Optimal route planning method for leveling operations based on raster topographic maps

Abstract

The present invention relates to the technical field of leveling operations, and in particular to an optimal operation route planning method for leveling operations based on a raster topographic map, comprising the steps of real-time acquisition of terrain data - data processing and rasterization - terrain analysis and classification - path planning method calculation - path navigation - leveling operations. In the present invention, three-dimensional terrain data of an operation area is acquired through a path planning navigation system, and when an operation vehicle body moves forward, terrain data is acquired in real time, and then an operation path is planned in real time according to the terrain data, so that a driver can be guided in real time according to the path, and the driver can perform land leveling according to the navigation information, and the technical requirements for the driver are relatively low, which is conducive to reducing the training cost of the driver. During the operation process, the driver does not need to determine the operation path by himself, which is conducive to improving the operation efficiency, avoiding repeated operations, and ensuring a relatively stable leveling effect and improving the leveling quality.

Description

Technical Field

[0001] The invention relates to the technical field of leveling operations, and in particular to an optimal operation route planning method for leveling operations based on a raster topographic map. Background Art

[0002] Land leveling is an important step in the process of land development and construction. The land surface is adjusted to a certain level by using land leveling machinery to facilitate subsequent construction work. However, the existing land leveling operation mainly relies on the experience and skills of the driver. During the land leveling process, the driver needs to judge the ups and downs according to the terrain map and decide the operation route by himself. This makes the technical requirements of the driver high and the training cost of the driver high. During the operation, the driver needs to spend a certain amount of time to determine the operation path. The operation efficiency is relatively low, and repeated operations may occur. In addition, the leveling process relies on personal experience, and the leveling effect is unstable. At the same time, the driver has high labor intensity and is prone to fatigue. Summary of the invention

[0003] The purpose of the present invention is to provide an optimal operation route planning method for leveling operations based on raster topographic maps to solve the problems raised in the above-mentioned background technology.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] The optimal operation route planning method for leveling operation based on raster terrain map includes the steps of real-time acquisition of terrain data—data processing and rasterization—terrain analysis and classification—path planning method calculation—path navigation—leveling operation. The specific steps are as follows:

[0006] Step (1) real-time terrain data collection: collecting three-dimensional terrain data of the operation area through a path planning navigation system installed on the vehicle body;

[0007] Step (2) data processing and rasterization: converting the collected three-dimensional point cloud data into a two-dimensional raster topographic map, rasterizing the three-dimensional topographic data, and then using image processing technology to extract topographic features;

[0008] Step (3) Terrain analysis and classification: Classify the terrain using machine learning algorithms based on terrain features;

[0009] Step (4) Path planning method calculation: Path planning is performed on the working terrain according to different working terrains;

[0010] Step (5) route navigation: providing real-time route navigation for the driver according to the planned route;

[0011] Step (6) ground leveling: the driver drives the vehicle to the working position according to the navigation information, and performs ground leveling through the bucket on the vehicle and the ground leveling structure.

[0012] Furthermore, during the process of the vehicle moving forward in step (1), the path planning and navigation system on the vehicle continues to collect three-dimensional terrain data of the operation area in real time, thereby updating the terrain data and the path in real time.

[0013] Furthermore, the path planning and navigation system includes a real-time terrain data acquisition module, a data processing module, and a path planning method module;

[0014] The real-time terrain data acquisition module can acquire three-dimensional terrain data, and the real-time terrain data acquisition module can transmit the acquired three-dimensional terrain data to the data processing module;

[0015] The data processing module can perform rasterization processing on the received three-dimensional terrain data, and convert the three-dimensional terrain data into a two-dimensional raster terrain map. Each pixel in the two-dimensional raster map represents a fixed-size plot of land, and the pixel value represents the average elevation of the plot of land. Then, terrain features are extracted according to the two-dimensional raster terrain map. The data processing module can pass the two-dimensional raster terrain map and the terrain feature data to the path planning method module;

[0016] The path planning method module can determine the terrain type of the leveling area according to the terrain feature data, and determine whether the terrain type of the leveling area is flat terrain, gully terrain, hilly terrain or composite terrain. The path planning method module can plan the operation path according to the terrain type.

