A path planning method and system for a new energy vehicle

Abstract

The application relates to the technical field of path planning, in particular to a path planning method and system for a new energy vehicle. The method comprises the following steps: acquiring the shortest distance from a starting point to an ending point and a chargeable node set in a path through a vehicle-mounted panoramic map of the new energy vehicle, estimating a theoretical driving distance, and simultaneously acquiring real-time road condition information; then, based on the real-time road condition information, simulating and deducing invalid energy consumption output in a congestion state, quantifying an energy consumption increment output ratio of road geometric characteristics, combining the invalid energy consumption output and the energy consumption increment output ratio to evaluate a composite index of energy consumption loss; finally, judging a limit driving radius according to the composite index of energy consumption loss and the theoretical driving distance, optimizing the chargeable node set based on the limit driving radius, and outputting path planning data. The application optimizes the path planning technology, so that the path planning technology is more accurate.

Description

Technical Field

[0001] This invention relates to the field of path planning technology, and in particular to a path planning method and system for new energy vehicles. Background Technology

[0002] Compared to traditional gasoline-powered vehicles, new energy vehicles not only have significant advantages in reducing carbon emissions and air pollution, but also excel in energy efficiency. However, the widespread adoption of new energy vehicles still faces some challenges, one of which is the limitation of driving range. Although the driving range of new energy vehicles is gradually increasing with the continuous development of battery technology, the limitations of battery capacity and the imperfect layout of charging infrastructure mean that energy consumption and charging issues during long-distance travel remain key factors influencing consumers' purchasing decisions. Therefore, optimizing route planning for new energy vehicles and rationally selecting charging stations have become important issues for improving the user experience of new energy vehicles.

[0003] In summary, traditional route planning methods for new energy vehicles still rely solely on mileage prediction to determine mileage, without fully considering actual road conditions and traffic congestion, resulting in low efficiency in route planning. Summary of the Invention

[0004] Therefore, it is necessary to provide a path planning method and system for new energy vehicles to solve at least one of the above-mentioned technical problems.

[0005] To achieve the above objectives, a path planning method for new energy vehicles is provided, the method comprising the following steps:

[0006] Step S1: Obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path using the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance.

[0007] Step S2: Simulate and deduce the ineffective energy consumption output of the congested state based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; evaluate the composite index of energy consumption loss based on the ineffective energy consumption output and the energy consumption increment output ratio to form the composite index of energy consumption loss.

[0008] Step S3: Determine the maximum driving radius of the new energy vehicle based on the composite energy loss index and the theoretical driving distance; make an optimal decision on the set of charging nodes based on the maximum driving radius, and output the path planning data.

[0009] Preferably, the present invention also provides a path planning system for new energy vehicles, used to execute the path planning method for new energy vehicles as described above, the path planning system for new energy vehicles comprising:

[0010] The road condition information acquisition module is used to obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path through the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance.

[0011] The energy loss analysis module is used to simulate and deduce the ineffective energy consumption output of congested conditions based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; and evaluate the energy loss composite index based on the ineffective energy consumption output and the energy consumption increment output ratio to form the energy loss composite index.

[0012] The path planning output module is used to determine the limit driving radius of the new energy vehicle based on the composite index of energy loss and the theoretical driving distance; to make an optimal decision on the set of charging nodes based on the limit driving radius, and to output the path planning data.

[0013] The beneficial effects of this invention are as follows: by obtaining the shortest distance path and the set of rechargeable nodes within it through an in-vehicle panoramic map, a complete basic information framework can be built at the initial stage of path planning, enabling the system to have a clear route structure and potential energy replenishment locations from the outset. Simultaneously, by estimating the theoretical driving distance through the shortest distance and obtaining real-time traffic information within that range, the planning process no longer relies on static maps but fully utilizes dynamic traffic data, improving the adaptability of path planning to real road environments and avoiding range prediction deviations caused by traffic changes, thus improving the accuracy and reliability of overall path planning from the source. By deriving ineffective energy consumption output under congested conditions through real-time traffic information, the system can accurately reflect the additional energy consumption generated by the vehicle under non-ideal operating conditions such as low speed and congestion, thereby avoiding overly idealistic energy estimations in traditional planning. Furthermore, by quantifying the incremental impact of road geometric features (such as slope, curvature, speed limits, etc.) on energy consumption, a more refined energy consumption prediction model can be achieved, making energy consumption estimation no longer solely dependent on driving speed and distance. By jointly evaluating ineffective energy consumption and geometric incremental energy consumption to form a composite energy loss index, the energy consumption risk of vehicles in actual roads can be comprehensively reflected, effectively improving the accuracy of energy consumption prediction and providing a more reliable basis for subsequent range assessment and charging node selection. Determining the limit driving radius of new energy vehicles based on the composite energy loss index and theoretical driving distance allows the system to assess the vehicle's reachable space in a way that more closely approximates real-world energy consumption performance, thus avoiding the risk of insufficient battery power due to inadequate energy consumption prediction in traditional methods. Optimizing the set of charging nodes based on this limit driving radius effectively selects the most suitable charging nodes, avoiding unnecessary detours and repeated charging, and ensuring that the vehicle completes its planned journey within its safe range. The final generated route planning data balances driving efficiency, energy safety, and charging convenience, providing new energy vehicle users with a more stable, intelligent, and energy-efficient travel solution. Therefore, this invention is an optimization of a traditional route planning method for new energy vehicles. It solves the problem that the traditional route planning method for new energy vehicles relies solely on a single mileage to predict the driving range, without fully considering actual road conditions and traffic congestion, resulting in low route planning efficiency. This invention strengthens the consideration of actual road conditions and traffic congestion, thereby improving the efficiency of route planning. Attached Figure Description

[0014] Figure 1 A flowchart illustrating the steps of a route planning method for an energy vehicle;

[0015] Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S2.

[0016] Figure 3 for Figure 1A detailed flowchart illustrating the implementation steps of step S3. Detailed Implementation

[0017] Please see Figure 1 A route planning method for an energy vehicle, the method comprising the following steps:

[0018] Step S1: Obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path using the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance.

[0019] In this embodiment of the invention, the remaining battery power of the new energy vehicle is first obtained. This remaining battery power is read in real time through the vehicle's battery management system. For example, if the total battery capacity of the new energy vehicle is 60 kWh and the remaining battery power percentage is 75%, then the remaining battery power is 45 kWh. Using the vehicle's onboard panoramic map, the shortest distance from the starting point to the destination and the set of charging nodes along the shortest path are obtained. The onboard panoramic map uses Dijkstra's algorithm or A* pathfinding algorithm to calculate the shortest path from the starting point to the destination. For example, if the user sets the starting point to city A and the destination to city B, the algorithm calculates the shortest distance to be 150 kilometers. Simultaneously, the location information of all charging stations along the path is extracted to form a set of charging nodes. This set of charging nodes includes three charging nodes: charging station A (30 kilometers from the starting point), charging station B (70 kilometers from the starting point), and charging station C (120 kilometers from the starting point). It should be noted that the theoretical driving distance of the new energy vehicle is estimated based on the remaining battery power and the shortest distance. This theoretical driving distance is calculated based on the vehicle's energy consumption per 100 kilometers calibration value. The specific calculation formula is: Theoretical driving distance = Remaining battery power ÷ Energy consumption per 100 kilometers calibration value × 100, where the remaining battery power represents the available battery power of the vehicle, in kilowatt-hours (kWh); the energy consumption per 100 kilometers calibration value represents the electricity consumed by the vehicle to travel 100 kilometers under standard operating conditions, in kilowatt-hours per 100 kilometers. For example, when the remaining battery power is 45 kWh and the energy consumption per 100 kilometers calibration value is 15 kWh per 100 kilometers, the theoretical driving distance is 45 divided by 15 and then multiplied by 100, resulting in a theoretical driving distance of 300 kilometers. Real-time traffic information is obtained from the theoretical driving distance through an in-vehicle panoramic map. This real-time traffic information is obtained through data interaction between the in-vehicle panoramic map and the traffic management cloud platform. The real-time traffic information includes traffic congestion status of each road segment, road gradient information, road curve information, and intersection information.

[0020] Step S2: Simulate and deduce the ineffective energy consumption output of the congested state based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; evaluate the composite index of energy consumption loss based on the ineffective energy consumption output and the energy consumption increment output ratio to form the composite index of energy consumption loss.

