Electric vehicle charging demand prediction method based on markov dynamic path decision

Abstract

A kind of electric vehicle charging demand prediction method based on Markov dynamic path decision, S1, collect the map of target area, vehicle trip origin-destination matrix, daily temperature, traffic congestion index and electric vehicle parameters;S2, analyze electric vehicle parameters, divide electric vehicle into family electric vehicle and rental electric vehicle, build trip chain model for family electric vehicle, obtain rental electric vehicle trip data;S3, build Markov dynamic path optimization model, obtain time-series path and energy consumption change data set;S4, build charging demand determination model, obtain electric vehicle charging demand;S5, charging load superposition and node load calculation, superimpose the charging demand of family electric vehicle and taxi on the corresponding road network node, generate the final electric vehicle charging demand space-time distribution map.The design not only improves the flexibility of path planning, but also effectively improves the prediction accuracy of traffic congestion.

Description

Technical Field

[0001] This invention relates to a method for predicting electric vehicle charging demand based on Markov dynamic path decision-making, specifically applicable to improving the accuracy of charging demand prediction, thereby enabling the rational planning of charging stations. Background Technology

[0002] With increasing awareness of environmental protection and energy conservation, electric vehicles (EVs) are gradually becoming the mainstream direction of future transportation. Compared with traditional fuel vehicles, EVs have advantages such as zero emissions, low noise, and high energy efficiency. Many countries and regions have promoted the development of the EV industry through policy support and subsidies. However, the popularization of EVs has also brought new problems, one of which is the management of charging demand, especially how to efficiently and conveniently find charging stations and rationally plan charging routes during driving. Solving this problem is crucial to improving the user experience of EVs and promoting their further development. Existing EV charging demand forecasting methods usually assume fixed routes and road network conditions, making it difficult to adapt to the dynamic route selection needs arising from changes in traffic conditions during actual driving. Traditional models often do not consider real-time traffic congestion when planning charging demand, resulting in inflexible route selection and potentially increasing travel time and energy consumption. Many methods fail to consider the psychological factors of drivers when choosing routes, ignoring the behavior of drivers choosing suboptimal routes in congested conditions. Therefore, by introducing a traffic congestion index and a route penalty coefficient, this paper incorporates the driver's tolerance for congested road sections and the psychological factors in route selection into the model, improving the problem of ignoring the impact of traffic congestion in traditional methods. By utilizing Markov decision processes and combining them with real-time traffic information, route selection can be dynamically adjusted according to changes in traffic conditions, thereby improving the flexibility of route planning and enabling it to better adapt to real-time changes in traffic conditions and rationally plan charging routes. Summary of the Invention

[0003] The purpose of this invention is to overcome the problem of unreasonable planning of electric vehicle charging stations in the prior art, and to provide a method for predicting electric vehicle charging demand based on Markov dynamic path decision that can reasonably plan charging stations.

[0004] To achieve the above objectives, the technical solution of the present invention is:

[0005] In a first aspect, the present invention provides a method for predicting electric vehicle charging demand based on Markov dynamic path decision-making, comprising the following steps:

[0006] S1. Collect maps of the target area, a matrix of vehicle origin and destination points, daily temperature (T), and traffic congestion index. And electric vehicle parameters;

[0007] S2. Construct a road-traffic node map of the target area based on the target area map; analyze electric vehicle parameters, classify electric vehicles into family electric vehicles and taxi electric vehicles, construct a travel chain model for family electric vehicles, and calculate travel time parameters to obtain family electric vehicle travel data; for taxi electric vehicles, sample departure point, destination, start time, and parking duration from the vehicle travel origin-destination matrix to obtain taxi electric vehicle travel data.

[0008] S3. Construct a Markov dynamic path optimization model. Input the travel data of family electric vehicles and rental electric vehicles into the Markov dynamic path optimization model, select the optimal path for each electric vehicle, and calculate the energy consumption change during the trip to obtain a time-series dataset of path and energy consumption change.

[0009] S4. Construct a charging demand determination model. For users of family electric vehicles, classify users into demand types to obtain the charging demand of family electric vehicles. For users of rental electric vehicles, construct a charging demand selection model to obtain the charging demand of rental electric vehicles.

[0010] S5. Charging load superposition and node load calculation: The charging demand of household electric vehicles and taxis is superimposed on the corresponding road network nodes to calculate the total node charging load, output the spatiotemporal distribution of charging demand, and generate the final spatiotemporal distribution map of electric vehicle charging demand based on the charging load data of all nodes.

[0011] Construct a road-traffic node map of the target area based on the target area map, extract the road map of the target area from the target area map, determine multiple traffic nodes according to the distribution of the road map, and obtain the road-traffic node map of the target area.

[0012] A travel chain model was constructed for family electric vehicles, and travel time parameters were calculated to obtain travel data for family electric vehicles.

[0013] The travel chain, which contains a variety of user information, is described using spatial and temporal features. Spatial features include the vehicle's origin and destination, stops along the way, and travel route information. Temporal features include the vehicle's departure time, destination stay time, travel time, and trip end time information. The city is divided into four areas: residential area, work area, entertainment area, and other areas, which are used to represent the coordinate partitioning of the spatial features.

[0014] Methods for calculating time characteristic quantities:

[0015] Vehicle departure time: The first departure time of a vehicle within a day follows a normal distribution, with the probability density function as follows:

[0016]

[0017] Where, μ T σ represents the average time since the first departure from home. T Let x be the variance of the time of first departure from home;

[0018] Dwell time at different destinations: The dwell time of a vehicle varies at different destinations, so it is represented by multiple probability density functions;

[0019] The probability of vehicle dwell time in residential areas follows a Weibull distribution, with the probability density function:

[0020]

[0021] Where k is the shape parameter of the Weibull distribution; λ is the scale parameter of the Weibull distribution;

[0022] The probability of vehicle dwell time within the work area follows a Type III extreme value distribution, with the probability density function:

[0023]

[0024] The probability of vehicle dwell time in entertainment areas and other areas follows a Type II extreme value distribution, with the probability density function:

[0025]

[0026] Where μ is the location parameter of the generalized extreme value distribution; δ is the scale parameter of the generalized extreme value distribution, δ>0; ε is the shape parameter of the generalized extreme value distribution, when ε>0 a=1 / ε, when ε<0 a=-1 / ε;

[0027] The start time of the next leg of the journey: After the electric vehicle arrives at its destination and ends its stop, the current journey is complete. The start time of the next leg of the journey is:

[0028]

[0029] Where J represents the total number of trips in the travel chain; t j+1 t0 is the time when the vehicle begins its (j+1)th segment of the journey; t0 is the time when the vehicle first leaves home. The total time taken for the vehicle to complete the first j segments of the journey; t stay,j Let be the dwell time of the vehicle at the destination in segment j of the journey; calculate the family electric vehicle travel data using the travel chain model.

[0030] In step S3, a Markov dynamic path optimization model is constructed. Travel data from family electric vehicles and taxi electric vehicles are input into the Markov dynamic path optimization model to select the optimal path for each electric vehicle.

[0031] Introducing the Traffic Congestion Index Itp The road traffic congestion index is used to quantitatively describe the operational status of road traffic. The value range is [0, 100], and the larger the value, the more congested the road is.