[0017] Furthermore, the path planning and navigation system also includes a human-computer interaction interface and a navigation module. The path planning method module can transmit the path information to the human-computer interaction interface, and display the path information through the human-computer interaction interface. The human-computer interaction interface can transmit the path information to the navigation module. The navigation module can determine the navigation information according to the path information. The navigation module can transmit the navigation information to the human-computer interaction interface, and display the navigation information through the human-computer interaction interface. The driver can turn and move forward according to the navigation information.

[0018] Furthermore, the real-time terrain data acquisition module includes an RTK locator and a laser radar. The laser radar is used to collect three-dimensional terrain data of the working area in real time. The RTK locator is used for positioning measurement. When the vehicle moves forward, the vehicle's own position is determined according to the RTK positioning data, and the distance and direction from the vehicle to the working area are determined. When the laser radar measures the terrain, the terrain data facing the edge of the working area and the distance between the measurement location and the vehicle is greater than the distance from the vehicle to the edge of the working area will be filtered.

[0019] Furthermore, in the process of rasterizing the data by the data processing module, an irregular triangulated network is constructed with the data collected by the RTK locator and the lidar, and then the TIN is converted into a regular grid using a linear interpolation method, and the grid resolution ranges from 0.5m to 1m.

[0020] Furthermore, in the process of extracting terrain features using image processing technology, a 3x3 moving window is used to calculate the elevation difference between the central pixel and the adjacent pixels, thereby obtaining the slope information. The Horn algorithm is used to calculate the slope aspect information based on the 3x3 window, and the local elevation standard deviation is calculated to represent the undulation of the terrain, thereby obtaining the terrain roughness. The second-order derivative is used to calculate the plane curvature and profile curvature of the terrain. The features are normalized and combined, the extracted terrain features are normalized, and the principal component analysis is used to reduce the dimensionality to generate a comprehensive terrain feature index.

[0021] Preferably, when leveling flat terrain, an S-shaped operation route is adopted as a whole, and local low-lying areas are processed first. In the process of multi-vehicle collaborative operation, multiple vehicles start leveling from multiple starting points. When a local low-lying area is detected, the nearest vehicle is selected to level it. After the local low-lying area is leveled, the leveling path is replanned on the spot for the vehicle, and the leveling information is wirelessly synchronized to the path planning navigation system on the remaining vehicles.

[0022] When leveling hilly terrain, the vehicle body works along the contour lines, processing in layers and gradually advancing from high to low;

[0023] When leveling gully terrain, first go around the edge of the leveling area to determine the main gullies. At the beginning of the operation, fill the main gullies first, and then use a zigzag route to advance horizontally. When multiple vehicles work together, one vehicle fills the main gullies, and the remaining vehicles fill the non-main gullies from multiple starting points.

[0024] When leveling composite terrain, sub-areas are divided according to terrain features, and different leveling methods are used for each area. During leveling, areas with larger elevation differences are given priority.

[0025] Preferably, in the process of path planning for terrain, the path planning method module converts the raster terrain map into a weighted graph, with each pixel as a node and the edge weight representing the elevation difference and distance between adjacent nodes. In the process of global path planning, the global path is planned based on minimizing the elevation difference, minimizing the path length and minimizing the number of turns.

[0026] Furthermore, based on the global planned path, the dynamic window method is used for local optimization, the path is adjusted in real time to adapt to terrain changes, and the path is smoothed using cubic spline curves to improve driving comfort.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] 1. The three-dimensional terrain data of the working area is collected through the path planning and navigation system. When the working vehicle moves forward, the terrain data is collected in real time, and then the working path is planned in real time according to the terrain data. In this way, the driver can be navigated in real time according to the path. The driver can level the land according to the navigation information. The technical requirements for the driver are relatively low, which is conducive to reducing the cost of driver training. During the operation, the driver does not need to determine the operation path by himself, which is conducive to improving the operation efficiency, avoiding repeated operations, and ensuring a relatively stable leveling effect and improving the leveling quality. At the same time, it is conducive to reducing the labor intensity of the driver and alleviating the burden on the driver.

