Intelligent control method and system for new energy motor home

Abstract

The invention provides an intelligent control method and system for a new energy motor home, and belongs to the technical field of power supply systems, and the method comprises the steps: generating a route grid topological matrix, obtaining a grid light energy vector based on weather forecast data and the route grid topological matrix, and obtaining a light energy optimal route according to the grid light energy vector and initial battery state data; a prediction photovoltaic power generation curve is obtained according to the optical energy optimal route, optical energy prediction calibration data is obtained according to actual photovoltaic power generation data, a user habit model is constructed, a load state transfer tensor is generated, and a man-machine cooperation correction factor is obtained based on a user real-time operation record and the load state transfer tensor; adjusting a grid light energy vector based on a light energy prediction calibration parameter, and adjusting and circulating a load state transfer tensor based on a man-machine cooperation correction factor; the problem that the utilization efficiency of a photovoltaic system of a traditional new energy motor home control system is low is solved while route selection and energy acquisition are prevented from being disjointed.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power supply systems, and more particularly relates to an intelligent control method and system for a new energy motor home. BACKGROUND

[0002] With the development of the integration of new energy technology and intelligent networking technology, new energy motor homes provide energy through energy storage batteries, realize zero emission, low noise and flexible travel experience, and become the best choice for leisure tourism and outdoor life.

[0003] The existing new energy motor homes equipped with photovoltaic systems can only roughly estimate the required power based on the distance when the user sets the destination or charging pile and plans the route, but cannot dynamically adjust in combination with the real-time power consumption of the equipment along the way. For example, if the user continuously turns on the air conditioner during driving in summer or uses an electric oven for cooking along the way, the energy consumption will increase. At this time, if the initial power calculation does not reserve redundancy, the remaining battery power may not be sufficient to support the journey to the destination. In this case, the photovoltaic system should supplement the energy, but if the user chooses a road that happens to be continuously rainy, foggy or long in dense forest sheltered sections, the photovoltaic system may not work effectively, further exacerbating the power shortage problem, and ultimately causing the vehicle to stop in the middle of the road.

[0004] In summary, the existing new energy motor homes equipped with photovoltaic systems do not predict the light energy along the route during route planning, so that the route selection is only based on traditional traffic factors, which is disconnected with the energy acquisition needs of the photovoltaic system, resulting in low utilization efficiency of the photovoltaic system. SUMMARY

[0005] To solve the above technical problems, the application provides an intelligent control method and system for a new energy motor home. The purpose and effect of the intelligent control method and system for a new energy motor home are achieved by the following specific technical means:

[0006] An intelligent control method for a new energy motor home, the method comprising:

[0007] S1: generating a route grid topology matrix and obtaining meteorological forecast data and initial battery state data, obtaining a grid light energy vector based on the meteorological forecast data and the route grid topology matrix, and obtaining an optimal light energy route based on the grid light energy vector and the initial battery state data;

[0008] S2: generating a predicted photovoltaic power generation curve based on the optimal light energy route and the grid light energy vector, obtaining actual photovoltaic power generation data, and obtaining light energy prediction calibration data based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve;

[0009] S3: constructing a user habit model and generating a load state transition tensor, establishing a load scheduling instruction set according to a predicted photovoltaic power generation curve and the load state transition tensor, obtaining user real-time operation records, and obtaining a man-machine collaborative correction factor based on the user real-time operation records and the load state transition tensor;

[0010] S4: adjusting the grid light energy vector based on the light energy prediction calibration parameter, adjusting the load state transition tensor based on the man-machine collaborative correction factor, and recycling steps S1 to S4.

[0011] According to a preferred embodiment, a route grid topology matrix is generated, meteorological forecast data and initial battery state data are obtained, a grid light energy vector is obtained based on the meteorological forecast data and the route grid topology matrix, and an optimal light energy route is obtained according to the grid light energy vector and the initial battery state data, including:

[0012] The starting point coordinates and the target point coordinates are obtained, a map grid is generated based on the starting point coordinates and the target point coordinates, satellite map data is obtained, and a route grid topology matrix is obtained based on the map grid and the satellite map data;

[0013] The meteorological forecast data is obtained, a light energy conversion operation is performed based on the meteorological forecast data and the route grid topology matrix, the grid light energy vector is obtained, and an initial route set is generated based on the satellite map data;

[0014] An energy gain operation is performed based on the initial route set, the grid light energy vector and the grid energy consumption matrix, and a route energy gain time sequence is obtained;

[0015] The initial battery state data is obtained, and an optimal light energy route is obtained based on the initial battery state data and the route energy gain time sequence.

[0016] According to a preferred embodiment, the starting point coordinates and the target point coordinates are obtained, a map grid is generated based on the starting point coordinates and the target point coordinates, satellite map data is obtained, and a route grid topology matrix is obtained based on the map grid and the satellite map data, including:

[0017] The starting point coordinates and the target point coordinates are converted into starting point plane rectangular coordinates and target point plane rectangular coordinates based on UTM projection, and a plane path vector is obtained based on the starting point plane rectangular coordinates and the target point plane rectangular coordinates;

[0018] The plane path vector is taken as the X-axis, and the plane path vector is rotated counterclockwise by 90 degrees as the Y-axis to generate a path principal axis coordinate system;

[0019] The map grid is generated by dividing along the X-axis direction at an interval of N meters and dividing along the Y-axis direction at an interval of N meters;

[0020] 9 sampling points are uniformly arranged in each cell of the map grid, elevation data is added to each sampling point according to the satellite map data, a terrain label is marked for each cell based on the elevation data of adjacent sampling points, and a route grid topology matrix is generated.

[0021] According to a preferred embodiment, meteorological forecast data is obtained, a light energy conversion operation is performed based on the meteorological forecast data and the route grid topology matrix, a grid light energy vector is obtained, and an initial passing route set is generated based on the satellite map data, including:

[0022] Based on the meteorological forecast data, solar irradiance reference data and cloud cover change curves of each cell in the route grid topology matrix are obtained, and the grid light energy vector of each cell is obtained according to the solar irradiance reference data and the cloud cover change curves;

[0023] Based on the satellite map data, all feasible road node coordinates are extracted, and initial attractive force values are attached to all feasible road node coordinates. Starting from the starting point coordinate, adjacent feasible road node coordinates are retrieved, and the initial attractive force values of the adjacent feasible road node coordinates are compared with the total value of the initial attractive force values. The next feasible road node coordinate is selected until the target point coordinate is reached, and an initial passing route set is generated.

[0024] According to a preferred embodiment, an energy gain operation is performed based on the initial passing route set, the grid light energy vector, and the grid energy consumption matrix, and a route energy gain time sequence is obtained, including:

[0025] Vehicle unit distance basic energy consumption data is obtained, the slope value and the elevation data of each cell in the route grid topology matrix are extracted, the vehicle unit distance basic energy consumption is corrected based on the slope value and the elevation data, vehicle unit distance corrected energy consumption data is obtained, and the vehicle unit distance corrected energy consumption data is marked in each cell in the route grid topology matrix, and a grid energy consumption matrix is obtained;

[0026] The initial passing route set is mapped to the route grid topology matrix, and a passing route in the initial passing route set is extracted. The passing route is represented as a plurality of line segments from the starting point coordinate to the destination point coordinate. An expected passing time period is attached to each line segment, and the light energy value of the corresponding period in the grid light energy vector is extracted according to the expected passing time period;

[0027] Based on the light energy value and the grid energy consumption matrix, a grid gain value is obtained. The grid gain value is positive when the light energy value is greater than the energy consumption value, and negative when the light energy value is less than the energy consumption value. The grid gain values are arranged in the order of vehicle passing time to obtain a route energy gain time sequence.

[0028] According to a preferred embodiment, the initial battery state data is acquired, and the light energy optimal route is acquired based on the initial battery state data and the route energy gain time sequence, including:

[0029] The path total light energy gain of each passing route in the initial passing route set is extracted based on the route energy gain time sequence, and the path total energy consumption of each passing route in the initial passing route set is extracted based on the grid energy consumption matrix;

[0030] The initial battery state data is acquired, the path evaluation operation is performed according to the path total light energy gain, the path total energy consumption and the initial battery state data, and the comprehensive energy score is acquired;

[0031] The initial attraction value of all feasible road node coordinates is corrected based on the comprehensive energy score, the candidate passing route set is generated based on the corrected feasible road node coordinates, and the comprehensive energy score of each path is recalculated;

[0032] The passing route with the highest comprehensive energy score after recalculation is selected as the light energy optimal route.

