Vehicle network interaction flow optimization method based on information and physical fusion

Abstract

The invention discloses a vehicle network interaction flow optimization method based on information and physical fusion, and belongs to the technical field of intelligent traffic and energy internet crossing. Comprising the steps that a vehicle network is divided into a plurality of grid areas, and each grid area comprises physical layer operation data and information layer operation data; constructing a vehicle network diagram structure, and calculating the schedulable potential of the EV in the stay duration of the charging station; according to the schedulable potential and the information layer operation data of the EV in the stay time of the charging station, the method aims at minimizing the peak-valley difference, maximizing the renewable energy consumption, the vehicle network interaction income and the operation income. Constructing an objective function of the multi-objective flow optimization model, an EV charging and discharging power constraint, an EV battery charge state constraint, a charging pile capacity constraint and a power grid safety constraint; and solving the multi-target traffic optimization model to obtain a vehicle network interaction traffic optimization result. According to the method, the prediction error of the EV arrival time is reduced, and the V2G flow regulation and control are refined.

Description

A Vehicle-to-Network Interaction Traffic Optimization Method Based on Information and Physical Fusion Technical Field

[0001] This invention relates to the field of intelligent transportation and energy internet technology, and in particular to a method for optimizing vehicle-to-grid (V2G) traffic flow based on information and physical fusion. Background Technology

[0002] Vehicle-to-grid (V2G) technology serves as a crucial link between electric vehicles (EVs) and smart grids (SGs). It enables bidirectional regulation of EV battery energy storage resources, which is of great significance for mitigating peak-valley differences in the power grid, promoting the consumption of renewable energy, and enhancing grid stability.

[0003] However, current V2G traffic optimization strategies have the following shortcomings:

[0004] Information and physical processes are disconnected: Existing optimization focuses on a single dimension such as grid-side load demand or user-side revenue, failing to fully integrate physical world information such as vehicle travel behavior and road traffic conditions with information world data such as grid operation status and market electricity price signals, resulting in deviations between optimization results and actual operating conditions.

[0005] Insufficient forecasting accuracy and dynamic response: The forecasting accuracy for EV user travel demand, renewable energy output, and real-time grid load is limited, and there is a lack of rapid response mechanisms to dynamic changes, making it difficult to cope with the complex and ever-changing vehicle-to-grid (V2G) environment. For example, if the forecasting of EV access times is based solely on historical data without considering real-time traffic congestion causing user return delays, it may lead to the failure of V2G scheduling plans and even affect user travel.

[0006] Challenges in multi-objective collaborative optimization: V2G involves multiple stakeholders, including grid operators, EV users, and aggregators. Existing strategies need to improve their collaborative optimization capabilities when balancing multiple objectives such as grid safety and stability, user travel convenience and economy, and aggregator revenue. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a vehicle-to-grid traffic optimization method based on information and physical fusion. This method can solve the technical problems of insufficient integration of information and physical processes, low prediction accuracy, slow dynamic response, and poor multi-objective collaborative optimization effect in existing V2G traffic optimization.

[0008] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0009] On the one hand, this invention provides a method for optimizing vehicle-to-grid (V2G) traffic based on information and physical fusion, including:

[0010] The vehicle network is divided into multiple grid regions, each of which includes physical layer operational data and information layer operational data.

[0011] Using the EV location in the physical layer operation data as the node feature of the graph structure and the traffic congestion index as the edge weight of the graph structure, a vehicle network graph structure is constructed for each grid region.

[0012] By fusing the spatiotemporal characteristics of the power grid load in the vehicle network diagram structure and physical layer operation data, regional-level EV activity thermal correlation characteristics with power grid load are obtained.

[0013] Based on the correlation characteristics between regional EV activity heat and grid load, the dispatchable potential of EVs during their stay at charging stations is obtained.

[0014] Based on the dispatchable potential of EVs during their stay at charging stations and information layer operation data, with the goal of minimizing peak-valley differences and maximizing renewable energy consumption, vehicle-grid interaction benefits, and operational benefits, a multi-objective flow optimization model is constructed, which includes objective functions for EV charging and discharging power, EV battery state of charge, charging pile capacity, and grid security constraints.

[0015] Solve the multi-objective traffic optimization model to obtain the vehicle-to-network interaction traffic optimization results.

[0016] Optionally, the physical layer operation data includes grid-side operation data, vehicle-side operation data, and traffic-side operation data, while the information layer operation data includes user travel plans, willingness to participate in vehicle-to-grid interaction and expected benefits, vehicle-to-grid interaction service prices released by aggregators, grid auxiliary service demand signals, and market signals.