[0032]

[0033] Among them, V tp C is the traffic flow on road p at time t; tp The maximum capacity of road p at time t; a penalty coefficient γ is introduced to ensure that the user's decision is not limited to the shortest path:

[0034]

[0035] Where, γ k,t This represents the penalty coefficient for road k at decision time t; It is the traffic congestion index of road k at decision time t; I tp,max It refers to the historical maximum traffic congestion index of the reference road segment; the steps for route decision-making for each vehicle are as follows:

[0036] Pre-departure planning: Before departure, users select their travel destination and plan the shortest route; the pre-planned shortest route is denoted as π. d Each segment of this path is called a reference segment L. d The set of paths from the unselected road network nodes is selected as the candidate path set, denoted as A. k ;

[0037] Reaching the decision node: When the user reaches road node i, they enter the decision state and select the reference road segment L. d Traffic Congestion Index I tp,d Using the penalty coefficient γ as a reference, the alternative path set A k For each road segment k in the given information, determine whether the following conditions are met: like This section of road will then be considered a high-congestion section and removed from the alternative routes.

[0038] Calculating state transition probabilities: When calculating the state transition probability for each road segment k in the remaining alternative paths, the length L is taken into account. k and penalty coefficient γ k,t Impact:

[0039]

[0040] Where P(k|i) represents the probability of transitioning from the current node i to the alternative path k; L k The length of alternative path k; γ k,tThe penalty coefficient of alternative path k at decision time t; the weighting factors of path length α and β and congestion penalty, which adjust the relative impact of the two on the decision;

[0041] Select the next path segment: Based on the state transition probability P(k|i), select the next path segment k with the highest state transition probability among the candidate paths. * Add to the final path π * Update the current node i to k. * Repeat the above steps until the target node is reached.

[0042] S3 calculates the energy consumption changes during the journey;

[0043] Establish an energy consumption model for electric vehicles:

[0044]

[0045] In the formula, ΔE represents the total electricity consumption of an electric vehicle during a single trip; P m Electricity consumption per km for electric vehicles in a transportation network under free-roaming conditions; L p To select the length of path p; P e The additional power consumption per unit time caused by the starting and stopping of electric vehicles due to traffic congestion; t e C is the additional time spent due to traffic congestion. EV The energy consumption model is modified to account for the battery capacity of the electric vehicle and the impact of ambient temperature on the operation of the air conditioning.

[0046]

[0047] In the formula, E represents the total electricity consumption of an electric vehicle during a single trip; For temperature T and road grade v k The energy consumption under temperature conditions is calculated using an electric vehicle energy consumption model to determine the energy consumption changes of an electric vehicle during its journey.

[0048] In step S3, based on the air conditioner usage curve, the start-up probability of the air conditioner is fitted with a normal distribution to obtain the corresponding normal distribution function for air conditioner start-up as follows:

[0049]

[0050] Among them, K open (T) represents the probability that the air conditioner will be turned on at temperature T; μ c and δ c These are the mean and variance of the heating start-up parameters, respectively; μ h and δ h These are the mean and variance of the cooling start-up parameters; when the temperature T is between 5 and 21°C, it is expressed as T0. cWhen the temperature is between 24 and 35°C, it is represented as T. h ;

[0051] Analyzing the air conditioning energy consumption at different temperatures T and its proportion during the trip, the relationship between temperature and the proportion of air conditioning energy consumption is as follows:

[0052] E ac (T) = 4.09 × 10 -6 T 3 +7.28×10 -5 T 2 -6.58×10 -3 T+0.0873

[0053] Determine different temperatures T and road grades v k Temperature energy consumption

[0054]

[0055] in, The energy consumption factors at time t represent different road classes, corresponding to the speeds of different road classes: v1, v2, v3, representing the driving speeds of Class I, II, and III roads, respectively; E p K represents the initial energy consumption per unit distance of an electric vehicle; oc Indicates the on / off status of the air conditioner, K oc =1 indicates that the air conditioner is on, K oc =0 indicates that the air conditioner is off: where K oc The possible values ​​are as follows:

[0056]

[0057] Here, r is a uniformly distributed random number in the interval [0,1], and its purpose is to compare the probability K of the air conditioner being turned on. open (T), and use this comparison to determine whether to turn on the air conditioner.

[0058] In step S4, for users of family electric vehicles, the user demand type is categorized, and the user's charging mode demand result is obtained based on the categorization:

[0059] By collecting and analyzing users' parking time, SOC, and expected electricity price information, users are classified into ordinary rigid users, emergency rigid users, high anchor elastic users, or low anchor elastic users.

[0060] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is greater than or equal to 6 hours, it is judged as a regular rigid user, the charging mode is slow charging, and the charging time is instant.

[0061] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is less than 6 hours, it is judged as an emergency user, the charging mode is fast charging, and the charging time is instant.

[0062] Electric vehicles' State of Charge (SOC) meets the needs of the next trip, and users' expectations of electricity prices are higher than the actual prices, making them users with high anchoring elasticity.

[0063] If the parking time is greater than or equal to 6 hours, the charging mode is slow charging and the charging time is instant.

[0064] If the parking time is less than 6 hours, the charging mode is fast charging and the charging time is instant.

[0065] Electric vehicles whose SOC meets the next trip's needs, and whose expected electricity price is lower than the actual price, are identified as low-anchor-elasticity users: the charging mode is slow charging, and the charging time is when the electricity price is low.

[0066] The charging needs of family electric vehicle users were obtained based on the classification.

[0067] For taxi charging demand: based on the current time t and the actual electricity price P at charging station c. c Calculate charging costs and evaluate them in conjunction with the driver's revenue needs:

[0068]

[0069] R c For the benefit trade-off index; F t Expected total benefits from operation after full charge; E t The energy increment required during charging, i.e., the full charge minus the remaining charge; P c The electricity price for charging stations;

[0070] The current SOC is sufficient to support the next trip, and drivers will then further consider revenue: when R c >R 阈值 Only then will the driver select the corresponding electricity price P. c You can charge at the designated charging station; otherwise, charging will be delayed or you can choose another station.

[0071] If the current SOC is insufficient to support the next trip, determine whether the driver's parking time allows for slow charging: if the parking time is greater than or equal to 6 hours, the driver will prioritize slow charging based on electricity price and charging cost, with charging time being immediate; if the parking time is less than 6 hours, the driver will choose fast charging.

[0072] The charging needs of taxi electric vehicle users were determined based on the above method.

[0073] Secondly, the present invention provides an electric vehicle charging demand prediction system based on Markov dynamic path decision-making, specifically including: a data acquisition module, a travel data construction module, a path optimization module, a charging demand determination module, and a spatiotemporal distribution map generation module;

[0074] Data acquisition module: Used to collect maps of the target area, a matrix of vehicle origin and destination points, daily temperature (T), and traffic congestion index. And electric vehicle parameters;

[0075] The travel data construction module is used to construct a road-traffic node map of the target area based on the target area map; analyze electric vehicle parameters, classify electric vehicles into private electric vehicles and taxi electric vehicles, construct a travel chain model for private electric vehicles, and calculate travel time parameters to obtain travel data for private electric vehicles; for taxi electric vehicles, sample the departure point, destination, start time, and parking duration from the vehicle travel origin-destination matrix to obtain travel data for taxi electric vehicles.