[0033] According to a preferred embodiment, the predicted photovoltaic power generation curve is generated according to the light energy optimal route and the grid light energy vector, the actual photovoltaic power generation data is acquired, and the light energy prediction calibration data is acquired based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve, including:

[0034] The grid sequence in the light energy optimal route is extracted, the light energy values of each grid corresponding period are spliced according to the time axis according to the vehicle expected passing time, the route light energy time sequence is generated, the photovoltaic panel parameters are acquired, and the predicted photovoltaic power generation curve is generated based on the photovoltaic panel parameters and the route light energy time sequence;

[0035] The actual photovoltaic power generation data of the current period is acquired, the actual photovoltaic power generation data of the current period is aligned with the predicted photovoltaic power generation curve period, the predicted photovoltaic power generation data of the current period is acquired, the period deviation rate vector is acquired according to the actual photovoltaic power generation data and the predicted photovoltaic power generation data, and the light energy prediction calibration parameter is acquired based on the period deviation rate vector.

[0036] According to a preferred embodiment, the user habit model is constructed and the load state transition tensor is generated, the load scheduling instruction set is established according to the predicted photovoltaic power generation curve and the load state transition tensor, the user real-time operation record is acquired, and the man-machine collaborative correction factor is acquired based on the user real-time operation record and the load state transition tensor, including:

[0037] Obtaining the in-vehicle device operation log, removing the interference items with single operation less than 1 second and power exceeding the upper limit of the device from the in-vehicle device operation log, extracting the high-frequency behavior characteristics and scene-related behavior characteristics in the in-vehicle device operation log, generating time-driven rules, event-triggered rules and device cooperation strategies, and constructing a user habit model based on the time-driven rules, event-triggered rules and device cooperation strategies;

[0038] Based on the user habit model, obtain a device behavior feature set, obtain device power demand data, obtain a device state power constraint table based on the device power demand data, obtain a state transition trigger condition according to the device behavior feature set and the device state power constraint table, establish a state switching logic based on the state transition trigger condition, and generate a load state transition tensor based on the state transition trigger condition and the state switching logic;

[0039] According to the predicted photovoltaic power generation curve and the load state transition tensor, obtain a device target state set, and establish a load scheduling instruction set according to the device target state set;

[0040] Obtaining real-time operation records of the user, matching the user operation records with the load scheduling instruction set, generating a strategy deviation event set, and obtaining a man-machine cooperation correction factor based on the strategy deviation event set and the load state transition tensor.

[0041] According to a preferred embodiment, according to the predicted photovoltaic power generation curve and the load state transition tensor, obtaining a device target state set, and establishing a load scheduling instruction set according to the device target state set, comprising:

[0042] Based on the predicted photovoltaic power generation curve, identifying power supply gaps or surplus periods, and generating power supply state feature identifiers, obtaining device current state data, matching the power supply state feature identifiers and the device current state data with the load state transition tensor, and obtaining a device target state set;

[0043] Mapping the device current state data to specific power values according to the device target state set and the device power demand data, and adding an execution time window and a priority label to the device, and generating an initial load scheduling instruction set according to the execution time window and the priority label of the device;

[0044] Obtaining real-time battery state data and real-time location coordinate data, generating a safe operation boundary threshold based on the real-time battery state data and the real-time location coordinate data, calculating a future 30-minute power decay rate and predicting a time point of reaching a safety threshold according to the safe operation boundary threshold, and obtaining a power change prediction vector;

[0045] Based on the initial load scheduling instruction set, obtaining a device power demand sum, adjusting the device current state data based on the power change prediction vector and the device power demand sum, and obtaining a load scheduling instruction set.

[0046] The application discloses an intelligent control system of a new energy motor home.

[0047] A data acquisition module is configured to acquire meteorological forecast data, initial battery state data, actual photovoltaic power generation data and user real-time operation records.

[0048] A route planning module is configured to acquire a grid light energy vector according to the meteorological forecast data and a route grid topology matrix, and select an optimal light energy route according to the grid light energy vector and the initial battery state data.

[0049] A power generation prediction module is configured to generate a predicted photovoltaic power generation curve according to the optimal light energy route and the grid light energy vector, and acquire a light energy prediction calibration parameter based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve.

[0050] A load scheduling module is configured to construct a user habit model and generate a load state transition tensor, establish a load scheduling instruction set according to the predicted photovoltaic power generation curve and the load state transition tensor, and acquire a man-machine collaborative correction factor according to the user real-time operation records and the load state transition tensor.

[0051] An adjustment circulation module is configured to adjust the grid light energy vector according to the light energy prediction calibration parameter, adjust the load state transition tensor based on the man-machine collaborative correction factor, and control the route planning module, the power generation prediction module, the user modeling module and the scheduling control module to cyclically execute corresponding operations.

[0052] Compared with the prior art, the application has the following beneficial effects:

[0053] 1. By coordinating the driving route planning and the photovoltaic energy utilization, the energy optimal adaptation of route selection is realized, the route grid topology matrix is generated based on satellite map data, the grid light energy vector is constructed in combination with the meteorological forecast data, and then the optimal light energy route is selected according to the initial battery state data, so that the problem that the traditional new energy motor home control system cannot control the energy consumption of the in-vehicle equipment due to the lack of prediction of light energy supplement along the route, and thus the utilization efficiency of the photovoltaic system is low, is solved, and the route planning no longer depends on the distance or the traffic condition, but can actively avoid the route section with insufficient light energy and high energy consumption, and preferentially select the path with sufficient light energy supplement and optimal energy consumption, thereby avoiding the disconnection between the route selection and the energy acquisition, and realizing the stable power supply in long-distance travel.

[0054] 2. The dynamic calibration of photovoltaic power generation prediction improves the reliability of energy supply planning, generates an initial predicted photovoltaic power generation curve based on the optimal route of light energy and the grid light energy vector, and obtains light energy prediction calibration parameters by combining actual photovoltaic power generation data, adjusts the grid light energy vector in real time, ensures that the prediction result can timely reflect the changes of light conditions, reduces the prediction error of photovoltaic power generation, and avoids the deviation caused by meteorological fluctuations in the traditional fixed model prediction, so that the system can more accurately predict the power supply capacity at different times, thereby providing a reliable basis for load scheduling and reducing the occurrence of energy supply shortage caused by inaccurate power supply estimation.

[0055] 3. By constructing a user habit model and combining a load state transition tensor, the matching of energy utilization and user demand is realized, the load scheduling module extracts behavior characteristics based on user operation logs, generates a load state transition tensor in line with user habits, and formulates a load scheduling instruction set in combination with the predicted photovoltaic power generation curve, and simultaneously acquires a man-machine collaborative correction factor according to real-time user operation records to dynamically adjust the load state transition tensor, which not only ensures that the energy distribution can conform to the user's use preference, but also dynamically adjusts the equipment operation state according to the power supply capacity, avoiding the problem of not meeting the user's demand caused by unified scheduling, and improving the energy utilization efficiency.

[0056] 4. Based on the light energy prediction calibration parameters and the man-machine collaborative correction factor, the grid light energy vector and the load state transition tensor are dynamically adjusted respectively, and the route planning, power generation prediction, load scheduling and other links are iterated, this continuous optimization mechanism enables the system to continuously adapt to changes in meteorological conditions, route environment differences and user habit evolution, avoiding the problem of insufficient adaptability of traditional fixed logic control in complex scenarios, ensuring that the new energy house car always maintains an efficient and stable energy management state in the long-term use process, realizing the continuous optimization of the system and improving the self-adaptation ability. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 It is a step flow chart of the intelligent control method of the new energy house car.

[0058] Figure 2 It is a step flow chart of the intelligent control method of the new energy house car.

[0059] Figure 3 It is a step flow chart of the intelligent control method of the new energy house car.

[0060] Figure 4 It is a schematic diagram of the intelligent control system of the new energy house car.

[0061] Figure 5is a schematic diagram of an electronic device. DETAILED DESCRIPTION

[0062] For a further understanding of the present application, preferred embodiments thereof will be described in conjunction with the accompanying drawings and examples, it being understood that the description is merely illustrative of the features and advantages of the present application, and not in limitation of the scope of the claims.