[0017] Optionally, the formula for calculating the dispatchable potential of the EV during its stay at the charging station is as follows:

[0018] ;

[0019] ;

[0020] ;

[0021] in, These represent the maximum discharge capacity and maximum recharge capacity of the k-th EV, respectively. This represents the rated battery capacity of the k-th EV. This indicates the time when the k-th EV arrives at the charging station. The state of charge; Indicates the target state of charge of the k-th EV; These represent battery discharge efficiency and battery charging efficiency, respectively. This represents the dispatchable potential of the k-th EV at time t; Let represent the maximum discharge power and maximum charging power of the kth EV at time t, respectively; These represent the start and end times of the discharge for the k-th EV, respectively. These represent the start and end times of charging for the k-th EV, respectively.

[0022] Optionally, the objective function of the multi-objective traffic optimization model is expressed as:

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] in, These represent minimization and maximization, respectively. These represent peak-valley difference, renewable energy consumption, vehicle-to-grid interaction revenue, and operating revenue, respectively. This represents the base load of the power grid at time t; This represents the total vehicle-to-network interaction traffic at time t; This represents the renewable energy consumption capacity at time t; Indicates the total number of time periods; This indicates that the vehicle generates revenue by discharging electricity to the outside world via V2G; This indicates that the vehicle is being charged via V2G and the charging fee has been paid. This represents the charging / discharging power of the k-th EV at time t; Let represent the discharge price and the charging price at time t, respectively; Indicates the length of the time period; This represents the ancillary service power provided to the grid by the aggregator at time t; The price of the ancillary service at time t; This represents the discharge subsidy or acquisition cost paid by the aggregator to the user at time t.

[0028] Optionally, the EV charging and discharging power constraint is expressed as:

[0029] ;

[0030] The EV battery state of charge constraint is expressed as follows:

[0031] ;

[0032] ;

[0033] The capacity constraint of the charging pile is expressed as follows:

[0034] ;

[0035] The power grid security constraints are expressed as follows:

[0036] or ;

[0037] ;

[0038] in, This represents the charging / discharging power of the k-th EV at time t; Let represent the minimum charging / discharging power and maximum charging / discharging efficiency of the k-th EV at time t, respectively. Indicates the time of the kth EV The state of charge; Indicates the target state of charge of the k-th EV; Indicates the permissible error range for the nuclear power plant's status; Let represent the state of charge of the k-th EV at time t and time t+1, respectively; Indicates the length of the time period; This represents the rated battery capacity of the k-th EV. These represent battery discharge efficiency and battery charging efficiency, respectively. Let represent the discharge power and charging power of the k-th EV at time t, respectively; Indicates charging station The capacity; This represents the real-time current of the l-th line; This indicates the maximum current that the l-th line can withstand; This represents the apparent power of the l-th line at time t; This represents the maximum apparent power of the l-th line; This represents the voltage of the i-th node at time t; These represent the lower limit of the voltage at node i and the upper limit of the voltage at node i, respectively.

[0039] Optional, also includes:

[0040] Based on the vehicle-to-network interaction traffic optimization results, calculate the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic, as well as the EV's execution deviation;

[0041] Based on the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic, the execution deviation of EVs, and the preset total power deviation threshold, the physical layer operation data and information layer operation data in the vehicle-to-network are continuously optimized to obtain the optimized operation data of the vehicle-to-network.

[0042] Optionally, the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic is expressed as:

[0043] ;

[0044] The execution deviation of the EV is expressed as:

[0045] ;

[0046] in, This represents the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic at time t. This represents the total vehicle-to-network interaction traffic at time t; This represents the total planned vehicle-to-everything (V2X) traffic at time t. This indicates the execution deviation of the k-th EV at time t; This represents the charging / discharging power of the k-th EV at time t; The value represents the actual charging / discharging power of the k-th EV at time t.

[0047] Optionally, based on the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic, the execution deviation of the EV, and a preset total power deviation threshold, rolling optimization is performed on the physical layer operation data and information layer operation data in the vehicle-to-network to obtain optimized vehicle-to-network operation data, including:

[0048] like , or If the physical layer and information layer operation data in the vehicle network are optimized, then the optimized operation data of the vehicle network will be obtained by rolling optimization; otherwise, the physical layer and information layer operation data in the vehicle network will continue to be used.

[0049] in, This represents the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic at time t. This indicates the execution deviation of the k-th EV at time t; This indicates the time it takes for the k-th EV to arrive at the charging station; This indicates the preset total power deviation threshold; This indicates the preset delay time threshold.

