A dynamic light-based path planning method and system

Abstract

The application discloses a path planning method and system based on dynamic light, and relates to the technical field of computers, and comprises the following steps: receiving DEM map data, dynamic light map data, a path planning starting point and a path planning target point, and generating slope map data based on the DEM map data; processing based on the slope map data and the dynamic light map data to generate a dynamic passability map; performing path search by using a three-dimensional path planning algorithm based on dynamic light based on the constraint of the dynamic passability map, and calculating the total cost of the planned path node based on the DEM map data, the slope map data and the dynamic light map data, wherein the total cost comprises a terrain cost, a distance cost and a light cost; determining the planned path node with the minimum total cost through multiple iterations, generating a path node sequence from the path planning starting point to the path planning target point based on the planned path node with the minimum total cost, taking the path node sequence as an optimal global path, and achieving the planning of the optimal path.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a path planning method and system based on dynamic light. BACKGROUND

[0002] Path planning algorithms suitable for lunar rover can be mainly divided into two categories: rule-based path planning algorithm and biological heuristic-based path planning algorithm. Among them, the rule-based path planning algorithm includes A* algorithm, D* algorithm, RRT algorithm, DWA algorithm, APF algorithm, etc. The biological heuristic-based path planning algorithm includes evolutionary learning related algorithms (such as GA algorithm, particle swarm optimization algorithm, etc.) and machine learning-based algorithms. According to the different use scenarios of the algorithm, the lunar rover path planning algorithm can also be divided into global path planning algorithm and local path planning algorithm. The global path planning algorithm uses prior knowledge such as digital elevation model (DEM), global slope map, and light map to generate a safe driving path for the rover to carry out long-distance multi-target detection activities, while the local path planning refers to the real-time perception of the surrounding environment state when the rover is actually driving on the lunar surface, and the dynamic obstacle avoidance and path generation are carried out in real time to deal with some unexpected situations.

[0003] The existing global path planning algorithm of the lunar rover considers the influence of dynamic constraints such as light in addition to static constraints such as terrain and slope, but few algorithms consider the quantitative influence of dynamic light conditions on path planning during the long-term movement of the lunar rover on the lunar surface. SUMMARY

[0004] To solve the problems mentioned in the background, the purpose of the present application is to provide a path planning method and system based on dynamic light.

[0005] In the first aspect, the purpose of the present application can be realized by the following technical scheme: a path planning method based on dynamic light, the method comprising the following steps:

[0006] receiving DEM map data, dynamic light map data, path planning starting point and path planning target point, generating slope map data based on the DEM map data; wherein the DEM map data is data generated by a pre-established digital elevation model, containing elevation values distributed according to grid size; processing based on the slope map data and the dynamic light map data to generate a dynamic passability map, wherein the dynamic passability map is a time-varying passability constraint of whether each position is passable in each time step;

[0007] The path planning starting point is taken as a starting position of path search, and the path planning target point is taken as a search target. A three-dimensional path planning algorithm based on dynamic lighting is used for path search based on the constraint of a dynamic passability graph. A total cost of a planned path node is calculated based on DEM graph data, slope graph data and dynamic lighting graph data, wherein the total cost includes a terrain cost, a distance cost and a lighting cost.

[0008] Through multiple iterations, a planned path node with a minimum total cost is determined. A path node sequence from the path planning starting point to the path planning target point is generated based on the planned path node with the minimum total cost. The path node sequence is taken as an optimal global path, so as to realize planning of the optimal path.

[0009] In combination with the first aspect, in some implementations of the first aspect, the method further includes: the process of generating the slope graph data based on the DEM graph data includes:

[0010] The variation rate of the elevation value distributed according to the grid size in the horizontal and numerical directions is calculated as the slope graph data.

[0011] In combination with the first aspect, in some implementations of the first aspect, the method further includes: the process of generating the dynamic passability graph based on the slope graph data and the dynamic lighting graph data includes:

[0012] The slope graph data and the dynamic lighting graph data are binarized to obtain a slope passability graph and a lighting passability graph of each hour. The slope passability graph and the lighting passability graph of each hour are respectively subjected to matrix point multiplication to obtain the dynamic passability graph.

[0013] The slope passability graph represents a passable region with a slope not exceeding a preset angle.

[0014] The lighting passability graph of each hour represents a passable region with a lighting rate not lower than a preset lighting threshold.

[0015] In combination with the first aspect, in some implementations of the first aspect, the method further includes: the three-dimensional path planning algorithm based on dynamic lighting is a heuristic path planning algorithm improved based on a Dijkstra algorithm. The heuristic function design of the three-dimensional path planning algorithm based on dynamic lighting includes:

[0016] For a current node n, a total cost function f(n) is defined as:

[0017] f(n) = g(n) + h(n) (1)

[0018] wherein g(n) represents the real cost from the start node to the current node n, and h(n) represents the cost estimated by the heuristic function from the current node n to the destination node.

