Power-traffic network charging load prediction method and system, and storage medium

Abstract

The invention discloses a power-traffic network charging load prediction method and system, and a storage medium. The method comprises the following steps: constructing an electric vehicle basic model, a road network basic model, a charging station basic model and a power distribution network basic model; generating dynamic traffic flow data through a generative adversarial network based on the basic model and the historical traffic data; constructing a traffic travel demand matrix based on a start-end point analysis method, and simulating the whole process of charging decision, queuing waiting and charging service of the electric vehicle in the charging station by using a queuing theory algorithm in combination with the dynamic traffic flow data and the basic model; constructing a space-time diagram attention network model, taking charging demand simulation data, a road network basic model and a charging station basic model as input, obtaining a space-time coupling relationship between a traffic network state and a charging demand through a multi-head diagram attention mechanism, and predicting to obtain space-time distribution of an electric vehicle charging load in a certain period of time in the future. And accurate prediction of the time-space part of the charging load of the electric vehicle is realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power grid and traffic network coupling, and relates to a power-traffic network charging load prediction method and system and a storage medium. BACKGROUND

[0002] With the rapid popularization of electric vehicles, the charging load introduced by electric vehicles in the power system presents significant spatio-temporal randomness and uncertainty, which brings severe challenges to the planning, safe operation and optimal scheduling of distribution networks. The charging load of electric vehicles is not only affected by complex spatio-temporal factors such as user travel behavior, road network and traffic flow state, but also closely related to the service capacity of charging stations.

[0003] Traditional load prediction methods mostly start from a single dimension of the power system or only focus on time series analysis, ignoring the deep coupling and interaction relationship between the traffic network and the power network and the dynamic propagation characteristics of the charging load in the spatial dimension, resulting in limited prediction accuracy. In addition, the existing methods still have deficiencies in accurately depicting the randomness of traffic flow and effectively mapping it to the spatio-temporal distribution of charging load, which is difficult to meet the high requirements of power-traffic coordination optimization for prediction accuracy. SUMMARY

[0004] The purpose of the present application is to provide a power-traffic network charging load prediction method, system and storage medium, to build a multi-source basic model framework integrating electric vehicle characteristics, road network, charging station operation parameters and distribution network topology, and to deeply mine the complex spatio-temporal coupling relationship between the traffic network and the charging demand based on the ST-GAT model, and finally to realize the accurate prediction of the spatio-temporal distribution of electric vehicle charging load and the quantitative evaluation of its impact on the carrying capacity of the distribution network through queuing theory simulation and power flow calculation.

[0005] The technical solution for achieving the purpose of the present application is as follows: A power-traffic network charging load prediction method, comprising the following steps: Building a basic model, the basic model including an electric vehicle basic model, a road network basic model, a charging station basic model, and a distribution network basic model; Based on the basic model and historical traffic data, generating dynamic traffic flow data through a generative adversarial network; Building a traffic travel demand matrix based on the origin-destination analysis method, and combining the dynamic traffic flow data and the basic model, superimposing the dynamic traffic flow data on the road network basic model to obtain a traffic network, and using a queuing theory algorithm to simulate the charging decision, queuing and charging service process of electric vehicles in the charging station; The space-time graph attention network model is constructed, and data of charging decisions, queuing and waiting and a whole charging service process of the simulated electric vehicle in the charging station, a road network basic model and a charging station basic model are taken as inputs, the space-time coupling relationship between the traffic network state and the charging demand is obtained through the multi-head graph attention mechanism to adaptively learn the spatial dependence strength between different nodes in the traffic network and the gate recurrent unit to capture the long-term regularity of the load change, and the space-time distribution of the electric vehicle charging load in a future period is predicted.

[0006] In the preferred technical solution, after predicting the space-time distribution of the electric vehicle charging load in a future period, the following steps are further included: The predicted space-time distribution data of the electric vehicle charging load is taken in the form of node injection power into the power distribution network basic model, the operation state of the power distribution network under each scenario is solved through power flow calculation, the influence of the electric vehicle charging load on the key operation indicators of the power distribution network is quantitatively evaluated, and the carrying capacity of the power distribution network is analyzed and obtained.

[0007] In the preferred technical solution, the electric vehicle basic model is constructed, and the following steps are further included: Different types of electric vehicles are simulated, and the types include private cars, taxis and buses, and different travel vehicle parameters are set for each type of vehicle; the vehicle parameters include battery capacity, initial state of charge, unit mileage power consumption, and rated charging power under fast charging and slow charging; The starting and ending points of the vehicle, the path selection and the stay time distribution of the set points in the path are simulated based on the trip chain data; Wherein, the state of charge of the vehicle changes with time during driving and satisfies the following formula:

[0008] Wherein, And S(t) and S(t+1) represent the state of charge of the vehicle at time t and t+1, respectively; is the battery charging and discharging efficiency coefficient; is the average power consumption per unit driving distance of the vehicle; is the driving distance of the vehicle in the interval (t, t+1), which is set according to dynamic traffic flow data; When the state of charge of the vehicle is lower than the preset charging threshold and is in a chargeable area, the charging demand is triggered.

[0009] In the preferred technical solution, the road network basic model is constructed, and the following steps are further included: The road network is abstracted as an undirected graph wherein U represents a set of network nodes, including road intersections, key traffic hubs, and geographic locations of charging stations; W represents a set of edges connecting the nodes, representing road segments, each edge is assigned a weight, which is obtained according to the actual distance, average travel time or travel cost between nodes; The connection relationship and weight between nodes are recorded by using an adjacency matrix D. If node i is directly connected to node j, then is the weight of the edge, otherwise 0 or infinity; The Dijkstra algorithm is used to calculate the shortest distance between any two points in the road network, which provides a basis for subsequent traffic flow allocation and travel time estimation.

