutes), and dispatch service personnel or service vehicles based on the instruction to respond to the charging request of the potential charging car that issued the instruction.

As a preferred implementation, according to the ranking of the charging demand probability of each potential charging vehicle, priority and actively communicate with the potential charging vehicle with a high charging demand probability, in order to predict and respond to the user's charging request as soon as possible, so as to provide a good user experience .

As a further improvement to the above-mentioned first embodiment, the method for predicting the charging demand of an electric vehicle further includes performing the following steps after step S16: if the potential charging vehicle does not accept the charging service through communication, or the potential charging vehicle has completed the charging service, then A number of charging vehicles with the highest charging demand probability in the monitoring area are added to the list of potential charging vehicles, and return to step S14 to continue execution.

The electric vehicle charging demand prediction method provided by the first embodiment above does not schedule services according to the user's charging order, but actively predicts the charging demand probability of each electric vehicle in the monitoring area to form a list of potential charging vehicles, and then communicates with each Potential charging vehicles communicate and schedule services in advance upon user requests or even without waiting for user requests. This scheduling method can significantly shorten the response time of user charging requests, relieve the peak pressure of the system, and provide a good user experience.

It can be foreseen that the charging demand probability calculated according to the above-mentioned first embodiment takes into account factors such as the user's historical selection, long-term habits, and the current battery capacity of the electric vehicle, so that this conditional probability is more suitable for the user's actual choice of whether to request charging . The list of potential charging vehicles generated based on the charging demand probability of each electric vehicle is more suitable for representing a series of electric vehicles that are currently expected to be charged. Efficiency of charging service system.

The second embodiment of the present invention provides a charging demand prediction system for electric vehicles, which includes an area monitoring module 201 , a probability calculation module 203 , an optional time prediction module 205 and an optional user communication module 207 .

Among them, the area monitoring module 201 respectively determines the battery power, the estimated parking time at the current location, and further determines the probability of charging at the current time for each electric vehicle in the monitoring area according to historical data.

The probability calculation module 203 is coupled with the area detection module 201, which respectively calculates the charging demand probability of each electric vehicle, and sorts the charging demand probability of each electric vehicle to form a list of potential charging vehicles. Wherein, the charging demand probability is the conditional probability that the electric vehicle will be charged when at least the first condition is satisfied, and the first condition may be expressed that the current battery power of the electric vehicle is lower than the first threshold.

As an improved implementation, the probability calculation module 203 calculates the conditional probability that the electric vehicle is charged when both the first condition and the second condition are satisfied, and takes the conditional probability as the charging demand probability of the electric vehicle; The current battery power of the car is lower than the first threshold, and the second condition is that the estimated parking time of the electric car at the current location exceeds the second threshold.

The time prediction module 205 and the user communication module 207 can be combined into the above-mentioned embodiments as additional modules.

The time estimation module 205 is coupled with the probability calculation module 203, and is used for estimating the estimated service time required for each potential charging vehicle in the list to be charged at the corresponding service point.

The user communication module 207 is coupled with the time estimation module, and communicates with each potential charging vehicle (user) on whether to implement the charging service.

The electric vehicle charging demand forecasting system provided by the above second embodiment actively predicts the potential charging demand of the electric vehicle (user), and then can respond to the user's charging request in advance to provide a good user experience. It is based on this prediction (prejudgment) that the system works more efficiently, can effectively avoid peak pressure, and is suitable for promotion in large and medium-sized cities.

It should be noted that the above method for predicting charging demand can be implemented with a computer program, therefore, it can be expected that a computer system and a computer-readable storage medium related to such a computer program should be included in the embodiments of the present invention.

As an example, the present invention also provides a computer system, which includes a memory, a processor, and a computer program stored on the memory and executed by the processor, wherein the following steps are implemented when the processor executes the computer program:

S21. For each electric vehicle in the monitoring area, determine its battery power, estimated parking time at the current location, and determine the probability of charging at the current time according to historical data.

S23. Calculate the charging demand probabilities of the electric vehicles respectively, and sort the charging demand probabilities of the electric vehicles to form a list of potential charging vehicles. Wherein, the charging demand probability is a conditional probability that the electric vehicle is charged when at least the first condition is met.

As a further improvement, in the above computer system, the following steps can be further implemented when the processor executes the computer program:

S25. Estimate the estimated service time required for each potential charging vehicle in the obtained list to be charged at the corresponding service point.

S27. Communicate separately with each potential charging vehicle on whether to implement the charging service.

In the above example, the charging demand probability is the conditional probability that the electric vehicle will be charged when both the first condition and the second condition are satisfied, wherein the first condition is that the current battery power of the electric vehicle is lower than the first threshold, and the second condition is that the electric vehicle The estimated parking time of the car at the current location exceeds a second threshold.

Preferably, the charging demand probability is calculated according to Bayes' law using the following formula: P(D)=P(C)*P(S|D)*P(T|D), where P(C) represents historical data The probability that an electric vehicle is charging at the corresponding current time, P(S|D) represents the probability that the corresponding current battery power of the electric vehicle is lower than the first threshold when the electric vehicle is charging in the historical data, P(T|D) Indicates the probability that the corresponding parking time of the electric vehicle exceeds the second threshold when charging in the historical data.

Further, various improvements of the method for predicting the charging demand of an electric vehicle provided in the first embodiment above may also be specifically implemented by a processor when executing a computer program on a memory.

As another example, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the following steps:

S31. For each electric vehicle in the monitoring area, determine its battery power, estimated parking time at the current location, and determine the probability of charging at the current time according to historical data.

S33. Calculate the charging demand probabilities of the electric vehicles respectively, and sort the charging demand probabilities of the electric vehicles to form a list of potential charging vehicles. Wherein, the charging demand probability is a conditional probability that the electric vehicle is charged when at least the first condition is satisfied.

