A new energy charging pile load prediction method

Abstract

The present invention belongs to the technical field of new energy load management, and in particular, relates to a new energy charging pile load prediction method. It includes the following steps: establishing a classification scheme of electric vehicle node types and a target selection probability scheme, determining the travel path of electric vehicles, and predicting the travel route and the next target node of electric vehicles; calculating and obtaining necessary data for prediction, and summarizing all electric vehicle data results to obtain the charging load of new energy charging piles in different target nodes. The new energy charging pile load prediction method of the present invention is based on the analysis of the travel path and charging behavior of electric vehicle owners at the source of power load in the region, and predicts the charging demand and charging load faced by different new energy charging pile nodes, providing a basis for realizing more comprehensive load demand calculation and power distribution under regional multi-nodes.

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy load management, and in particular relates to a new energy charging pile load prediction method. Background Art

[0002] With the massive increase in new energy electric vehicles, people have an increasing demand for convenient charging of electric vehicles. Based on the early unified planning and construction of the national power grid and the access of a large number of private charging piles in recent years, a huge electric vehicle charging pile network has been established in the domestic market. Due to the decentralized nature of new energy charging piles, whether they are charging and the charging time are not stable, so the actual load status of new energy charging stations is also uncertain, which provides new challenges for the power supply management of the power grid. Summary of the invention

[0003] The purpose of the present invention is to provide a new energy charging pile load prediction method based on the distribution network load control demand in order to improve the charging compliance of regional new energy charging piles.

[0004] To achieve the above purpose, the present invention adopts the following technical solution.

[0005] A new energy charging pile load prediction method comprises the following steps:

[0006] S1. Establishing a node type classification scheme for electric vehicles and a target selection probability scheme, specifically: establishing classification data of necessary nodes and possible nodes in the driving path of electric vehicles; wherein necessary nodes are fixed starting nodes and target nodes of electric vehicle owners, and possible nodes are optional nodes that may be passed in the periodic action route of electric vehicle owners;

[0007] Assuming that the set of all necessary nodes is M, and the set of all possible nodes is N, then at time t, the probability ρ of each electric vehicle owner going from the current node i to the target node j is t (i,j) can be expressed as:

[0008]

[0009] Where P t (i,j) refers to the congestion index of the route from the current node i to the target node j;

[0010] t j It refers to the expected time when the electric car owner arrives at the must-pass node j;

[0011] S2. Determine the travel path of the electric vehicle, specifically: determine the target node at each time based on step S1, map each target node to the nearest one or several line nodes in the road network, and use the current road network to build a directed network to determine the travel distance; establish a minimum travel cost function, and determine the travel path of the electric vehicle by traversing the directed road network;

[0012] S3, based on the charging and discharging operation data of electric vehicles in the area, determine the starting node of the electric vehicles in the area to be analyzed; predict the travel route and the next target node of the electric vehicles according to steps S1 and S2; calculate the charging probability, battery power, and charging time after reaching the target node, and the specific steps include:

[0013] Establish the driving status of electric vehicles at all times k Model expression sta=a k ∪b k , where a k Indicates that the electric vehicle is in online charging state, b k Indicates that the electric vehicle is in an offline charging state (including driving or parking); the probability of the electric vehicle being in two states is:

[0014]

[0015] ρ(a k )+ρ(b k )=1k∈[0,Bat max ]

[0016] Among them, ρ(a k ) represents the probability that the electric vehicle is in the online charging state, ρ(b k ) represents the probability that the electric vehicle is in an offline charging state; Bat max represents the maximum allowable capacity percentage of the battery; k is the current power percentage of the electric vehicle; ρ k represents the probability of the electric vehicle being connected to the grid for charging when the current power percentage of the electric vehicle is k, σ k represents the probability of the electric vehicle being offline when the current power percentage of the electric vehicle is k; γ k represents the probability of the electric vehicle stopping when the current battery percentage of the electric vehicle is k;

[0017] According to the different charging pressures of electric vehicle owners, the charging time of electric vehicle owners can be expressed as The driving time of an electric vehicle can be expressed as The time that the electric vehicle is parked and not charged can be expressed as T3 = T out -T in -T1;