[0010] In the preferred technical solution, the construction of the charging station basic model comprises: In the road network basic model, the charging station is modeled as a coupling node of the power distribution network and the road network; Each charging station node is associated with the following attributes in addition to the basic geographic location attribute: service capacity; rated power of each charging pile; device utilization S, wherein S is the sum of the rated power of all charging piles of the charging station; and the number information of the access node of the power distribution network connected thereto, to establish a mapping relationship between the traffic network and the power distribution network; The key operating parameters of the charging station are defined, including: the number of charging piles, which is used to determine the service channel number of the system; the rated charging power of each charging pile; the device utilization; the charging service rate, which sets the charging service time of each vehicle to obey a negative exponential distribution, and the mean value is determined by the vehicle battery capacity, charging power and initial state of charge.

[0011] In the preferred technical solution, the construction of the power distribution network basic model comprises: The electrical connection relationship and parameters of the power distribution network are established by using the IEEE standard power system topology, the line impedance and transformer parameters are set based on standard values, and the power distribution network is coupled with the road network through geographic location information.

[0012] In the preferred technical solution, the dynamic traffic flow data is generated by the generative adversarial network, comprising: A feature tensor composed of multi-source data is constructed, which comprises: a historical traffic flow feature matrix, the historical traffic flow feature matrix comprising the traffic flow and speed of each road segment at the same time of the past N days; an adjacency matrix of the road network basic model, used to introduce spatial structure information; a time period feature vector, the time period feature vector comprising the time of the day, weekdays, morning peak, evening peak; and weather condition data; The feature tensor composed of the multi-source data is input into a generator of a generative adversarial network; the generator adopts a neural network structure comprising a graph convolution layer and a full connection layer, and is used for learning a mapping of realistic traffic flow data from random noise and the feature tensor composed of the multi-source data; The discriminator of the generative adversarial network adopts a graph convolution network structure, and is used for discriminating the authenticity of the single node traffic flow data and the difference between the generated traffic flow and the real traffic flow in spatial correlation; The training process is trained by using a Wasserstein GAN framework, and an optimization objective is to minimize the Wasserstein distance between a generated data distribution and a real data distribution, and a loss function is defined as follows:

[0013] Among them, The optimization objective of the generator G is to minimize the entire expression, The optimization objective of the discriminator is to maximize the entire expression, is a real traffic flow data distribution, is a data distribution output by the generator G, and D is a 1-Lipschitz function set, The discriminator D must be from a function set satisfying the 1-Lipschitz continuity condition, Indicates that the expectation value is calculated for all samples x sampled from the real data distribution , Indicates that the expectation value is calculated for all generated samples sampled from the generated data distribution , is a score of the discriminator D on the real sample x, is a score of the discriminator D on the generated sample ; through the adversarial training, the generator can generate dynamic traffic flow data which retains the real road network space-time features.

[0014] In the preferred technical solution, the traffic travel demand matrix is constructed based on the start and end point analysis method, and the queuing theory algorithm is used to simulate the charging decision, queuing and waiting and the whole charging service process of the electric vehicle in the charging station in combination with the dynamic traffic flow data and the basic model, including: Traffic travel demand matrix construction: based on historical traffic data, a start and end point traffic travel demand matrix in different time periods is constructed , indicating the number of vehicles traveling from the start point o to the end point d in the time period t; Traffic flow distribution and path selection: load the traffic demand matrix to the road network base model, get the dynamic road section travel time from the dynamic traffic flow data, use the stochastic user equilibrium path selection model to distribute the travel demand to each path of the road network, so as to get the traffic flow, speed and travel time of each vehicle in the road network at different time; Charging behavior triggering and queuing simulation: when the simulation vehicle is driving, its SOC is lower than the threshold value, the charging station is selected based on distance, current queuing condition and charging price; the service process in the charging station is simulated by using M / M / C queuing model; Vehicle arrival process: set the process of vehicle arrival at each charging station to follow the Poisson process with parameter λ, and the value of λ is dynamically determined by the number of vehicles arriving at the station; Service process: the service time of each charging pile obeys the negative exponential distribution with parameter μ, which is related to the charging power and the required charging capacity; Get the charging request queue length, average waiting time, and the time of starting charging and the required charging capacity of each charging station at different time periods.

[0015] In the preferred technical scheme, the space-time graph attention network model comprises a space attention module and a time attention module, the space attention module adopts a multi-head graph attention mechanism, and the feature vector representation of each node i at the t-th time step in the graph is represented as:

[0016] wherein, is a splicing operator of K vectors, K is the total number of attention heads, is a nonlinear activation function, and j ∑ i is a summation operation on all nodes j in the neighbor set of node i, is a normalized attention weight coefficient calculated by the kth attention head, representing the importance of node j to node i, is a trainable weight matrix exclusive to the kth attention head, is the input feature vector of neighbor node j at the t-th time step; l The time attention module adopts a gated recurrent unit to capture the time dependence; The input of the space-time graph attention network model is a space-time sequence tensor with dimensions of time steps x node number x features, wherein the features include predicted charging vehicle number, queue length, average speed of road section and time features of each charging station; the output of the space-time graph attention network model is the predicted charging load value of each charging station in the future T time steps; The mean square error is used as the loss function for supervised learning, and the formula is as follows:​​

[0017] wherein, is the true load value, is the model predicted value, N is the number of samples, is all trainable parameters of the model, the loss function is minimized using the Adam optimizer.

[0018] The application further discloses a power-traffic network charging load prediction system for realizing the power-traffic network charging load prediction method. a basic model construction module, which constructs a basic model, the basic model comprising an electric vehicle basic model, a road network basic model, a charging station basic model and a power distribution network basic model; a dynamic traffic flow generation module, which generates dynamic traffic flow data based on the basic model and historical traffic data through a generative adversarial network; a charging demand simulation module, which constructs a traffic travel demand matrix based on an origin-destination analysis method, and combines the dynamic traffic flow data and the basic model to superimpose the dynamic traffic flow data on the road network basic model to obtain a traffic network, and simulates the charging decision, queuing and waiting and charging service whole process of the electric vehicle in the charging station by using a queuing theory algorithm; a charging load prediction module, which constructs a spatio-temporal graph attention network model, takes the simulated data of the charging decision, queuing and waiting and charging service whole process of the electric vehicle in the charging station, the road network basic model and the charging station basic model as input, adaptively learns the spatial dependence strength between different nodes in the traffic network through a multi-head graph attention mechanism, and combines a gated recurrent unit to capture the long-term regularity of load variation, to obtain the spatio-temporal coupling relationship between the traffic network state and the charging demand, and to predict the spatio-temporal distribution of the electric vehicle charging load in a certain period in the future.