[0063] Embodiment:

[0064] Referring to FIG. Figures 1 to 3 The present application provides an intelligent control method for a new energy motor home, comprising the following steps:

[0065] Step S1, generating a route grid topology matrix and obtaining meteorological forecast data and initial battery state data, obtaining a grid light energy vector based on the meteorological forecast data and the route grid topology matrix, and obtaining an optimal light energy route based on the grid light energy vector and the initial battery state data.

[0066] In this embodiment, obtaining the optimal light energy route comprises the following steps:

[0067] Step S10, obtaining a starting point coordinate and a target point coordinate, generating a map grid based on the starting point coordinate and the target point coordinate, obtaining satellite map data, and obtaining a route grid topology matrix based on the map grid and the satellite map data.

[0068] Specifically, the starting point coordinate and the target point coordinate are obtained by a data acquisition module, which are in the format of longitude and latitude, and are used to determine the starting point and the ending point of the trip. Satellite map data is called from a navigation data or a vehicle-mounted map database, and is converted into a starting point plane rectangular coordinate and a target point plane rectangular coordinate based on UTM projection. Conversion into a plane rectangular coordinate can eliminate the difference in distance conversion of longitude and latitude in different areas, facilitating subsequent grid division. A two-point connection vector is calculated based on the starting point plane rectangular coordinate and the target point plane rectangular coordinate. The vector serves as a plane path vector to obtain the general direction from the starting point coordinate to the target point coordinate, and provides a reference main axis for subsequent grid division. The starting point plane rectangular coordinate and the target point plane rectangular coordinate are taken as boundaries, the direction of the plane path vector is taken as the X-axis, and the direction of the plane path vector rotated counterclockwise by 90 degrees is taken as the Y-axis. In this way, a path main axis coordinate system is established. Through the setting of the path main axis coordinate system, the calculation of irrelevant areas far from the path can be reduced, and the efficiency can be improved. The X-axis direction is divided at intervals of N meters, and the Y-axis direction is divided at intervals of N meters, forming a map grid composed of a plurality of square cells.

[0069] Wherein, N meters is the interval value determined according to the terrain complexity of the route area, the vehicle's expected driving speed and the system data processing capacity. In the area where the terrain is relatively simple, the road is flat and the vehicle's driving speed is high, the N value can be selected as a larger value to reduce the number of single grids and thus reduce the data processing amount. In the area where the terrain is complex, contains more steep slopes or curves and the vehicle's driving speed is low, the N value can be selected as a smaller value to improve the analysis accuracy of the local terrain and light conditions.

[0070] Further, 9 sampling points are uniformly arranged in each single grid of the map grid. The elevation data of each sampling point is added by calling the elevation information of the corresponding position in the satellite map data. The elevation difference is calculated based on the elevation data of the adjacent sampling points, and the terrain label is labeled for the single grid according to the elevation difference. The terrain label includes flat label, gentle slope label, steep slope label and forest label. The average elevation difference is obtained based on all the elevation differences in the single grid. When the average elevation difference is less than the range interval, the flat label is labeled. When the average elevation is in the range interval, the gentle slope label is labeled. When the average elevation difference is greater than the range interval, the steep slope label is labeled. The range interval is represented as [5%N, 15%N]. If the satellite map data shows that there is a large amount of vegetation coverage in the single grid, the forest label is additionally added. These terrain labels are used for subsequent evaluation of the driving energy consumption of the vehicle in different grids and the light energy receiving condition of the photovoltaic panel. Finally, the spatial index and the terrain label of all single grids are integrated to generate the route grid topology matrix.

[0071] For example, assuming that the starting point is a campsite in the suburbs and the target point is a scenic spot in the mountains, after UTM projection conversion, the planar rectangular coordinates of the starting point are (X1, Y1), the planar rectangular coordinates of the target point are (X2, Y2), and the planar path vector is the vector from (X1, Y1) to (X2, Y2). The X axis is established along the general direction from the campsite to the scenic spot, and the Y axis is perpendicular to the direction. The map grid is divided along the X axis and the Y axis at an interval of 500 meters. The elevation data of the 9 sampling points in each grid is obtained through the satellite map. The sampling points in a certain grid have small elevation difference and no vegetation coverage, and are labeled as flat label. The sampling points in another grid have large elevation difference and steep slope, and are labeled as steep slope label. These labels and the spatial information of the grid jointly constitute the route grid topology matrix, which provides a basis for subsequent calculation of the energy consumption and light energy utilization in the area.

[0072] In some possible embodiments, when the value of N is large, the geographical range covered by a single grid is wide, and the interior can contain various terrains such as flat land, gentle slope, and steep slope. At this time, if the label of a single grid is determined only by calculating the average of the elevation differences of all adjacent sampling points, the influence of local steep terrain on vehicle energy consumption will be hidden, and the label based on the average elevation difference cannot accurately represent the overall terrain characteristics in the single grid. Therefore, on the basis of retaining 9 basic sampling points, the number of sampling points is increased to 16 by adding the edge midpoint and the center subdivision point of the single grid, so as to improve the capture accuracy of local terrain. The single grid is divided into 4 quadrant sub-regions according to the latitude and longitude, each sub-region contains 4 sampling points, the elevation difference and the corresponding slope of adjacent sampling points in each sub-region are calculated, and a sub-label is labeled for each sub-region. The sub-label is represented as a quadrant sub-region + terrain. The main label of the single grid is determined according to the area proportion of each sub-region. If the area of a certain type of sub-region accounts for more than 50% of the total area of the single grid, the label is taken as the main label. At the same time, a secondary label and a proportion are appended to the main label. In the energy consumption calculation, the overall label of the single grid is no longer used as the basis, but the slope parameters of the corresponding sub-label are called according to the sub-region passed by the actual driving track of the vehicle to correct the energy consumption.

[0073] For example, it is assumed that the vehicle drives to the northeast quadrant-steep slope, and the energy consumption is calculated according to the steep slope standard. When the vehicle drives to the northwest quadrant-flat land, the energy consumption is calculated according to the flat land standard. Through the subdivision of the region and the layered label, the overall terrain feature description of the single grid is retained, and the distribution of the local steep terrain is reflected, so that the high-influence terrain is not hidden due to the averaging process. When the vehicle driving route passes through a certain quadrant, the energy consumption correction coefficient corresponding to the quadrant-terrain is automatically called to ensure that the influence of the local steep terrain on the energy consumption is accurately taken into account, and the energy consumption prediction deviation caused by the actual passing terrain being steep while the overall label being flat is reduced.

[0074] It needs to be further explained that when there are multiple passable roads in a single grid, the driving directions of different roads can form different angles with the slope direction in the single grid, which directly affects the vehicle energy consumption and passing time. Therefore, the slope direction is added to the terrain label based on the elevation data of the 9 sampling points, so that the influence of the terrain label on the energy consumption and the passing time can be analyzed based on the driving direction and the slope direction on the passable road.

[0075] It can be understood that the elevation difference direction of adjacent points is calculated based on the elevation data of 9 sampling points, the highest proportion of the slope direction in the single grid is counted, and the direction is taken as an additional attribute of the terrain label. When there are multiple roads in the single grid, the driving directions of the roads are extracted, and the driving directions of the roads are compared with the slope direction. If the driving direction of the road is consistent with the slope direction, the vehicle energy consumption is reduced, and the passing time is shortened. If the driving direction is opposite to the slope direction, the energy consumption is increased, and the passing time is prolonged.

[0076] In step S11, meteorological forecast data is acquired, light energy conversion operation is performed based on the meteorological forecast data and the route grid topology matrix, a grid light energy vector is acquired, and an initial passing route set is generated based on satellite map data.