[0029] 2. Determine the terrain type of the leveling area through the path planning method module, and build an overall operation plan for different terrains in advance within the path planning navigation system. In this way, when the terrain type is determined, the overall path planning can be quickly determined according to the overall plan, which is conducive to reducing the planning time. In the process of multi-machine collaborative operation, the operation information can be synchronized through wireless communication, which is convenient for rapid updating of the operation path in a complex operating environment, which is conducive to the rapid leveling of the land. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the overall flow of the algorithm of the present invention;

[0031] Figure 2 It is a schematic diagram of the process flow of the path planning navigation system in the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] See also Figure 1 In an embodiment of the present invention, the optimal operation route planning method for leveling operation based on raster terrain map includes the steps of real-time acquisition of terrain data - data processing and rasterization - terrain analysis and classification - path planning method calculation - path navigation - leveling operation, and the specific steps are as follows:

[0034] Step (1) real-time terrain data collection: collecting three-dimensional terrain data of the operation area through a path planning navigation system installed on the vehicle body;

[0035] Step (2) data processing and rasterization: converting the collected three-dimensional point cloud data into a two-dimensional raster topographic map, rasterizing the three-dimensional topographic data, and then using image processing technology to extract topographic features;

[0036] Step (3) Terrain analysis and classification: Classify the terrain using machine learning algorithms based on terrain features;

[0037] Step (4) Path planning method calculation: Path planning is performed on the working terrain according to different working terrains;

[0038] Step (5) route navigation: providing real-time route navigation for the driver according to the planned route;

[0039] Step (6) ground leveling: the driver drives the vehicle to the working position according to the navigation information, and performs ground leveling through the bucket on the vehicle and the ground leveling structure;

[0040] During the process of the vehicle moving forward in step (1), the path planning and navigation system on the vehicle continues to collect three-dimensional terrain data of the operation area in real time, thereby updating the terrain data and path in real time.

[0041] Specifically, the working path is planned in real time according to the terrain data, so that the driver can be navigated in real time according to the path. The driver can level the land according to the navigation information. The technical requirements for the driver are relatively low, which is conducive to reducing the cost of driver training. The working path is automatically calculated and determined by the path planning navigation system, which is conducive to improving work efficiency, avoiding repeated operations, and ensuring a relatively stable leveling effect, improving the leveling quality, and at the same time helping to reduce the labor intensity of the driver and alleviate the burden on the driver.

[0042] Embodiment 1

[0043] like Figure 2 As shown, in this embodiment, the path planning navigation system includes a real-time terrain data acquisition module, a data processing module, and a path planning method module;

[0044] The real-time terrain data acquisition module can acquire three-dimensional terrain data, and the real-time terrain data acquisition module can transmit the acquired three-dimensional terrain data to the data processing module;

[0045] The data processing module can rasterize the received three-dimensional terrain data and convert the three-dimensional terrain data into a two-dimensional raster terrain map. Each pixel in the two-dimensional raster map represents a fixed-size plot of land, and the pixel value represents the average elevation of the plot of land. Then, terrain features are extracted based on the two-dimensional raster terrain map. The data processing module can pass the two-dimensional raster terrain map and terrain feature data to the path planning method module.

[0046] The path planning method module can determine the terrain type of the leveling area according to the terrain feature data, and determine whether the terrain type of the leveling area is flat terrain, gully terrain, hilly terrain or composite terrain. The path planning method module can plan the operation path according to the terrain type.

[0047] In specific implementation, the path planning method module classifies the terrain according to the terrain features and determines the terrain type;

[0048] Among them, the terrain classification algorithm procedure is as follows:

[0049] Python

[0050] from sklearn.ensemble import RandomForestClassifier

[0051] from sklearn.model_selection import train_test_split

[0052] def train_terrain_classifier(features,labels):

[0053] X_train,X_test,y_train,y_test=train_test_split(features,labels,test_size=0.2,random_state=42)

[0054] clf=RandomForestClassifier(n_estimators=100,random_state=42)

[0055] clf.fit(X_train,y_train)

[0056] accuracy=clf.score(X_test,y_test)

[0057] print(f"Terrain classification accuracy:{accuracy}")

[0058] return clf

[0059] def classify_terrain(features,classifier):

[0060] return classifier.predict(features)

[0061] An overall operation plan for different terrains can be constructed in advance within the path planning and navigation system. In this way, when the terrain type is determined, the corresponding overall plan can be quickly called to determine the global path planning, which is conducive to reducing the planning time. Multi-machine collaborative operation can be adopted, and a path planning and navigation system is set on each machine. The path planning and navigation systems on different machines can synchronize operation information through wireless communication, which is convenient for real-time updating of the operation path in a complex operating environment, which is conducive to the collaborative operation of multiple machines.