[0019] The application further discloses a computer storage medium, which stores a computer program, and the computer program is executed by a computer to realize the power-traffic network charging load prediction method.

[0020] The application further discloses an electronic device, which comprises a memory and a processor, the memory stores a computer program, the processor runs the computer program stored on the memory, and the computer program is executed to realize the power-traffic network charging load prediction method.

[0021] Compared with the prior art, the application has the following advantages: 1. By constructing a multi-source basic model framework integrating electric vehicle characteristics, road network, charging station operation parameters and power distribution network topology, dynamic traffic flow data conforming to real distribution and having spatio-temporal randomness is generated by using the WGAN model to overcome data deficiency or noise interference, and based on the ST-GAT, the complex spatio-temporal coupling relationship between the traffic network and charging demand is deeply mined, and finally the precise prediction of the spatio-temporal distribution of electric vehicle charging load and the quantitative evaluation of the impact of the electric vehicle charging load on the power distribution network carrying capacity are realized through queuing theory simulation and power flow calculation, so as to effectively solve the problems of the existing prediction methods, such as the insufficient description of the power-traffic network coupled system and the weak processing ability of uncertainty, and provide reliable technical support for the orderly charging of electric vehicles and the evaluation of the carrying capacity of the power distribution network.

[0022] 2. The method adaptively learns the spatial dependence strength between different nodes in the traffic network through the multi-head graph attention mechanism, and combines the gated recurrent unit to capture the long-term regularity of load changes, so as to realize the high-precision prediction of the spatio-temporal distribution of charging load in the complex coupled system, and the root mean square error of the prediction result is reduced by about 15% compared with the existing mainstream methods, effectively improving the accuracy and practicability of the prediction model. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The flow chart of the electric power-traffic network charging load prediction method of the embodiment; Figure 2 The structure diagram of the coupling model of the road network and the power distribution network topology of the embodiment; Figure 3 The structure diagram of the WGAN model for generating dynamic traffic flow of the embodiment; Figure 4 The effect comparison diagram of the WGAN of the embodiment and other traffic flow simulation methods; Figure 5 The structure diagram of the ST-GAT model for predicting charging load of the embodiment; Figure 6 The average MSE comparison diagram of the prediction method based on the ST-GAT of the embodiment and other prediction models; Figure 7 The voltage out-of-limit node details of the power flow calculation in a specific case of the embodiment; Figure 8 The principle block diagram of the electric power-traffic network charging load prediction system of the embodiment. DETAILED DESCRIPTION

[0024] The principle of the present application is: by constructing a multi-source basic model framework integrating the characteristics of electric vehicles, road network, charging station operation parameters and power distribution network topology, using WGAN to generate dynamic traffic flow data conforming to the real distribution and having spatio-temporal randomness to overcome data deficiency or noise interference, and based on ST-GAT to deeply mine the complex spatio-temporal coupling relationship between the traffic network and charging demand, finally realizing the accurate prediction of the spatio-temporal distribution of electric vehicle charging load and the quantitative evaluation of the impact of the electric vehicle charging load on the carrying capacity of the power distribution network through queuing theory simulation and power flow calculation, thereby effectively solving the problems of the existing prediction methods, such as the insufficient description of the power-traffic network coupled system and the weak ability to deal with uncertainty, and providing reliable technical support for the orderly charging of electric vehicles and the evaluation of the carrying capacity of the power distribution network.

[0025] Embodiments: As shown in Figure 1 A power-traffic network charging load prediction method, comprising the following steps: Constructing a basic model, the basic model comprising an electric vehicle basic model, a road network basic model, a charging station basic model, and a power distribution network basic model; Based on the basic model and historical traffic data, generating dynamic traffic flow data through a generative adversarial network; Based on the origin-destination analysis method, constructing a traffic travel demand matrix, and combining the dynamic traffic flow data and the basic model, superimposing the dynamic traffic flow data on the road network basic model to obtain a traffic network, and using a queuing theory algorithm to simulate the charging decision, queuing and waiting, and charging service whole process of electric vehicles in the charging station; Constructing a spatio-temporal graph attention network model, taking the data of the simulated charging decision, queuing and waiting, and charging service whole process of electric vehicles in the charging station, the road network basic model, and the charging station basic model as input, adaptively learning the spatial dependence strength between different nodes in the traffic network through a multi-head graph attention mechanism, and combining a gated recurrent unit to capture the long-term regularity of load changes, obtaining the spatio-temporal coupling relationship between the traffic network state and the charging demand, and predicting the spatio-temporal distribution of the electric vehicle charging load in a certain period in the future.

[0026] In a preferred embodiment, after predicting the spatio-temporal distribution of the electric vehicle charging load in a certain period in the future, the method further comprises: Connecting the predicted spatio-temporal distribution data of the electric vehicle charging load to the power distribution network basic model in the form of node injection power, solving the operating state of the power distribution network under each scenario through power flow calculation, quantitatively evaluating the impact of the electric vehicle charging load on the key operating indicators of the power distribution network, and analyzing the carrying capacity of the power distribution network.

[0027] The key operating indicators include: Voltage deviation: calculate the deviation of each node voltage from the rated voltage in the distribution network, when the node voltage deviation exceeds the rated voltage or it is considered as out of limits; Line overload: calculate the load rate (actual current / allowable current) of each branch, when the line load rate continuously exceeds 100% it is considered as overload; System peak-valley difference rate: calculate the ratio of the difference between the maximum load and the minimum load to the maximum load, which is used to evaluate the volatility of the load.

[0028] By analyzing the changes of the above indicators before and after the predicted charging load is accessed, the influence of electric vehicle charging load on the carrying capacity of the distribution network is quantitatively evaluated.