[0077] Specifically, the meteorological forecast data is acquired by the data acquisition module, wherein the meteorological forecast data includes solar irradiance reference data within 24 hours in the future and a cloud cover change curve, the solar irradiance reference data is represented as the unit area illumination intensity without cloud layer shielding, the cloud cover change curve takes time as the horizontal axis and the proportion of cloud layer shielding the sun as the vertical axis, 0 represents no cloud, and 1 represents complete shielding; in combination with the terrain label of each single grid in the route grid topology matrix, the light energy value of each single grid in the vehicle expected passing period is calculated, the light energy values are arranged in the order of the spatial index of the single grid to form a grid light energy vector, which provides a quantitative basis for subsequent evaluation of the light energy supplement capacity of different routes; based on the road layer in the satellite map data, all road intersections, turning points and the like available for vehicle passing are extracted as feasible road node coordinates, each node coordinate corresponds to a specific position on the actual road, the same initial attraction value is added to all feasible road node coordinates, the initial attraction value of each adjacent node coordinate is calculated based on the proportion of the initial attraction value of each adjacent node coordinate in the total initial attraction value of all adjacent node coordinates, the next feasible road node coordinate is randomly selected according to the proportion, the process is repeated until the selected node reaches the target point coordinate, thereby generating a plurality of different paths to form an initial passing route set, which provides a basis for subsequent selection of the optimal light energy route.

[0078] For example, if the solar irradiance reference data of a single grid in the route grid topology matrix in the vehicle passing period is 1000 W / m2, the cloud cover change curve in the corresponding period shows that the cloud layer shielding proportion is 0.3, and the terrain label of the single grid is flat, then the light energy value of the single grid is 700 W / m2, all the above data are hypothetical and the specific data are determined according to the actual situation; the light energy values of a plurality of single grids are arranged in the order of the index to form a grid light energy vector; at the same time, the adjacent road nodes around the starting point are extracted from the satellite map data, and an initial attraction value is added to each adjacent road node, the next node is selected according to the proportion of the initial attraction value of each adjacent node in the total initial attraction value of all adjacent nodes, when the initial attraction values of the adjacent nodes are the same, the next node is randomly selected and the selection is repeated until the target point is reached, thereby generating an initial passing route, and the process is repeated multiple times to generate an initial passing route set including a plurality of different paths.

[0079] In step S12, energy gain operation is performed based on the initial passing route set, the grid light energy vector and the grid energy consumption matrix to acquire a route energy gain time sequence.

[0080] Specifically, the energy consumption of the vehicle per 100 kilometers on flat road is first retrieved as the vehicle unit distance basic energy consumption data, the slope value and the altitude data of each single grid in the route grid topology matrix are extracted, the preset correction coefficient table is queried based on the slope value, and the vehicle unit distance basic energy consumption is corrected in combination with the altitude data to obtain vehicle unit distance correction energy consumption data. These correction energy consumption data are marked in the route grid topology matrix according to the single grid index to form a grid energy consumption matrix, which serves to provide quantitative data for the actual energy consumption of each road section and ensure that the energy consumption calculation is in line with the actual terrain.

[0081] The generation of the correction coefficient table is based on multidimensional data collection, statistical analysis and verification adjustment, and the specific process is as follows: first, the core variables affecting the energy consumption of the vehicle are determined, including the terrain slope and the driving direction, and these variables are combined into specific condition items; second, field data collection is performed, typical road sections covering all the above condition items are selected, more than 3 test road sections of different lengths are matched for each condition item, under the same environmental conditions, a new energy house car of the corresponding type is arranged to drive back and forth at a constant speed, the actual energy consumption per unit distance is recorded through the energy consumption monitoring system of the vehicle, the test of each condition item is repeated more than 10 times to reduce the error caused by accidental factors; then the average energy consumption of all tests under the same condition item is calculated, the average energy consumption is compared with the baseline energy consumption to obtain the initial correction coefficient of the condition item, and then verification is carried out, road sections and vehicles not involved in the initial test are selected to test the energy consumption under the same condition item, the actual energy consumption is compared with the calculation result of the corrected energy consumption, if the deviation exceeds 5%, the test data of the condition item is re-collected and the coefficient is adjusted until the deviation of all condition items is controlled within 5%; finally, all verified correction coefficients are classified according to variable combination and entered into a table to form a preset correction coefficient table.

[0082] Further, the initial passing route set is mapped into the route grid topology matrix, that is, the specific single grid passed by each passing route is determined through coordinate matching, the passing route composed of these single grids is extracted, the expected passing period of the vehicle passing each line segment is estimated according to the length of each line segment and the terrain label of the single grid through which the line segment passes, and then the light energy value of the single grid at the corresponding period is extracted from the grid light energy vector according to the expected passing period; based on the extracted light energy value and the unit distance corrected energy consumption data of the corresponding single grid in the grid energy consumption matrix, the grid gain value of each single grid is obtained, wherein the grid gain value is represented as the size relationship between the light energy value and the unit distance corrected energy consumption data, when the light energy value is greater than the unit distance corrected energy consumption data, the grid gain value is positive; when the light energy value is less than the unit distance corrected energy consumption data, the grid gain value is negative, and the grid gain values are arranged in time sequence according to the time sequence of the vehicle passing each single grid, to form a route energy gain time sequence; the route energy gain time sequence is used to intuitively show the energy surplus or deficit of the route at different time nodes, and provide energy change basis in time dimension for subsequent screening of the light energy optimal route.

[0083] For example, assuming that a certain initial passing route in the initial passing route set passes through 3 single grids, wherein the first single grid is flat ground with a slope of 0 degrees, an altitude of 100 meters, a unit distance basic energy consumption of 10 kWh / 100km, a corrected energy consumption of 10 kWh / 100km, a road segment length of 0.5 kilometers, a total energy consumption of 0.5 kWh, and an expected passing period of 9:00-9:00.5 in the morning, the light energy value corresponding to the grid light energy vector at the period is 0.8 kWh, and the grid gain value is 0.3 kWh; the second single grid is steep slope with a slope of 5 degrees, an altitude of 200 meters, a unit distance basic energy consumption corrected to 12 kWh / 100km, a road segment length of 0.8 kilometers, a total energy consumption of 0.96 kWh, and an expected passing period of 9:00.5-9:01.3, the corresponding light energy value is 0.7 kWh, and the grid gain value is -0.26 kWh; the third single grid is downhill with a slope of -3 degrees, an altitude of 150 meters, a corrected energy consumption of 8 kWh / 100km, a road segment length of 0.6 kilometers, a total energy consumption of 0.48 kWh, and an expected passing period of 9:01.3-9:02, the corresponding light energy value is 0.6 kWh, and the grid gain value is 0.12 kWh; the three gain values are arranged in time sequence to obtain a route energy gain time sequence [0.3 kWh, -0.26 kWh, 0.12 kWh], which clearly shows the energy change of the vehicle passing through the route, and the above data is only used for logical reference, and the specific data needs to be determined according to the specific situation.

[0084] In step S13, initial battery state data is obtained, and the light energy optimal route is obtained based on the initial battery state data and the route energy gain time sequence.

[0085] Specifically, the battery power is acquired as initial battery state data, the path total light energy gain of each passing route in the initial passing route set is extracted based on the algebraic sum of each grid gain value in the route energy gain time sequence, that is, the cumulative sum of positive values in all grid gain values, representing the total amount of additional light energy supplement that can be obtained throughout the route; at the same time, the path total energy consumption of each passing route is accumulated based on the unit distance correction energy consumption data of the single grid through which the corresponding route passes in the grid energy consumption matrix, that is, the total amount of energy consumption throughout the vehicle travel; the current power in the initial battery state data is acquired, and the path evaluation operation is performed according to the path total light energy gain, the path total energy consumption and the initial battery state data, to calculate the comprehensive energy score;

[0086] Wherein, the path evaluation operation is represented as the initial battery state data plus the path total light energy gain minus the path total energy consumption, if the result is positive and greater than the minimum reserved power required for the vehicle to reach the target point, then the comprehensive energy score is the result, if the result is less than the minimum reserved power, then the comprehensive energy score is the difference between the result minus the minimum reserved power, based on the comprehensive energy score, the initial attractive force value of all feasible road node coordinates is modified, for the nodes contained in the passing route with higher comprehensive energy score, the attractive force value is increased, and for the nodes contained in the route with lower score, the attractive force value is decreased, based on the modified feasible road node coordinates, the candidate passing route set is generated from the starting point node according to the same search method as generating the initial passing route set, and the comprehensive energy score of each candidate route is recalculated, and iteration is performed in this way until the passing route with the highest comprehensive energy score and the result being positive appears as the light energy optimal route, which can ensure that the energy reserve of the vehicle is always higher than the minimum reserved power during the travel process, avoiding depletion of power halfway;