[0021] In this embodiment of the invention, after obtaining real-time traffic information, road segments marked as low-speed traffic and key traffic congestion nodes are first extracted from the path. The average start-stop frequency of congested segments is analyzed using a time-series analysis method with a fixed sliding window width. This frequency is mapped to the rated torque and speed range of the vehicle's drive motor to determine the decreasing gradient of motor energy conversion efficiency under start-stop conditions. Efficiency sample data within 30 minutes is recorded using a time series method, and the actual trend of motor energy conversion efficiency decreasing over time is obtained using a weighted difference method. Based on the trend curve, the additional motor energy consumption in that road segment is calculated. Then, the real-time traffic information is extracted... The system includes road geometric feature data, sampling the slope, radius of curvature, and intersection density of each road segment. A discrete weighted cumulative method is used to calculate the incremental contribution ratio of road geometric features to vehicle energy consumption. For slopes exceeding 5 degrees, the weighting coefficients are increased in the weight matrix. For turning sections with a radius of curvature less than 7.2m, the incremental ratio is adjusted according to turning frequency and centrifugal force components. After all real-time energy consumption factors are calculated, the output of ineffective energy consumption and the incremental energy consumption output ratio of each road segment are unified. A weighted clustering algorithm is then used to calculate the composite energy consumption loss index, which is obtained by weighting the statistical average ineffective energy consumption and the geometric incremental energy consumption. The energy loss composite index, obtained after normalization, describes the comprehensive energy loss characteristics that vary with time and space along the path. This output is used for the next step of determining the maximum driving radius. For example, in a 100km path, real-time traffic information shows that the section from 20km to 35km is highly congested, with an average speed of 12km / h, a traffic flow of 650 vehicles / h, and an average start-stop frequency of 5 times / min in the congested area. By collecting sample data of the drive motor's output efficiency at a start-stop frequency of 5 times / min, the motor efficiency decrease gradient is found to be -0.15 / kWh. This is then calculated using time-weighted regression. The additional energy loss in this section was calculated to be 2.4 kWh. Simultaneously, the average slope of 4°, range of 14°, turning radius of 5m, and intersection density of 2 / km were extracted from road geometry information. The weighting coefficient for road sections with slopes higher than 10° was set to 1.3, and the weighting coefficient for road sections with turning radii less than 7.2m was set to 1.2. The energy consumption increment ratio was quantified to obtain an output ratio of 1.18 related to road geometry features. Finally, the additional energy loss under congestion and the output ratio of energy consumption increment based on geometric features were normalized through aggregation to obtain a composite energy loss index of 0.92, which was used for subsequent determination of the maximum driving radius.

[0022] Step S3: Determine the maximum driving radius of the new energy vehicle based on the composite energy loss index and the theoretical driving distance; make an optimal decision on the set of charging nodes based on the maximum driving radius, and output the path planning data.

[0023] In this embodiment of the invention, after obtaining the composite energy loss index, the theoretical driving distance is first corrected based on the index. The correction coefficient is set to the reciprocal of the composite energy loss index. The theoretical distance is multiplied by the correction coefficient to obtain the limit driving radius. This limit driving radius represents the maximum sustainable driving range of the vehicle under real-time traffic conditions and road geometric constraints. Then, using this radius as the search area radius, spatial domain filtering is performed on the previously generated set of charging nodes. The filtering process uses a spherical distance algorithm to calculate the geographic straight-line distance from each charging node to the vehicle's current location. All nodes with a distance less than or equal to the limit driving radius are included in the secondary optimization set. Then, data interaction is performed with the charging station cloud management system through the vehicle network interface to receive the real-time status of each charging node. The parameters include the number of currently available charging piles, the output power of a single pile, the connection compatibility protocol, and the estimated queuing time. A multi-objective minimization optimization algorithm is used to calculate the comprehensive priority value with the shortest total time and the maximum power matching degree as the objective functions. All nodes are sorted and the node with the highest comprehensive priority value is selected as the target charging node. On this basis, the path planning subsystem is called again to perform a secondary path redraw. The A* heuristic path algorithm is used in combination with the corrected energy consumption weight to regenerate the complete path information from the current vehicle location through the target charging node to the destination. The path planning results are output as a path planning data file in the form of a geographic coordinate sequence and a mileage interval cost table. This data file is used for the dynamic path command generation and actual driving path control of the vehicle's onboard navigation system.

[0024] Step S1 includes the following steps:

[0025] Step S11: Obtain the current remaining battery power of the new energy vehicle;

[0026] Step S12: Obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest path using the onboard panoramic map of the new energy vehicle;

[0027] Step S13: Estimate the theoretical driving distance of the new energy vehicle based on the remaining battery power and the shortest distance.

[0028] Step S14: Obtain real-time road condition information for the theoretical driving distance using the in-vehicle panoramic map.

[0029] In this embodiment of the invention, after the new energy vehicle is started, the battery management system outputs the current state of charge (SOC) value of the battery by sampling voltage, current, and remaining capacity signals. The instantaneous current is collected by the milliampere metering unit and the energy change per unit time is integrated and statistically analyzed. The remaining power data in percentage form is obtained by comparing the cumulative output power with the battery's rated capacity. The sampling period is set to 1 second. When the battery's nominal capacity is 75 kWh, the current remaining power is detected to be 42%. The value is transmitted to the energy consumption calculation and control unit. Under the arrangement of control instructions, the value is written into the memory identifier area. When it is called, the information is numbered according to the sequence number, including timestamp, battery temperature, voltage distribution curve, and single cell balance coefficient. All parameters are stored and converted into the basic numerical source for calculation. In the power acquisition and calculation process, the current sampling accuracy is not less than 0.05A and the voltage measurement accuracy is not less than 0.01V to ensure the accuracy of the subsequent theoretical driving distance assessment.

[0030] The system reads the latitude and longitude data of the starting and ending points through the path analysis interface of the vehicle-mounted panoramic map system, and constructs a path grid index in the high-precision map database. The Dijkstra shortest path search algorithm is used, with the inverse ratio of road mileage and average travel speed as the single-sided weight. Weighted minimum cost calculation is performed on each connected node. After traversal, the path with the minimum total distance is output. This path is determined as the shortest path after the path re-examination mechanism verifies the consistency of the weights. During the path scanning stage, the charging facility data interface is called to extract all charging pile nodes with DC output power greater than 60kW within the path. Nodes with charging interfaces that conform to the national standard GB / T20234 are selected to form a set of charging nodes. This set of nodes contains charging station name, latitude and longitude coordinates, charging power, output voltage range and real-time operating status information, which is cached in the form of structured data to provide basic node coordinate information for subsequent path planning.

[0031] The remaining power obtained in step S11 is input into the calculation process through the energy consumption calculation formula node. It is matched with the average energy consumption value of 15.2kWh / 100km under the rated load of 1.8t and constant speed of 90km / h. The theoretical driving distance is obtained after conversion. During the data processing, the temperature compensation factor and battery degradation correction factor are introduced. When the detected ambient temperature is 25℃ and the battery degradation rate is 4%, the theoretical driving distance is corrected according to the compensation parameters. The calculation result is theoretical driving distance = 75kWh×42% / (15.2kWh / 100km)×(1-0.04). In the programming process, each value is stored in floating point format and the unit is set to km to form the theoretical driving distance data output, ensuring that the output data can be directly read and called by the path planning main process.

[0032] The system calls upon the real-time traffic information interface of the in-vehicle panoramic map to collect road condition data for the road sections covered by the theoretical driving distance. The collection range is set to all networked road nodes within a theoretical distance of 115km. The system obtains real-time dynamic information from the traffic flow data center, including six dimensions: average vehicle speed, traffic density, weather type, road slipperiness coefficient, traffic light cycle, and lane occupancy rate. The data update interval is 5 minutes, and the network transmission rate is no less than 10Mbps. After receiving the data, the system uses the road ID as an index to build a weighted array table of traffic flow characteristics such as the real-time vehicle speed ratio, the rate of missed vehicle passage, and the proportion of large vehicles for each road segment. Then, it combines the local congestion index collected by radar sensors to confirm the current road congestion level. The real-time traffic information of each road segment is stored in the path analysis cache in the form of a matrix structure and time stamp is retained as the set of input variables for subsequent energy consumption simulation derivation.

[0033] refer to Figure 2 The aforementioned step S2 includes the following steps:

[0034] Step S21: Extract congested road section information and road slope and intersection structure information from real-time traffic information;

[0035] In this embodiment of the invention, traffic information data streams are invoked during the real-time traffic data synchronization cycle to perform data indexing and retrieval on the entire road segment of the target path. Road segments with vehicle speeds below 20 km / h and durations exceeding 180 seconds are extracted as congested road segment identification results. Simultaneously, slope and road intersection structure data within the same geographical area are extracted from the 3D map terrain information. The slope is calculated by the ratio of the elevation difference between the two road nodes to the horizontal projection distance, with a calculation accuracy maintained at 0.1°. Road intersection structure data is extracted from the intersection topology stored in the map vector layer, recording the number of road connections, lane distribution, traffic light control cycle, and traffic flow distribution ratio at each intersection point. The above data are uniformly classified into the congested road segment information set and the road slope and intersection structure information set. A structured record is established for each road segment with fields such as timestamp, latitude and longitude, slope value, intersection type, vehicle flow, and average vehicle speed. During data processing, a dynamic real-time information table is formed with a data update cycle of 5 minutes to serve as the input basis for the subsequent energy consumption calculation stage.