[0050] Secondly, the present invention provides a vehicle-to-grid (V2G) traffic optimization device based on information and physical fusion, comprising:

[0051] The region division module is used to divide the vehicle network into multiple grid regions, each grid region including physical layer operation data and information layer operation data;

[0052] The graph structure construction module is used to: use the EV location in the physical layer runtime data as the node feature of the graph structure and the traffic congestion index as the edge weight of the graph structure to construct the vehicle network graph structure for each grid region.

[0053] The feature fusion module is used to: perform spatiotemporal feature fusion of the power grid load in the vehicle network diagram structure and physical layer operation data to obtain the regional-level EV activity thermal and power grid load correlation features;

[0054] The schedulable calculation module is used to: obtain the schedulable potential of EVs during their stay at charging stations based on the correlation characteristics between regional EV activity heat and grid load;

[0055] The model building module is used to: construct the objective function of a multi-objective flow optimization model based on the schedulable potential of EVs during their stay at charging stations and information layer operation data, with the goal of minimizing peak-valley differences and maximizing renewable energy consumption, vehicle-grid interaction benefits, and operational benefits; as well as EV charging and discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid security constraints.

[0056] The model solving module is used to solve the multi-objective traffic optimization model and obtain the vehicle-to-network interaction traffic optimization results.

[0057] Thirdly, the present invention provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implements the steps of the vehicle-to-grid traffic optimization method based on information and physical fusion described in the first aspect.

[0058] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0059] This invention reduces the prediction error of EV arrival time by deeply integrating information and physical data with spatiotemporal features. The refined V2G traffic control can reduce the peak-valley difference of the power grid in the pilot area, fully consider the travel needs of users, and maximize benefits while ensuring travel. The multi-objective optimization of vehicle-to-grid interaction traffic balances the interests of power grid operators, EV users, and aggregators. For example, aggregators can obtain additional revenue by providing ancillary services, users can obtain economic returns by participating in V2G, and the power grid can obtain peak-shaving resources and renewable energy consumption channels. Attached Figure Description

[0060] Figure 1 shows a flowchart of one embodiment of the vehicle-to-grid traffic optimization method based on information and physical fusion of the present invention. Detailed Implementation

[0061] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0062] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0063] Example 1

[0064] As shown in Figure 1, this embodiment introduces a vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion, including the following steps:

[0065] Step 1: Cyber-physical multi-source data sensing and acquisition, specifically:

[0066] Real-time collection of physical layer and information layer operation data of the vehicle-to-grid network. Physical layer operation data includes grid-side operation data, vehicle-side operation data, and traffic-side operation data. Information layer operation data includes user travel plans, vehicle-to-grid (V2G) participation intentions and expected benefits, V2G service prices released by aggregators, grid ancillary service demand signals, and market signals.

[0067] Among them, the grid-side operation data includes real-time active load, reactive load, and node voltage of each feeder area collected through the Supervisory Control and Data Acquisition (SCADA) system, with a sampling frequency of ≤5 seconds / time; as well as the time-of-use electricity price signal (unit: yuan / kWh) and the output forecast curve of renewable energy such as wind power / photovoltaic (unit: kW) for the next 24 hours obtained through the Energy Management System (EMS).

[0068] Vehicle-side operational data includes the real-time latitude and longitude of the vehicle's location reported by the on-board diagnostics (OBD) or telematics box (T-BOX) of the electric vehicle (EV), the original sampling frequency, the average value after spatiotemporal filtering, the current state of charge (SOC) (accuracy ±2%), and the battery health status; the planned charging time window and target SOC (in %) submitted by the user's app or on-board system, combined with historical behavior analysis to predict demand curves, with an update frequency of ≤15 minutes / time; and the extraction of the EV user's origin and destination points (OD points), route sequence, and average trip duration over the past 30 days from the cloud database, with a sampling interval of 5 minutes, for behavioral pattern modeling and prediction.

[0069] Traffic-side operational data includes real-time data from road sensors or floating cars acquired through the city's traffic management center, namely, average road speed (unit: km / h) and traffic flow (unit: veh / h), with a sampling frequency of ≤1 minute / time and an accuracy of ±5%; congestion levels for each area are calculated based on historical and real-time data, with values ​​ranging from 0 to 10 (0 indicating smooth traffic and 10 indicating severe congestion), and updates are pushed out in real-time via a mobile app, with a frequency of ≤5 minutes / time; as well as parking lot / charging station location and capacity: integrating geographic information system data, including charging station coordinates, number of available charging piles, rated power (unit: kW), and idle status, with an update frequency of ≤10 minutes / time, supporting dynamic reservation and occupancy prediction.

[0070] The user's travel plan for the kth EV was collected via the user's app. ,in These represent the departure time and arrival time of the k-th EV, respectively. Let represent the starting point coordinates and the destination coordinates of the kth EV, respectively.