[0087] For example, the current power in the initial battery state data is 50 kWh, the minimum reserved power is 10 kWh, the initial set of passable routes includes two routes, the total path light energy gain of route one is 20 kWh, the total path energy consumption is 55 kWh, and the comprehensive energy score is 50+20-55=15 kWh. The total path light energy gain of route two is 15 kWh, the total path energy consumption is 60 kWh, and the comprehensive energy score is 50+15-60=5 kWh. Based on the score correction node attraction value, the node attraction value of route one is increased, and the node attraction value of route two is decreased. In the generated candidate passable route set, the revised version of the comprehensive energy score of route one is increased to 18 kWh, and the revised version of the comprehensive energy score of route two is decreased to 3 kWh. Therefore, route one is selected as the optimal light energy route. In this way, the driving route planning and photovoltaic energy utilization are coordinated, the route grid topology matrix provides a structured analysis basis for geographic space, the grid light energy vector quantifies the light energy potential of different road sections, and the optimal light energy route selected in combination with the initial battery state data enables the motor home to preferentially select a path with more sufficient light energy supplement and more reasonable energy consumption during driving, solving the problem of traditional new energy motor home control systems that only consider distance or road conditions for route selection without considering geographical and meteorological correlations, reducing the disconnection between route selection and energy acquisition, and helping to maintain the stability of battery endurance and provide support for stable power supply during long-distance travel.

[0088] In step S2, a predicted photovoltaic power generation curve is generated according to the optimal light energy route and the grid light energy vector, actual photovoltaic power generation data is obtained, light energy prediction calibration data is obtained based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve, and the power generation prediction result is dynamically corrected.

[0089] Specifically, all grids passed by the optimal light energy route are grouped into a grid sequence according to the driving order, each grid corresponds to a predicted passing time period, the light energy values of the grids in the time period are sequentially spliced according to the time axis order to form a continuous route light energy time sequence, wherein the route light energy time sequence takes time as the horizontal axis and light energy value as the vertical axis, and is used to show the trend of light energy change with time during the whole driving process of the vehicle; the conversion efficiency in the photovoltaic panel parameters is obtained, and the predicted power generation in each time period is calculated based on the conversion efficiency and the route light energy time sequence, and the predicted power generation is connected according to the time axis to generate a predicted photovoltaic power generation curve, which is used to predict the photovoltaic power generation in different time periods during the driving process of the vehicle; the actual photovoltaic power generation data of the photovoltaic panel in the current time period is obtained through the data acquisition module, the actual photovoltaic power generation data in the current time period is aligned with the part in the same time period of the predicted photovoltaic power generation curve, and the predicted photovoltaic power generation data in the time period is obtained; the deviation rate of each corresponding time period is calculated according to the actual photovoltaic power generation data and the predicted photovoltaic power generation data, and the ratio of the actual photovoltaic power generation data minus the predicted photovoltaic power generation data to the predicted photovoltaic power generation data is obtained, so as to obtain a plurality of deviation rates in different time periods, and the deviation rates in different time periods are grouped into a time period deviation rate vector, and the light energy prediction calibration parameter is obtained based on the time period deviation rate vector; wherein the light energy prediction calibration parameter is represented as the average value of each deviation rate in the time period deviation rate vector, and is used to correct the predicted photovoltaic power generation data in the subsequent time period.

[0090] For example, the grid sequence of the optimal light energy route contains 3 grids, the first grid is predicted to pass through the time period of 10:00-10:10, the light energy value is 5kWh, the second grid is 10:10-10:20, the light energy value is 6kWh, and the third grid is 10:20-10:30, the light energy value is 4kWh, and the route light energy time sequence is generated by splicing according to the time axis [10:00-10:10:5kWh, 10:10-10:20:6kWh, 10:20-10:30:4kWh]; the conversion efficiency in the photovoltaic panel parameters is 20%, and the predicted power generation of the predicted photovoltaic power generation curve corresponding to the time period is 1kWh, 1.2kWh and 0.8kWh; the actual photovoltaic power generation data collected at 10:00-10:10 is 1.1kWh, and compared with the predicted value 1kWh, the deviation rate is 0.1, 10:10-10:20 is 1.14kWh, the deviation rate is-0.05, the time period deviation rate vector is [0.1, -0.05], the average deviation rate is 0.025, the light energy prediction calibration parameter is 0.025, which is used to correct the predicted value of 10:20-10:30, and the corrected value is 0.8x(1+0.025)=0.82kWh. The above data is only used for logical reference, and the specific data should be determined according to the specific situation.

[0091] It can be understood that, through the above-mentioned mode, the initial predicted photovoltaic power generation curve generated based on the optimal route of light energy and the grid light energy vector, combined with the calibration parameters obtained from the actual data, can adjust the grid light energy vector in real time, reduce the prediction error caused by meteorological changes and other factors, avoid the problem that the equipment may run with insufficient power after running at high power supply due to high predicted power generation but low actual power generation, or waste power generation due to low predicted power generation but high actual power generation, and ensure the matching of energy supply and demand, providing support for stable power supply.

[0092] Step S3, constructing a user habit model and generating a load state transition tensor, establishing a load scheduling instruction set according to the predicted photovoltaic power generation curve and the load state transition tensor, obtaining user real-time operation records, and obtaining a man-machine cooperative correction factor based on the user real-time operation records and the load state transition tensor.

[0093] Specifically, an operation log of an in-vehicle device is acquired, wherein the operation log at least includes switch time, running power, and duration data of a load device on the vehicle, false touch data with single operation less than 1 second and interference data with power exceeding the rated upper limit of the device are removed from the operation log, high-frequency behavior features and scene-related behavior features in the remaining valid records are extracted; wherein the high-frequency behavior features represent that the user repeatedly performs the device operation behavior with a higher frequency in a fixed time period in the long-term use process, and the stable use habit of the user is reflected through the repetition frequency in the time dimension, and the scene-related behavior features represent the association between the device operation behavior of the user and the specific scene state of the vehicle, and the adaptation relationship between the user operation and the environmental condition is reflected through the change of the scene state; a time-driven rule is generated based on the high-frequency behavior features, an event trigger rule and a device cooperation strategy are generated based on the scene-related behavior features, and the time-driven rule, the event trigger rule, and the device cooperation strategy are integrated to build a user habit model, the user habit model can be constructed by using convolutional neural network, machine learning, big data analysis technology, etc., and assuming that a supervised learning algorithm in machine learning is selected for construction: the extracted high-frequency behavior features and scene-related behavior features are converted into structured input feature vectors, the corresponding device operation results are taken as output labels, a certain proportion of data is selected from the valid records as a training set, and the rest is taken as a validation set, the training set is trained by the supervised learning algorithm, the mapping relationship between the features and the labels is calculated through iteration, the model parameters are adjusted to gradually reduce the prediction error to a preset range; the prediction accuracy of the model is evaluated by using the validation set, if the prediction accuracy of a certain type of operation reaches a preset standard, the training of the user habit model is completed, which is used to predict the use demand of the user for the device in different time periods, and in the subsequent process, a certain amount of new operation logs are accumulated, the model is retrained with the new data to update the parameters, so as to ensure that the prediction of the model on the user behavior is close to the actual operation; a device behavior feature set is extracted based on the user habit model, device power demand data is obtained through a device parameter table, a device state power constraint table is generated according to the device power demand data, the state transition trigger condition is determined by combining the device behavior feature set and the device state power constraint table, the state switching logic is established based on the trigger condition, the state transition trigger condition, the state switching logic, and the corresponding power change are integrated to generate a load state transition tensor, and the tensor is used to describe the state conversion relationship and power demand of the device under different conditions.

[0094] The time-driven rule is represented as an operation rule based on a fixed time period in the user high-frequency behavior feature, and is a pre-set device automatic operation rule, that is, when the system detects that the current time falls into a time period in which the user usually operates a device, the device is automatically triggered to operate in a conventional mode; the event-triggered rule is represented as a pre-set device response rule based on the association between the scene state and the operation in the user scene-associated behavior feature, that is, when the system detects that the house or car enters a specific scene state, the device is automatically triggered to operate according to the associated habit; and the device coordination strategy is represented as a pre-set cooperation rule between devices based on the behavior feature of multi-device linkage in the user operation, that is, when a certain device is started or switched, the associated device adjusts the operating state according to the conventional linkage logic.