[0062] like Figure 2 As shown, in this embodiment, the path planning navigation system also includes a human-computer interaction interface and a navigation module. The path planning method module can transmit the path information to the human-computer interaction interface, and display the path information through the human-computer interaction interface. The human-computer interaction interface can transmit the path information to the navigation module, and the navigation module can determine the navigation information according to the path information. The navigation module can transmit the navigation information to the human-computer interaction interface, and display the navigation information through the human-computer interaction interface. The driver can turn and move forward according to the navigation information.

[0063] During specific implementation, the current position, planned path and topographic map can be displayed in real time on the display screen of the human-computer interaction interface. Different colors can be used to mark elevation differences, and the next turning point and distance information can be displayed according to the navigation information. A voice prompt module can be set to remind the driver by voice according to the navigation information. The driver can manually adjust the path. After a path is manually adjusted, the next leveling route is automatically planned according to the last working position. The driver's operations can be recorded for algorithm optimization.

[0064] like Figure 2 As shown, in this embodiment, the real-time terrain data acquisition module includes an RTK locator and a laser radar. The laser radar is used to collect three-dimensional terrain data of the working area in real time, and the RTK locator is used for positioning measurement. When the vehicle moves forward, the position of the vehicle itself is determined according to the RTK positioning data, and the distance and direction from the vehicle to the working area are determined. When the laser radar measures the terrain, the terrain data facing the edge of the working area and the distance between the measurement location and the vehicle is greater than the distance from the vehicle to the edge of the working area is filtered.

[0065] In specific implementation, an attitude sensor can be configured to measure the direction of the laser radar, and the real-time terrain data acquisition module can be fixed on the vehicle body. In this way, the direction of the vehicle body can be determined according to the angle of the laser radar relative to the vehicle body, and the position of the vehicle body can be determined according to the RTK locator. When the vehicle body is near the edge of the working area, if the laser radar is facing the edge of the working area, the data exceeding the working area is automatically filtered to avoid processing irrelevant data and affecting the planning of the working path in the working area.

[0066] The laser radar can be rotatably set on the vehicle body, and a power mechanism for driving the laser radar to rotate can be set. In this way, when the vehicle body reaches the working point to perform land leveling operations, the angle of the laser radar can be adjusted through the power mechanism to rotate the laser radar to measure the terrain around the vehicle body. When the vehicle body moves to the edge of the working area, the direction of the laser radar can be adjusted to make the laser radar face the inside of the working area, so as to avoid measuring outside the working area as much as possible and quickly measure the working area. At the same time, through the rotation of the laser radar, the detected area can be re-detected, and some measurement data can be compared and corrected. When multiple machines work together, if a vehicle body passes through a leveled area, the angle of the laser radar can be adjusted so that one vehicle body can measure the leveled area and detect the leveling effect.

[0067] Embodiment 2

[0068] Based on the first embodiment, Figure 1 As shown, in this embodiment, during the rasterization process of the data processing module, the data collected by the RTK locator and the laser radar are used to construct an irregular triangulated network, which can better represent the details of complex terrain, especially steep slopes and sudden changes in terrain. Then, the linear interpolation method is used to convert the TIN into a regular grid. The grid resolution can be adjusted according to the accuracy requirements of the operation, and the grid resolution range is 0.5m-1m.

[0069] In the process of extracting terrain features using image processing technology, a 3x3 moving window is used to calculate the elevation difference between the central pixel and the adjacent pixels to obtain the slope information. The Horn algorithm is used to calculate the slope information based on the 3x3 window, and the local elevation standard deviation is calculated to indicate the degree of terrain undulation and obtain the terrain roughness. The second-order derivative is used to calculate the plane curvature and profile curvature of the terrain. The features are normalized and combined, the extracted terrain features are normalized, and the principal component analysis is used to reduce the dimension to generate a comprehensive terrain feature index.

[0070] In specific implementation, image processing technology is used to extract terrain features such as slope, slope aspect, and elevation difference;

[0071] The terrain analysis algorithm is as follows:

[0072] Python

[0073] def analyze_terrain(raster_map):

[0074] features=extract_features(raster_map)

[0075] terrain_type=classify_terrain(features,trained_classifier)

[0076] return terrain_type,features

[0077] def extract_features(raster_map):

[0078] #Extract features such as slope, aspect, elevation difference, etc.