[0036] Step S22: Simulate and deduce the additional energy loss under congestion conditions based on congested road segment information;

[0037] In this embodiment of the invention, after extracting congested road segment information, energy consumption loss is calculated for each congested segment based on the sampled congestion mileage, average vehicle speed, and number of start-stop cycles. Assuming a road segment is 3.2km long, the average vehicle speed is 10km / h, and the vehicle start-stop frequency is 5 times / min, the energy output efficiency reduction rate of the drive motor in this frequency range is -0.14 / kWh as the calculation basis. The energy consumption increase is calculated based on the time series statistics of vehicle speed fluctuation amplitude and motor output power change. At the same time, the power output status of auxiliary equipment is collected: air conditioning system output power 1.2kW, vehicle infotainment system 0.3kW, lighting system 0.15kW. Within the average congestion travel time of 19.2min, the additional equipment energy output is 0.0256kWh. The additional energy consumption loss in congestion is obtained by adding the equipment energy consumption and drive energy loss. The data, including road segment number, energy loss, number of start-stop cycles, and traffic density, are recorded in a numerical table for the next step of nonlinear energy consumption fitting processing.

[0038] Step S23: Perform nonlinear energy loss fitting on the additional energy loss to obtain nonlinear energy loss fitting data;

[0039] In this embodiment of the invention, after the calculation in S22 is completed, the energy loss of each congested section is arranged in time series and an energy loss curve is constructed. 600 sampling points are formed with a sampling time interval of 1 second. The sampling sequence is subjected to a second difference operation to reflect the progressive nonlinear change trend of energy consumption over time. Then, the gradient of the energy consumption growth rate is calculated by the least squares regression method. After continuous time smoothing of the gradient, a smooth curve sequence is output. Piecewise exponential regression is performed on the curve to identify the attenuation intensity of energy conversion efficiency in different time periods. The relative change rate of adjacent time moments is extracted from the attenuation sequence and the average time intensity is calculated. Nonlinear energy loss fitting data is output and a data field structure is established. The fields include energy consumption gradient, fitting accuracy, time window index, and cumulative energy loss, which are used for synchronous analysis with the geometric feature increment ratio.

[0040] Step S24: Based on the road slope and intersection structure information, quantify the energy consumption increment output ratio of road geometric features to obtain the energy consumption increment output ratio associated with road geometric features;

[0041] In this embodiment of the invention, based on the road slope and intersection structure information extracted in S21, the slope and turning radius data measured for each slope segment are quantitatively calculated. The slope range is set to 0° to 12°, and the turning radius range is set to 5.8m to 7.2cm. For sections with a slope exceeding 5°, an energy consumption increment correction is performed using a linear weighting coefficient of 1.2. For sections with a slope exceeding 10°, an energy consumption increment correction is performed using a weighting coefficient of 1.5. At the same time, a coefficient of 1.1 is added to curved roads with a turning radius of less than 7.2m. For nodes with intersection structures containing three or more connecting roads, the waiting signal loss time is calculated based on the vehicle flow distribution ratio. The signal loss time is multiplied by the vehicle idling power consumption of 0.9kW to obtain the additional energy consumption value. This value is used as the additional energy consumption of intersections and weighted into the total road segment energy consumption. In the full path analysis, 100m is used as the smallest segment unit. The slope, curvature, and intersection factor energy consumption of all road segments are superimposed and averaged to form the road geometric feature-related energy consumption increment output ratio.

[0042] Step S25: Evaluate the composite index of energy loss based on the nonlinear energy loss fitting data and the energy loss increment output ratio to form the composite index of energy loss.

[0043] In this embodiment of the invention, the nonlinear energy loss fitting data obtained in S23 and the energy consumption increment output ratio associated with road geometric features obtained in S24 are matched one-to-one according to timestamps and mileage coordinates. After normalization processing of the data within the same mileage interval, a two-dimensional energy consumption matrix is ​​formed. With energy loss intensity as the vertical axis and geometric increment ratio as the horizontal axis, the composite energy consumption value is calculated by a weighted aggregation algorithm and a rolling average operation is performed over the entire path range. The average value of the energy consumption matrix is ​​updated every 100 seconds, and the composite energy loss index is output. This index is recorded as a dimensionless value and is used to quantitatively represent the degree of vehicle energy loss under real-time road conditions. The record includes path number, time interval, average composite index and corresponding mileage segment information, providing an input basis for subsequent extreme driving radius assessment and path optimization decision-making.

[0044] Step S22 includes the following steps:

[0045] Step S221: Extract the congestion kilometers, estimated travel time, traffic volume, and average vehicle speed from the congested road segment information;

[0046] Step S222: Calculate the average vehicle start-stop frequency based on the congestion kilometers, estimated travel time, traffic flow, and average vehicle speed; simulate and derive the energy conversion efficiency reduction gradient of the drive motor based on the average vehicle start-stop frequency, traffic flow, and average vehicle speed.

[0047] Step S223: Perform time-series intensity regression fitting on the energy conversion efficiency descent gradient to obtain efficiency loss intensity fitting data;

[0048] Step S224: Obtain the energy consumption output status of the vehicle's built-in equipment through the vehicle system; predict the energy consumption output increment ratio of the vehicle equipment based on the estimated travel time, and obtain the equipment energy consumption output increment ratio; wherein the built-in equipment of the vehicle includes the vehicle system, air conditioning and vehicle electrical appliances;

[0049] Step S225: Based on the efficiency loss intensity fitting data and the ratio of equipment energy consumption output increment, simulate and deduce the additional energy consumption loss in the congestion state.

[0050] In this embodiment of the invention, after extracting real-time traffic data, the road segment numbering system traverses all sections, identifies road segments marked as congested, and extracts the start and end mileage coordinates to calculate the congestion kilometers. In this implementation, the total length of the congested section is 5.6 km, the average traffic flow is 780 vehicles / h, the average speed is 14 km / h, and the estimated travel time is 24 min. During the data acquisition phase, each parameter is recorded by traffic sensing nodes at a sampling interval of 60 seconds, and speed anomalies are removed by a filtering algorithm. For small fluctuations less than 10%, a weighted smoothed average is taken. All data is stored in the traffic buffer in chronological order for subsequent driving energy efficiency calculations, with a time accuracy of 1 second and a speed accuracy of 0.1 km / h. After obtaining the above data, the average start-stop frequency of vehicles was first calculated. Using the average vehicle speed of 14 km / h and traffic flow of 780 vehicles / h in the road segment as input variables, the average distance time series period was estimated based on a vehicle travel distance of 25m. The average number of accelerations and decelerations per unit time of the vehicle was measured. The start-stop frequency was found to be 4.8 times / min according to the time series statistics. Then, using the start-stop frequency, traffic flow density, and average vehicle speed as input quantities, the ratio of output power to motor input power was analyzed by looking up the table through the motor efficiency characteristic curve. The gradient of the energy conversion efficiency of the drive motor was identified. When the start-stop frequency was 4.8 times / min and the vehicle speed was below 15 km / h, the highest conversion efficiency of the drive motor dropped from 94% to 80%. The gradient was calculated to be -0.14 / kWh. This value was obtained by point-by-point linear interpolation and formed into an efficiency decline time series dataset for output. Based on the energy conversion efficiency decline gradient sequence calculated in step S222, a time series function set with a sampling interval of 1s is constructed. Time series intensity analysis is performed on the data recorded for 300 consecutive seconds. The mean and variance of the efficiency decline intensity within each minute are calculated using a weighted regression algorithm. Further sliding window regression smoothing is performed with a window length of 60s and a step size of 10s to reduce short-term fluctuation interference. Then, exponential decay regression is performed to fit the efficiency decline gradient sequence. The time constant of the fitted curve is compared with the decay exponent to obtain the mean efficiency loss intensity of 0.13. The fitting accuracy and time scale index are recorded to finally generate the efficiency loss intensity fitting data result. This dataset is stored in the form of time series numbering for the next stage of energy consumption superposition calculation.

[0051] During road driving, the real-time power output status of the air conditioning system, vehicle infotainment system, and vehicle electrical appliances is obtained by reading the power sensor signals of various electrical systems inside the vehicle through the vehicle bus. When the ambient temperature detected by the in-vehicle temperature sensor is 30℃ and the in-vehicle set temperature is 23℃, the power of the air conditioning system compressor is maintained at 1.3kW, the power of the blower output current is 0.15kW, the power of the vehicle infotainment system is 0.4kW, and the power of the lighting system is 0.2kW. Using the estimated travel time of 24 minutes as the input parameter, the energy consumption output is calculated based on the power-time integral. The energy consumption of the air conditioning system is 0.52kWh, the energy consumption of the vehicle infotainment system is 0.16kWh, and the energy consumption of lighting and auxiliary electrical appliances is 0.08kWh. The total equipment energy output is 0.76kWh. The difference is compared with the normal steady-state operating power of 0.6kWh to form the energy consumption output increment ratio of 1.27. This value is marked as the equipment energy consumption output increment ratio.

[0052] Combining the efficiency loss intensity fitting data obtained in step S223 with the equipment energy consumption output increment ratio obtained in step S224, point-to-point correlation calculation is performed using timestamp matching. In the energy consumption superposition calculation of the drive motor efficiency loss and the equipment's additional output energy occurring simultaneously within each minute, the weight allocation is based on the drive system's proportion of 0.8 and the equipment system's proportion of 0.2. When the efficiency loss intensity fitting data is 0.13 and the equipment energy consumption increase ratio is 1.27, the additional energy consumption multiplier is 0.13×0.8+0.27×0.2, which results in 0.186. Multiplying this multiplier by the vehicle's theoretical energy consumption value of 14.8kWh / 100km and combining it with the congested mileage of 5.6km, the additional energy consumption loss is calculated to be 1.55kW / h. The data is written into the energy consumption database using the road segment number and time index to form the additional energy consumption loss data output under congested conditions, which serves as the basic input for the composite analysis of energy consumption loss.