[0071] Collect the V2G participation intention of the kth EV at time t through user app or in-vehicle system. This flag is a binary variable (1 indicates willingness to participate, 0 indicates refusal), and is dynamically predicted by combining user historical preference data (such as participation frequency and time period preference); the expected return of the k-th EV at time t. V2G service pricing based on the time released by the aggregator. (Unit: Yuan / kWh) and the available battery capacity of the user in the kth EV at time t. (Unit: kWh) Calculation, the calculation formula is as follows ,in, The charge / discharge efficiency coefficient (default value 0.9) should be updated at a frequency of ≤10 minutes / time.

[0072] V2G service pricing released by aggregator t Due to the influence of grid load, electricity pricing policies and real-time market supply and demand, the time resolution should be ≤5 minutes, and a geographically differentiated pricing strategy should be supported.

[0073] Power grid ancillary service demand signal at time t This includes demand types, such as frequency regulation, voltage support, and the time period for grid ancillary service demand. Target power at time t (Unit: kW) and priority level, the signal update frequency should be ≤ 5 minutes / time, and broadcast to all connected EV users through the vehicle network; These represent the start time and end time of the demand for grid ancillary services, respectively.

[0074] Market signals, namely V2G aggregator platforms releasing V2G ancillary service pricing, such as peak-shaving discharge rates at time t. Rotational spare price at time t wait.

[0075] Step 2: Cyber-physical data preprocessing and fusion, specifically:

[0076] The collected physical layer operation data and information layer operation data are cleaned, standardized, and completed before preprocessing. The deep learning spatiotemporal fusion model adopts a graph neural network algorithm based on long short-term memory or a spatiotemporal fusion model similar to spatiotemporal graph convolutional network to mine the hidden correlation between EV travel behavior, traffic conditions and power grid load.

[0077] First, Kalman filtering is used to smooth the physical layer operation data and the information layer operation data;

[0078] Then, for the missing SOC data, a time-series prediction model based on the Long Short-Term Memory (LSTM) network is used to complete it;

[0079] Finally, data of different dimensions, such as electricity price and congestion index, are normalized to the [0,1] interval.

[0080] The vehicle network is divided into multiple N×N grid regions, each of which includes physical layer operation data and information layer operation data;

[0081] By using the EV locations in the physical layer operational data as node features of the graph structure and the traffic congestion index as edge weights of the graph structure, a vehicle network graph structure is constructed for each grid region.

[0082] Spatio-Temporal Graph Convolutional Networks (STGCN) are used to fuse the spatio-temporal features of the power grid load in the vehicle network graph structure and physical layer operation data. This minimizes the weighted sum of reconstruction and prediction errors between the fused data and each source data. The temporal convolutional layer of STGCN captures the evolution of features in each region over time, while the graph convolutional layer captures the spatial dependencies between regions. Finally, the fused regional-level EV activity thermal and power grid load correlation features are output.

[0083] Step 3: Predict the schedulable potential of EVs, specifically:

[0084] By utilizing the correlation characteristics between regional EV activity heat and power grid load, and based on the constructed EV travel behavior prediction model, the dwell time of EVs at charging stations is obtained.

[0085] Input the user's historical travel trajectory sequence for the kth EV The current location of the kth EV User travel plan for the kth EV Regional-level EV activity thermodynamic characteristics at time t Characteristics related to power grid load Output the time when the kth EV arrives at the charging station. (Error controlled within ±15 minutes); This represents the historical travel trajectory of the user of the k-th EV. The m represents the historical travel trajectory of the user of the k-th EV;

[0086] The dwell time of the kth EV at the charging station is then: ,in The time it takes for the kth EV to leave the charging station.

[0087] Based on the EV's dwell time at the charging station, current SOC, battery characteristics, and target SOC, the dispatchable potential of the EV within its dwell time at the charging station is obtained:

[0088] The maximum discharge capacity of the kth EV Represented as:

[0089] ;

[0090] The discharge period is ;

[0091] The maximum rechargeable capacity of the kth EV Represented as:

[0092] ;

[0093] The charging period is ;

[0094] The dispatchable potential of each EV is represented as a time-varying power matrix or vector, indicating the maximum charge / discharge power available at time t, and the dispatchable potential of the k-th EV at time t. Represented as:

[0095] ;

[0096] in, This indicates the time when the k-th EV arrives at the charging station. The state of charge; Indicates the target state of charge of the k-th EV; These represent the maximum discharge power and maximum charging power of the kth EV at time t, respectively, which are limited by the battery rate and the charging pile power. These represent the start and end times of the discharge for the k-th EV, respectively. These represent the start and end times of charging for the k-th EV, respectively. Positive power indicates vehicle-to-grid (V2G) discharge, while negative power indicates grid-to-vehicle (G2V) charging.