[0076] Route optimization module: Used to build Markov dynamic route optimization model. Input family electric vehicle travel data and taxi electric vehicle travel data into Markov dynamic route optimization model, select the optimal route for each electric vehicle, and calculate the energy consumption change during the trip to obtain time series route and energy consumption change dataset.

[0077] Charging demand determination module: Used to build a charging demand determination model. For users of private electric vehicles, it classifies users into demand types to obtain the charging demand of private electric vehicles; for users of rental electric vehicles, it builds a charging demand selection model to obtain the charging demand of rental electric vehicles.

[0078] Spatiotemporal distribution map generation module: used for charging load superposition and node load calculation, superimposing the charging demand of household electric vehicles and taxis onto the corresponding road network nodes to calculate the total node charging load, output the spatiotemporal distribution of charging demand, and generate the final spatiotemporal distribution map of electric vehicle charging demand based on the charging load data of all nodes.

[0079] In the travel data construction module, a road-traffic node map of the target area is constructed based on the target area map. The road map of the target area is extracted from the target area map, and multiple traffic nodes are determined according to the distribution of the road map to obtain the road-traffic node map of the target area.

[0080] A travel chain model was constructed for family electric vehicles, and travel time parameters were calculated to obtain travel data for family electric vehicles.

[0081] The travel chain, which contains a variety of user information, is described using spatial and temporal features. Spatial features include the vehicle's origin and destination, stops along the way, and travel route information. Temporal features include the vehicle's departure time, destination stay time, travel time, and trip end time information. The city is divided into four areas: residential area, work area, entertainment area, and other areas, which are used to represent the coordinate partitioning of the spatial features.

[0082] Methods for calculating time characteristic quantities:

[0083] Vehicle departure time: The first departure time of a vehicle within a day follows a normal distribution, with the probability density function as follows:

[0084]

[0085] Where, μ T σ represents the average time since the first departure from home. T Let x be the variance of the time of first departure from home;

[0086] Dwell time at different destinations: The dwell time of a vehicle varies at different destinations, so it is represented by multiple probability density functions;

[0087] The probability of vehicle dwell time in residential areas follows a Weibull distribution, with the probability density function:

[0088]

[0089] Where k is the shape parameter of the Weibull distribution; λ is the scale parameter of the Weibull distribution;

[0090] The probability of vehicle dwell time within the work area follows a Type III extreme value distribution, with the probability density function:

[0091]

[0092] The probability of vehicle dwell time in entertainment areas and other areas follows a Type II extreme value distribution, with the probability density function:

[0093]

[0094] Where μ is the location parameter of the generalized extreme value distribution; δ is the scale parameter of the generalized extreme value distribution, δ>0; ε is the shape parameter of the generalized extreme value distribution, when ε>0 a=1 / ε, when ε<0 a=-1 / ε;

[0095] The start time of the next leg of the journey: After the electric vehicle arrives at its destination and ends its stop, the current journey is complete. The start time of the next leg of the journey is:

[0096]

[0097] Where J represents the total number of trips in the travel chain; t j+1 t0 is the time when the vehicle begins its (j+1)th segment of the journey; t0 is the time when the vehicle first leaves home. The total time taken for the vehicle to complete the first j segments of the journey; t stay,j Let be the dwell time of the vehicle at the destination in segment j of the journey; calculate the family electric vehicle travel data using the travel chain model.

[0098] The route optimization module constructs a Markov dynamic route optimization model. Travel data from family electric vehicles and taxi electric vehicles are input into the Markov dynamic route optimization model to select the optimal route for each electric vehicle.

[0099] Introducing the Traffic Congestion Index I tp The road traffic congestion index is used to quantitatively describe the operational status of road traffic. The value range is [0, 100], and the larger the value, the more congested the road is.

[0100]

[0101] Among them, V tp C is the traffic flow on road p at time t; tp The maximum capacity of road p at time t; a penalty coefficient γ is introduced to ensure that the user's decision is not limited to the shortest path:

[0102]

[0103] Where, γ k,t This represents the penalty coefficient for road k at decision time t; It is the traffic congestion index of road k at decision time t; I tp,max It refers to the historical maximum traffic congestion index of the reference road segment; the steps for route decision-making for each vehicle are as follows:

[0104] Pre-departure planning: Before departure, users select their travel destination and plan the shortest route; the pre-planned shortest route is denoted as π. d Each segment of this path is called a reference segment L. d The set of paths from the unselected road network nodes is selected as the candidate path set, denoted as A. k ;

[0105] Reaching the decision node: When the user reaches road node i, they enter the decision state and select the reference road segment L. d Traffic Congestion Index I tp,d Using the penalty coefficient γ as a reference, the alternative path set A k For each road segment k in the given information, determine whether the following conditions are met: like This section of road will then be considered a high-congestion section and removed from the alternative routes.

[0106] Calculating state transition probabilities: When calculating the state transition probability for each road segment k in the remaining alternative paths, the length L is taken into account. k and penalty coefficient γ k,t Impact:

[0107]

[0108] Where P(k|i) represents the probability of transitioning from the current node i to the alternative path k; L k The length of alternative path k; γ k,t The penalty coefficient of alternative path k at decision time t; the weighting factors of path length α and β and congestion penalty, which adjust the relative impact of the two on the decision;

[0109] Select the next path segment: Based on the state transition probability P(k|i), select the next path segment k with the highest state transition probability among the candidate paths. * Add to the final path π * Update the current node i to k. * Repeat the above steps until the target node is reached.

[0110] Calculate the changes in energy consumption during the journey;

[0111] Establish an energy consumption model for electric vehicles:

[0112]

[0113] In the formula, ΔE represents the total electricity consumption of an electric vehicle during a single trip; P m Electricity consumption per km for electric vehicles in a transportation network under free-roaming conditions; L p To select the length of path p; P e The additional power consumption per unit time caused by the starting and stopping of electric vehicles due to traffic congestion; t e C is the extra time spent due to traffic congestion. EV The energy consumption model is modified to account for the battery capacity of the electric vehicle and the impact of ambient temperature on the operation of the air conditioning.

[0114]

[0115] In the formula, E represents the total electricity consumption of an electric vehicle during a single trip; For temperature T and road grade v k The energy consumption under temperature conditions is calculated using an electric vehicle energy consumption model to determine the energy consumption changes of an electric vehicle during its journey.

[0116] Based on the air conditioner usage curve, a normal distribution is fitted to the air conditioner's start-up probability to obtain the corresponding normal distribution function for air conditioner start-up as follows:

[0117]

[0118] Among them, K open (T) represents the probability that the air conditioner will be turned on at temperature T; μ c and δ c These are the mean and variance of the heating start-up parameters, respectively; μ h and δ h These are the mean and variance of the cooling start-up parameters; when the temperature T is between 5 and 21°C, it is expressed as T0. c When the temperature is between 24 and 35°C, it is represented as T. h ;

[0119] Analyzing the air conditioning energy consumption at different temperatures T and its proportion during the trip, the relationship between temperature and the proportion of air conditioning energy consumption is as follows:

[0120] E ac (T) = 4.09 × 10 -6 T 3 +7.28×10 -5 T 2 -6.58×10 -3 T+0.0873

[0121] Determine different temperatures T and road grades v k Temperature energy consumption

[0122]

[0123] in, The energy consumption factors at time t represent different road classes, corresponding to the speeds of different road classes: v1, v2, v3, representing the driving speeds of Class I, II, and III roads, respectively; E p K represents the initial energy consumption per unit distance of an electric vehicle; oc Indicates the on / off status of the air conditioner, K oc =1 indicates that the air conditioner is on, K oc =0 indicates that the air conditioner is off: where K oc The possible values ​​are as follows:

[0124]

[0125] Here, r is a uniformly distributed random number in the interval [0,1], and its purpose is to compare the probability K of the air conditioner being turned on. open (T), and use this comparison to determine whether to turn on the air conditioner.