[0053] Step S223 includes:

[0054] Plot the conversion efficiency decline curve based on the energy conversion efficiency decline gradient; extract the monotonically decreasing curve from the conversion efficiency decline curve.

[0055] The decreasing relative numerical variance under time progression is calculated based on the monotonically decreasing curve.

[0056] By fitting the monotonically decreasing curve with an exponential rate decay, the decay rate exponent is obtained.

[0057] Point process accumulation is identified based on decreasing relative numerical variance and decay rate exponent to obtain conversion efficiency decay point process data.

[0058] Time-series intensity regression fitting is performed based on the process data of conversion efficiency decay point to obtain efficiency loss intensity fitting data.

[0059] In this embodiment of the invention, based on the energy conversion efficiency decline gradient data of the drive motor obtained in step S222, an efficiency decline curve is first plotted on the time axis at a sampling interval of 1 second. The efficiency value of each sampling point is arranged in an orderly manner with time as the horizontal axis and efficiency decline gradient as the vertical axis, forming 300 sets of recording points within a 300-second sampling period. Linear interpolation is used to smooth the data between adjacent sampling points, and abnormal increases caused by instantaneous peaks during motor startup are deleted within the time interval to ensure the overall trend of the curve is continuous. Then, monotonicity detection is performed on the curve, and the positive and negative change intervals are determined by the derivative sign. All continuous negative derivative intervals are extracted as the main body of the monotonically decreasing curve. The average number of segments in the monotonically decreasing curve is 5, and the average duration of each segment is about 60 seconds. The efficiency value data of the monotonically decreasing curve is re-indexed to form a time series matrix as the basis for subsequent calculations.

[0060] Based on the extracted monotonically decreasing curve, the decreasing relative numerical variance under time progression is calculated in each time interval. In this process, the efficiency change value of adjacent 5 seconds is calculated by moving difference operation, the mean variance is taken and the relative change amplitude is recorded to obtain the decreasing relative numerical variance dataset. In order to reduce the impact of short-term oscillations, a weighting coefficient is set in the variance calculation. The weighting coefficient for the difference value within the time interval of 2 seconds to 10 seconds is set to 0.8, and the weighting coefficient for the interval of more than 10 seconds is set to 0.4. After weight normalization, the mean of the decreasing relative numerical variance of each segment is output. The average value is controlled between 0.002 and 0.006, which represents the stable rate range of efficiency decay.

[0061] The above monotonically decreasing curve sequence was fitted with an exponential rate decay to obtain the decay rate index. During the fitting process, a stepwise regression method was used to calculate the curve descent rate segment by segment. The initial efficiency value and the final efficiency value of each curve segment were updated as exponential regression input samples, which were converted into paired data of time and efficiency loss intensity. After solving by the least squares method, the decay rate index result was output. Experimental data showed that the decay rate index ranged from 0.012 to 0.021 under normal congestion conditions, and the index was higher than 0.025 for road segments that were stagnant for a longer period of time. The calculation results were stored in the decay curve index table in chronological order for subsequent pointer access.

[0062] Subsequently, based on the decreasing relative variance and decay rate exponent, the process accumulation and identification operation is performed. The time axis is divided into 10-second detection windows. Efficiency decay abrupt change points are identified within each window. A decay event point is determined by a numerical threshold when the 10-second variance value is higher than 1.5 times the average variance of the entire region and the instantaneous decay rate value is higher than 1.3 times the average rate exponent. All event point index numbers and time records are accumulated to form the process data of conversion efficiency decay points. The data fields include event number, timestamp, local variance value, instantaneous decay rate value, and location index.

[0063] After obtaining the aforementioned data on the efficiency decay point process, time-series intensity regression calculation is performed using this data sequence as input. The number of event points is counted within each 60-second window, the event density function per unit time is calculated, and a moving regression fit is performed. Anomalies are smoothed using a weighted moving average. The event point density of each window is matched with the decay rate change amplitude to calculate the average efficiency loss intensity. After normalization, the average value of the output efficiency loss intensity fitted data is 0.13, in kWh. Each update interval is 10 seconds. The output format includes the time-series index, window number, event count, and fitted intensity value. This data serves as the core input for the subsequent congestion energy consumption derivation stage.

[0064] Step S24 includes the following steps:

[0065] Step S241: Determine the road segment with a ramp turning structure based on the road ramp and intersection structure information; calculate the slope range of the road ramp;

[0066] Step S242: Determine the additional slope resistance based on the slope range and evaluate the additional energy output ratio associated with the slope difference;

[0067] Step S243: Perform a regression analysis on the energy consumption growth ratio of ramp turns based on the ramp turn structure section to obtain the energy consumption growth regression ratio between ramp turns;

[0068] Step S244: Based on the energy consumption additional output ratio and the energy consumption growth regression ratio, quantify the energy consumption increment output ratio of road geometric features to obtain the energy consumption increment output ratio associated with road geometric features.

[0069] In this embodiment of the invention, after obtaining path information through a high-precision road geographic data interface, the slope and curvature of the road are sampled in steps of 100m. Road segments with a continuous length exceeding 300m and a non-zero slope are marked as slope segments. Then, road network nodes with a turning radius of less than 5m in adjacent segments are extracted to form a set of slope-turning structure road segments. A total of 46 segments containing slope and turning features are detected in the road samples, with an average slope of 4.2°, a maximum slope of 12.8°, and a minimum slope of 0.3°. The slope range is calculated by the difference between the maximum and minimum slope of each segment. This value reflects the energy change of the vehicle during the climbing and descending process. In the data processing, a segmented accumulation method is used to avoid local calculation errors caused by abrupt changes in terrain. Finally, the slope range is calculated to be from 3° to 12.5°. The average slope, range, and corresponding turning radius of each segment are recorded together in the road geometric feature data table, providing a basis for subsequent energy consumption characteristic quantification.

[0070] After obtaining the gradient range data, additional gradient resistance was calculated for each section. The resistance sources include the projection of the gravity component and the deviation of rolling resistance. The calculations were performed under standard conditions: a vehicle curb weight of 1800 kg, a rolling resistance coefficient of 0.01, an air resistance parameter of 0.32, and a frontal area of ​​2.2 m². When the gradient range was 10°, the increase in the additional gravity projection component due to the gradient change was 0.173, and the increase in rolling resistance was 0.006. The overall additional gradient resistance was equivalent to 1.179 times the vehicle's theoretical driving resistance. The additional energy output ratio is determined based on the resistance ratio. The calculation of the additional energy output ratio takes into account the motor conversion efficiency and vehicle transmission efficiency. When the vehicle's climbing output power is stable at 45kW and the efficiency is 0.9, the power demand increases to 52.3kW due to the gradient difference, and the efficiency decreases to 0.84. The calculated additional energy output ratio is 1.162. The calculation results of this type for each slope section are output in tabular form, with fields including the average gradient, gradient difference, additional resistance ratio, and additional energy output ratio for subsequent energy consumption growth ratio regression analysis.

[0071] Based on the ramp turning structure determined in step S241, an energy consumption growth regression analysis was conducted using the average turning radius and vehicle speed of each turning segment as the main variables. It was assumed that the vehicle turning speed ranged from 20 km / h to 45 km / h, and the turning radius ranged from 6 m to 8 m. Each data set was sampled 30 times, and the changes in lateral acceleration, centrifugal force, and tire slip angle during turning were calculated. The measured vehicle driving power was compared with the power data during straight-line driving, and the additional energy consumption increase per unit turn was calculated. The correlation coefficient of the curve obtained through multivariate least squares regression was 0.91, and the energy consumption growth ratio ranged between 1.05 and 1.21. Simultaneously, considering the slope effect, a slope resistance factor of 0.08 was superimposed in areas with a slope exceeding 5° and a turning radius below 7.2 m to reflect the coupling effect of gravity and centrifugal force. Finally, the energy consumption growth ratio of each ramp turning segment was calculated, and the values ​​were numbered and stored in the energy consumption growth database. The record fields included turning radius, slope, vehicle speed, additional energy consumption ratio, and regression residuals for further quantitative and comprehensive processing.

[0072] The energy consumption additional output ratio obtained in step S242 and the slope turning energy consumption growth ratio obtained in step S243 are matched and integrated according to road number. For every 100m segment, a weighting factor is quantitatively calculated based on the weighting relationship between the two ratios. The weighting factor is determined according to the ratio coefficient of slope and turning length within the segment. When the slope length ratio is higher than 70%, the weighting is 0.7 for the energy consumption additional output ratio and 0.3 for the turning growth ratio. When the turning length ratio is higher than 50%, the weighting is reversed. After weighted summation, a composite energy consumption increment output ratio is formed. In the calculation, the time continuity of the data is maintained by using a sliding window with a 5s movement step size of 1s to update the average ratio, avoiding abrupt jumps in the value. After standardization and normalization, the output result is in the range of 1.00 to 1.25. The output result is recorded as the road geometric feature associated energy consumption increment output ratio and stored as a list data structure. The column fields include time index, road number, composite energy consumption increment value, current slope, current turning radius, and slope range for reference in the energy loss composite index calculation stage.