[0079] #Use numpy and scipy libraries for calculation

[0080] #...

[0081] return features

[0082] def classify_terrain(features,classifier):

[0083] return classifier.predict(features)

[0084] During the vehicle's advancement, terrain information is continuously collected, and then the newly collected data is continuously processed through the data processing module, thereby continuously updating the topographic map of the operating area to complete the topographic map of the operating area. At the same time, the terrain data detected in the leveled area can be processed to detect whether the target leveling effect has been achieved.

[0085] like Figure 1 As shown, in this embodiment, during the path planning process of the path planning method module for the terrain, the raster terrain map is converted into a weighted graph, each pixel is regarded as a node, and the weight of the edge represents the elevation difference and distance between adjacent nodes. During the global path planning process, the global path is planned based on minimizing the elevation difference, minimizing the path length, and minimizing the number of turns;

[0086] Based on the global planning path, the dynamic window method is used for local optimization, the path is adjusted in real time to adapt to terrain changes, and the path is smoothed using cubic spline curves to improve driving comfort.

[0087] In the specific implementation, the improved A* algorithm is used for global path planning;

[0088] The improved A* algorithm is as follows:

[0089]

[0090]

[0091]

[0092] Among them, the dynamic window optimization algorithm is as follows:

[0093]

[0094] Among them, the spline curve smoothing algorithm is as follows:

[0095]

[0096] x,y=zip(*path)

[0097] t = np.arange(len(x))

[0098] cs_x=CubicSpline(t,x)

[0099] cs_y=CubicSpline(t,y)

[0100] t_new=np.linspace(0,len(x)-1,num=len(x)*3)

[0101] smoothed_path=list(zip(cs_x(t_new),cs_y(t_new)))

[0102] return smoothed_path

[0103] Finally, the real-time update algorithm is as follows:

[0104] Python

[0105] def update_terrain_model(current_position,sensor_data,terrain_model):

[0106] new_data=process_sensor_data(sensor_data)

[0107] updated_model=integrate_new_data(terrain_model,new_data,current_position)

[0108] return updated_model

[0109] def adaptive_path_adjustment(current_path, updated_terrain, current_position):

[0110] if significant_terrain_change(updated_terrain):

[0111] new_end = current_path[-1]

[0112] new_path = plan_path(updated_terrain, current_position, new_end)

[0113] return smooth_transition(current_path, new_path)

[0114] return current_path

[0115] def smooth_transition(old_path, new_path):

[0116] transition_point = find_closest_point(old_path, new_path[0])

[0117] return old_path[:transition_point] + new_path

[0118] def find_closest_point(path, point):

[0119] return min(range(len(path)), key=lambda i: np.linalg.norm(np.array(path[i]) - np.array(point)))

[0120] If Figure 1As shown, in this embodiment, when leveling flat terrain, an S-shaped operation route is adopted as a whole, and local low-lying areas with a large degree of depression are processed first. In the process of multi-vehicle collaborative operation, multiple vehicles start leveling from multiple starting points. When a local low-lying area with a large degree of depression is detected, the nearest vehicle is selected to level it. After the leveling of the local low-lying area with a large degree of depression is completed, the leveling path is replanned on the spot for the vehicle, and the leveling information is wirelessly synchronized to the path planning navigation system on the remaining vehicles;

[0121] When leveling hilly terrain, the vehicle body works along the contour lines, processing in layers and gradually advancing from high to low;

[0122] When leveling gully terrain, first go around the edge of the leveling area to determine the main gullies. At the beginning of the operation, fill the main gullies first, and then use a zigzag route to advance horizontally. When multiple vehicles work together, one vehicle fills the main gullies, and the remaining vehicles fill the non-main gullies from multiple starting points.

[0123] When leveling composite terrain, sub-areas are divided according to terrain features, and different leveling methods are used for each area. During leveling, areas with larger elevation differences are given priority.