[0126] In the charging demand determination module, for users of household electric vehicles, users are categorized into demand types, and the charging mode demand results are obtained based on the categorization:

[0127] By collecting and analyzing users' parking time, SOC, and expected electricity price information, users are classified into ordinary rigid users, emergency rigid users, high anchor elastic users, or low anchor elastic users.

[0128] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is greater than or equal to 6 hours, it is judged as a regular rigid user, the charging mode is slow charging, and the charging time is instant.

[0129] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is less than 6 hours, it is judged as an emergency user, the charging mode is fast charging, and the charging time is instant.

[0130] Electric vehicles' State of Charge (SOC) meets the needs of the next trip, and users' expectations of electricity prices are higher than the actual prices, making them users with high anchoring elasticity.

[0131] If the parking time is greater than or equal to 6 hours, the charging mode is slow charging and the charging time is instant.

[0132] If the parking time is less than 6 hours, the charging mode is fast charging and the charging time is instant.

[0133] Electric vehicles whose SOC meets the next trip's needs, and whose expected electricity price is lower than the actual price, are identified as low-anchor-elasticity users: the charging mode is slow charging, and the charging time is when the electricity price is low.

[0134] The charging needs of family electric vehicle users were obtained based on the classification.

[0135] For taxi charging demand: based on the current time t and the actual electricity price P at charging station c. c Calculate charging costs and evaluate them in conjunction with the driver's revenue needs:

[0136]

[0137] R c For the benefit trade-off index; F t Expected total benefits from operation after full charge; E t The energy increment required during charging, i.e., the full charge minus the remaining charge; P c The electricity price for charging stations;

[0138] The current SOC is sufficient to support the next trip, and drivers will then further consider revenue: when R c >R 阈值 Only then will the driver select the corresponding electricity price P. cYou can charge at the designated charging station; otherwise, charging will be delayed or you can choose another station.

[0139] If the current SOC is insufficient to support the next trip, determine whether the driver's parking time allows for slow charging: if the parking time is greater than or equal to 6 hours, the driver will prioritize slow charging based on electricity price and charging cost, with charging time being immediate; if the parking time is less than 6 hours, the driver will choose fast charging.

[0140] The charging needs of taxi electric vehicle users were determined based on the above method.

[0141] Thirdly, the present invention provides an electric vehicle charging demand prediction device based on Markov dynamic path decision, including a memory and a processor, wherein the memory is used to store computer program code and transmit the computer program code to the processor;

[0142] The processor is configured to execute the aforementioned electric vehicle charging demand prediction method based on Markov dynamic path decision according to instructions in the computer program code.

[0143] Fourthly, the present invention provides a computer program product, including a computer program that is executed by a processor to perform the aforementioned electric vehicle charging demand prediction method based on Markov dynamic path decision.

[0144] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0145] 1. This invention addresses the issue of electric vehicle charging demand forecasting based on Markov dynamic path decision-making. Traditional models often assume fixed routes and road network conditions, making it difficult to adapt to the dynamic path selection demands arising from changing traffic conditions during actual driving. Traditional models frequently fail to consider real-time traffic congestion when planning charging demand, leading to inflexible route selection and potentially increased travel time and energy consumption. This invention introduces dynamic path selection, utilizing Markov decision processes and real-time traffic information to dynamically adjust path selection according to changes in traffic conditions, thereby improving the flexibility of path planning. By introducing a traffic congestion index and a path penalty coefficient, the model incorporates drivers' tolerance for congested road sections and psychological factors in path selection, overcoming the problem of neglecting the impact of traffic congestion in traditional methods.

[0146] 2. This invention, a Markov dynamic path decision-making system for electric vehicle charging demand forecasting, demonstrates superior performance in considering path dynamics and driver decision-making psychology. It can better adapt to real-time changes in traffic conditions and rationally plan charging routes. This improvement enhances the accuracy and adaptability of electric vehicle charging demand forecasting, helps alleviate urban traffic congestion, and promotes the more efficient adoption and use of electric vehicles.

[0147] 3. The electric vehicle charging demand forecasting system based on Markov dynamic path decision-making of the present invention includes: a data acquisition module, a travel data construction module, a route optimization module, a charging demand determination module, and a spatiotemporal distribution map generation module. This system is used to implement the steps of the electric vehicle charging demand forecasting method based on Markov dynamic path decision-making provided in any of the above technical solutions. Therefore, this system simultaneously includes all the beneficial effects of the electric vehicle charging demand forecasting method based on Markov dynamic path decision-making provided in any of the above technical solutions, which will not be elaborated further here.

[0148] 4. The electric vehicle charging demand prediction device based on Markov dynamic path decision-making of the present invention includes a processor and a memory. The memory is used to store computer program code and transmit the computer program code to the processor. The processor is used to execute the electric vehicle charging demand prediction method based on Markov dynamic path decision-making provided in any of the above-mentioned technical solutions according to the instructions in the computer program code. Therefore, this device simultaneously includes all the beneficial effects of the electric vehicle charging demand prediction method based on Markov dynamic path decision-making provided in any of the above-mentioned technical solutions, which will not be repeated here.

[0149] 5. The present invention provides a computer program product that, when executed by a processor, implements the steps of the electric vehicle charging demand forecasting method based on Markov dynamic path decision-making provided in any of the above-described technical solutions. Therefore, this computer program product simultaneously includes all the beneficial effects of the electric vehicle charging demand forecasting method based on Markov dynamic path decision-making provided in any of the above-described technical solutions, which will not be elaborated further here. Attached Figure Description

[0150] Figure 1 This is a flowchart of the method of the present invention.

[0151] Figure 2 This is a system diagram of the present invention.

[0152] Figure 3 This is a diagram of the device of the present invention.

[0153] Figure 4 This is a flowchart of the steps in Example 1. Detailed Implementation

[0154] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0155] Example 1:

[0156] See Figure 1 , Figure 4 A method for predicting electric vehicle charging demand based on Markov dynamic path decision-making includes the following steps:

[0157] S1. Collect maps of the target area, a matrix of vehicle origin and destination points, daily temperature (T), and traffic congestion index. And electric vehicle parameters;

[0158] S2. Construct a road-traffic node map of the target area based on the target area map; analyze electric vehicle parameters, classify electric vehicles into family electric vehicles and taxi electric vehicles, construct a travel chain model for family electric vehicles, and calculate travel time parameters to obtain family electric vehicle travel data; for taxi electric vehicles, sample departure point, destination, start time, and parking duration from the vehicle travel origin-destination matrix to obtain taxi electric vehicle travel data.