[0095] Specifically, the surplus power supply period and the gap period are identified according to the predicted photovoltaic power generation curve, the device target state set is matched in combination with the load state transition tensor, the load scheduling instruction set is generated according to the device target state set, and the load scheduling instruction set at least includes the planned operation period, the power range and the priority of each device; the user real-time operation record is obtained, the user real-time operation record is compared with the load scheduling instruction set, when the user operation is inconsistent with the instruction, the inconsistent operation record is combined into a strategy deviation event set, the frequency of the inconsistent operation record is compared with the load state transition tensor, and a man-machine cooperation correction factor is obtained, which is used to adjust the user habit model.

[0096] In this embodiment, an example is given in a simple and easy-to-understand manner, and the specific data content depends on the specific situation. It is assumed that the in-vehicle device operation log shows that the user uses the electric oven continuously from 19:00 to 20:00 for one week, and the electric oven is turned on within 10 minutes after parking. After removing the interference items, the high-frequency behavior feature is extracted, which is represented as 19:00-20:00 electric oven heating, and the scene association feature is represented as electric oven starting after parking. The time-driven rule is generated, which is represented as 19:00-20:00 allowing the electric oven to be in a heatable state, the event trigger rule is "after detecting the parking state, the electric oven enters the standby mode", and the user habit model is constructed. Based on the model, the device behavior feature set includes the time period and power of the electric oven, and the device power demand data shows that the electric oven heating mode is 2000W and the standby mode is 50W. The device state power constraint table is generated, and the state transition trigger condition can be represented as 19:00 and the parking state, and the electric oven is switched from standby to heating. The state switching logic can be represented as the electric oven is switched from heating to standby when the power is less than 25%. Thus, the load state transition tensor is generated. The predicted photovoltaic power generation curve shows that 19:00-19:30 is a power supply surplus period and 19:30-20:00 is a power supply gap period. The target state set of the device is matched with the tensor, which can be represented as 19:00-19:30 electric oven heating and standby after 19:30. The load scheduling instruction set is generated. If the user's real-time operation record shows that the electric oven is still heated at 19:30-20:00 for three consecutive days, a strategy deviation event is generated, a correction factor is calculated to adjust the state transition trigger condition, and the electric oven can maintain heating during the power supply gap period if the current power is greater than or equal to 30%. This makes the scheduling more in line with the user's actual use requirements.

[0097] It can be understood that intelligent load scheduling is achieved by constructing a user habit model and combining a load state transition tensor. The load scheduling module extracts behavior features based on user operation logs, generates a load state transition tensor that conforms to user habits, and then formulates a load scheduling instruction set in combination with a predicted photovoltaic power generation curve. At the same time, a human-machine collaborative correction factor is obtained according to the user's real-time operation record to dynamically adjust the load state transition tensor. This process enables the load scheduling to match the supply rhythm of photovoltaic energy and to conform to the user's actual use preferences, avoiding the problem of disconnection between traditional unified scheduling strategies and user needs. In addition, it improves energy utilization efficiency while enhancing user convenience and comfort.

[0098] In this embodiment, the establishment of the load scheduling instruction set includes the following steps:

[0099] Step S30, based on the predicted photovoltaic power generation curve, identify the power supply gap or surplus period, and generate a power supply state feature identifier. Obtain the current state data of the device, match the power supply state feature identifier and the current state data of the device with the load state transition tensor, and obtain the target state set of the device.

[0100] Specifically, the power generation of each period in the predicted photovoltaic power generation curve is analyzed and compared with the preset basic load power. When the power generation continuously falls below the basic load power for more than 15 minutes, the period is identified as a power supply gap period. When the power generation continuously rises above the sum of the basic load power and the maximum charging power of the energy storage device for more than 15 minutes, the period is identified as a power supply surplus period, and the corresponding power supply state feature identifier is generated. The current state data of the device is obtained, which at least includes the switching state, operating mode, and current power of the device. The power supply state feature identifier and the current state data of the device are input into the load state transition tensor. The tensor is matched according to the preset state transition rule, and the target state set of the device is output. The set contains the state that each device should be in under the current power supply state.

[0101] For example, assuming that the predicted photovoltaic power generation curve shows that 14:00-15:00 is a power supply gap period, and the current time is 13:50, the device state data shows that the microwave oven is in an open state with a power of 800W. After matching the load state transition tensor, the output device target state set is "microwave oven is closed before 14:00". The above example is expressed in a simple and understandable way, and the specific data content depends on the specific situation.

[0102] Step S31, mapping the current state data of the device to a specific power value according to the target state set of the device and the power demand data of the device, and adding an execution time window and a priority label to the device, generating an initial load scheduling instruction set according to the execution time window and the priority label of the device.

[0103] Specifically, based on the target state set of the device and the power demand data of the device, the current state of the device is mapped to a specific power value, an execution time window is added to each device operation, and the determination of the time window is based on the predicted photovoltaic power generation curve and the user habit model. Priority labels are assigned to load devices, wherein the priority labels are determined according to the importance of device functions and the frequency of user use. The target power value, execution time window and priority label of the device are integrated to generate an initial load scheduling instruction set, which specifies the operating parameters of each device at different periods.

[0104] Step S32, obtaining real-time battery state data and real-time location coordinate data, generating a safety operation boundary threshold value based on the real-time battery state data and the real-time location coordinate data, calculating the power decay rate in the next 30 minutes and predicting the time point of reaching the safety threshold according to the safety operation boundary threshold value, and obtaining the power change prediction vector.

[0105] Specifically, the current battery power is obtained as real-time battery state data and real-time position coordinate data; a battery safe operation boundary threshold is generated based on the real-time battery state data, and the threshold is dynamically adjusted in combination with the real-time position coordinate data; the power decay rate of the next 30 minutes is calculated based on the current device running state and the predicted photovoltaic power generation curve; based on the decay rate and the safe operation boundary threshold, the time point of reaching the safety threshold is predicted, and the power prediction value of each minute in the next 30 minutes is arranged in time sequence to form a power change prediction vector.

[0106] For example, assuming that the real-time battery state data shows that the current power is 45%, the real-time position coordinate shows that the distance to the charging station is 30 kilometers, the safety threshold is 25%, and the power decay rate of the next 30 minutes is calculated to be 1% per minute, it is predicted that the safety threshold will be reached after 10 minutes, and the power change prediction vector is [44%, 43%,..., 35%].

[0107] Step S33, based on the initial load scheduling instruction set, the total power demand of the device is obtained, and the current state data of the device is adjusted based on the power change prediction vector and the total power demand of the device, to obtain a load scheduling instruction set.

[0108] Specifically, the power demands of each device in the same period in the initial load scheduling instruction set are added to obtain the total power demand of the device; the total power demand of the device is compared with the power change prediction vector, when the predicted power of a period is lower than the safety threshold, the device state is adjusted according to the device priority, and the adjusted total power demand of the device is recalculated, and compared again with the power change prediction vector to ensure that the power of each period is not lower than the safety threshold; the adjusted device state data is updated to the initial load scheduling instruction set to obtain the final load scheduling instruction set.

[0109] For example, assuming that the initial instruction set shows that the total demand of all devices from 14:00 to 15:00 is 800W, but the power change prediction vector shows that the power will drop to 18% at the end of the period, which is lower than the safety threshold of 25%, then the microwave oven with priority 3 and the water heater with priority 2 are turned off, and necessary devices such as refrigerators and lighting are retained, and the total demand is reduced to 200W after adjustment, ensuring that the power is maintained above the safety threshold. The above data is only used for logical explanation, and the specific data is determined according to the specific situation.

[0110] Step S4, based on the light energy prediction calibration parameter, the grid light energy vector is adjusted, based on the human-computer collaborative correction factor, the load state transfer tensor is adjusted, and steps S1 to S4 are cycled.

[0111] Specifically, the light energy prediction calibration parameter is applied to the light energy value of the corresponding period in the grid light energy vector, and the correction is performed through multiplication operation, and a confidence level is marked for each calibrated grid according to the absolute value of the deviation rate, and an adjusted grid light energy vector is formed, which can better fit the actual light conditions and provide more accurate light energy data for subsequent route planning;Based on the human-computer collaborative correction factor, extract high-frequency deviated device data, adjust the state transition trigger condition of these deviated devices according to the correction factor, update the state switching logic, and generate an adjusted load state transition tensor, so that the tensor is more in line with the actual use habits of users;The adjusted grid light energy vector and the load state transition tensor are input into steps S1 to S4 for cyclic iteration, so that the new energy house car can continuously adapt to the changes of external environment and user behavior during driving.