As a further improvement, the computer program stored on the readable storage medium can further implement the following steps when being executed:

S35 , estimating the estimated service time required for each potential charging vehicle in the obtained list to be charged at the corresponding service point.

S37. Communicate separately with each potential charging vehicle on whether to implement the charging service.

In the above example, similarly, the charging demand probability is the conditional probability that the electric vehicle is charged when both the first condition and the second condition are satisfied, wherein the first condition is that the current battery power of the electric vehicle is lower than the first threshold, and the second The condition is that the estimated parking time of the electric vehicle at the current location exceeds the second threshold.

Preferably, the charging demand probability is calculated according to Bayes' law using the following formula: P(D)=P(C)*P(S|D)*P(T|D), where P(C) represents historical data The probability that an electric vehicle is charging at the corresponding current time, P(S|D) represents the probability that the corresponding current battery power of the electric vehicle is lower than the first threshold when the electric vehicle is charging in the historical data, P(T|D) Indicates the probability that the corresponding parking time of the electric vehicle exceeds the second threshold when charging in the historical data.

Further, various improvement solutions of the method for predicting the charging demand of electric vehicles provided by the above first embodiment may also be specifically implemented by the processor when executing the computer program on the storage medium.

The above description is only aimed at preferred embodiments of the present invention, and is not intended to limit the scope of protection of the present invention. Those skilled in the art can make various deformation designs without departing from the idea of the present invention and the appended claims.

Claims (14)

1. A method for forecasting electric vehicle charging demand, comprising the steps of: a) For each electric vehicle in the monitoring area, determine its battery power, estimated parking time at the current location, and determine the probability of charging at the current time based on historical data; and b) respectively calculating the charging demand probability of each of the electric vehicles, and sorting the charging demand probabilities of each of the electric vehicles to form a list of potential charging vehicles; wherein, the charging demand probability is that the electric vehicle is at least The conditional probability of charging when the first condition is met. 2. The method according to claim 1, characterized in that the method further comprises: c), estimating the estimated service time required for each of the potential charging vehicles in the list to be charged at the corresponding service point; and d) Communicate with each of the potential charging vehicles on whether to implement the charging service. 3. The method according to claim 1, characterized in that, in the step b), the charging demand probability is a condition for the electric vehicle to be charged when the first condition and the second condition are both satisfied probability, wherein the first condition is that the current battery power of the electric vehicle is lower than a first threshold, and the second condition is that the estimated parking time of the electric vehicle at the current location exceeds a second threshold. 4. The method according to claim 3, wherein the charging demand probability is calculated according to Bayes' law using the following formula: P(D)=P(C)*P(S|D)*P(T|D), wherein, P(C) represents the probability that a certain electric vehicle in the historical data is charged at the corresponding current time, P (S|D) represents the probability that the corresponding current battery power of the electric vehicle in the historical data is lower than the first threshold when charging, and P(T|D) represents the corresponding probability of the electric vehicle in the historical data when charging. The probability that the parking time exceeds the second threshold. 5. The method according to claim 3, wherein the first threshold is the average value of the corresponding battery power when each electric vehicle is charged, and the second threshold is the corresponding parking time of each electric vehicle when charging average value. 6. The method according to claim 2, wherein in the step c), the estimated service time means that service personnel or service vehicles respond to the charging request of the potential charging vehicle until the charging of the potential charging vehicle is completed Time to charge and return to user. 7. The method according to claim 2, wherein in the step d), if the estimated service time meets the service standard, the user is actively reminded to charge; if the estimated service time cannot meet the service standard standard, inform the user of the expected service time, wait for the user's instruction within the first time window, and dispatch the service personnel or service vehicle based on the instruction to respond to the potential charging car that issued the instruction. Charging request. 8. The method according to claim 2, characterized in that, in the step d), according to the ranking of the charging demand probabilities of the potential charging vehicles, preferentially and proactively contact those with high charging demand probabilities. The potentially charging car communicates. 9. The method according to claim 2, further comprising step e): If the potential charging vehicle does not accept the charging service through communication, or the potential charging vehicle has completed the charging service, then add several charging vehicles with the highest charging demand probability in the monitoring area to the potential charging vehicle list, and return to step c) to continue execution. 10. An electric vehicle charging demand prediction system, comprising: The area monitoring module, for each electric vehicle in the monitoring area, respectively determines its battery power, the estimated parking time at the current location, and determines the probability of charging at the current time based on historical data; and A probability calculation module, configured to separately calculate the charging demand probability of each of the electric vehicles, and sort the charging demand probabilities of each of the electric vehicles to form a list of potential charging vehicles; wherein, the charging demand probability is the electric vehicle The conditional probability that the car will charge when at least the first condition is met. 11. The system according to claim 10, further comprising: a time estimation module, configured to estimate the estimated service time required for each of the potential charging vehicles in the list to be charged at the corresponding service point; and The user communication module communicates with each of the potential charging vehicles on whether to implement the charging service. 12. The system according to claim 10, wherein the probability calculation module calculates the conditional probability that the electric vehicle is charged when the first condition and the second condition are satisfied as the charging condition of the electric vehicle. Demand probability, wherein the first condition is that the current battery power of the electric vehicle is lower than a first threshold, and the second condition is that the estimated parking time of the electric vehicle at the current location exceeds a second threshold. 13. A computer system, comprising a memory, a processor, and a computer program stored on the memory and run by the processor, characterized in that, when the processor executes the computer program, claims 1 to 9 The method described in any one is carried out. 14. A computer-readable storage medium on which a computer program is stored, wherein when the computer program is executed by a processor, the method according to any one of claims 1 to 9 is performed.