[0097] Step 4: Multi-objective V2G traffic optimization decision, specifically:

[0098] The decision variable is the charging / discharging power of the k-th EV at time t (e.g., 15 minutes as a time period). ,in Indicates discharge. This indicates that the device is charging.

[0099] Based on the dispatchable potential of EVs during their stay at charging stations and information layer operation data, a multi-objective traffic optimization model is constructed with the goal of minimizing peak-valley differences and maximizing renewable energy consumption, vehicle-to-grid interaction benefits, and operational benefits.

[0100] The objective function of the multi-objective traffic optimization model is expressed as:

[0101] ;

[0102] ;

[0103] ;

[0104] ;

[0105] in, These represent minimization and maximization, respectively. These represent peak-valley difference, renewable energy consumption, vehicle-to-grid interaction revenue, and operating revenue, respectively. This represents the base load of the power grid at time t; This represents the total vehicle-to-grid interaction traffic at time t, with discharge being positive and charging being negative. Indicates the number of EVs; This represents the renewable energy consumption capacity at time t; This indicates the total number of time periods (e.g., 24h / 0.25h = 96 time periods). This indicates that the vehicle generates revenue by discharging electricity to the outside world via V2G; This indicates that the vehicle is being charged via V2G and the charging fee has been paid. This represents the charging / discharging power of the k-th EV at time t; These represent the discharge price (including subsidies) and the charging price at time t, respectively. Indicates the length of the time period; This represents the ancillary service power provided to the grid by the aggregator at time t; The price of the ancillary service at time t; This represents the discharge subsidy or acquisition cost paid by the aggregator to the user at time t.

[0106] Based on the schedulable potential of EVs during their stay at charging stations and information layer operation data, a multi-objective flow optimization model is constructed, including EV charging and discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid security constraints.

[0107] The EV charging and discharging power constraint is expressed as:

[0108] ;

[0109] The EV battery state of charge constraint is expressed as:

[0110] ;

[0111] ;

[0112] The capacity constraint of charging piles is expressed as:

[0113] ;

[0114] Power grid security constraints are expressed as follows:

[0115] or ;

[0116] ;

[0117] in, Let represent the minimum charging / discharging power and maximum charging / discharging efficiency of the k-th EV at time t, respectively, and let represent the dispatchable potential of the k-th EV at time t. Decide; Indicates the time of the kth EV The state of charge; This indicates the permissible error range of the nuclear power plant's state of matter, representing the maximum deviation of the actual state of charge (SOC) from the target SOC. Let represent the state of charge of the k-th EV at time t and time t+1, respectively; Let represent the discharge power and charging power of the k-th EV at time t, respectively; Indicates charging station The capacity; This represents the real-time current of the l-th line; This indicates the maximum current that the l-th line can withstand; This represents the apparent power of the l-th line at time t; This represents the maximum apparent power of the l-th line; This represents the voltage of the i-th node at time t; These represent the lower limit of the voltage at node i and the upper limit of the voltage at node i, respectively.

[0118] An improved Non-dominated Sorting Genetic Algorithm III (NSGA-III) is employed to solve the multi-objective traffic optimization model. An elite retention strategy is used to select a compromise solution from the Pareto optimal set as the scheduling scheme, yielding the optimized vehicle-to-network traffic flow. .

[0119] Step 5: Dynamic execution and feedback adjustments, specifically:

[0120] By continuously monitoring actual performance and system status changes, when the deviation exceeds a preset threshold, a rolling optimization mechanism is triggered to re-initiate data fusion, prediction, and optimization decisions.

[0121] Optimize the results of vehicle network interaction traffic The data is sent to the corresponding EV charging pile or on-board V2G control unit to obtain the actual charging / discharging power of the k-th EV at the time t uploaded by the charging pile in real time. , as well as the updated SOC and location information uploaded by the vehicle terminal;

[0122] Based on the vehicle-to-network (V2N) interaction traffic optimization results, calculate the deviation between the actual total V2N interaction traffic and the planned total V2N interaction traffic, as well as the EV execution deviation. Represented as:

[0123] ;

[0124] EV execution deviation Represented as:

[0125] ;

[0126] like , or If the physical layer and information layer operation data in the vehicle network are optimized, then the optimized operation data of the vehicle network will be obtained by rolling optimization; otherwise, the physical layer and information layer operation data in the vehicle network will continue to be used.

[0127] in, This indicates the time it takes for the k-th EV to arrive at the charging station; This indicates the preset total power deviation threshold; This indicates the preset delay time threshold.