[0018] The amount of electricity after the electric vehicle reaches the target node is Soc t =Soc t-1 -dis t E0;

[0019] When the user arrives at a node where a charging pile is located, the remaining power of the vehicle needs to be updated in order to predict the possibility of the electric car owner charging at this node; if the current power cannot meet the power demand of the next trip, charging is necessary, and within a certain range, the gap between the power demand of the next trip and the current power directly affects the charging decision of the electric car owner, and the charging probability prediction model of the electric car owner can be obtained. Among them, Soc min It refers to the minimum amount of electricity that electric car owners can accept before each trip, σ socmin Soc min The standard deviation of

[0020] Bat1 represents the capacitance corresponding to each power percentage, Δk + represents the expected percentage increase in power during charging; η represents the charging efficiency, μ represents the rated charging power; T out represents the time when the electric vehicle leaves the charging station, T in Indicates the time when the electric vehicle enters the charging station; Δk - It represents the percentage of electricity consumed during driving time, E0 represents the electricity consumption per unit distance, and v represents the average driving speed of the electric vehicle; Soc t-1 represents the initial power of the electric vehicle when it starts from the previous node, dis t Indicates the distance from the previous node to the current node;

[0021] S4. Continuously obtain the driving route, battery power and charging status of the electric vehicle, predict the target node of the electric vehicle at the next moment and the charging status after reaching the target node, and summarize all electric vehicle data results to obtain the charging load of new energy charging piles in different target nodes.

[0022] For further optimization or specific implementation steps of the aforementioned new energy charging pile load prediction method, the necessary nodes are composed of residential places, workplaces, dining places, and children's teaching places; the possible nodes are composed of consumer and entertainment places.

[0023] Further optimization or specific implementation steps of the above-mentioned new energy charging pile load prediction method, in step S2, for areas with commercial entertainment areas and tourist consumption areas as core areas, the driving path of electric vehicles is predicted and analyzed according to time and space constraints, specifically including the following steps:

[0024] S20. Establish a regional road directed network model G = (V, E, W)

[0025] Where G represents a directed road network, V is a set of road nodes, E refers to a set of road routes, and W refers to the road resistance. Generally speaking, there are multiple feasible paths from the initial node A to the target node B. For each feasible path A→B, its passing time is composed of the total travel cost of all routes in the path, using the cost variable To express; where Pr(v r-1 ,v r ) represents the travel cost of the rth route in the feasible path A→B, v r represents the end point of the rth line; the electric vehicle route planning strives to minimize the travel cost, which may be the minimum driving time or the shortest driving path, etc., which is determined according to the actual state and preference of the electric vehicle; based on the minimum travel cost, the optimal path L(A,B)=min{Pr(a→b)} from the initial node A to the target node B is determined.

[0026] Further optimization or specific implementation steps of the above-mentioned new energy charging pile load prediction method, when the current power of the electric vehicle owner exceeds the power required for the next trip, that is, Soc t ≥Soc t+1,start When , the probability of the electric car owner continuing to charge is estimated by the following function Among them, Soc max It refers to the amount of electricity that the electric car owner expects to reach when charging is finished.

[0027] Its beneficial effects are:

[0028] The new energy charging pile load prediction method of the present invention predicts the charging demand and charging load faced by different new energy charging pile nodes based on the analysis of the travel routes and charging behaviors of electric car owners at the source of power load in the region, thereby providing a basis for realizing more comprehensive load demand calculation and power distribution under multiple nodes in the region. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the new energy charging pile load prediction method. DETAILED DESCRIPTION

[0030] The present invention is described in detail below in conjunction with specific embodiments.

[0031] The new energy charging pile load prediction method of the present invention is mainly used to provide a method for analyzing and predicting the charging demand and real-time load conditions that the new energy charging piles at each node in the network may face according to the dynamic operation changes of electric vehicles in the region, and is used to provide auxiliary reference basis for construction planning such as new energy charging station construction, power supply and load distribution, and realize analysis and processing of regional distribution load.