[0112] For example, assume that the new energy house car drives to a section of suddenly overcast road, the light energy prediction calibration parameter shows that the light energy value of the current grid needs to be reduced by 30%, the adjusted grid light energy vector corrects the grid light energy value from 800Wh to 560Wh, and marks it as "medium confidence"; At the same time, the user manually turns on the air conditioner during the power gap period in the past three days, and the human-computer collaborative correction factor records this behavior, and the adjusted load state transition tensor adjusts the state transition trigger condition of the air conditioner from "prohibit opening during power gap period" to "allow opening during 17:30-18:30"; Through the cycle of S1 to S4, the system re-plans the route based on the corrected grid light energy vector, and generates new load scheduling instructions based on the adjusted tensor, so that the route selection and load control are more in line with the actual light and user operation;This continuous optimization mechanism enables the system to adapt to real-time changes in weather conditions, differences in route environment, and evolution of user habits, avoiding energy management mismatch problems caused by changes in environment or behavior in traditional fixed logic control, ensuring that the new energy house car always adapts to actual demand in long-term use, and maintains efficient and stable power supply state.

[0113] Please refer to Figure 4 As shown in the figure, the present application also provides an intelligent control system for a new energy house car, comprising:

[0114] The data acquisition module can obtain meteorological forecast data within the next 24 hours through the vehicle-mounted meteorological receiving unit, including solar irradiance and cloud cover curve, for evaluating the light conditions at different times;Obtain initial battery state data and real-time battery state data through the battery management system;The photovoltaic panel parameters of the photovoltaic panel installed on the new energy house car are input in advance, and the actual photovoltaic power generation data of the photovoltaic panel can be collected;Connected with the in-vehicle control panel of the new energy house car to obtain real-time operation records of the user.

[0115] The route planning module generates a route grid topology matrix based on the weather forecast data provided by the data acquisition module and satellite map data, calculates light energy values of each grid in a predicted passing period to form a grid light energy vector, and then compares total light energy gain and total energy consumption of different routes based on initial battery state data to select a route with the most energy surplus as an optimal light energy route.

[0116] The power generation prediction module calculates predicted power generation of each period based on photovoltaic panel parameters provided by the data acquisition module, the optimal light energy route determined by the route planning module and the grid light energy vector, and splices a prediction photovoltaic power generation curve according to a time axis; meanwhile, the actual photovoltaic power generation data is received, and the actual value and the predicted value of the same period are compared to calculate a deviation rate, and the deviation rate of multiple periods is used to generate a light energy prediction calibration parameter.

[0117] The load scheduling module extracts high-frequency behavior features and scene association features after removing interference items with a single operation less than 1 second from user historical operation logs collected by the data acquisition module, constructs a user habit model, generates a device behavior feature set based on the user habit model, forms a device state power constraint table combined with device power demand data, and then determines state transition trigger conditions and switching logic to generate a load state transition tensor; the prediction photovoltaic power generation curve of the power generation prediction module is combined to match the target state of the device in different power supply periods to generate a load scheduling instruction set, and a man-machine collaborative correction factor is generated by comparing the user real-time operation record and the load scheduling instruction set.

[0118] The adjustment cycle module receives the light energy prediction calibration parameter of the power generation prediction module, corrects the grid light energy vector in the route planning module, updates the state transition trigger condition in the load state transition tensor based on the man-machine collaborative correction factor of the load scheduling module, and then controls the route planning module to re-plan the route based on the corrected grid light energy vector, the power generation prediction module to generate a corrected prediction curve based on the new route, and the load scheduling module to adjust the scheduling instruction set based on the updated tensor, and the cycle is repeated.

[0119] As shown in Figure 5 , the present application also provides an electronic device, which comprises:

[0120] at least one processor; and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method proposed in Embodiment One of the present application.

[0121] The various components of the electronic device will be described in detail as follows:

[0122] The processor is the control center of the electronic device, and can be one processor or a combination of multiple processing elements. For example, the processor is one or more central processing units (CPUs), application specific integrated circuits (ASICs), or one or more integrated circuits configured to implement an embodiment of the present application, such as one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs).

[0123] The processor can execute various functions of the electronic device by running or executing software programs stored in the memory and calling data stored in the memory.

[0124] The memory is used to store software programs for implementing the present application, and is controlled by the processor to execute. The specific implementation manner can refer to the above method embodiments, and will not be described here.

[0125] The memory can be a read-only memory (ROM) or other type of static storage device that can store static information and instructions, a random access memory (RAM) or other type of dynamic storage device that can store information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only (CD-ROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto. The memory can be integrated with the processor or exist independently and coupled to the processor through the interface circuit of the electronic device, and the present application is not limited in this regard.

[0126] The above-described embodiments can be implemented in whole or in part by software, hardware (such as a circuit), firmware, or any combination thereof. When implemented in software, the above-described embodiments can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are wholly or partially generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transferred from one website, computer, server or data center to another website, computer, server or data center by limited (for example, infrared, wireless, microwave, etc.) mode. The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, data center, etc. containing one or more available medium collections. The available medium can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), an optical medium (for example, DVD), or a semiconductor medium. The semiconductor medium can be a solid state disk.

[0127] It should be understood that the term "and / or" herein merely describes an association relationship of associated objects, which means that there can be three relationships, for example, A and / or B, which can represent three cases of A alone, A and B together, and B alone, where A and B can be singular or plural. In addition, the character " / " herein generally represents an "or" relationship between the front and rear associated objects, but can also represent an "and / or" relationship. The specific meaning can be understood according to the context before and after.

[0128] It should be understood that in the embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution, and the execution order of the processes should be determined according to their functions and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0129] The above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for intelligent control of a new energy motor home, characterized in that, The method comprises: S1: generating a route grid topology matrix and obtaining meteorological forecast data and initial battery state data, obtaining a grid light energy vector based on the meteorological forecast data and the route grid topology matrix, and obtaining an optimal light energy route based on the grid light energy vector and the initial battery state data; S2: generating a predicted photovoltaic power generation curve based on the optimal light energy route and the grid light energy vector, obtaining actual photovoltaic power generation data, and obtaining light energy prediction calibration data based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve; S3: constructing a user habit model and generating a load state transition tensor, establishing a load scheduling instruction set based on the predicted photovoltaic power generation curve and the load state transition tensor, obtaining user real-time operation records, and obtaining a man-machine collaborative correction factor based on the user real-time operation records and the load state transition tensor; S4: adjusting the grid light energy vector based on the light energy prediction calibration parameter, adjusting the load state transition tensor based on the man-machine collaborative correction factor, and repeating steps S1 to S4.

2. The intelligent control method of the new energy recreational vehicle according to claim 1, characterized in that, Generating a route grid topology matrix and obtaining meteorological forecast data and initial battery state data, obtaining a grid light energy vector based on the meteorological forecast data and the route grid topology matrix, and obtaining an optimal light energy route based on the grid light energy vector and the initial battery state data, comprising: Obtaining a starting point coordinate and a target point coordinate, generating a map grid based on the starting point coordinate and the target point coordinate, obtaining satellite map data, and obtaining a route grid topology matrix based on the map grid and the satellite map data; Obtaining meteorological forecast data, performing light energy conversion based on the meteorological forecast data and the route grid topology matrix, obtaining a grid light energy vector, and generating an initial passing route set based on the satellite map data; Performing energy gain based on the initial passing route set, the grid light energy vector, and the grid energy consumption matrix, and obtaining a route energy gain time sequence; Obtaining initial battery state data, and obtaining an optimal light energy route based on the initial battery state data and the route energy gain time sequence.