[0128] Based on the latest information layer and physical layer operation data, new vehicle-to-network interaction traffic optimization results are generated and dynamic closed-loop adjustments are achieved.

[0129] Example 2

[0130] Based on Example 1, this example introduces an experimental example of a vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion:

[0131] Taking a city-level V2G network including 1,000 EVs, 5 20kV feeder areas, several public charging stations, and private charging piles as an example:

[0132] Step 1: Obtain physical layer operation data and information layer operation data, specifically:

[0133] Physical layer:

[0134] The power grid side collects real-time active / reactive load and node voltage of each feeder through the distribution SCADA system every 5 seconds; and obtains the time-of-use electricity price signal and wind / solar power output prediction curve for the next 24 hours through the EMS system.

[0135] The vehicle reports its GPS coordinates in real time via the EV's onboard terminal once per second. After processing, the 5-minute average, current SOC, and SOH are calculated. The system also collects the user's travel plan for the next day via the user's APP, such as "depart from community A to company B at 8:00 am tomorrow and return at 6:00 pm".

[0136] The traffic management department obtains real-time congestion indices and average vehicle speeds on major road sections from the city's traffic management bureau's intelligent transportation platform.

[0137] Information layer: V2G aggregator platforms recently released V2G ancillary service pricing, including peak-shaving electricity prices and standby capacity prices.

[0138] Step 2: Data preprocessing and fusion, specifically:

[0139] Kalman filtering was used to denoise the GPS drift data; a time-series prediction model based on LSTM was used to complete the missing SOC data; data such as electricity price and congestion index were normalized to the [0,1] interval; the EV location in the physical layer operation data was used as the node feature of the graph structure and the traffic congestion index was used as the edge weight of the graph structure to construct the vehicle network graph structure for each grid area; a spatiotemporal graph convolutional network was used to fuse the spatiotemporal features of the vehicle network graph structure and the power grid load in the physical layer operation data to obtain the regional-level EV activity heat and power grid load correlation features.

[0140] Step 3: Predict the EV schedulable potential based on fused data, specifically as follows:

[0141] For each EV, its historical travel trajectory, current location, user-inputted departure / return time, and fused regional EV activity heat map and grid load correlation characteristics are used as inputs. Based on the constructed EV travel behavior prediction model, the precise time of its arrival at the charging station and its expected stay duration are predicted, with the error controlled within ±15 minutes.

[0142] Based on the predicted dwell time, current SOC, battery rated capacity, and user-set target SOC (e.g., 80% SOC upon returning home), the maximum discharge capacity and acceptable minimum charging capacity, along with their time distribution, of the EV during the dwell time can be calculated. For example, an EV with a current SOC of 60%, a battery capacity of 50kWh, and a planned return home at 18:00, remaining there until 7:00 the next day with a target SOC of 80%, has a dispatchable potential of up to 10kWh of discharge capacity between 19:00 and 23:00, and up to 10kWh of charging capacity between 2:00 and 6:00 AM.

[0143] Step 4: Multi-objective V2G traffic optimization decision, specifically:

[0144] Based on the dispatchable potential of EVs within the dwell time, and with the goals of minimizing peak-valley difference and maximizing renewable energy consumption, vehicle-grid interaction benefits, and operational benefits, a multi-objective flow optimization model is constructed, along with EV charging and discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid security constraints of the multi-objective flow optimization model.

[0145] Under the constraints, the power flow of each feeder shall not exceed 90% of its rated capacity; the node voltage shall be maintained between 0.95 and 1.05 pu; the charging and discharging power of a single charging pile shall not exceed its rated power; and the charging and discharging power of an EV shall not exceed the allowable charging and discharging rate of its battery.

[0146] An improved NSGA-III algorithm is adopted, with a population size of 100, 50 iterations, a crossover probability of 0.8, and a mutation probability of 0.1. An elite retention strategy is used to select a compromise solution from the Pareto optimal set as the scheduling scheme.

[0147] Step 5: Dynamic execution and feedback adjustments, specifically:

[0148] After an EV arrives at a charging station, the charging pile performs charging and discharging according to the dispatch instructions. If an EV is expected to arrive 2 hours late due to a sudden traffic congestion, the on-board terminal will upload the updated estimated arrival time to the fusion center. If the system detects that the actual dispatchable window of the EV deviates from the prediction by more than 1 hour (a preset threshold), it will immediately trigger rolling optimization. Based on the latest status of all EVs and grid data, it will re-predict the dispatchable potential of EVs, the dispatchable capacity and time period, and make multi-objective V2G traffic optimization decisions, generating new dispatch instructions for issuance.