[0029] In a preferred embodiment, the construction of the electric vehicle basic model includes: simulate different types of electric vehicles, including private cars, taxis and buses, and set different travel vehicle parameters for each type of vehicle; the vehicle parameters include battery capacity (unit: kWh), initial state of charge (SOC, range 0-1), unit mileage power consumption (unit: kWh / km), rated charging power (kW) under fast charging and slow charging; Based on the trip chain data, simulate the origin and destination (OD) of the vehicle, the path selection and the stay time distribution at the set points (such as residential areas, commercial areas and work sites) in the path; wherein the state of charge of the vehicle changes with time during driving and satisfies the following formula:

[0030] wherein, and represent the state of charge of the vehicle at time t and t+1 respectively; is the battery charging and discharging efficiency coefficient (usually , when driving represents energy loss); is the average power consumption per unit of vehicle driving distance (unit: kWh / km); is the driving distance of the vehicle in the interval (t, t+1) (unit: km), which is set according to dynamic traffic flow data; When the state of charge of the vehicle is lower than the preset charging threshold and is in a chargeable area, the charging demand is triggered.

[0031] In a preferred embodiment, the construction of the road network basic model includes: The road network is abstracted as an undirected graph wherein U represents a set of network nodes, including road intersections, key transportation hubs, and geographic locations of charging stations; W represents a set of edges connecting the nodes, representing road segments, and each edge is assigned a weight, which is obtained according to the actual distance, average travel time, or travel cost between nodes; The connection relationship and weight between nodes are recorded by using an adjacency matrix D, which has a dimension of If node i is directly connected to node j, then is the weight of the edge, otherwise it is 0 or infinity; Dijkstra algorithm is used to calculate the shortest distance between any two points in the road network, which provides a basis for subsequent traffic flow allocation and travel time estimation.

[0032] In a preferred embodiment, the construction of the charging station basic model includes: In the road network basic model, the charging station is modeled as a coupling node of the power distribution network and the road network; Each charging station node is associated with the following attributes in addition to the basic geographic location attribute: service capacity (such as the number of vehicles that can be served simultaneously, i.e., the number of charging piles); rated power of each charging pile; device utilization / S, wherein S is the sum of the rated powers of all charging piles in the charging station; and the number information of the access node of the power distribution network connected thereto, to establish a mapping relationship between the traffic network and the power distribution network; The key operating parameters of the charging station are defined, including: the number of charging piles, which is used to determine the service channel number of the system; the rated charging power of each charging pile; the device utilization; the charging service rate, which sets the charging service time of each vehicle to follow a negative exponential distribution, and the mean value is determined by the vehicle battery capacity, charging power, and initial state of charge.

[0033] In a preferred embodiment, the construction of the power distribution network basic model includes: The electrical connection relationship and parameters of the power distribution network are established by using the IEEE standard power system topology, the line impedance and transformer parameters are set based on standard values, and the power distribution network is coupled with the road network through geographic location information.

[0034] In a preferred embodiment, the generation of dynamic traffic flow data by the generative adversarial network includes: A multi-source data composed feature tensor is constructed, and the multi-source data composed feature tensor includes: a historical traffic flow feature matrix, the historical traffic flow feature matrix including traffic flows and speeds of each road section at the same time on N past days; an adjacency matrix of a road network basic model, used for introducing spatial structure information; a time period feature vector, the time period feature vector including a time of day, a weekday, an early peak, and a late peak; and weather condition data (such as a sunny day, a rainy day, a haze, or a specific meteorological index). The multi-source data composed feature tensor is input into a generator of a generative adversarial network; the generator adopts a neural network structure including a graph convolution layer and a full connection layer, and is used for learning a mapping of realistic traffic flow data from random noise and the multi-source data composed feature tensor. A discriminator of the generative adversarial network adopts a graph convolution network structure, and is used for discriminating authenticity of single node traffic flow data, and a difference (that is, a congestion propagation mode such as a cooperative change mode of flow between road network nodes) between generated traffic flow and real traffic flow in spatial correlation. A training process is trained by using a Wasserstein GAN framework, and an optimization target is to minimize a Wasserstein distance between a generated data distribution and a real data distribution, and a loss function is defined as follows:

[0035] wherein, an optimization target of the generator G is to minimize the entire expression, an optimization target of the discriminator is to maximize the entire expression, is a real traffic flow data distribution, is a data distribution output by the generator G, and D is a 1-Lipschitz function set, the discriminator D must be from a function set satisfying a 1-Lipschitz continuity condition, represents calculation of an expected value for all samples x sampled from the real data distribution , represents calculation of an expected value for all generated samples sampled from the generated data distribution , is a score of the discriminator D on the real sample x, is a score of the discriminator D on the generated sample ; through adversarial training, the generator can generate dynamic traffic flow data that retains real road network space-time features.

[0036] In a preferred embodiment, a traffic travel demand matrix is constructed based on an origin-destination analysis method, and combined with the dynamic traffic flow data and a basic model, the dynamic traffic flow data is superimposed on the road network basic model to obtain a traffic network, and a queuing theory algorithm is used to simulate the charging decision, queuing and charging service process of the electric vehicle in the charging station, including: Traffic travel demand matrix construction: based on historical traffic data, an origin-destination (OD) traffic travel demand matrix in different time periods (in hours) is constructed , which represents the number of vehicles traveling from origin o to destination d in period t; Traffic flow distribution and path selection: load the traffic travel demand matrix onto the road network basic model, obtain the dynamic link travel time from the dynamic traffic flow data, and use the stochastic user equilibrium path selection model to distribute the travel demand to each path in the road network, thereby obtaining the traffic flow, speed and travel time of each vehicle in the road network at different times in each time period; Charging behavior triggering and queuing simulation: when the simulated vehicle's SOC is lower than the threshold during driving, select a charging station based on distance, current queuing situation and charging price; the service process in the charging station is simulated using the M / M / C queuing model; Vehicle arrival process: set the vehicle arrival process at each charging station to follow a Poisson process with parameter λ, and the value of λ is dynamically determined by the number of vehicles arriving at the station; Service process: the service time (i.e. charging time) of each charging pile follows a negative exponential distribution with parameter μ, which is related to the charging power and the required charging amount; Obtain the charging request queue length, average waiting time, and the time of starting charging and the required charging amount of each charging station in different time periods.