[0159] In step S2, a road-traffic node map of the target area is constructed based on the target area map. The road map of the target area is extracted from the target area map, and multiple traffic nodes are determined according to the distribution of the road map to obtain the road-traffic node map of the target area.

[0160] A travel chain model was constructed for family electric vehicles, and travel time parameters were calculated to obtain travel data for family electric vehicles.

[0161] The travel chain, which contains a variety of user information, is described using spatial and temporal features. Spatial features include the vehicle's origin and destination, stops along the way, and travel route information. Temporal features include the vehicle's departure time, destination stay time, travel time, and trip end time information. The city is divided into four areas: residential area, work area, entertainment area, and other areas, which are used to represent the coordinate partitioning of the spatial features.

[0162] Methods for calculating time characteristic quantities:

[0163] Vehicle departure time: The first departure time of a vehicle within a day follows a normal distribution, with the probability density function as follows:

[0164]

[0165] Where, μ T The average time to first departure from home is 7.32; σ T Let x be the variance of the first time away from home, taken as 1.34, where x is the first time away from home;

[0166] Dwell time at different destinations: The dwell time of a vehicle varies at different destinations, so it is represented by multiple probability density functions;

[0167] The probability of vehicle dwell time in residential areas follows a Weibull distribution, with the probability density function:

[0168]

[0169] Where k is the shape parameter of the Weibull distribution, which is taken as 14.05; λ is the scale parameter of the Weibull distribution, which is taken as 7.16;

[0170] The probability of vehicle dwell time within the work area follows a Type III extreme value distribution, with the probability density function:

[0171]

[0172] The probability of vehicle dwell time in entertainment areas and other areas follows a Type II extreme value distribution, with the probability density function:

[0173]

[0174] Where μ is the location parameter of the generalized extreme value distribution; δ is the scale parameter of the generalized extreme value distribution, δ>0; ε is the shape parameter of the generalized extreme value distribution, when ε>0 a=1 / ε, when ε<0 a=-1 / ε;

[0175] The start time of the next leg of the journey: After the electric vehicle arrives at its destination and ends its stop, the current journey is complete. The start time of the next leg of the journey is:

[0176]

[0177] Where J represents the total number of trips in the travel chain; t j+1 t0 is the time when the vehicle begins its (j+1)th segment of the journey; t0 is the time when the vehicle first leaves home. The total time taken for the vehicle to complete the first j segments of the journey; t stay,j Let be the dwell time of the vehicle at the destination in segment j of the journey; calculate the family electric vehicle travel data using the travel chain model.

[0178] S3. Construct a Markov dynamic path optimization model. Input the travel data of family electric vehicles and rental electric vehicles into the Markov dynamic path optimization model, select the optimal path for each electric vehicle, and calculate the energy consumption change during the trip to obtain a time-series dataset of path and energy consumption change.

[0179] In step S3, a Markov dynamic path optimization model is constructed. Travel data from family electric vehicles and taxi electric vehicles are input into the Markov dynamic path optimization model to select the optimal path for each electric vehicle.

[0180] Detailed description of travel choices: During travel, drivers primarily consider the distance traveled and the degree of traffic congestion when choosing a route. Therefore, Dijkstra's algorithm is typically used to select the shortest route. Regarding traffic conditions, under the same congestion level, due to the complexity of urban road networks, people may have different experiences under different road network conditions. Therefore, a traffic congestion index I is introduced. tp The road traffic congestion index quantifies the operational status of road traffic. The value range of the road traffic congestion index is [0,100], and the larger the value, the more congested the road is.

[0181]

[0182] Among them, V tp C is the traffic flow (number of vehicles / hour) on road p at time t; tp The maximum capacity of road p at time t (number of vehicles per hour);

[0183] A penalty coefficient γ is introduced to ensure that users' decisions are not limited to the shortest path:

[0184]

[0185] Where, γ k,t This represents the penalty coefficient for road k at decision time t; It is the traffic congestion index of road k at decision time t; I tp,max It refers to the historical maximum traffic congestion index of the reference road segment; the steps for route decision-making for each vehicle are as follows:

[0186] Pre-departure planning: Before departure, users select their travel destination and plan the shortest route; the pre-planned shortest route is denoted as π. d Each segment of this path is called a reference segment L. d The set of paths from the unselected road network nodes is selected as the candidate path set, denoted as A. k ;

[0187] Reaching the decision node: When the user reaches road node i, they enter the decision state and select the reference road segment L. d Traffic Congestion Index I tp,d Using the penalty coefficient γ as a reference, the alternative path set A k For each road segment k in the given information, determine whether the following conditions are met: like This section of road will then be considered a high-congestion section and removed from the alternative routes.

[0188] Calculating state transition probabilities: When calculating the state transition probability for each road segment k in the remaining alternative paths, the length L is taken into account. k and penalty coefficient γk,t Impact:

[0189]

[0190] Where P(k|i) represents the probability of transitioning from the current node i to the alternative path k; L k The length of alternative path k; γ k,t The penalty coefficient of alternative path k at decision time t; the weighting factors of path length α and β and congestion penalty, which adjust the relative impact of the two on the decision;

[0191] Select the next path segment: Based on the state transition probability P(k|i), select the next path segment k with the highest state transition probability among the candidate paths. * Add to the final path π * Update the current node i to k. * Repeat the above steps until the target node is reached.

[0192] S3 calculates the energy consumption changes during the journey;

[0193] Establish an energy consumption model for electric vehicles:

[0194]

[0195] In the formula, ΔE represents the total electricity consumption of an electric vehicle during a single trip; P m The energy consumption per km for electric vehicles under free-roaming conditions in the transportation network is taken as 0.2 kWh / km; L p To select the length of path p, in km; P e The additional power consumption per unit time caused by the starting and stopping of electric vehicles due to traffic congestion; t e C is the additional time spent due to traffic congestion. EV The energy consumption model is modified to account for the battery capacity of the electric vehicle and the impact of ambient temperature on the operation of the air conditioning.

[0196]

[0197] In the formula, E represents the total electricity consumption of an electric vehicle during a single trip; For temperature T and road grade v k The energy consumption under temperature conditions is calculated using an electric vehicle energy consumption model to determine the energy consumption changes of an electric vehicle during its journey.