[0019] In combination with the first aspect, in some implementations of the first aspect, the method further comprises:

[0020] Suppose the rover is located at position P(x, y), and its corresponding slope is θ x，y . The specific process is as follows:

[0021]

[0022] f H and f V represent the change rate of the elevation in the horizontal and vertical directions respectively, and are calculated as follows:

[0023]

[0024] In the above formulae, H1-H9 represent the elevation values of the corresponding positions in the 3x3 grid centered on H5, and ΔH and ΔV are the resolutions of the elevation map in the horizontal and vertical directions respectively;

[0025] When calculating the terrain cost, three factors are mainly considered, i.e., the change of terrain elevation, the change of slope, and the roughness of terrain, which are represented by Cost h , Cost s , and Cost r respectively, and the total terrain cost is represented by Cost T . The passable area in the dynamic passability map is marked as Area yes , and the impassable area is marked as Area no . Suppose the current position of the rover is P, and the next position is P', then the terrain-related cost is calculated as follows:

[0026]

[0027] Cost T = ω1Cost h + ω2Cost s + ω3Cost r (8)

[0028] In formula (5), H P and H P′ represent the elevation values of positions P and P' respectively, and K h represents the influence coefficient of the change of elevation. In formula (6), θ P and θ P′ represent the slope values of positions P and P' respectively, and K sK represents an influence coefficient of slope change, in formula (7) P and R P′ respectively represent terrain roughness values of positions P and P', K r represents an influence coefficient of roughness change, In formula (8), parameters ω1, ω2, ω3 satisfy a condition of ω1+ω2+ω3=1.

[0029] In combination with the first aspect, in some implementations of the first aspect, the method further includes: a calculation process of the distance cost is:

[0030] The distance cost is calculated using Euclidean distance or diagonal distance, assuming that a current time of the patroller is located at a P point, and next time moves to a P' point, then the distance cost Cost D can be calculated as follows:

[0031]

[0032] In the above formula, dist (P, P') is a distance between positions P and P', K D is an influence coefficient of distance factor.

[0033] In combination with the first aspect, in some implementations of the first aspect, the method further includes: a calculation process of the light cost is:

[0034] Assuming that a light intensity at the P position at the t time is a light intensity at the P' position at the t+1 time is then the light cost Cost I can be defined as:

[0035]

[0036] In the above formula, K I is an influence coefficient of the light factor on the light cost function, and the overall light cost decreases as the light intensity at the t+1 time becomes larger.

[0037] In combination with the first aspect, in some implementations of the first aspect, the method further includes: a heuristic function of the three-dimensional path planning algorithm based on dynamic light is defined as:

[0038] Cost heuristic = α * Cost T + β * Cost D + γ * Cost I (11)

[0039] In the above formula, parameters α, β, γ satisfy a condition of α+β+γ=1.

[0040] In a second aspect, to achieve the above object, the application discloses a path planning system based on dynamic light, comprising:

[0041] A data processing module is configured to receive DEM map data, dynamic light map data, a path planning starting point and a path planning target point, generate slope map data based on the DEM map data, wherein the DEM map data is data generated by a pre-established digital elevation model and contains elevation values distributed according to grid size, and process the slope map data and the dynamic light map data to generate a dynamic passability map, wherein the dynamic passability map is a time-varying passability constraint indicating whether each position is passable at each time step.

[0042] A cost acquisition module is configured to take the path planning starting point as a starting position of path search, take the path planning target point as a search target, perform path search based on a constraint of the dynamic passability map by using a three-dimensional path planning algorithm based on dynamic light, and calculate a total cost of a planned path node based on the DEM map data, the slope map data and the dynamic light map data, wherein the total cost includes a terrain cost, a distance cost and a light cost.

[0043] A path planning module is configured to determine a planned path node with the minimum total cost by multiple iterations, generate a path node sequence from the path planning starting point to the path planning target point based on the planned path node with the minimum total cost, take the path node sequence as an optimal global path, and thus realize planning of the optimal path.

[0044] In another aspect of the application, to achieve the above object, a terminal device is disclosed, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, the memory stores a computer program capable of running on the processor, and the processor loads and executes the computer program, and a path planning method based on dynamic light as described above is adopted.