[0124] In specific implementation, in flat terrain operations, the system adopts an S-shaped or roundabout route, giving priority to local low-lying areas. The path planning method considers minimizing the number of turns to improve work efficiency. In hilly terrain, the system first levels the terrain from low to high along a straight line upward, and first levels the upper and lower passages in the work area to facilitate subsequent leveling and transportation of soil. At the same time, in the process of leveling along a straight line upward, the terrain on both sides of the straight line route is detected and contour lines are divided. After the vehicle body reaches the highest point, an operation route along the contour line is generated, giving priority to high-altitude areas, and then gradually advancing to low-altitude areas. In gully terrain operations, the vehicle body is first moved along the edge of the gully perpendicular to the working area, and the gully is measured by laser radar. The system first plans a path to fill the main gully, and then uses a zigzag route to advance horizontally to ensure that the gully is evenly filled.

[0125] In the multi-vehicle collaborative operation, when operating on flat terrain, the vehicles can start from different points, and the low-lying values ​​can be pre-set to give priority to low-lying places. When vehicle one detects a place with a larger low-lying value first, the route of vehicle one is modified so that vehicle one processes the low-lying place first. During the processing, if other vehicles detect a new low-lying place, such as vehicle two detects a new low-lying place, if the new low-lying place is more than a certain distance from vehicle one, vehicle two is transferred to process it. If it is not more than a certain distance, vehicle one continues to process the new low-lying place, while vehicle two continues to process it according to the original route. After vehicle one processes the low-lying place, a new flat ground path is planned for vehicle one on the spot, and vehicle one can choose to process the low-lying place only. In this way, during the leveling process, experienced drivers can give priority to processing difficult low-lying places, which can reduce the requirements for other drivers. At the same time, when vehicle one goes to the low-lying area, a relatively flat route is planned according to the terrain information detected by vehicle one to bypass the overly rugged area.

[0126] In the multi-machine collaborative operation, when operating on gully terrain, multiple vehicles can be used to process different gullies at the same time, so that one of the vehicles can operate on the main gully first, and the remaining vehicles can operate on the remaining gullies in a zigzag pattern. Before the operation, the vehicle can first level the middle of several gullies along the vertical direction of the gully, detect the middle of the leveling area through the laser radar, and level a path in the middle of the operation area, so that the gully can be leveled from the middle to both ends. In this way, multiple vehicles can be used to level a main gully at the same time;

[0127] In multi-machine collaborative operation, when working on hilly terrain, a path is first leveled to facilitate the vehicles to go up and down the hill. At the same time, the terrain is detected, and several vehicles are evenly spaced along the corresponding contour lines for leveling. One vehicle can first level the land on the contour lines with more low-lying areas, and the remaining vehicles can level the land down along the contour lines in sequence;

[0128] When operating on complex terrain, the vehicle is first moved along the edge of each terrain to detect each terrain area and divide it into sub-areas. The vehicle determines the operating range based on the divided sub-areas. During the operation, only the terrain within its own operating range is detected.

[0129] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0130] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. The optimal operation route planning method for leveling operation based on raster topographic map is characterized by: It includes the steps of real-time collection of terrain data - data processing and rasterization - terrain analysis and classification - path planning method calculation - path navigation - leveling operation. The specific steps are as follows: Step (1) real-time terrain data collection: collecting three-dimensional terrain data of the operation area through a path planning navigation system installed on the vehicle body; Step (2) data processing and rasterization: converting the collected three-dimensional point cloud data into a two-dimensional raster topographic map, rasterizing the three-dimensional topographic data, and then using image processing technology to extract topographic features; Step (3) Terrain analysis and classification: Classify the terrain using machine learning algorithms based on terrain features; Step (4) Path planning method calculation: Path planning is performed on the working terrain according to different working terrains; Step (5) route navigation: providing real-time route navigation for the driver according to the planned route; Step (6) ground leveling: the driver drives the vehicle to the working position according to the navigation information, and performs ground leveling through the bucket on the vehicle and the ground leveling structure.

2. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 1, characterized in that: During the process of the vehicle moving forward in step (1), the path planning and navigation system on the vehicle continues to collect three-dimensional terrain data of the operation area in real time, thereby updating the terrain data and path in real time.

3. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 1, characterized in that: The path planning and navigation system includes a real-time terrain data acquisition module, a data processing module, and a path planning method module; The real-time terrain data acquisition module can acquire three-dimensional terrain data, and the real-time terrain data acquisition module can transmit the acquired three-dimensional terrain data to the data processing module; The data processing module can perform rasterization processing on the received three-dimensional terrain data, and convert the three-dimensional terrain data into a two-dimensional raster terrain map. Each pixel in the two-dimensional raster map represents a fixed-size plot of land, and the pixel value represents the average elevation of the plot of land. Then, terrain features are extracted according to the two-dimensional raster terrain map. The data processing module can pass the two-dimensional raster terrain map and the terrain feature data to the path planning method module; The path planning method module can determine the terrain type of the leveling area according to the terrain feature data, and determine whether the terrain type of the leveling area is flat terrain, gully terrain, hilly terrain or composite terrain. The path planning method module can plan the operation path according to the terrain type.

4. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 3 is characterized in that: The path planning and navigation system also includes a human-computer interaction interface and a navigation module. The path planning method module can transmit path information to the human-computer interaction interface and display the path information through the human-computer interaction interface. The human-computer interaction interface can transmit the path information to the navigation module. The navigation module can determine navigation information based on the path information. The navigation module can transmit the navigation information to the human-computer interaction interface and display the navigation information through the human-computer interaction interface. The driver can turn and move forward based on the navigation information.

5. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 3, characterized in that: The real-time terrain data acquisition module includes an RTK locator and a laser radar. The laser radar is used to collect three-dimensional terrain data of the working area in real time. The RTK locator is used for positioning measurement. When the vehicle moves forward, the vehicle's own position is determined according to the RTK positioning data, and the distance and direction from the vehicle to the working area are determined. When the laser radar measures the terrain, the terrain data facing the edge of the working area and the distance between the measurement location and the vehicle is greater than the distance from the vehicle to the edge of the working area will be filtered.

6. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 5, characterized in that: In the process of rasterizing data in the data processing module, the data collected by the RTK locator and the lidar are used to construct an irregular triangulated network, and then the linear interpolation method is used to convert the TIN into a regular grid with a grid resolution range of 0.5m-1m.

7. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 6, characterized in that: In the process of extracting terrain features using image processing technology, a 3x3 moving window is used to calculate the elevation difference between the central pixel and the adjacent pixels to obtain the slope information. The Horn algorithm is used to calculate the slope information based on the 3x3 window, and the local elevation standard deviation is calculated to indicate the degree of terrain undulation and obtain the terrain roughness. The second-order derivative is used to calculate the plane curvature and profile curvature of the terrain. The features are normalized and combined, the extracted terrain features are normalized, and the principal component analysis is used to reduce the dimension to generate a comprehensive terrain feature index.

8. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 6, characterized in that: When leveling flat terrain, an S-shaped operation route is adopted as a whole, and local low-lying areas are processed first. In the process of multi-vehicle collaborative operation, multiple vehicles start leveling from multiple starting points. When a local low-lying area is detected, the nearest vehicle is selected to level it. After the local low-lying area is leveled, the leveling path is replanned on the spot for the vehicle, and the leveling information is wirelessly synchronized to the path planning navigation system on the remaining vehicles; When leveling hilly terrain, the vehicle body works along the contour lines, processing in layers and gradually advancing from high to low; When leveling gully terrain, first go around the edge of the leveling area to determine the main gullies. At the beginning of the operation, fill the main gullies first, and then use a zigzag route to advance horizontally. When multiple vehicles work together, one vehicle fills the main gullies, and the remaining vehicles fill the non-main gullies from multiple starting points. When leveling composite terrain, sub-areas are divided according to terrain features, and different leveling methods are used for each area. During leveling, areas with larger elevation differences are given priority.

9. The method for planning the optimal operation route for leveling operations based on raster topographic maps according to claim 6, characterized in that: In the process of path planning for terrain, the path planning method module converts the raster terrain map into a weighted graph, each pixel is regarded as a node, and the weight of the edge represents the elevation difference and distance between adjacent nodes. In the process of global path planning, the global path is planned based on minimizing the elevation difference, minimizing the path length, and minimizing the number of turns.

10. The method for planning the optimal operation route for leveling operation based on raster topographic map according to claim 9, characterized in that: Based on the global planning path, the dynamic window method is used for local optimization, the path is adjusted in real time to adapt to terrain changes, and the path is smoothed using cubic spline curves to improve driving comfort.

Images