[0032] Based on the current rules of electric vehicle operation and maintenance management, in an area with a relatively stable number of electric vehicles, each electric vehicle owner who has been active in the area for a long time has their own relatively stable starting point and target node. According to the daily electric vehicle load forecast analysis, the starting point and target location can be divided into residential places, workplaces, and consumer entertainment places. Generally speaking, the above nodes can be divided into necessary nodes and possible nodes according to the probability of electric vehicle owners passing through.

[0033] Among them, the must-go nodes are the fixed starting nodes and target nodes of the electric car owners. During the fixed time period of each day, the electric car owners will inevitably reach the corresponding nodes. The must-go nodes are generally composed of residential places, teaching places, workplaces, and a small number of consumption and entertainment places, such as catering places, etc.

[0034] Possible nodes are optional nodes that electric vehicle owners may pass through in their periodic action routes, and are generally composed of places for consumption and entertainment;

[0035] The selection of the necessary nodes is generally not directly affected by various factors, while the probability of the optional nodes being selected is often directly affected by the road congestion between the current node and the node route;

[0036] Assuming that the set of all necessary nodes is M, and the set of all possible nodes is N, then at time t, the probability ρ of each electric vehicle owner going from the current node i to the target node j is t (i,j) can be expressed as:

[0037]

[0038] Where P t (i,j) refers to the congestion index of the route from the current node i to the target node j;

[0039] t j It refers to the expected time when the electric car owner arrives at the must-pass node j;

[0040] Based on step S1, the target node at each moment is determined, and each target node is mapped to the nearest one or several route nodes in the road network. Then, the current road network can be used to construct a directed network (because there are one-way routes such as one-way streets) to determine the travel distance; the distance between adjacent road nodes can be simplified to its Euclidean distance; usually, an electric car owner needs to pass through several road nodes to go from the current node i to the target node j, and there are often multiple different routes to reach a certain target node. In general, electric car owners are more inclined to choose the route with the shortest mileage; therefore, the travel route to each different target node generally remains unchanged, and a feasible route is determined by traversing the directed road network, that is, the sum of the Euclidean distances of all nodes on the way is used to represent its mileage;

[0041] Furthermore, for areas with commercial entertainment areas, tourist consumption areas, etc. as core areas, the driving route characteristics of vehicles are different from those of traditional cities. In order to better determine the travel routes and driving distances of electric vehicles in such areas, the driving paths of electric vehicles are predicted and analyzed based on time and space constraints, which specifically includes the following steps:

[0042] S20. Establish a regional road directed network model G = (V, E, W)

[0043] Where G represents a directed road network, V is a set of road nodes, E refers to a set of road routes, and W refers to the road resistance. Generally speaking, there are multiple feasible paths from the initial node A to the target node B. For each feasible path A→B, its passing time is composed of the total travel cost of all routes in the path, using the cost variable To express; where Pr(v r-1 ,v r ) represents the travel cost of the rth route in the feasible path A→B, v r represents the end point of the rth route; the electric vehicle route planning strives to minimize the travel cost, which may be the minimum driving time or the shortest driving path, etc., determined according to the actual state and preference of the electric vehicle; based on the minimum travel cost, the optimal path L(A,B)=min{Pr(a→b)} from the initial node A to the target node B is determined;

[0044] Based on the charging and discharging operation data of electric vehicles in the region, the starting node of the electric vehicles in the area to be analyzed is determined; the travel route and the next target node of the electric vehicles are predicted according to steps S1 and S2; the charging probability, battery power and charging time after reaching the target node are calculated. The specific steps include:

[0045] Establish the driving status of electric vehicles at all times k Model expression sta=a k ∪b k, where a k Indicates that the electric vehicle is in online charging state, b k Indicates that the electric vehicle is in an offline charging state (including driving or parking); the probability of the electric vehicle being in two states is:

[0046]

[0047] ρ(a k )+ρ(b k )=1k∈[0,Bat max ]