[0073] Step S243 includes the following steps:

[0074] The average turning radius and slope angle are calculated based on the road section with a ramp-turn structure; the minimum passing speed of vehicles is derived based on the average turning radius and slope angle.

[0075] Based on the vehicle's minimum speed, average turning radius, and slope angle, the gravity and centrifugal force vectors for lateral turning are simulated and derived; the tilted resultant force sector is identified by analyzing the gravity and centrifugal force vectors to obtain the tilted dynamic resultant force sector;

[0076] Based on the tilted dynamic resultant force sector, an additional yaw moment imbalance analysis was performed to obtain additional yaw moment imbalance data.

[0077] Based on the aforementioned additional yaw moment imbalance data, a dynamic deviation analysis of the vector space of gravity drag and lateral drag is performed to obtain the dynamic deviation of the vector space of gravity drag and lateral drag.

[0078] Based on the dynamic deviation in the vector space, a dynamic cancellation numerical torque analysis is performed on the power drive output to obtain the dynamic cancellation numerical torque.

[0079] Based on the dynamically offset numerical torque, a regression analysis of the energy consumption growth ratio during ramp turns is performed to obtain the energy consumption growth regression ratio between ramp turns.

[0080] In this embodiment of the invention, based on the sample data of the slope turning road segment extracted in step S241, the spatial geometry of the road is analyzed in detail. Equally spaced sampling points are established along the segment with a sampling interval of 10m. The rate of change of tangent direction Δθ between adjacent sampling points is calculated by the geometric coordinate difference of the road centerline. The average turning angle change rate is calculated based on the cumulative arc length. The average turning radius is obtained by back-calculating the angle change per unit arc length. Verified by field data, the average turning radius of this typical urban uphill turning road segment is 6.8m. In the same area, the elevation coordinates of the sampling points are extracted by digital elevation data. The ratio of elevation change to horizontal distance is calculated, and the slope inclination of this road segment is 5.7°. In other similar road segment data obtained by sampling, the slope range is 3° to 9°. Thus, a basic dataset containing the average turning radius, slope inclination, turning angular rate and sampling time sequence is generated.

[0081] Based on the average turning radius of 6.8m and the slope angle of 5.7°, a vehicle stability calculation range was established. The vehicle's curb weight was 1800kg, the wheelbase was 1.62m, the center of gravity height was 0.55m, and the tire friction coefficient was 0.82. According to the balance constraint that simultaneously satisfies the critical conditions of rollover and sideslip, the calculation was performed by decomposing the gravity and centrifugal force onto the slope plane. The minimum passing speed of the vehicle on this road section was obtained from the balance condition solution as 21km / h. Insufficient passing speed will cause the lateral adhesion requirement to exceed the tire limit, while exceeding the passing speed will cause the centrifugal acceleration to exceed the gravity component balance range. The minimum passing speed calculation result was controlled within the safe range by the stability boundary algorithm, and the passing speed of 21km / h was recorded as the speed threshold parameter of this range. Subsequent analyses all used this speed value as the input variable.

[0082] After obtaining the minimum passing speed, the lateral force and gravity vector decomposition was derived by combining the average turning radius and slope angle. The vehicle forces were first divided into two parts: a three-dimensional gravity vector and a horizontal centrifugal force vector. The gravity was calculated by multiplying the vehicle's curb weight by the gravitational acceleration, resulting in a total of 17.64 kN. Under a slope angle of 5.7°, the gravity component along the slope direction was 17.64 × sin5.7° = 1.75 kN, and the tangential component was approximately 17.56 kN. The magnitude of the centrifugal force was calculated based on the vehicle speed of 21 km / h = 5.83 m / s and a radius of 6.8 m. The resultant force was 2.52 kN. Subsequently, a resultant force vector was constructed in the slope coordinate system. The centrifugal force and gravity components were added together to obtain a resultant force modulus of approximately 18.4 kN. Its direction was skewed relative to the vertical axis by an angle of approximately 7.6°. Then, the vector scanning method was used to divide the area into 120 regions with an angle step of 3° and a 360° sector. The force vector projection intensity was calculated for each region. The sections with force projection greater than 1.15 times the average force intensity of the entire region were classified as the main tilted dynamic resultant force sectors. The force modulus, direction angle and time index information of each sector were output to form the tilted dynamic resultant force sector data.

[0083] Subsequently, based on the tilted dynamic resultant force sector data, the yaw moment imbalance was calculated. The normal support forces of the front and rear axles of the vehicle were obtained separately. The normal force of the front axle was corrected by the resultant force direction, reducing the normal force of the front axle to 8.1 kN and increasing the normal force of the rear axle to 9.2 kN. The force difference between the two axles was 1.1 kN. Using a wheelbase of 1.62 m as the lever arm length, the yaw moment was calculated to be 0.891 kN·m. The moment acts in the inward tilting direction, causing the vehicle to rotate around the vertical axis. Its instantaneous yaw angular acceleration is 0.25 rad / s². Within the 0.4 s integration time window, the instantaneous yaw angular velocity is obtained to be approximately 0.1 rad / s. This yields additional yaw moment imbalance data, and the time index, horizontal offset angle, yaw angular acceleration, and front and rear axle force ratio parameters are recorded. This dataset serves as the input basis for subsequent slope turning energy consumption growth regression analysis.

[0084] Based on the obtained additional yaw moment imbalance data, a dynamic deviation analysis of the vector space of gravity drag and lateral drag is first performed. Instantaneous force data of the vehicle during turning is taken, using the coordinate system of the vehicle's center of gravity as the analysis reference. The gravity drag vector is defined along the longitudinal direction of the slope, and the lateral drag vector is defined along the centrifugal direction of the turn. In a typical sample, the vehicle's curb weight is 1800 kg, the speed is 21 km / h (equivalent to 5.83 m / s), the slope angle is 5.7°, the turning radius is 6.8 m, the instantaneous centrifugal force is 2.52 kN, and the gravitational component is 17.64 kN. This is performed according to real-time sampling steps. The directional change angle and modulus change rate of the two vectors were calculated at intervals of 0.01s. After fifth-order difference smoothing, the average directional deviation angle was 7.3° and the maximum deviation angle was 9.8°. The resulting instantaneous dynamic deviation in the vector space was 1.54kN. This deviation indicates that the direction of gravity drag was dynamically shifted due to the intrusion of lateral inertial force when the vehicle was turning. Subsequently, the deviation at all times in the entire road segment was statistically integrated over the interval. The average dynamic deviation value and peak value of each sampling window were output, and the corresponding time index and position coordinates were recorded to form the dynamic deviation in the vector space of gravity drag and lateral drag.

[0085] Based on the aforementioned dynamic deviation data in vector space, a dynamic offset numerical torque analysis of the power drive output is performed. The vehicle drive torque time series and the vector deviation angle series are matched point by point. Within the sampling window, the instantaneous component of the drive motor output torque is calculated with a time step of 10ms. The asymmetric drive force caused by the deviation angle is converted into a torque loss component. Under the comprehensive working condition of a slope of 5.7° and a turning radius of 6.8m, the nominal output torque of the motor is 320N·m and the speed is 2700r / min. After angle correction, the effective drive torque is 320×cos7.3°=317.5N·m, resulting in an offset ratio of 0.0078 and a corresponding energy loss rate of 0.6%. During a continuous 60s of driving state sampling, 21 dynamic offset peaks were recorded, with each peak lasting no more than 0.36s. The total energy consumption of dynamic offset throughout the process is calculated to be 0.064kWh using the integral method, and a dynamic offset numerical torque sequence data record is generated. The data record fields include time index, angle deviation, instantaneous torque value, correction coefficient, and power loss.

[0086] Based on the dynamic offset torque data, a regression analysis of the energy consumption growth ratio during slope turning was conducted. During the analysis, a multivariate regression data table was constructed using the drive power correction at each moment and the vehicle's real-time speed. The energy consumption growth rate was used as the dependent variable, and the dynamic offset torque, slope angle, and turning radius were used as independent variables. Multiple least squares regression analysis was performed, fitting 120 sets of data under different steering conditions, resulting in a coefficient of determination R² of 0.89. The slope of the regression function corresponded to an energy consumption growth ratio range of 1.06 to 1.22. A weighted correction factor of 0.05 was applied to the additional regression residuals in high-slope, small-radius sections to eliminate calculation bias caused by low-speed torque jitter. The output of the calculation results is the regression ratio of energy consumption growth during slope turning. The resulting data set includes columns for average slope, average radius, vehicle speed, dynamic offset torque, average energy loss, and regression ratio, and records the data timestamp as input for calculating the road geometric energy consumption increment output ratio. This calculation process is updated every second, achieving continuous quantification of the dynamic growth trend of energy consumption during slope turning.