3. The intelligent control method of the new energy house car according to claim 2, characterized in that, Obtaining a starting point coordinate and a target point coordinate, generating a map grid based on the starting point coordinate and the target point coordinate, obtaining satellite map data, and obtaining a route grid topology matrix based on the map grid and the satellite map data, comprising: Converting the starting point coordinate and the target point coordinate into a starting point plane rectangular coordinate and a target point plane rectangular coordinate based on UTM projection, and obtaining a plane path vector based on the starting point plane rectangular coordinate and the target point plane rectangular coordinate; Generating a path main axis coordinate system by taking the plane path vector as the X-axis and rotating the plane path vector counterclockwise by 90 degrees as the Y-axis; Dividing along the X-axis direction at intervals of N meters and dividing along the Y-axis direction at intervals of N meters to generate a map grid; Uniformly setting 9 sampling points in each single grid of the map grid, adding elevation data to each sampling point according to the satellite map data, labeling each single grid with a terrain label based on the elevation data of adjacent sampling points, and generating a route grid topology matrix.

4. The intelligent control method of the new energy house car according to claim 2, characterized in that, Obtaining meteorological forecast data, performing light energy conversion based on the meteorological forecast data and the route grid topology matrix, obtaining a grid light energy vector, and generating an initial passing route set based on the satellite map data, comprising: Based on the meteorological forecast data, solar radiation reference data and cloud variation curve of each single grid in the route grid topology matrix are obtained, and a grid light energy vector of each single grid is obtained according to the solar radiation reference data and the cloud variation curve; Based on satellite map data, all feasible road node coordinates are extracted, and initial attractive force values are added to all feasible road node coordinates. Starting from the starting point coordinate, adjacent feasible road node coordinates are retrieved, and the next feasible road node coordinate is selected by comparing the initial attractive force value of the adjacent feasible road node coordinate with the total initial attractive force value until the target point coordinate is reached, thereby generating an initial passing route set.

5. The intelligent control method of the new energy recreational vehicle according to claim 4, characterized in that, Based on the initial passing route set, the grid light energy vector and the grid energy consumption matrix, an energy gain operation is performed to obtain a route energy gain time sequence, including: Vehicle unit distance basic energy consumption data is obtained, the slope value and the altitude data of each single grid in the route grid topology matrix are extracted, the vehicle unit distance basic energy consumption is corrected based on the slope value and the altitude data, vehicle unit distance corrected energy consumption data is obtained, and the vehicle unit distance corrected energy consumption data is marked in each single grid in the route grid topology matrix, thereby obtaining a grid energy consumption matrix; The initial passing route set is mapped to the route grid topology matrix, and a passing route in the initial passing route set is extracted. The passing route is represented as a plurality of line segments from the starting point coordinate to the destination point coordinate. An expected passing time period is added to each line segment, and the light energy value of the corresponding period in the grid light energy vector is extracted according to the expected passing time period; Based on the light energy value and the grid energy consumption matrix, a grid gain value is obtained. The grid gain value is positive when the light energy value is greater than the energy consumption value, and negative when the light energy value is less than the energy consumption value. The grid gain values are arranged in the order of vehicle passing time to obtain a route energy gain time sequence.

6. The intelligent control method of the new energy recreational vehicle according to claim 5, characterized in that: Initial battery state data is obtained, and an optimal light energy route is obtained based on the initial battery state data and the route energy gain time sequence, including: The total path light energy gain of each passing route in the initial passing route set is extracted based on the route energy gain time sequence, and the total path energy consumption of each passing route in the initial passing route set is extracted based on the grid energy consumption matrix; Initial battery state data is obtained, and a path evaluation operation is performed according to the total path light energy gain, the total path energy consumption and the initial battery state data to obtain a comprehensive energy score; Based on the comprehensive energy score, the initial attractive force values of all feasible road node coordinates are corrected, a candidate passing route set is generated based on the corrected feasible road node coordinates, and the comprehensive energy scores of each path are recalculated; The passing route with the highest comprehensive energy score after recalculation is selected as the optimal light energy route.

7. The intelligent control method of the new energy recreational vehicle according to claim 1, characterized in that, A predicted photovoltaic power generation curve is generated according to the optimal light energy route and the grid light energy vector, actual photovoltaic power generation data is obtained, and light energy prediction calibration data is obtained based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve, including: The grid sequence in the optimal light energy route is extracted, the light energy values of each grid corresponding to the time period are spliced according to the time axis to generate a route light energy time sequence, photovoltaic panel parameters are obtained, and a predicted photovoltaic power generation curve is generated based on the photovoltaic panel parameters and the route light energy time sequence; The actual photovoltaic power generation data of the current period is obtained, the actual photovoltaic power generation data of the current period is aligned with the predicted photovoltaic power generation curve period, the predicted photovoltaic power generation data of the current period is obtained, the period deviation rate vector is obtained according to the actual photovoltaic power generation data and the predicted photovoltaic power generation data, and the light energy prediction calibration parameter is obtained based on the period deviation rate vector.

8. The intelligent control method of the new energy recreational vehicle according to claim 1, characterized in that, A user habit model is constructed and a load state transition tensor is generated, a load scheduling instruction set is established according to the predicted photovoltaic power generation curve and the load state transition tensor, real-time operation records of users are obtained, and a man-machine collaborative correction factor is obtained based on the real-time operation records of users and the load state transition tensor, including: An in-vehicle device operation log is obtained, interference items with single operation less than 1 second and power exceeding the upper limit of the device in the in-vehicle device operation log are removed, high-frequency behavior features and scene-related behavior features in the in-vehicle device operation log are extracted, time-driven rules, event-triggered rules, and device collaboration strategies are generated, and a user habit model is constructed based on the time-driven rules, the event-triggered rules, and the device collaboration strategies; A device behavior feature set is obtained based on the user habit model, device power demand data is obtained, a device state power constraint table is obtained based on the device power demand data, a state transition trigger condition is obtained according to the device behavior feature set and the device state power constraint table, a state switching logic is established based on the state transition trigger condition, and a load state transition tensor is generated based on the state transition trigger condition and the state switching logic; According to the predicted photovoltaic power generation curve and the load state transition tensor, a device target state set is obtained, and a load scheduling instruction set is established according to the device target state set. Real-time operation records of users are obtained, user operation records are matched with the load scheduling instruction set, a strategy deviation event set is generated, and a man-machine collaborative correction factor is obtained based on the strategy deviation event set and the load state transition tensor.

9. The intelligent control method of the new energy recreational vehicle according to claim 8, characterized in that, According to the predicted photovoltaic power generation curve and the load state transition tensor, a device target state set is obtained, and a load scheduling instruction set is established according to the device target state set, including: A power supply gap or surplus period is identified based on the predicted photovoltaic power generation curve, and a power supply state feature identifier is generated, current state data of the device is obtained, the power supply state feature identifier and the current state data of the device are matched with the load state transition tensor, and a device target state set is obtained; The current state data of the device is mapped to a specific power value according to the device target state set and the device power demand data, and an execution time window and a priority label are added to the device, and an initial load scheduling instruction set is generated according to the execution time window and the priority label of the device; Real-time battery state data and real-time location coordinate data are obtained, a safety operation boundary threshold is generated based on the real-time battery state data and the real-time location coordinate data, a power decay rate for the next 30 minutes is calculated, and a time point of reaching a safety threshold is predicted according to the safety operation boundary threshold, and a power change prediction vector is obtained; A device power demand sum is obtained based on the initial load scheduling instruction set, the current state data of the device is adjusted based on the power change prediction vector and the device power demand sum, and a load scheduling instruction set is obtained.

10. An intelligent control system of a new energy motor home, characterized in that, Including: The data acquisition module is configured to acquire meteorological forecast data, initial battery state data, actual photovoltaic power generation data, and user real-time operation records. The route planning module is configured to acquire a grid light energy vector according to the meteorological forecast data and a route grid topology matrix, and select an optimal light energy route according to the grid light energy vector and the initial battery state data. The power generation prediction module is configured to generate a predicted photovoltaic power generation curve according to the optimal light energy route and the grid light energy vector, and acquire a light energy prediction calibration parameter based on the actual photovoltaic power generation data and the predicted photovoltaic power generation curve. The load scheduling module is configured to construct a user habit model and generate a load state transition tensor, establish a load scheduling instruction set according to the predicted photovoltaic power generation curve and the load state transition tensor, and acquire a man-machine collaborative correction factor according to the user real-time operation records and the load state transition tensor. The adjustment cycle module is configured to adjust the grid light energy vector according to the light energy prediction calibration parameter, adjust the load state transition tensor based on the man-machine collaborative correction factor, and control the route planning module, the power generation prediction module, the user modeling module, and the scheduling control module to cyclically perform corresponding operations.

Images