[0149] Compared with traditional optimization methods, this embodiment improves the accuracy of EV dispatchable potential prediction by 15-20%, reduces the peak-valley difference of the power grid by 8-12%, increases the average V2G revenue of users by 10-15%, and maintains user travel satisfaction above 95%.

[0150] Example 3

[0151] Based on Embodiment 1 or 2, this embodiment introduces a vehicle-to-grid (V2G) traffic optimization device based on information and physical fusion, comprising:

[0152] The region division module is used to divide the vehicle network into multiple grid regions, each grid region including physical layer operation data and information layer operation data;

[0153] The graph structure construction module is used to: use the EV location in the physical layer runtime data as the node feature of the graph structure and the traffic congestion index as the edge weight of the graph structure to construct the vehicle network graph structure for each grid region.

[0154] The feature fusion module is used to: perform spatiotemporal feature fusion of the power grid load in the vehicle network diagram structure and physical layer operation data to obtain the regional-level EV activity thermal and power grid load correlation features;

[0155] The schedulable calculation module is used to: obtain the schedulable potential of EVs during their stay at charging stations based on the correlation characteristics between regional EV activity heat and grid load;

[0156] The model building module is used to: construct the objective function of a multi-objective flow optimization model based on the schedulable potential of EVs during their stay at charging stations and information layer operation data, with the goal of minimizing peak-valley differences and maximizing renewable energy consumption, vehicle-grid interaction benefits, and operational benefits; as well as EV charging and discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid security constraints.

[0157] The model solving module is used to solve the multi-objective traffic optimization model and obtain the vehicle-to-network interaction traffic optimization results.

[0158] The specific functions of each module described above are explained in the relevant content of the method in Embodiment 1, and will not be repeated here.

[0159] Example 4

[0160] This embodiment introduces a computer-readable storage medium storing a computer program / instruction, which, when executed by a processor, implements the steps of the vehicle-to-grid traffic optimization method based on information and physical fusion as described in Embodiment 1 or 2.

[0161] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0162] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0163] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0164] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0165] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A method for optimizing vehicle-to-grid (V2G) traffic interaction based on information and physical fusion, characterized in that, include: The vehicle network is divided into multiple grid regions, each of which includes physical layer operational data and information layer operational data. Using EV locations from physical layer operational data as node features and traffic congestion index as edge weights, a vehicle-to-grid (V2G) graph structure is constructed for each grid region. The V2G graph structure and the grid load from the physical layer operational data are then fused spatiotemporally to obtain regional-level EV activity heatmap and grid load correlation characteristics. Based on these regional-level EV activity heatmap and grid load correlation characteristics, the dispatchable potential of EVs within their dwell time at charging stations is obtained. Based on the dispatchable potential of EVs within their dwell time at charging stations and the information layer operational data, with the objectives of minimizing peak-valley differences and maximizing renewable energy consumption, V2G interaction benefits, and operational benefits, a multi-objective flow optimization model is constructed, including objective functions, EV charging / discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid safety constraints. Solving the multi-objective flow optimization model yields the V2G interaction flow optimization results.

2. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 1, characterized in that, The physical layer operation data includes grid-side operation data, vehicle-side operation data, and traffic-side operation data. The information layer operation data includes user travel plans, willingness to participate in vehicle-to-grid interaction and expected benefits, vehicle-to-grid interaction service prices released by aggregators, grid auxiliary service demand signals, and market signals.

3. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 1, characterized in that, The formula for calculating the dispatchable potential of the EV during its stay at the charging station is as follows: ； ； ;in, These represent the maximum discharge capacity and maximum recharge capacity of the k-th EV, respectively. This represents the rated battery capacity of the k-th EV. This indicates the time when the k-th EV arrives at the charging station. The state of charge; Indicates the target state of charge of the k-th EV; These represent battery discharge efficiency and battery charging efficiency, respectively. This represents the dispatchable potential of the k-th EV at time t; Let represent the maximum discharge power and maximum charging power of the kth EV at time t, respectively; These represent the start and end times of the discharge for the k-th EV, respectively. These represent the start and end times of charging for the k-th EV, respectively.

4. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 1, characterized in that, The objective function of the multi-objective traffic optimization model is expressed as: ； ； ； ;in, These represent minimization and maximization, respectively. These represent peak-valley difference, renewable energy consumption, vehicle-to-grid interaction revenue, and operating revenue, respectively. This represents the base load of the power grid at time t; This represents the total vehicle-to-network interaction traffic at time t; This represents the renewable energy consumption capacity at time t; Indicates the total number of time periods; This indicates that the vehicle generates revenue by discharging electricity to the outside world via V2G; This indicates that the vehicle is being charged via V2G and the charging fee has been paid. This represents the charging / discharging power of the k-th EV at time t; Let represent the discharge price and the charging price at time t, respectively; Indicates the length of the time period; This represents the ancillary service power provided to the grid by the aggregator at time t; The price of the ancillary service at time t; This represents the discharge subsidy or acquisition cost paid by the aggregator to the user at time t.

5. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 1, characterized in that, The EV charging and discharging power constraint is expressed as follows: The EV battery state-of-charge constraint is expressed as follows: ； The capacity constraint of the charging pile is expressed as: The power grid security constraints are expressed as follows: or ； ;in, This represents the charging / discharging power of the k-th EV at time t; Let represent the minimum charging / discharging power and maximum charging / discharging efficiency of the k-th EV at time t, respectively. Indicates the time of the kth EV The state of charge; Indicates the target state of charge of the k-th EV; Indicates the permissible error range for the nuclear power plant's status; Let represent the state of charge of the k-th EV at time t and time t+1, respectively; Indicates the length of the time period; This represents the rated battery capacity of the k-th EV. These represent battery discharge efficiency and battery charging efficiency, respectively. Let represent the discharge power and charging power of the k-th EV at time t, respectively; Indicates charging station The capacity; This represents the real-time current of the l-th line; This indicates the maximum current that the l-th line can withstand; This represents the apparent power of the l-th line at time t; This represents the maximum apparent power of the l-th line; This represents the voltage of the i-th node at time t; These represent the lower limit of the voltage at node i and the upper limit of the voltage at node i, respectively.

6. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 1, characterized in that, Also includes: Based on the vehicle-to-network interaction traffic optimization results, calculate the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic, as well as the EV's execution deviation; Based on the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic, the execution deviation of EVs, and the preset total power deviation threshold, the physical layer operation data and information layer operation data in the vehicle-to-network are continuously optimized to obtain the optimized operation data of the vehicle-to-network.

7. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 6, characterized in that, The deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic is expressed as follows: The execution deviation of the EV is expressed as: ;in, This represents the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic at time t. This represents the total vehicle-to-network interaction traffic at time t; This represents the total planned vehicle-to-everything (V2X) traffic at time t. This indicates the execution deviation of the k-th EV at time t; This represents the charging / discharging power of the k-th EV at time t; The value represents the actual charging / discharging power of the k-th EV at time t.

8. The vehicle-to-grid (V2G) traffic optimization method based on information and physical fusion according to claim 6, characterized in that, Based on the deviation between the actual and planned total vehicle-to-network (V2N) interaction traffic, the EV's execution deviation, and the preset total power deviation threshold, rolling optimization is performed on the physical layer and information layer operational data in the V2N network to obtain optimized V2N network operational data, including: if 、 or If the physical layer and information layer operational data in the vehicle network are optimized, then the optimized operational data for the vehicle network is obtained through rolling optimization; otherwise, the physical layer and information layer operational data in the vehicle network continue to be used. This represents the deviation between the actual total vehicle-to-network interaction traffic and the planned total vehicle-to-network interaction traffic at time t. This indicates the execution deviation of the k-th EV at time t; This indicates the time it takes for the k-th EV to arrive at the charging station; This indicates the preset total power deviation threshold; This indicates the preset delay time threshold.

9. A vehicle-to-grid (V2G) traffic optimization device based on information and physical fusion, characterized in that, include: The region division module is used to divide the vehicle network into multiple grid regions, each grid region including physical layer operation data and information layer operation data; The graph structure construction module is used to: use the EV location in the physical layer runtime data as the node feature of the graph structure and the traffic congestion index as the edge weight of the graph structure to construct the vehicle network graph structure for each grid region. The feature fusion module is used to: perform spatiotemporal feature fusion of the power grid load in the vehicle network diagram structure and physical layer operation data to obtain the regional-level EV activity thermal and power grid load correlation features; The schedulable calculation module is used to: obtain the schedulable potential of EVs within the dwell time of charging stations based on the correlation characteristics between regional EV activity heat and grid load; the model building module is used to: construct the objective function of a multi-objective flow optimization model based on the schedulable potential of EVs within the dwell time of charging stations and information layer operation data, with the goal of minimizing peak-valley difference and maximizing renewable energy consumption, vehicle-grid interaction benefits, and operational benefits, as well as EV charging and discharging power constraints, EV battery state of charge constraints, charging pile capacity constraints, and grid security constraints. The model solving module is used to solve the multi-objective traffic optimization model and obtain the vehicle-to-network interaction traffic optimization results.

10. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the vehicle-to-grid traffic optimization method based on information and physical fusion as described in any one of claims 1-8.

Images