[0087] refer to Figure 3 As stated, step S3 includes the following steps:

[0088] Step S31: Perform composite energy loss characteristic analysis based on the claimed composite energy loss index to obtain composite energy loss characteristics;

[0089] In this embodiment of the invention, composite energy loss characteristics are analyzed based on the composite energy loss index. First, the composite energy loss index is processed by time-series segmentation, dividing the entire driving path into analysis units of 5 kilometers each. The composite energy loss index value corresponding to each analysis unit is extracted to form an energy loss composite index sequence. For example, a 150-kilometer driving path is divided into 30 analysis units, and each analysis unit corresponds to one composite energy loss index value. Statistical features are extracted from the composite energy loss index sequence to calculate the mean energy loss, variance energy loss, peak energy loss, and skewness of energy loss distribution. The formula for calculating the mean energy loss is that the mean energy loss equals the sum of the composite energy loss indices of all analysis units divided by the total number of analysis units. The formula for calculating the variance energy loss is that the variance energy loss equals the sum of the squares of the differences between the composite energy loss index and the mean of each analysis unit divided by the total number of analysis units. The peak energy loss is the maximum composite energy loss index value in the sequence. The skewness of energy loss distribution is used to measure the symmetry of energy loss distribution along the path.

[0090] In one implementation of this invention, trend characteristic analysis is performed on the energy loss composite index sequence to identify the changing trend of energy loss along the driving path, including an increasing trend segment, a decreasing trend segment, and a stable trend segment. A moving average method is used to smooth the energy loss composite index sequence, with the moving window size set to three analysis units. The calculation formula is that the smoothed energy loss value equals the sum of the energy loss composite indices of the current analysis unit and its adjacent units divided by 3. Based on the smoothed energy loss sequence, the rate of change of energy loss between adjacent analysis units is calculated using the formula: energy loss... The rate of change of energy loss is equal to the energy loss value of the next analysis unit minus the energy loss value of the previous analysis unit, and then divided by the energy loss value of the previous analysis unit. When the rate of change of energy loss is greater than 0.05, it is determined to be an increasing trend segment; when the rate of change of energy loss is less than -0.05, it is determined to be a decreasing trend segment; and when the rate of change of energy loss is between -0.05 and 0.05, it is determined to be a stationary trend segment. It should be noted that the mean energy loss, variance energy loss, peak energy loss, skewness of energy loss distribution, and distribution ratio of each trend segment are combined to form a composite energy loss characteristic, which is used for subsequent energy loss optimization analysis.

[0091] Step S32: Perform gradient descent optimization of energy loss based on the characteristics of composite energy loss to obtain optimized energy loss data;

[0092] In this embodiment of the invention, an energy loss objective function is constructed based on the composite energy loss characteristics. This objective function comprehensively considers the impact of the mean energy loss, the variance of energy loss, and the peak energy loss on the overall energy consumption. The mathematical expression of the objective function is: the energy loss objective function equals the mean energy loss multiplied by the mean weight coefficient, plus the variance of energy loss multiplied by the variance weight coefficient, plus the peak energy loss multiplied by the peak weight coefficient. The mean weight coefficient represents the degree of influence of the mean energy loss on the overall energy consumption and has a value of 0.5; the variance weight coefficient represents the degree of influence of the volatility of energy loss on the overall energy consumption and has a value of 0.3; the peak weight coefficient represents the degree of influence of extreme energy loss on the overall energy consumption and has a value of 0.2. The sum of the three weight coefficients is 1 and is obtained by calibration using historical driving data. The gradient descent optimization parameters are initialized with a learning rate of 0.01, a maximum number of iterations of 8, and a convergence threshold of 0.0001. The initial energy loss optimization parameter vector includes a driving speed adjustment factor (including the influence of road conditions and road structure status), an energy recovery intensity factor, and an air conditioning power adjustment factor.

[0093] In one implementation of this invention, an iterative optimization process using gradient descent is performed. In each iteration, the partial derivatives of the energy loss objective function with respect to each optimization parameter are calculated to form a gradient vector. The gradient vector is calculated using numerical differentiation, with the formula being: the partial derivative of the current parameter (i.e., the initial energy loss optimization parameter vector) equal to the function value of the objective function when the parameter increases by a small increment minus the function value of the objective function when the parameter decreases by a small increment, divided by twice the small increment, where the small increment is 0.001. The optimization parameter vector is updated based on the gradient vector, with the update formula being: the new parameter value equals the original parameter value minus the learning rate multiplied by the gradient value corresponding to that parameter. For example, when the initial value of the driving speed adjustment factor is... When the value is 1.0 and the corresponding gradient value is 0.5, the updated driving speed adjustment factor is 1.0 minus 0.01 multiplied by 0.5, which equals 0.995. It should be noted that during the iteration process, the convergence condition is judged. When the absolute value of the difference between the objective function values ​​of two adjacent iterations is less than the convergence threshold of 0.0001 or the maximum number of iterations of 8 is reached, the iteration stops and the final optimized parameter vector is output. The energy consumption loss optimization coefficient is calculated based on the final optimized parameter vector. The calculation formula is that the energy consumption loss optimization coefficient is equal to 1 minus the optimized objective function value divided by the initial objective function value. The energy consumption loss optimization coefficient, the optimized mean energy consumption loss, the energy consumption loss variance, and the energy consumption loss peak value are combined to form the energy consumption loss optimization data.

[0094] In another embodiment, the average gradient direction of the composite exponent is determined, i.e., the trend of energy loss along the path increasing with distance. Then, the direction of minimum energy loss is calculated using a numerical gradient descent algorithm. The optimization algorithm iteratively corrects the difference between the energy consumption value of each continuous road segment and the target minimum energy consumption value, updating the energy consumption gradient weighting coefficient during the iteration process to gradually flatten the energy consumption distribution. The adjustment range of energy consumption for each road segment is controlled by setting a convergence threshold. When the gradient change is less than the convergence threshold, the optimization process stops. The final output energy loss optimization data includes the optimized energy consumption value, energy consumption gradient vector, and total path correction ratio for each road segment. This data maintains the consistency of the original path data index in structure, so as to be jointly calculated with the theoretical driving distance in the future, thereby achieving accurate determination of the limit driving radius.

[0095] Step S33: Determine the maximum driving radius of the new energy vehicle based on the energy loss optimization data and the theoretical driving distance;

[0096] In this embodiment of the invention, the energy consumption loss optimization coefficient and the optimized average energy consumption loss are extracted based on the energy consumption loss optimization data. The actual drivable distance is then calculated by combining this with the theoretical driving distance. The calculation formula is: the actual drivable distance equals the theoretical driving distance multiplied by 1 minus the difference between the optimized average energy consumption loss and the theoretical driving distance. Here, the theoretical driving distance represents the maximum driving range supported by the vehicle's remaining battery power under ideal operating conditions, and the optimized average energy consumption loss represents the average energy consumption loss ratio after gradient descent optimization. For example, if the theoretical driving distance is 300 kilometers and the optimized average energy consumption loss is 0.12, the actual drivable distance is the difference between 300 kilometers multiplied by 1 and 0.12. The actual drivable distance is 264 kilometers. It should be noted that, in order to ensure driving safety and reserve sufficient charging buffer margin, a safety margin correction is required for the actual drivable distance. The safety margin coefficient is dynamically adjusted according to the variance of energy consumption loss. When the variance of energy consumption loss is large, it means that the energy consumption fluctuates greatly and more safety margin needs to be reserved. The formula for calculating the safety margin coefficient is: the safety margin coefficient equals the basic safety coefficient minus the variance of energy consumption loss multiplied by the variance adjustment factor. The basic safety coefficient is 0.95 and the variance adjustment factor is 0.1. For example, when the variance of energy consumption loss is 0.3, the safety margin coefficient is 0.95 minus 0.3 multiplied by 0.1, which equals 0.92.

[0097] In one implementation of this invention, the maximum driving radius of the new energy vehicle is calculated based on the actual drivable distance and the safety margin coefficient. The calculation formula is that the maximum driving radius equals the actual drivable distance multiplied by the safety margin coefficient. For example, if the actual drivable distance is 264 kilometers and the safety margin coefficient is 0.92, the maximum driving radius is 264 multiplied by 0.92, which equals 242.88 kilometers. After rounding, the maximum driving radius is 242 kilometers. It should be noted that the impact of peak energy consumption loss on the maximum driving radius also needs to be considered. When the optimized peak energy consumption loss exceeds a preset peak threshold of 0.25, adjustments need to be made. The extreme driving radius is further corrected using the formula: the corrected extreme driving radius equals the extreme driving radius multiplied by 1 minus half the difference between the peak energy loss and the peak threshold. This correction ensures that the vehicle still has sufficient charge to reach the charging node when passing through high-energy-consumption road sections. For example, when the peak energy loss is 0.35, the corrected extreme driving radius is 242 multiplied by 1 minus half the difference between 0.35 and 0.25, which equals 242 multiplied by 0.95, or 229.9 kilometers. After rounding, the corrected extreme driving radius is 229 kilometers. Finally, the corrected extreme driving radius is used as the output extreme driving radius of the new energy vehicle.