[0037] In a preferred embodiment, the space-time graph attention network model includes a space attention module and a time attention module.

[0038] The space attention module adopts a multi-head graph attention mechanism, and the feature vector representation of each node i at the layer in the graph is:

[0039] wherein, is a concatenation operator of K vectors, K is the total number of attention heads, is a nonlinear activation function, and j ∑ i ∈N( ) is the summation operation of all nodes j in the neighbor set (including itself) of node i, a trainable weight matrix specific to the k-th attention head, input feature vector of the k-th layer for neighbor node j; l input feature vector of the k-th layer for neighbor node j; The temporal attention module adopts a Gated Recurrent Unit (GRU) to capture temporal dependencies. The input of the spatio-temporal graph attention network model is a spatio-temporal sequence tensor with dimensions of time steps x node number x features, where the features include predicted charging vehicle number, queue length, average speed of road segment, and time features of each charging station. The output of the spatio-temporal graph attention network model is the predicted charging load value (unit: kW) of each charging station in the future T time steps (for example, 24 hours in the future, with an interval of 15 minutes, a total of 96 time points). The mean square error is used as the loss function for supervised learning, and the formula is as follows:

[0040] wherein, is the true load value, is the model prediction value, and N is the number of samples, is all trainable parameters of the model, and the Adam optimizer is used to minimize the loss function.

[0041] Another embodiment, as shown in Figure 8 A power-traffic network charging load prediction system for implementing the power-traffic network charging load prediction method described above, comprising: a basic model construction module for constructing a basic model, wherein the basic model includes an electric vehicle basic model, a road network basic model, a charging station basic model, and a power distribution network basic model; a dynamic traffic flow generation module for generating dynamic traffic flow data based on the basic model and historical traffic data through a generative adversarial network; a charging demand simulation module for constructing a traffic trip demand matrix based on the origin-destination analysis method, and combining the dynamic traffic flow data and the basic model to simulate the charging decision, queuing waiting, and entire charging service process of electric vehicles in charging stations using a queuing theory algorithm; a charging load prediction module for constructing a spatio-temporal graph attention network model, taking the simulated data of the charging decision, queuing waiting, and entire charging service process of electric vehicles in charging stations, the road network basic model, and the charging station basic model as inputs, adaptively learning the spatial dependency strength between different nodes in the traffic network through a multi-head graph attention mechanism, and combining a Gated Recurrent Unit to capture the long-term regularity of load changes, to obtain the spatio-temporal coupling relationship between the traffic network state and the charging demand, and to predict the spatio-temporal distribution of electric vehicle charging load in a certain period of time in the future.

[0042] Preferably, the power-traffic network charging load prediction system can further comprise a power distribution network carrying capacity evaluation module, which is configured to call a power flow calculation method, receive the predicted charging load, evaluate the influence of the predicted charging load on the power distribution network (such as key operation indexes of power distribution network node voltage deviation, line overload, and system peak-valley difference rate), and obtain an analysis result.

[0043] The modules are sequentially connected to form a closed-loop process from basic data input to final power distribution network carrying capacity evaluation.

[0044] In a specific implementation, a typical area of a city is taken as a simulation object, which contains 37 road nodes, 15 of which correspond to charging stations, and an IEEE 33-node power distribution network system is taken as a power grid topology basis. The following steps are included: S1, a basic model construction step: respectively constructing basic models for describing electric vehicle characteristics, road network topology structure, charging station operation parameters, and power distribution network topology and electrical parameters, as input basis and data carrier of the entire prediction framework.

[0045] In this step, four types of basic models need to be constructed, including the following steps: S1.1, constructing an electric vehicle basic model: simulating three types of electric vehicles: private cars (70%), taxis (20%), and buses (10%). The parameter settings are as follows: Battery capacity: 60 kWh for private cars, 80 kWh for taxis, and 200 kWh for buses; Initial SOC: subject to uniform distribution U(0.4, 0.8); Unit mileage power consumption: 0.15-0.25 kWh / km; Charging power: fast charging 60 kW, slow charging 7 kW.

[0046] The state of charge of the vehicle changes with time during driving and satisfies the following formula:

[0047] Wherein, the charging and discharging efficiency coefficient 0.98 (charging) and 0.95 (driving), Output by the dynamic traffic flow generation step (S2).

[0048] When the vehicle SOC is less than or equal to 30%, the charging demand is triggered, and the charging station selection strategy is based on the shortest path algorithm.

[0049] S1.2, constructing a road network basic model: abstracting the road network as an undirected graph Where the node set U contains 37 nodes, of which 15 are charging station nodes, and the weight of the edge set W is the road link travel time (seconds). The shortest path between OD pairs is calculated using Dijkstra's algorithm, and the adjacency matrix D is generated from OpenStreetMap data.

[0050] In this embodiment, the road network of a typical area in a city is obtained and modeled, as shown in Figure 2 The blue part.

[0051] S1.3, build a charging station basic model: define the key operating parameters of the charging station, mainly including: Each charging station configuration: the number of fast charging piles is 4-6, and the number of slow charging piles is 6-15; the device utilization rate is calculated as / S, where is the total power actually used by the charging station at time t, and S is the sum of the rated power of all charging piles of the charging station.

[0052] The charging service time obeys a negative exponential distribution, and the mean μ is determined by the vehicle battery capacity, charging power and initial state of charge.

[0053] In addition, the geographical location attribute of the charging station needs to be configured, and the number information of the road access node connected thereto and the number information of the power distribution network access node connected thereto are configured according to the geographical location.

[0054] S1.4, build a power distribution network basic model: adopt the IEEE standard power system topology to establish the electrical connection relationship and parameters of the power distribution network for subsequent carrying capacity analysis.

[0055] In this embodiment, the IEEE 33-node standard power system topology is adopted, the line impedance and transformer parameters are set based on standard values, and the geographical location information is coupled with the road network, as shown in Figure 2 The black part, where b represents the bus and g represents the generator.

[0056] S2, dynamic traffic flow generation step: adopt Wasserstein generative adversarial network (WGAN) to generate dynamic traffic flow data that can reflect spatio-temporal randomness and uncertainty based on the road network model and historical traffic data.