[0048] Among them, ρ(a k ) represents the probability that the electric vehicle is in the online charging state, ρ(b k ) represents the probability that the electric vehicle is in an offline charging state; Bat max represents the maximum allowable capacity percentage of the battery; k is the current power percentage of the electric vehicle; ρ k represents the probability of the electric vehicle being connected to the grid for charging when the current power percentage of the electric vehicle is k, σ k represents the probability of the electric vehicle being offline when the current power percentage of the electric vehicle is k; γ k represents the probability of the electric vehicle stopping when the current battery percentage of the electric vehicle is k;

[0049] According to the different charging pressures of electric vehicle owners, the charging time of electric vehicle owners can be expressed as The driving time of an electric vehicle can be expressed as The time that the electric vehicle is parked and not charged can be expressed as T3 = T out -T in -T1;

[0050] The amount of electricity after the electric vehicle reaches the target node is Soc t =Soc t-1 -dis t E0;

[0051] When the user arrives at a node where a charging pile is located, the remaining power of the vehicle needs to be updated in order to predict the possibility of the electric vehicle owner charging at this node; if the current power cannot meet the power demand of the next trip, charging is necessary, and within a certain range, the power demand of the next trip Soc t+1,start And the current power Soc t The difference directly affects the charging decision of electric car owners, and the charging probability C of electric car owners can be obtained. t Prediction Model:

[0052]

[0053] Bat1 represents the capacitance corresponding to each power percentage, Δk + represents the expected percentage increase in power during charging; η represents the charging efficiency, μ represents the rated charging power; T out represents the time when the electric vehicle leaves the charging station, T in Indicates the time when the electric vehicle enters the charging station; Δk - It represents the percentage of electricity consumed during driving time, E0 represents the electricity consumption per unit distance, and v represents the average driving speed of the electric vehicle; Soc t-1 represents the initial power of the electric vehicle when it starts from the previous node, dis t Indicates the distance from the previous node to the current node; Soc min It refers to the minimum amount of electricity that electric car owners can accept before each trip, σ socmin Soc min The standard deviation of

[0054] In particular, the charging probability C of the electric car owner mentioned above t When the current power level is close to or exceeds the power required for the next trip, the prediction model often cannot accurately determine the true probability of actual charging behavior based on the situation at different nodes. The actual situation is that even if the current power level can meet the power demand for the next trip, out of the mentality of ensuring sufficient power and the trend of expenditure balance, when electric car owners face lower charging prices or stay longer at the node, they will take the initiative to charge without being affected by the current power level and the power required for the next trip.

[0055] According to data statistics, when the current power of the electric car owner exceeds the power required for the next trip, the Soc t ≥Soc t+1,start When , the probability that the electric car owner continues to charge can be estimated by the following function Among them, Soc max Refers to the amount of electricity that the electric car owner expects to reach when charging is finished.

[0056] Continuously obtain the driving route, battery power and charging status of electric vehicles, predict the target node of electric vehicles at the next moment and the charging status after reaching the target node, and summarize all electric vehicle data results to obtain the charging load of new energy charging piles in different target nodes;

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the essence and scope of the technical solution of the present invention.