[0045] The application has the following beneficial effects:

[0046] The application proposes a global path planning algorithm Dynamic-3DA* which comprehensively considers the terrain slope, distance and time-varying light of the lunar south pole. Compared with path planning algorithms which only consider the terrain, distance and light factors, the algorithm has better dynamic light adaptability and lower terrain undulation index, which is beneficial to the fast and safe arrival of a lunar rover powered by solar energy at a detection destination. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0048] Figure 1 is a schematic diagram of the method of the present application;

[0049] Figure 2 is a slope distribution map and a slope direction distribution map of the target area of the present application;

[0050] Figure 3 is a distribution of average light intensity in the target area of the present application during a preset time period;

[0051] Figure 4 is a schematic diagram of the average light intensity distribution in the 8 groups of time periods of the present application;

[0052] Figure 5 is a dynamic passability map generated by using the slope map and the dynamic light map of the present application;

[0053] Figure 6 is a schematic diagram of the mobile model of the rover of the present application;

[0054] Figure 7 is a schematic diagram of the paths generated by the four path planning algorithms of the present application;

[0055] Figure 8 is a trend chart of various parameters of the path generated by the path planning algorithm of the present application;

[0056] Figure 9 is a schematic diagram of the paths generated by the algorithm starting at different starting times of the present application;

[0057] Figure 10 is a comparison schematic diagram of the experimental results of path planning starting at different times in the time-varying light environment of the present application;

[0058] Figure 11 is a schematic diagram of the system structure of the present application. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort belong to the scope of protection of the present application.

[0060] Embodiment one:

[0061] As Figure 1 shown, a dynamic light-based path planning method, the method comprising the following steps:

[0062] S101: receiving DEM map data, dynamic light map data, path planning starting point and path planning target point, generating slope map data based on DEM map data; wherein the DEM map data is data generated by a pre-established digital elevation model, containing elevation values distributed by grid size; processing based on slope map data and dynamic light map data to generate dynamic passability map, wherein the dynamic passability map is a time-varying passability constraint of whether each location is passable in each time step;

[0063] The process of generating slope map data based on DEM map data includes:

[0064] By calculating the change rate of elevation values distributed by grid size in the horizontal and numerical directions as slope map data.

[0065] The process of processing based on slope map data and dynamic light map data to generate dynamic passability map includes:

[0066] The slope map data and the dynamic light map data are binarized to obtain a slope passability map and an hourly light passability map, and the slope passability map and the hourly light passability map are respectively subjected to matrix point multiplication operation to obtain a dynamic passability map.

[0067] The slope passability map represents a passable area with a slope not exceeding a preset angle.

[0068] The hourly light passability map represents a passable area with a light rate not lower than a preset light threshold.

[0069] S102: taking the path planning starting point as the starting position of path search, taking the path planning target point as the search target, using a dynamic light-based three-dimensional path planning algorithm to perform path search based on the constraint of the dynamic passability map, and calculating the total cost of the planned path node based on DEM map data, slope map data and dynamic light map data, wherein the total cost includes terrain cost, distance cost and light cost.

[0070] A* algorithm is an improved heuristic path planning algorithm based on Dijkstra algorithm, which improves the calculation efficiency compared with the conventional graph search-based algorithm. In the scene of the present application, the light changes with time, so when the inspector moves in the time-varying environment, it is equivalent to doing path planning in a three-dimensional environment. As Figure 6As shown, the state of the rover at each time can be represented by a triple (x, y, t), and the red dot in the figure represents the current position of the rover, and the black dot represents the eight positions adjacent to it. The rover can choose a total of nine states at the next time, including moving to the other eight positions adjacent to it and still staying in the original position, as shown by the green dashed line in the figure.

[0071] The three-dimensional path planning algorithm based on dynamic lighting is a heuristic path planning algorithm improved based on the Dijkstra algorithm, and the heuristic function design of the three-dimensional path planning algorithm based on dynamic lighting includes:

[0072] For the current node n, the total cost function f(n) is defined as:

[0073] f(n) = g(n) + h(n) (1)

[0074] In the formula: g(n) represents the real cost from the starting node to the current node n, and h(n) represents the cost estimated by the heuristic function from the current node n to the destination node.

[0075] The calculation process of the terrain cost is as follows:

[0076] When the rover is located at position P(x, y), the corresponding slope is θ x，y . Specifically as follows:

[0077]

[0078] f H and f V respectively represent the change rate of the elevation in the horizontal and vertical directions, and are calculated as follows:

[0079]

[0080] In the above formula, H1-H9 represent the elevation values of the corresponding positions in the 3x3 grid centered on H5, and ΔH and ΔV are the resolutions of the elevation map in the horizontal and vertical directions, respectively;

[0081] When calculating the terrain cost, three factors are mainly considered, namely the change of terrain elevation, the change of slope, and the roughness of the terrain, which are represented by Cost h , Cost s , and Cost r respectively, and the total terrain cost is represented by Cost T . The passable area in the dynamic passability map is marked as Area yes , and the non-passable area is marked as Area no . Let the current position of the rover be P and the next position be P', then the terrain-related cost is calculated as follows:

[0082]

[0083] Cost T =ω1Cost h +ω2Cost s +ω3Cost r (8)

[0084] In formula (5), H P and H P′ respectively represent the elevation values of positions P and P', K h represents the influence coefficient of the elevation change, in formula (6), θ P and θ P′ respectively represent the slope values of positions P and P', K s represents the influence coefficient of the slope change, in formula (7), R P and R P′ respectively represent the terrain roughness values of positions P and P', K r represents the influence coefficient of the roughness change. In formula (8), the parameters ω1, ω2, ω3 satisfy the condition ω1+ω2+ω3=1.