[0057] In this step, the modeling of the Wasserstein generative adversarial network specifically includes the following steps: S2.1 Generator Input: Includes historical traffic flow data: average speed matrix of road segments at 15-minute intervals over the past 30 days (dimension 37 nodes × 96 time steps); road network adjacency matrix (dimension 37 nodes × 37 nodes); external features: time period (morning peak, evening peak), weather (sunny, rainy, snowy).

[0058] S2.2 Generator Structure: Contains a 3-layer Graph Convolutional Network (GCN), with each layer outputting dimensions of [64, 128, 256], using the ReLU activation function. It is used to learn a mapping from a feature tensor composed of random noise and multi-source data to realistic traffic flow data.

[0059] S2.3 Discriminator Structure: The WGAN-GP framework with gradient penalty is adopted. The discriminator is a 2-layer GCN with a hidden layer dimension of 128. A gradient penalty term (λ=10) is added to the loss function.

[0060] S2.4 Training process: The Wasserstein GAN (WGAN) framework is used for training, and its optimization objective is to minimize the Wasserstein distance between the generated data distribution and the real data distribution.

[0061] In this embodiment, the structure diagram of the WGAN model used to generate dynamic traffic flow is shown below. Figure 3 To verify the superiority of this method, a comparison chart of the performance of the WGAN method of this invention with other traffic flow simulation methods is shown below. Figure 4 Applying this method, the Wasserstein distance between the generated data and the real data decreased from the initial value of 1.52 to 0.31, and the spatial correlation coefficient (Pearson correlation coefficient) of the generated traffic flow reached 0.96.

[0062] S3. Charging demand simulation steps: Construct a traffic demand matrix based on origin-destination (OD) analysis and combine it with dynamic traffic flow data (to form a traffic network). Use queuing theory algorithms to simulate the entire process of electric vehicles' charging decisions, queuing, and charging services at charging stations.

[0063] In this step, an OD (Original Demand) matrix needs to be constructed based on historical traffic data, and travel demand needs are allocated to various paths on the road network. Charging behavior is triggered by factors such as the electric vehicle's SOC (State of Charge) status, simulating the entire charging process. The specific steps are as follows: S3.1 OD Demand Matrix Construction: Based on historical traffic data, generate the OD demand matrix for the morning peak (7:00-9:00), in which commuter traffic accounts for 65%.

[0064] S3.2 Traffic flow assignment and route selection: Use a stochastic user equilibrium model to assign OD demand to the road network, and calculate travel time by combining the speed data of dynamic traffic flow generated in step S2.

[0065] S3.3 Charging Behavior Triggering and Queue Simulation: When the vehicle's SOC is below the threshold, the nearest charging station is selected based on the actual queuing situation. The charging behavior is simulated using the M / M / C queuing model (C is the number of charging piles).

[0066] Vehicle arrival process: Assume that the process of a vehicle arriving at each charging station follows a Poisson process with parameter λ, the value of which is dynamically determined by the number of vehicles arriving at that station.

[0067] Service process: The service time (i.e., charging time) of each charging station follows a negative exponential distribution with parameter μ, which is related to the charging power and the amount of charging required.

[0068] This model is used to simulate the entire process of an electric vehicle arriving at a charging station, queuing, and receiving charging services. It outputs the charging request queue length, average waiting time, and the time and amount of charging required for each charging station at different times.

[0069] The transportation network consists of a static road network basic model and dynamic traffic flow superimposed on it. The interaction between the two determines the spatiotemporal distribution of electric vehicles in the road network, thus providing a basis for charging load prediction.

[0070] S4. Charging load prediction step: Construct a spatiotemporal graph attention network (ST-GAT) model, taking the output of the charging demand simulation step, road network topology and charging station information as input, and predicting the spatiotemporal distribution of electric vehicle charging load in a certain period of time by capturing the complex spatiotemporal coupling relationship between traffic network status and charging demand.

[0071] This step, ST-GAT model construction and training, specifically includes the following steps: S4.1 ST-GAT Model Construction: The spatiotemporal graph attention network model consists of a spatial attention module and a temporal attention module.

[0072] Spatial Attention Module: Employs a multi-head graph attention mechanism, focusing on the attention of each node i in the graph at the first... The feature vector of a layer is represented as:

[0073] in, This is the concatenation operator for K vectors, where K is the total number of attention heads. For nonlinear activation functions, ∑ j ∈N( i The expression () represents the summation operation on all nodes j in the neighbor set (including itself) of node i. These are the normalized attention weight coefficients, calculated from the k-th attention head, representing the importance of node j to node i. Let be the trainable weight matrix specific to the k-th attention head. Let be the input feature vector of neighbor node j at layer l.

[0074] Time Attention Module: Uses Gated Cyclic Units (GRUs) with a hidden layer dimension of 64 to capture daily periodicity and trends.

[0075] S4.2 Model Input and Output: The model input is a spatiotemporal sequence tensor with dimensions of (96 time steps × 15 charging stations × 5 features), where the features include the number of charging vehicles, queue length, average road speed, time period encoding, and weather encoding. The model output is the load curve (unit: kW) of each charging station for the next 24 hours (96 time points at 15-minute intervals).

[0076] S4.3 Model Training: Supervised learning is performed using mean squared error (MSE) as the loss function, as shown in the following formula:

[0077] in, This is the actual load value. These are the model predictions, where N is the number of samples. These are all trainable parameters of the model, and the Adam optimizer is used to minimize this loss function. The optimizer's initial learning rate is set to 0.001, and the number of training epochs is set to 200.

[0078] In this embodiment, the ST-GAT model structure used for charging load prediction is shown below. Figure 5 To verify the superiority of this method, a comparison chart of the performance of the ST-GAT of this invention with other prediction models is shown below. Figure 6 The average MSE of this method is 32.43 kW, which is better than other methods.

[0079] S5. Distribution network carrying capacity assessment steps: The predicted spatiotemporal distribution data of electric vehicle charging load is input into the distribution network topology model in the form of node injected power. By performing power flow calculation, the impact of electric vehicle charging load on key operating indicators of the distribution network is quantitatively assessed, thereby analyzing the carrying capacity of the distribution network.