[0198] In S3,

[0199] Since air conditioning is one of the main energy-consuming devices in electric vehicles, its on / off status and temperature changes directly affect the charging needs of the electric vehicle. The energy consumption of the vehicle's air conditioning is directly drawn from the battery, and its energy consumption varies significantly with temperature and on / off status. Under different temperature conditions, the desired temperature value inside the vehicle will also differ. Based on the air conditioning usage curve, a normal distribution is fitted to the air conditioning start probability to obtain the corresponding normal distribution function for air conditioning start-up as follows:

[0200]

[0201] Among them, K open (T) represents the probability that the air conditioner will be turned on at temperature T; μ c and δ c These are the mean and variance of the heating start-up parameters, μ. c The value is 35.15, δ c The value is 6.13; μ h and δ h These are the mean and variance of the cooling start-up parameters, μ h The value is 4.56, δ h The value is 7.54. Temperatures T between 5 and 21°C are represented as T. c When the temperature is between 24 and 35°C, it is represented as T. h The analysis of air conditioning energy consumption at different temperatures T and its proportion during the trip yields the following relationship between temperature and air conditioning energy consumption:

[0202] E ac (T) = 4.09 × 10 -6 T 3 +7.28×10 -5 T 2 -6.58×10 -3 T+0.0873

[0203] Determine different temperatures T and road grades v k Temperature energy consumption

[0204]

[0205] in, The energy consumption factors at time t represent different road classes, corresponding to different road classes speeds, denoted as v1, v2, and v3 (unit: km / h), representing Class I, Class II, and Class III roads respectively; E p K represents the initial energy consumption per unit distance of an electric vehicle; oc Indicates the on / off status of the air conditioner, K oc =1 indicates that the air conditioner is on, K oc =0 indicates that the air conditioner is off: where K oc The possible values ​​are as follows:

[0206]

[0207] Here, r is a uniformly distributed random number in the interval [0,1], and its purpose is to compare the probability K of the air conditioner being turned on. open (T), and use this comparison to determine whether to turn on the air conditioner.

[0208] S4. Construct a charging demand determination model. For users of family electric vehicles, classify users into demand types to obtain the charging demand of family electric vehicles. For users of rental electric vehicles, construct a charging demand selection model to obtain the charging demand of rental electric vehicles.

[0209] In step S4, for users of family electric vehicles, the user demand type is categorized, and the user's charging mode demand result is obtained based on the categorization:

[0210] By collecting and analyzing users' parking time, SOC (State of Charge), and expected electricity prices (obtained from electric vehicle parameters), users are categorized into ordinary rigid users, emergency rigid users, high-anchor-elasticity users, or low-anchor-elasticity users. Based on user type and parking time parameters, the system matches an appropriate charging mode (fast charging or slow charging) to maximize charging efficiency and meet travel needs. By default, fast charging prices are higher than slow charging prices during the same time period.

[0211] The concepts of high anchor and low anchor in the psychological concept of anchoring effect are used to judge users' acceptance of electricity prices; high anchor indicates that users' expected value of electricity price is higher than the actual price, and users are more willing to charge; low anchor indicates that users' expected electricity price is lower than the actual price.

[0212] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is greater than or equal to 6 hours, it is judged as a regular rigid user, the charging mode is slow charging, and the charging time is instant.

[0213] If the SOC of an electric vehicle does not meet the needs of the next trip and the parking time is less than 6 hours, it is judged as an emergency user, the charging mode is fast charging, and the charging time is instant.

[0214] Electric vehicles' State of Charge (SOC) meets the needs of the next trip, and users' expectations of electricity prices are higher than the actual prices, making them users with high anchoring elasticity.

[0215] If the parking time is greater than or equal to 6 hours, the charging mode is slow charging and the charging time is instant.

[0216] If the parking time is less than 6 hours, the charging mode is fast charging and the charging time is instant.

[0217] Electric vehicles whose SOC meets the next trip's needs, and whose expected electricity price is lower than the actual price, are identified as low-anchor-elasticity users: the charging mode is slow charging, and the charging time is when the electricity price is low.

[0218] The charging needs of family electric vehicle users were obtained based on the classification.

[0219] Taxi charging demand: based on the current time t and the actual electricity price P at charging station c. c Calculate charging costs and evaluate them in conjunction with the driver's revenue needs:

[0220]

[0221] R c For the benefit trade-off index; F t Expected total revenue from operation after a full charge (e.g., revenue from the order's origin to its destination); E t The energy increment required during charging, i.e., the full charge minus the remaining charge; P c The electricity price for charging stations;

[0222] The current SOC is sufficient to support the next trip, and drivers will then further consider revenue: when R c >R 阈值 Only then will the driver select the corresponding electricity price P. c You can charge at the designated charging station; otherwise, charging will be delayed or you can choose another station.

[0223] If the current SOC is insufficient to support the next trip, determine whether the driver's parking time allows for slow charging: if the parking time is greater than or equal to 6 hours, the driver will prioritize slow charging based on electricity price and charging cost, with charging time being immediate; if the parking time is less than 6 hours, the driver will choose fast charging.

[0224] During periods of low electricity prices (such as at night), even if the State of Charge (SOC) is sufficient, drivers may proactively slow down their charging to reduce charging costs during future peak hours. During peak hours or periods of high electricity prices, drivers may delay charging to prioritize completing their operational tasks.

[0225] The charging needs of taxi electric vehicle users were determined based on the above method.

[0226] S5. Charging load superposition and node load calculation: The charging demand of household electric vehicles and taxis is superimposed on the corresponding road network nodes to calculate the total node charging load, output the spatiotemporal distribution of charging demand, and generate the final spatiotemporal distribution map of electric vehicle charging demand based on the charging load data of all nodes.

[0227] This invention achieves dynamic route selection through real-time traffic data and path penalty coefficients. First, the method adjusts the route in real-time during the journey to avoid congested sections, thereby reducing energy consumption. Simultaneously, a Markov decision process is used to comprehensively evaluate current traffic conditions, the vehicle's state of charge (SOC), and the driver's actual decision-making behavior, simulating the user's "adverse selection" tendency in complex traffic environments, making the route decision more realistic. Furthermore, the method intelligently recommends charging modes based on parking time, selecting slow charging to extend battery life when time is ample, and prioritizing fast charging to meet needs in emergencies. This method not only improves the flexibility and accuracy of route planning but also optimizes the user charging experience, saving charging costs, extending battery life, and enhancing the adaptability of electric vehicles in complex urban traffic environments.

[0228] Example 2:

[0229] An electric vehicle charging demand prediction system based on Markov dynamic path decision-making, the system being used to execute the aforementioned electric vehicle charging demand prediction method based on Markov dynamic path decision-making, specifically including: a data acquisition module, a travel data construction module, a path optimization module, a charging demand determination module, and a spatiotemporal distribution map generation module;

[0230] Data acquisition module: Used to collect maps of the target area, a matrix of vehicle origin and destination points, daily temperature (T), and traffic congestion index. And electric vehicle parameters;

[0231] The travel data construction module is used to construct a road-traffic node map of the target area based on the target area map; analyze electric vehicle parameters, classify electric vehicles into private electric vehicles and taxi electric vehicles, construct a travel chain model for private electric vehicles, and calculate travel time parameters to obtain travel data for private electric vehicles; for taxi electric vehicles, sample the departure point, destination, start time, and parking duration from the vehicle travel origin-destination matrix to obtain travel data for taxi electric vehicles.

[0232] Route optimization module: Used to build Markov dynamic route optimization model. Input family electric vehicle travel data and taxi electric vehicle travel data into Markov dynamic route optimization model, select the optimal route for each electric vehicle, and calculate the energy consumption change during the trip to obtain time series route and energy consumption change dataset.

[0233] Charging demand determination module: Used to build a charging demand determination model. For users of private electric vehicles, it classifies users into demand types to obtain the charging demand of private electric vehicles; for users of rental electric vehicles, it builds a charging demand selection model to obtain the charging demand of rental electric vehicles.