[0098] Step S34: Based on the maximum driving radius, make an optimal decision on the set of charging nodes to select charging nodes, and output the path planning data.

[0099] In this embodiment of the invention, the set of charging nodes is filtered by distance based on the maximum driving radius. All charging nodes in the set are traversed, and the path distance between each charging node and the starting point is calculated. Charging nodes whose path distance is less than the maximum driving radius are selected to form a distance-adapted set of charging nodes. For example, if the maximum driving radius is 229 kilometers, and the set of charging nodes includes five charging nodes (charging station A: 30 kilometers from the starting point, charging station B: 70 kilometers from the starting point, charging station C: 120 kilometers from the starting point, charging station D: 180 kilometers from the starting point, and charging station E: 250 kilometers from the starting point), then the distance-adapted set of charging nodes includes four charging stations: charging station A, charging station B, charging station C, and charging station D. Charging nodes: By establishing a data connection between the vehicle's infotainment system and the charging station's cloud platform, the system can obtain real-time operational status information of all charging nodes in the distance-adaptive charging node cluster. It can also extract data on the number of available charging piles, charging power, and charging compatibility from the charging node status. For example, charging station A has 4 available charging piles with a maximum charging power of 60 kW and supports national standard fast charging; charging station B has 2 available charging piles with a maximum charging power of 120 kW and supports national standard fast charging; charging station C has 6 available charging piles with a maximum charging power of 180 kW and supports national standard fast charging and super fast charging; and charging station D has 3 available charging piles with a maximum charging power of 90 kW and supports national standard fast charging.

[0100] In one implementation of this invention, a comprehensive evaluation model for charging nodes is constructed to make optimal decisions on the set of distance-adaptable charging nodes. The comprehensive evaluation model adopts a multi-attribute decision-making method, and the evaluation indicators include the number of available charging piles, charging power, charging compatibility, and distance adaptability. First, each evaluation indicator is normalized. The normalized value of the number of available charging piles is equal to the number of available charging nodes divided by the maximum number of available charging nodes among all candidate nodes. The normalized value of charging power is equal to the charging power of the charging node divided by the maximum charging power among all candidate nodes. Charging compatibility is quantified based on the number of compatible interface types, with a value of 1 for two interface types and a value of 0.6 for one interface type. Distance adaptability is equal to 1 minus the distance of the charging node divided by the maximum driving radius. The formula for calculating the comprehensive evaluation score is: Where S represents the overall evaluation score; N represents the number of available charging piles at the charging node; This indicates the maximum number of available nodes among all candidate nodes; P represents the charging power of the charging node. C represents the maximum charging power among all candidate nodes; D represents the charging compatibility quantification value; and D represents the distance between the charging node and the starting point. Indicates the maximum driving radius; This indicates the available quantity weight and has a value of 0.2; This indicates the charging power weight and has a value of 0.35. δ represents the compatibility weight and has a value of 0.15; δ represents the distance weight and has a value of 0.3. It should be noted that the charging node with the highest comprehensive evaluation score is selected as the optimal charging node, and complete route planning data from the starting point to the destination via the optimal charging node is generated. This route planning data includes segmented navigation route information, estimated travel time for each segment, estimated remaining battery percentage when reaching the optimal charging node, suggested charging time, and estimated range after charging. Finally, the route planning data is visualized and output through the navigation interface of the vehicle system and voice prompts are provided.

[0101] It needs to be explained that between steps S31 and S34, firstly, based on a comprehensive consideration of the composite index of congestion energy consumption and road geometric feature energy consumption, the characteristics and patterns of energy loss under different scenarios are analyzed in depth to clarify the core influencing factors and changing trends of energy loss; then, gradient descent optimization is performed on these characteristics to eliminate the influence of ineffective energy consumption and correct energy consumption estimation bias, resulting in more accurate energy consumption optimization data; subsequently, combining this optimization data with the vehicle's theoretical driving distance, the actual maximum driving range of the vehicle under the current battery level and road conditions is scientifically defined to avoid range anxiety caused by misjudgment of energy consumption; finally, based on this limit range, suitable charging nodes are selected, and key information such as the availability of charging piles, charging efficiency, and compatibility of the nodes are obtained through network connection, and the optimal selection of charging nodes and the final route are planned by comprehensively considering these practical indicators. The entire process begins with energy consumption characteristic analysis, optimizes and corrects to achieve accurate judgment of driving range, and then combines the actual status of charging nodes to complete closed-loop path planning. Logically, it is progressive and data supports each other, which not only ensures the driving range safety of the path planning, but also improves the energy replenishment efficiency by selecting the best charging nodes, effectively solving the problem of path planning and energy replenishment decision-making for long-distance driving of new energy vehicles.

[0102] Step S34 includes the following steps:

[0103] Step S341: Select a distance-adaptive set of rechargeable nodes based on the maximum driving radius;

[0104] Step S342: Connect the vehicle system to the charging station cloud and obtain the status of all charging nodes in the distance-adaptable charging node set;

[0105] Step S343: Extract the number of available charging piles, charging power, and charging compatibility from the charging node status;

[0106] Step S344: Based on the number of available charging piles, charging power, and charging compatibility, optimize the selection of charging nodes for the maximum driving radius, and output the route planning data.

[0107] In this embodiment of the invention, distance filtering of the set of charging nodes is performed based on the maximum driving radius. First, the location coordinates of all charging nodes in the set are obtained. The actual path distance between each charging node and the current starting point is calculated using an in-vehicle panoramic map. This actual path distance is calculated using the shortest path algorithm of the road network, rather than a straight-line distance. For example, if the set of charging nodes includes five charging nodes: charging station A, charging station B, charging station C, charging station D, and charging station E, the path calculation shows that charging station A is 30 kilometers from the starting point, charging station B is 70 kilometers from the starting point, charging station C is 120 kilometers from the starting point, charging station D is 180 kilometers from the starting point, and charging station E is 250 kilometers from the starting point. The path distance of each charging node is compared with the maximum driving radius, and the distance filtering condition is set as follows: the path distance of the charging node is less than or equal to the maximum driving radius.

[0108] The vehicle's infotainment system establishes a data connection with the charging station operator's cloud server via its wireless communication module. This module supports 4G and 5G cellular network communication protocols. The system sends a charging node status query request to the cloud server, carrying a unique identifier code for all charging nodes in the distance-adaptive charging node set in the request data packet. The cloud server retrieves the real-time operating status data of the corresponding charging station based on the identifier code and returns it to the vehicle's infotainment system. It is important to note that to ensure the timeliness of the acquired charging node status data, the system sets a data refresh cycle of 60 seconds, continuously updating the charging node status information during navigation. When the network connection is interrupted, the system will use the most recently cached charging node status data and mark the data update time on the interface.

[0109] In one implementation of this invention, the charging node status data returned by the cloud server adopts a structured data format. The status data of each charging node includes the charging station name, charging station address, total number of charging piles, number of available charging piles, number of occupied charging piles, number of faulty charging piles, rated charging power of each charging pile, supported charging interface type, number of vehicles currently waiting in the queue, and estimated average waiting time. For example, the status data of charging station A shows that the total number of charging piles is 6, the number of available charging piles is 4, the number of occupied charging piles is 1, the number of faulty charging piles is 1, the rated charging power is 60 kW, it supports the national standard fast charging interface, the number of vehicles currently waiting in the queue is 0, and the estimated average waiting time is 0 minutes. After receiving the status data of all distance-adaptable charging nodes, the vehicle system caches the data locally and verifies the data integrity. If the status data of a certain charging node is missing or abnormal, the status of the node is marked as unknown and its priority is reduced in subsequent evaluations.

[0110] The data from the charging node status is extracted as follows: The number of available charging piles is extracted, indicating the number of charging piles currently in idle standby mode and available for normal use. The number of available charging piles directly affects the waiting time for users after arriving at the charging station; a higher number indicates more abundant charging resources. The charging power field is also extracted, representing the maximum rated charging power of the charging piles configured at the charging station, measured in kilowatts. Charging power determines the vehicle's charging speed; higher power results in shorter charging time. Common charging power levels include 60 kW, 120 kW, 180 kW, and 250 kW. Finally, the charging compatibility field is extracted, indicating the types of charging interfaces and charging protocol standards supported by the charging station, including national standard fast charging interfaces, national standard slow charging interfaces, and super-fast charging interfaces.

[0111] In one implementation of this invention, the extracted charging compatibility data is quantified to facilitate subsequent comprehensive evaluation calculations. The charging compatibility quantification rules are as follows: when the charging station supports both the national standard fast charging interface and the super fast charging interface, the charging compatibility quantification value is 1.0; when the charging station only supports the national standard fast charging interface, the charging compatibility quantification value is 0.7; when the charging station only supports the national standard slow charging interface, the charging compatibility quantification value is 0.3; and when the charging interface type is incompatible with the vehicle, the charging compatibility quantification value is 0. It should be noted that the vehicle system determines the charging compatibility value based on the vehicle's specifications. The charging interface configuration automatically matches compatible charging stations, filtering out charging nodes with a charging compatibility quantification value of 0. For example, after data extraction, it is found that the available number of charging stations A is 4, the charging power is 60 kW, and the compatibility quantification value is 0.7; the available number of charging stations B is 2, the charging power is 120 kW, and the compatibility quantification value is 0.7; the available number of charging stations C is 6, the charging power is 180 kW, and the compatibility quantification value is 1.0; and the available number of charging stations D is 3, the charging power is 90 kW, and the compatibility quantification value is 0.7.