[0080] In this step, the predicted charging load of each charging station at each future time needs to be injected into the corresponding access node of the distribution network. The forward-backward substitution method or the Newton-Raphson method is used to calculate the power flow and solve the operating state of the distribution network under each scenario.

[0081] Key performance indicators include: S5.1 Voltage Deviation: Calculate the deviation between the voltage at each node in the distribution network and the rated voltage. When the node voltage deviation exceeds the rated voltage... or This is considered exceeding the limit; S5.2 Line overload: Calculate the load rate (actual current / allowable current carrying capacity) of each branch. When the line load rate continuously exceeds 100%, it is considered an overload. S5.3 System Peak-Valley Difference Rate: The ratio of the difference between the daily maximum load and the minimum load to the maximum load is used to assess load fluctuation.

[0082] By analyzing the changes of the above indicators before and after the connection of predicted charging load, the impact of electric vehicle charging load on the carrying capacity of the distribution network is quantitatively assessed.

[0083] In this embodiment, the output node voltage exceedance status is determined based on the calculation results of the charging load carrying capacity of the distribution network. (See...) Figure 7 .

[0084] In another embodiment, a computer storage medium stores a computer program thereon, which, when executed by a computer, implements the power-transportation network charging load forecasting method described in any of the preceding embodiments.

[0085] The specific implementation method is the one described above, and will not be repeated here.

[0086] In another embodiment, an electronic device includes a memory and a processor, wherein the memory stores a computer program, and the processor runs the computer program stored in the memory, wherein the computer program, when executed, implements the power-transportation network charging load forecasting method described in any of the preceding embodiments.

[0087] The specific implementation method is the one described above, and will not be repeated here.

[0088] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for predicting charging load in a power-transportation network, characterized in that, Includes the following steps: Construct basic models, which include basic models of electric vehicles, road networks, charging stations, and power distribution networks; Based on the basic model and historical traffic data, dynamic traffic flow data is generated through generative adversarial networks. Based on the origin-end point analysis method, a traffic demand matrix is ​​constructed. Combined with the dynamic traffic flow data and the basic model, the dynamic traffic flow data is superimposed on the road network basic model to obtain the traffic network. The queuing theory algorithm is used to simulate the entire process of electric vehicles charging decision-making, queuing, and charging service at charging stations. A spatiotemporal graph attention network model is constructed. The model takes as input data the entire process of electric vehicle charging decision-making, queuing, and charging service in charging stations, as well as basic road network and charging station models. The model adaptively learns the spatial dependency strength between different nodes in the traffic network through a multi-head graph attention mechanism. Combined with gated cyclic units to capture the long-term pattern of load changes, the spatiotemporal coupling relationship between traffic network state and charging demand is obtained, and the spatiotemporal distribution of electric vehicle charging load in a certain period of the future is predicted.

2. The power-transportation network charging load forecasting method according to claim 1, characterized in that, After predicting the spatiotemporal distribution of electric vehicle charging load over a certain period of time, the following steps are also included: The predicted spatiotemporal distribution data of electric vehicle charging load is input into the basic model of the distribution network in the form of node injected power. By performing power flow calculations, the operating status of the distribution network under various scenarios is solved, the impact of electric vehicle charging load on key operating indicators of the distribution network is quantitatively evaluated, and the carrying capacity of the distribution network is analyzed.

3. The power-transportation network charging load forecasting method according to claim 1, characterized in that, Building a basic model for electric vehicles includes: The simulation covers different types of electric vehicles, including private cars, taxis, and buses, and sets different travel vehicle parameters for each type of vehicle. These vehicle parameters include battery capacity, initial state of charge, energy consumption per unit distance, and rated charging power under fast and slow charging conditions. Based on travel chain data, the origin and destination of vehicles, route selection, and the distribution of dwell time at designated points along the route are simulated. The state of charge of a vehicle during operation changes with time according to the following formula: in, and These represent the vehicle's state of charge at times t and t+1, respectively. The battery charge / discharge efficiency coefficient; This refers to the average power consumption per unit distance traveled by the vehicle. This represents the distance traveled by a vehicle within the interval (t, t+1), and its value is set based on dynamic traffic flow data. When the vehicle's state of charge is below the preset charging threshold and it is in a rechargeable area, a charging request is triggered.

4. The power-transportation network charging load forecasting method according to claim 1, characterized in that, Constructing a basic road network model includes: Abstract the road network as an undirected graph. , where U represents the set of network nodes, including the geographical locations of road intersections, key transportation hubs, and charging stations; W represents the set of edges connecting these nodes, representing road segments, with each edge assigned a weight based on the actual distance between nodes, average travel time, or travel cost; The adjacency matrix D records the connection relationships and weights between nodes. If node i is directly connected to node j, then... The weight of the edge is 0 or infinity if it is not. Dijkstra's algorithm is used to calculate the shortest distance between any two points in the road network, providing a basis for subsequent traffic flow allocation and travel time estimation.

5. The power-transportation network charging load forecasting method according to claim 1, characterized in that, The basic model for building a charging station includes: In the basic model of the road network, the charging station is modeled as a coupled node between the power distribution network and the road network; In addition to the basic geographical location attribute, each charging station node is also associated with the following attributes: service capacity; rated power of each charging pile; and equipment utilization rate. / S, where Let t be the total power actually used by the charging station, S be the sum of the rated power of all charging piles in the charging station; and let S be the numbering information of the distribution network access node connected to it, to establish the mapping relationship between the transportation network and the distribution network. Define the key operating parameters of the charging station, including: the number of charging piles, used to determine the number of service channels in the system; the rated charging power of each charging pile; the equipment utilization rate; and the charging service rate, which sets the charging service time for each vehicle to follow a negative exponential distribution, with the mean determined by the vehicle's battery capacity, charging power, and initial state of charge.

6. The power-transportation network charging load forecasting method according to claim 5, characterized in that, Constructing a basic model of a power distribution network includes: The electrical connection relationships and parameters of the distribution network are established by adopting the IEEE standard power system topology, and the line impedance and transformer parameters are set based on standard values. The network is coupled with the road network through geographical location information.