[0234] Spatiotemporal distribution map generation module: used for charging load superposition and node load calculation, superimposing the charging demand of household electric vehicles and taxis onto the corresponding road network nodes to calculate the total node charging load, output the spatiotemporal distribution of charging demand, and generate the final spatiotemporal distribution map of electric vehicle charging demand based on the charging load data of all nodes.

[0235] Example 3:

[0236] An electric vehicle charging demand prediction device based on Markov dynamic path decision includes a memory and a processor. The memory stores computer program code and transmits the computer program code to the processor. The processor executes the aforementioned electric vehicle charging demand prediction method based on Markov dynamic path decision according to the instructions in the computer program code.

[0237] Example 4:

[0238] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned electric vehicle charging demand prediction method based on Markov dynamic path decision.

[0239] A computer program product includes a computer program that is executed by a processor to perform the aforementioned electric vehicle charging demand forecasting method based on Markov dynamic path decision.

Claims

1. A method for electric vehicle charging demand prediction based on Markov dynamic path decision, characterized in that: The method comprises the following steps: S1, collect the map of the target area, the matrix of the starting and ending points of the vehicle trip, the daily temperature T, the traffic congestion index and the electric vehicle parameters; S2, constructing a road-traffic node graph of the target region based on a target region map; analyzing electric vehicle parameters, dividing electric vehicles into family electric vehicles and rental electric vehicles, constructing a trip chain model for the family electric vehicles, calculating trip time parameters, and obtaining family electric vehicle trip data; for the rental electric vehicles, sampling the starting point, destination, starting time, and parking duration from a vehicle trip origin-destination matrix to obtain rental electric vehicle trip data; S3, constructing a Markov dynamic path optimization model, inputting the family electric vehicle trip data and the rental electric vehicle trip data into the Markov dynamic path optimization model, selecting an optimal path for each electric vehicle, and calculating energy consumption changes in the trip to obtain a time-series path and energy consumption change dataset; S3, constructing a Markov dynamic path optimization model, inputting the family electric vehicle trip data and the rental electric vehicle trip data into the Markov dynamic path optimization model, selecting an optimal path for each electric vehicle, Traffic congestion index , which quantitatively describes the running state of road traffic, the value range of the road traffic congestion index is [0, 100], and the larger the value is, the more congested the road is; a penalty coefficient is introduced to make the user's decision not limited to the shortest path: ； wherein, denotes the decision time the lower road a penalty coefficient; denotes the decision time the lower road a traffic congestion index; is the historical maximum traffic congestion index of the reference road segment; the steps of the path decision for each car are as follows: Pre-departure planning: Before departure, the user selects a destination and plans a shortest path; the pre-planned shortest path is denoted as Each segment of the path is called a reference road segment The path set of unselected road network nodes is selected as the candidate path set, denoted as ; Reaching the decision node: When the user reaches the road node Enter decision-making state and select reference road segment. Traffic congestion index and penalty coefficient For reference, the set of alternative paths Each section of the road Determine whether the following conditions are met: ,like If so, the road segment is considered a high-congestion segment and removed from the alternative routes; Calculate state transition probability: for each link in the remaining candidate paths , consider the length and the penalty coefficient when calculating the state transition probability. ； wherein, represents the probability of shifting from the current node to an alternative path ; the length of the alternative path ; the penalty factor of the alternative path at the decision moment ; , the weight factor of the path length and the congestion penalty, adjusting the relative influence of both on the decision. selecting the next segment of the path according to the state transition probabilities selecting the next segment of the path according to the state transition probabilities , adding the final path ; updating the current node to , repeating the above steps until the target node is reached; S4, constructing a charging demand determination model, dividing users of the family electric vehicles into demand types to obtain charging demand of the family electric vehicles; for users of the rental electric vehicles, constructing a charging demand selection model to obtain charging demand of the rental electric vehicles; S5, superimposing charging loads and calculating node loads, superimposing charging demand of the family electric vehicles and the rental vehicles on corresponding road network nodes to calculate total node charging loads, outputting a time-space distribution of charging demand, and generating a final time-space distribution map of electric vehicle charging demand based on charging load data of all nodes. 2.The electric vehicle charging demand prediction method based on Markov dynamic path decision according to claim 1, characterized in that: In the S2 Constructing a road-traffic node graph of the target region based on a target region map, extracting a road graph of the target region in the target region map, determining a plurality of traffic nodes according to distribution of the road graph, and obtaining the road-traffic node graph of the target region; Constructing a trip chain model for the family electric vehicles and calculating trip time parameters to obtain family electric vehicle trip data; Describing the trip chain containing various user information with spatial feature quantities and temporal feature quantities; the spatial feature quantities include vehicle trip start and end points, passing stopover sites, and travel path information; the temporal feature quantities include vehicle start time, destination stopover time, travel time, and trip end time information; the city is divided into four regions: residential areas, work areas, entertainment areas, and other areas, and the coordinate partitions of the spatial feature quantities are represented by this; The calculation method of the temporal feature quantities is as follows: Vehicle start time: the initial time of the vehicle leaving home within a day satisfies a normal distribution, and the probability density function is: ； wherein is the mean of the first departure time; is the variance of the first departure time, x is the first departure time; The stopover time of different destination sites: the stopover time of the vehicle at different destination sites is different, so a plurality of probability density functions are used to represent it; The stopover duration probability of the vehicle in the residential area satisfies a Weibull distribution, and the probability density function is: ； wherein is a shape parameter of the Weibull distribution; is a scale parameter of the Weibull distribution; The stopover duration probability of the vehicle in the work area satisfies a III-type extreme value distribution, and the probability density function is: ； The stopover duration probability of the vehicle in the entertainment area and the other area satisfies a II-type extreme value distribution, and the probability density function is: ； wherein is a location parameter of the generalized extreme value distribution; δ is a scale parameter of the generalized extreme value distribution, δ > 0; ε is a shape parameter of the generalized extreme value distribution, ε > 0 when = 1 / ε, ε < 0 when = -1 / ε. Next trip start time: after the electric vehicle arrives at the destination, the end of the stay is the completion of this trip, and the start time of the next trip is: ； wherein J is the total number of trips in the trip chain; is the time when the vehicle starts the j+1th trip; is the time when the vehicle leaves home for the first time; is the total time spent by the vehicle in completing the first j-1th trips; is the total time spent by the vehicle in completing the first j-1th trips; is the time spent by the vehicle at the destination of the jth trip; and is the time spent by the vehicle at the destination of the jth trip; and using the trip chain model to calculate the electric vehicle trip data of a family. 3.The electric vehicle charging demand prediction method based on Markov dynamic path decision according to claim 1, characterized in that: The energy consumption change in the trip is calculated in S3; An electric vehicle energy consumption model is established: ； In the formula, is the total power consumption of the electric vehicle in a trip; is the power consumption per km of the electric vehicle in a free traffic network; is the length of the selected path p; is the additional power consumption per unit time caused by the start-stop of the electric vehicle due to traffic congestion; is the additional time consumption caused by traffic congestion; is the battery capacity of the electric vehicle; considering that the ambient temperature will affect the opening and closing of the air conditioner of the electric vehicle, the energy consumption model is corrected: ； In the formula, E is the total power consumption of the electric vehicle in a one-way trip; For the temperature energy consumption at temperature T and road level v k The energy consumption of the electric vehicle in the trip is calculated by using the electric vehicle energy consumption model. 4.The method of claim 3, wherein: In S3, According to the air conditioner usage curve, the starting probability of the air conditioner is fitted with a normal distribution to obtain the corresponding normal distribution function of the air conditioner starting as follows: ； wherein represents the probability that the air conditioner is on at temperature ; and are the mean and variance of the heating start parameter, respectively; and are the mean and variance of the cooling start parameter; temperature is represented as at 5 to 21 °C and as at 24 to 35 °C; The air conditioner energy consumption and its proportion in the trip at different temperatures T are analyzed to obtain the relationship between temperature and air conditioner energy consumption proportion as follows: ； Determine temperature at different temperatures and road grade v k under temperature energy consumption : ( )( )； wherein, represents the energy consumption factor of different road levels at time , corresponding to the speed of different road levels: , , , respectively represent the driving speed of first, second and third class roads; represents the initial energy consumption per unit mileage of the electric vehicle; represents the start-stop state of the air conditioner, represents that the air conditioner is on, represents that the air conditioner is off: wherein the value of is as follows: ； wherein, is a uniformly distributed random number in the interval [0,1] and its purpose is to be used for comparing the probability of the air conditioner to be on and by this comparison to decide whether the air conditioner is on or not.