[0112] A comprehensive evaluation model for charging nodes is constructed to make optimal decisions on the set of charging nodes with suitable distance. The comprehensive evaluation model adopts a weighted comprehensive scoring method, and the evaluation indicators include four dimensions: the number of available charging piles, charging power, charging compatibility, and distance adaptability. First, each evaluation indicator is normalized to eliminate the difference in dimensions. The formula for calculating the comprehensive evaluation score is the formula in the embodiment of step S34 above. The comprehensive evaluation scores of all candidate charging nodes are compared, and the charging station with the highest score is selected as the optimal charging node. Complete path planning data from the starting point to the destination via the charging station is generated. The path planning data includes the first segment of the navigation path from the starting point to the charging station. Finally, the path planning data is visualized through the vehicle system navigation interface and voice navigation is started.

[0113] This invention also provides a path planning system for new energy vehicles, used to execute the path planning method for new energy vehicles as described above. The path planning system for new energy vehicles includes:

[0114] The road condition information acquisition module is used to obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path through the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance.

[0115] The energy loss analysis module is used to simulate and deduce the ineffective energy consumption output of congested conditions based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; and evaluate the energy loss composite index based on the ineffective energy consumption output and the energy consumption increment output ratio to form the energy loss composite index.

[0116] The path planning output module is used to determine the limit driving radius of the new energy vehicle based on the composite index of energy loss and the theoretical driving distance; to make an optimal decision on the set of charging nodes based on the limit driving radius, and to output the path planning data.

[0117] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A path planning method for new energy vehicles, characterized in that, Includes the following steps: Step S1: Obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path using the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance. Step S2: Simulate and deduce the ineffective energy consumption output of the congested state based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; evaluate the composite index of energy consumption loss based on the ineffective energy consumption output and the energy consumption increment output ratio to form the composite index of energy consumption loss. Step S3: Determine the maximum driving radius of the new energy vehicle based on the composite energy loss index and the theoretical driving distance; make an optimal decision on the set of charging nodes based on the maximum driving radius, and output the path planning data. Step S2 includes: Step S21: Extract congested road section information and road slope and intersection structure information from real-time traffic information; Step S22: Simulate and deduce the additional energy loss under congestion conditions based on congested road segment information; Step S23: Perform nonlinear energy loss fitting on the additional energy loss to obtain nonlinear energy loss fitting data; Step S24: Based on the road slope and intersection structure information, quantify the energy consumption increment output ratio of road geometric features to obtain the energy consumption increment output ratio associated with road geometric features; Step S25: Evaluate the composite index of energy loss based on the nonlinear energy loss fitting data and the energy loss increment output ratio to form the composite index of energy loss; Step S24 includes: Step S241: Determine the slope turning structure section based on the road slope and intersection structure information; calculate the slope range of the road slope; Step S242: Determine the additional slope resistance based on the slope range and evaluate the additional energy output ratio associated with the slope difference; Step S243: Perform a regression analysis on the energy consumption growth ratio of ramp turns based on the ramp turn structure section to obtain the energy consumption growth regression ratio between ramp turns; Step S244: Based on the energy consumption additional output ratio and the energy consumption growth regression ratio, quantify the energy consumption increment output ratio of road geometric features to obtain the energy consumption increment output ratio associated with road geometric features; Step S243 includes: The average turning radius and slope angle are calculated based on the road section with a ramp-turn structure; the minimum passing speed of vehicles is derived based on the average turning radius and slope angle. Based on the vehicle's minimum speed, average turning radius, and slope angle, the gravity and centrifugal force vectors for lateral turning are simulated and derived; the tilted resultant force sector is identified by analyzing the gravity and centrifugal force vectors to obtain the tilted dynamic resultant force sector; Based on the tilted dynamic resultant force sector, an additional yaw moment imbalance analysis was performed to obtain additional yaw moment imbalance data. Based on the aforementioned additional yaw moment imbalance data, a dynamic deviation analysis of the vector space of gravity drag and lateral drag is performed to obtain the dynamic deviation of the vector space of gravity drag and lateral drag. Based on the dynamic deviation in the vector space, a dynamic cancellation numerical torque analysis is performed on the power drive output to obtain the dynamic cancellation numerical torque. Based on the dynamically offset numerical torque, a regression analysis of the energy consumption growth ratio during ramp turns is performed to obtain the energy consumption growth regression ratio between ramp turns.

2. The path planning method for new energy vehicles according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Obtain the current remaining battery power of the new energy vehicle; Step S12: Obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest path using the onboard panoramic map of the new energy vehicle; Step S13: Estimate the theoretical driving distance of the new energy vehicle based on the remaining battery power and the shortest distance. Step S14: Obtain real-time road condition information for the theoretical driving distance using the in-vehicle panoramic map.

3. The path planning method for new energy vehicles according to claim 1, characterized in that, Step S22 includes the following steps: Step S221: Extract the congestion kilometers, estimated travel time, traffic volume, and average vehicle speed from the congested road segment information; Step S222: Calculate the average vehicle start-stop frequency based on the congestion kilometers, estimated travel time, traffic flow, and average vehicle speed; simulate and derive the energy conversion efficiency reduction gradient of the drive motor based on the average vehicle start-stop frequency, traffic flow, and average vehicle speed. Step S223: Perform time-series intensity regression fitting on the energy conversion efficiency descent gradient to obtain efficiency loss intensity fitting data; Step S224: Obtain the energy consumption output status of the vehicle's built-in equipment through the vehicle system; predict the energy consumption output increment ratio of the vehicle equipment based on the estimated travel time, and obtain the equipment energy consumption output increment ratio; wherein the built-in equipment of the vehicle includes the vehicle system, air conditioning and vehicle electrical appliances; Step S225: Based on the efficiency loss intensity fitting data and the ratio of equipment energy consumption output increment, simulate and deduce the additional energy consumption loss in the congestion state.

4. The path planning method for new energy vehicles according to claim 3, characterized in that, Step S223 includes: Plot the conversion efficiency decline curve based on the energy conversion efficiency decline gradient; extract the monotonically decreasing curve from the conversion efficiency decline curve. The decreasing relative numerical variance under time progression is calculated based on the monotonically decreasing curve. By fitting the monotonically decreasing curve with an exponential rate decay, the decay rate exponent is obtained. Point process accumulation is identified based on decreasing relative numerical variance and decay rate exponent to obtain conversion efficiency decay point process data. Time-series intensity regression fitting is performed based on the process data of conversion efficiency decay point to obtain efficiency loss intensity fitting data.

5. The path planning method for new energy vehicles according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Perform composite energy loss characteristic analysis based on the claimed composite energy loss index to obtain composite energy loss characteristics; Step S32: Perform gradient descent optimization of energy loss based on the characteristics of composite energy loss to obtain optimized energy loss data; Step S33: Determine the maximum driving radius of the new energy vehicle based on the energy loss optimization data and the theoretical driving distance; Step S34: Based on the maximum driving radius, make an optimal decision on the set of charging nodes to select charging nodes, and output the path planning data.

6. The path planning method for new energy vehicles according to claim 5, characterized in that, Step S34 includes the following steps: Step S341: Select a distance-adaptive set of rechargeable nodes based on the maximum driving radius; Step S342: Connect the vehicle system to the charging station cloud and obtain the status of all charging nodes in the distance-adaptable charging node set; Step S343: Extract the number of available charging piles, charging power, and charging compatibility from the charging node status; Step S344: Based on the number of available charging piles, charging power, and charging compatibility, optimize the selection of charging nodes for the maximum driving radius, and output the route planning data.

7. A path planning system for a new energy vehicle, characterized in that, For executing the path planning method for a new energy vehicle as described in claim 1, the path planning system for the new energy vehicle includes: The road condition information acquisition module is used to obtain the shortest distance from the starting point to the destination and the set of charging nodes along the shortest distance path through the onboard panoramic map of the new energy vehicle; estimate the theoretical driving distance of the new energy vehicle based on the shortest distance, and obtain real-time road condition information along the theoretical driving distance. The energy loss analysis module is used to simulate and deduce the ineffective energy consumption output of congested conditions based on real-time traffic information; quantify the energy consumption increment output ratio of road geometric features based on the real-time traffic information to obtain the energy consumption increment output ratio associated with road geometric features; and evaluate the energy loss composite index based on the ineffective energy consumption output and the energy consumption increment output ratio to form the energy loss composite index. The path planning output module is used to determine the limit driving radius of the new energy vehicle based on the composite index of energy loss and the theoretical driving distance; to make an optimal decision on the set of charging nodes based on the limit driving radius, and to output the path planning data.

Images