7. The power-transportation network charging load forecasting method according to claim 1, characterized in that, Generating dynamic traffic flow data through generative adversarial networks includes: A feature tensor composed of multi-source data is constructed. The feature tensor composed of multi-source data includes: a historical traffic flow feature matrix, which includes the traffic flow and speed of each road segment at the same time in the past N days; an adjacency matrix of the road network basic model, which is used to introduce spatial structure information; a time period feature vector, which includes the time of day, weekdays, morning peak, and evening peak; and weather condition data. The feature tensor composed of multi-source data is input into the generator of the generative adversarial network; the generator adopts a neural network structure containing graph convolutional layers and fully connected layers to learn a mapping of realistic traffic flow data from the feature tensor composed of random noise and multi-source data. The discriminator of the generative adversarial network adopts a graph convolutional network structure to distinguish the authenticity of traffic flow data of individual nodes and the difference between generated traffic flow and real traffic flow in spatial correlation. The training process uses the Wasserstein GAN framework, and its optimization objective is to minimize the Wasserstein distance between the generated data distribution and the real data distribution. The loss function is defined as follows: in, This indicates that the optimization objective of generator G is to minimize the entire expression. The goal of the discriminator is to maximize the entire expression. It is the distribution of real traffic flow data. G is the data distribution output by the generator, and D is the set of 1-Lipschitz functions. The discriminator D must come from a set of functions that satisfy the 1-Lipschitz continuity condition. This indicates that the calculation is performed for all data from the true data distribution. The expected value of sample x from the sampled sample. This indicates that the computation is performed for all data distributions generated. Mid-sample generation Expected value It is the score given by the discriminator D to the real sample x. The discriminator D is used to generate samples The scoring; through adversarial training, the generator is able to produce dynamic traffic flow data that retains the spatiotemporal characteristics of the real road network.

8. The power-transportation network charging load forecasting method according to claim 1, characterized in that, The queuing theory algorithm is used to simulate the entire process of electric vehicles' charging decision-making, queuing, and charging service at a charging station, including: Construction of Transportation Demand Matrix: Based on historical traffic data, construct a transportation demand matrix for origin and destination at different time periods. , representing the number of vehicles traveling from the starting point o to the ending point d within the time period t; Traffic flow assignment and route selection: The traffic demand matrix is ​​loaded onto the basic road network model. Dynamic road segment travel time is obtained from dynamic traffic flow data. Using a stochastic user equilibrium route selection model, travel demand is assigned to each path in the road network, thereby obtaining the traffic flow, speed and vehicle travel time at different times in the road network. Charging behavior triggering and queuing simulation: When the SOC of the simulated vehicle is lower than the threshold during driving, the charging station is selected based on distance, current queuing status, and charging price; the service process within the charging station is simulated using the M / M / C queuing model. Vehicle arrival process: The process of a vehicle arriving at each charging station is set to follow a Poisson process with parameter λ, the value of which is dynamically determined by the number of vehicles arriving at that station. Service process: The service time of each charging station follows a negative exponential distribution with parameter μ, where μ is related to the charging power and the amount of charging required; The system obtains the charging request queue length, average waiting time, and the final start time and required charging amount for each charging station at different time periods.

9. The power-transportation network charging load forecasting method according to claim 1, characterized in that, The spatiotemporal graph attention network model includes a spatial attention module and a temporal attention module. The spatial attention module employs a multi-head graph attention mechanism, which performs attention on each node i in the graph at the first... The feature vector of a layer is represented as: in, This is the concatenation operator for K vectors, where K is the total number of attention heads. For nonlinear activation functions, ∑ j ∈N( i The expression () represents the summation operation on all nodes j in the neighbor set of node i. These are the normalized attention weight coefficients, calculated from the k-th attention head, representing the importance of node j to node i. Let be the trainable weight matrix specific to the k-th attention head. For neighbor node j in the th... l The input feature vector of the layer; The time attention module uses gated loop units to capture time dependencies; The input to the spatiotemporal graph attention network model is a spatiotemporal sequence tensor with dimensions of time step × number of nodes × features. The features include the predicted number of charging vehicles at each charging station, queue length, average speed of the road segment, and time features. The output of the spatiotemporal graph attention network model is the predicted charging load value of each charging station in the next T time steps. Supervised learning uses mean squared error as the loss function, as shown in the following formula: in, This is the actual load value. These are the model predictions, where N is the number of samples. These are all the trainable parameters of the model, and the loss function is minimized using the Adam optimizer.

10. A power-transportation network charging load forecasting system, characterized in that, The method for predicting charging load in a power-transportation network as described in any one of claims 1-9 includes: The basic model building module constructs basic models, which include basic models of electric vehicles, basic models of road networks, basic models of charging stations, and basic models of power distribution networks. The dynamic traffic flow generation module generates dynamic traffic flow data based on the base model and historical traffic data through a generative adversarial network. The charging demand simulation module constructs a traffic travel demand matrix based on the origin-end point analysis method. It combines the dynamic traffic flow data and the basic model, overlays the dynamic traffic flow data onto the road network basic model to obtain the traffic network, and uses queuing theory algorithm to simulate the entire process of electric vehicles charging decision-making, queuing, and charging service at charging stations. The charging load prediction module constructs a spatiotemporal graph attention network model. It takes data from the entire process of electric vehicles' charging decisions, queuing, and charging services at charging stations, as well as basic road network and charging station models, as inputs. Through a multi-head graph attention mechanism, it adaptively learns the spatial dependency strength between different nodes in the traffic network and combines it with gated cyclic units to capture long-term patterns of load changes. This yields the spatiotemporal coupling relationship between traffic network status and charging demand, and predicts the spatiotemporal distribution of electric vehicle charging load over a certain period in the future.

11. A computer storage medium having a computer program stored thereon, characterized in that, When the computer executes the computer program, it implements the power-transportation network charging load prediction method according to any one of claims 1-9.

12. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor runs the computer program stored in the memory. When the computer program is executed, it implements the power-transportation network charging load forecasting method according to any one of claims 1-9.

Images