5. The method of claim 1, wherein: In S4, for the users of the family electric vehicle, the users are divided into demand types, and the charging mode demand results of the users are obtained according to the division types: By collecting and analyzing the parking time, SOC, and expected information of the user, the user is classified into an ordinary rigid user, an emergency rigid user, a high-anchor elastic user, or a low-anchor elastic user; The SOC of the electric vehicle does not meet the next trip demand, the parking time is greater than or equal to 6 hours, and the charging mode is slow charging, and the charging time is immediate; The SOC of the electric vehicle does not meet the next trip demand, the parking time is less than 6 hours, and the charging mode is fast charging, and the charging time is immediate; The SOC of the electric vehicle meets the next trip demand, and the user's expected value of the electricity price is higher than the actual price, which is a high-anchor elastic user: The parking time is greater than or equal to 6 hours, the charging mode is slow charging, and the charging time is immediate; The parking time is less than 6 hours, the charging mode is fast charging, and the charging time is immediate; The SOC of the electric vehicle meets the next trip demand, and the user's expected electricity price is lower than the actual price, which is a low-anchor elastic user: the charging mode is slow charging, and the charging time is when the electricity price is low; The charging demand results of the family electric vehicle users are obtained according to the division types.

6. The electric vehicle charging demand prediction method based on Markov dynamic path decision according to claim 1, characterized in that: Taxi charging demand: according to current time and charging station Actual electricity price Calculate charging cost and evaluate in combination with driver's income demand ； a benefit-to-balance index; a total benefit expected from running on a full charge; an incremental energy demand at charging time, i.e. full charge minus remaining charge; a price for the charging station; The current SOC is enough to support the next trip, the driver will further consider the benefit: when , the driver will choose the corresponding charging station with the electricity price , otherwise, the driver will delay charging or choose other sites; If the current SOC is insufficient to support the next trip, it is determined whether the driver's parking time allows slow charging: if the parking time is greater than or equal to 6 hours, the driver will prefer slow charging in combination with the electricity price and charging cost, and the charging time is immediate; If the parking time is less than 6 hours, the driver chooses fast charging; The charging demand results of the rental electric vehicle users are obtained according to the above determination method.

7. A Markov dynamic path decision based electric vehicle charging demand prediction system, characterized in that, The system is used to perform the electric vehicle charging demand prediction method based on Markov dynamic path decision as claimed in any one of claims 1 to 6, and specifically includes a data acquisition module, a trip data construction module, a path optimization module, a charging demand determination module, and a time-space distribution map generation module; Data collection module: for collecting the target area map, vehicle trip origin-destination matrix, daily temperature T, traffic congestion index and electric vehicle parameters; The trip data construction module is used to construct a road-traffic node graph of a target region based on a target region map; analyze electric vehicle parameters, divide electric vehicles into family electric vehicles and rental electric vehicles, construct a trip chain model for family electric vehicles, calculate trip time parameters, and obtain family electric vehicle trip data; for rental electric vehicles, sample the starting place, destination, and starting time and parking duration from a vehicle trip origin-destination matrix to obtain rental electric vehicle trip data; The path optimization module is used for constructing a Markov dynamic path optimization model, inputting the family electric vehicle travel data and the taxi electric vehicle travel data into the Markov dynamic path optimization model, selecting an optimal path for each electric vehicle, and calculating the energy consumption change in the journey to obtain a time-series path and energy consumption change data set. The Markov dynamic path optimization model is constructed, the family electric vehicle travel data and the taxi electric vehicle travel data are input into the Markov dynamic path optimization model, and an optimal path is selected for each electric vehicle, Traffic congestion index , which quantitatively describes the running state of road traffic, the value range of the road traffic congestion index is [0, 100], and the larger the value is, the more congested the road is; a penalty coefficient is introduced to make the user's decision not limited to the shortest path: ； wherein, denotes the decision time the lower road a penalty coefficient; is the traffic congestion index of the lower road at the decision time the lower road is the historical maximum traffic congestion index of the reference road; the steps of the path decision for each car are as follows: Pre-departure planning: Before departure, the user selects a destination and plans a shortest path; the pre-planned shortest path is denoted as Each segment of the path is called a reference road segment The path set of unselected road network nodes is selected as the candidate path set, denoted as ; Reaching the decision node: When the user reaches the road node Enter decision-making state and select reference road segment. Traffic congestion index and penalty coefficient For reference, the set of alternative paths Each section of the road Determine whether the following conditions are met: ,like If so, the road segment is considered a high-congestion segment and removed from the alternative routes; Calculate state transition probability: for each link in the remaining candidate paths , consider the length and the penalty coefficient when calculating the state transition probability: ； in, Indicates starting from the current node Switch to alternative path The probability of; Alternative routes Length; Decision-making moment Alternative paths The penalty coefficient; , The weighting factors for path length and congestion penalty are adjusted to moderate their relative impact on the decision. selecting the next segment of the path according to the state transition probabilities selecting the next segment of the path according to the state transition probabilities , adding the final path ; updating the current node to , repeating the above steps until the target node is reached; The charging demand determination module is used for constructing a charging demand determination model, dividing the users of the family electric vehicles into demand types to obtain the charging demand of the family electric vehicles, and constructing a charging demand selection model for the users of the taxi electric vehicles to obtain the charging demand of the taxi electric vehicles. The space-time distribution map generation module is used for charging load superposition and node load calculation, superimposes the charging demands of the family electric vehicles and the taxis on corresponding road network nodes to calculate the total node charging load, outputs the space-time distribution of the charging demand, and generates a final space-time distribution map of the electric vehicle charging demand according to the charging load data of all nodes.

8. An electric vehicle charging demand prediction device based on Markov dynamic path decision, characterized by, The memory is used for storing computer program code and transmitting the computer program code to the processor. The processor is used for executing the electric vehicle charging demand prediction method based on the Markov dynamic path decision according to the instructions in the computer program code.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to execute the electric vehicle charging demand prediction method based on the Markov dynamic path decision.

Images