Claims

1. A new energy charging pile load prediction method, characterized in that: The steps include: S1. Establishing a node type classification scheme for electric vehicles and a target selection probability scheme, specifically: establishing classification data of necessary nodes and possible nodes in the driving path of electric vehicles; wherein necessary nodes are fixed starting nodes and target nodes of electric vehicle owners, and possible nodes are optional nodes that may be passed in the periodic action route of electric vehicle owners; Assuming that the set of all necessary nodes is M, and the set of all possible nodes is N, then at time t, the probability ρ of each electric vehicle owner going from the current node i to the target node j is t (i,j) can be expressed as: Where P t (i,j) refers to the congestion index of the route from the current node i to the target node j; t j It refers to the expected time when the electric car owner arrives at the must-pass node j; S2. Determine the travel path of the electric vehicle, specifically: determine the target node at each time based on step S1, map each target node to the nearest one or several line nodes in the road network, and use the current road network to build a directed network to determine the travel distance; establish a minimum travel cost function, and determine the travel path of the electric vehicle by traversing the directed road network; S3, based on the charging and discharging operation data of electric vehicles in the area, determine the starting node of the electric vehicles in the area to be analyzed; predict the travel route and the next target node of the electric vehicles according to steps S1 and S2; calculate the charging probability, battery power, and charging time after reaching the target node, and the specific steps include: Establish the driving status of electric vehicles at all times k Model expression sta=a k ∪b k , where a k Indicates that the electric vehicle is in online charging state, b k Indicates that the electric vehicle is in an offline charging state; the probability that the electric vehicle is in two states is: p(a k )+ρ(b k )=1,k∈[0,Bat max ] Among them, ρ(a k ) represents the probability that the electric vehicle is in the online charging state, ρ(b k ) represents the probability that the electric vehicle is in an offline charging state; Bat max represents the maximum allowable capacity percentage of the battery; k is the current power percentage of the electric vehicle; ρ k represents the probability of the electric vehicle being connected to the grid for charging when the current power percentage of the electric vehicle is k, σ k represents the probability of the electric vehicle being offline when the current power percentage of the electric vehicle is k; γ k represents the probability of the electric vehicle stopping when the current battery percentage of the electric vehicle is k; According to the different charging pressures of electric vehicle owners, the charging time of electric vehicle owners can be expressed as The driving time of an electric vehicle can be expressed as The time that the electric vehicle is parked and not charged can be expressed as T3 = T out -T in -T1; The amount of electricity after the electric vehicle reaches the target node is Soc t =Soc t-1 -dis t E0; When the user arrives at a node where a charging pile is located, the remaining power of the vehicle needs to be updated in order to predict the possibility of the electric car owner charging at this node; if the current power cannot meet the power demand of the next trip, charging is necessary, and within a certain range, the gap between the power demand of the next trip and the current power directly affects the charging decision of the electric car owner, and the charging probability prediction model of the electric car owner can be obtained. Among them, Soc min It refers to the minimum amount of electricity that electric car owners can accept before each trip, σ socmin Soc min The standard deviation of Bat1 represents the capacitance corresponding to each power percentage, Δk + represents the expected percentage increase of power during charging; η represents the charging efficiency, μ represents the rated charging power; T out represents the time when the electric vehicle leaves the charging station, T in Indicates the time when the electric vehicle enters the charging station; Δk - It represents the percentage of electricity consumed during driving time, E0 represents the electricity consumption per unit distance, and v represents the average driving speed of the electric vehicle; Soc t-1 represents the initial power of the electric vehicle when it starts from the previous node, dis t Indicates the distance from the previous node to the current node; S4. Continuously obtain the driving route, battery power and charging status of the electric vehicle, predict the target node of the electric vehicle at the next moment and the charging status after reaching the target node, and summarize all electric vehicle data results to obtain the charging load of new energy charging piles in different target nodes.

2. The new energy charging pile load prediction method according to claim 1 is characterized in that: The necessary nodes are composed of residential places, teaching places, and working places; the possible nodes are composed of consumption and entertainment places.

3. The new energy charging pile load prediction method according to claim 1, characterized in that: In step S2, for areas with commercial entertainment areas and tourist consumption areas as core areas, a prediction analysis of the driving path of the electric vehicle is performed according to time and space constraints, specifically including the following steps: S20. Establish a regional road directed network model G=(V,E,W) Where G represents a directed road network, V is a set of road nodes, E refers to a set of road routes, and W refers to the road resistance. There are multiple feasible paths from the initial node A to the target node B. For each feasible path A→B, its passing time is composed of the total travel cost of all routes in the path, using the cost variable To express; where Pr(v r-1 ,v r ) represents the travel cost of the rth route in the feasible path A→B, v r represents the end point of the rth line; the electric vehicle route planning strives to achieve the lowest travel cost, which is the minimum driving time or the shortest driving path target, determined according to the actual state and preference of the electric vehicle; based on the minimum travel cost, the optimal path L(A,B)=min{Pr(a→b)} from the initial node A to the target node B is determined.

4. The new energy charging pile load prediction method according to claim 1, characterized in that: When the electric car owner's current power exceeds the power required for the next trip, the Soc t ≥Soc t+1,start When , the probability of the electric car owner continuing to charge is estimated by the following function Among them, Soc max It refers to the amount of electricity that the electric car owner expects to reach when charging is finished.

Images