[0085] The calculation process of the distance cost:

[0086] The distance cost is calculated using the Euclidean distance or the diagonal distance, assuming that the current time of the inspector is located at P point, and the next time is moved to P' point, then the distance cost Cost D can be calculated as follows:

[0087]

[0088] In the above formula, dist(P, P') is the distance between positions P and P', K D is the influence coefficient of the distance factor.

[0089] The calculation process of the illumination cost:

[0090] Assuming that the illumination intensity at P position at t time is and the illumination intensity at P' position at t+1 time is then the illumination cost Cost I can be defined as:

[0091]

[0092] In the above formula, K I is the influence coefficient of the illumination factor on the illumination cost function, and the overall illumination cost decreases as the illumination intensity at t+1 time becomes larger.

[0093] The heuristic function of the three-dimensional path planning algorithm based on dynamic light is defined as follows when considering the terrain cost, distance cost and light cost:

[0094] Cost heuristic = α * Cost T + β * Cost D + γ * Cost I (11)

[0095] In the above formula, the parameters α, β and γ satisfy the condition α + β + γ = 1.

[0096] S103: Determine the planned path node with the minimum total cost through multiple iterations, generate a path node sequence of the path planning starting point to the path planning target point based on the planned path node with the minimum total cost, take the path node sequence as an optimal global path, and thus realize planning of the optimal path.

[0097] Specifically, the present application is further described through embodiments as follows:

[0098] A 1300m x 1300m slope map and a slope direction map with a resolution of 20 meters / pixel are generated near the CE-7 mission potential landing area using the DEM map, as shown in FIG. 2. Figure 2 Then, the present application simulates the dynamic light condition of the area from November 1, 2026 to December 31, 2026, with a time resolution of 1 hour. Since there are 61 days from November to December in 2026, the present application obtains 1464 hours of dynamic light maps. Figure 3 The average light condition of the research area within 2 months is shown. In order to intuitively reflect the dynamic change of the light condition of the lunar south polar region, the 1464 hours of light maps are divided into 8 groups in time order, Figure 4 The average light rate in the 8 time periods is shown. It can be obviously seen that the light intensity in the area also dynamically changes with time.

[0099] Considering the engineering constraints, the rover needs to move in the area with a slope ≤ 15° and a light rate ≥ 60%. Therefore, the above slope map and dynamic light map are first binarized, and then a dynamic passability map is constructed, as shown in FIG. 4. Figure 5The slope passability map is shown in FIG. 6. The white pixel block in the map indicates that the slope value of the current position satisfies the constraint of ≤15°, so the rover can pass through the position. The black pixel block indicates that the slope value of the current position exceeds the maximum allowable slope value, and the current position is not passable. When generating the light passability map, 1464 light maps in the time range of the study area need to be processed respectively. When the light rate of the corresponding pixel block is ≥60%, it is marked as white, indicating that the light intensity of the position satisfies the light condition constraint, and the rover can pass through the position; otherwise, the pixel block is marked as black, indicating that it is not passable. Finally, the light passability map corresponding to each hour of light map is calculated. Finally, the slope passability map and the light passability map of each hour are respectively subjected to matrix point multiplication operation to obtain the dynamic passability map. The map reflects the passability of the rover in each hour of the study area from November 1, 2026 to December 31, 2026. The white pixel indicates that the rover can pass through the position, and the black pixel indicates that the position is not passable. The dynamic passability map generated by the slope map and the dynamic light map can guide the subsequent path planning generation.

[0100] In the present application, based on the designed heuristic function, an improved three-dimensional A* path planning algorithm considering dynamic light (Dynamic-3DA*) is proposed for the lunar polar rover. The algorithm is a heuristic search algorithm based on A*, which can finally generate the optimal global path on the DEM map or dynamic light map. At the same time, in each step of the search of the Dynamic-3DA* algorithm, the calculation of the state transition cost considers the terrain, distance, and dynamic light, and can adapt to different application scenarios through the setting of different weight parameters.

[0101] In order to quantitatively evaluate the effect of the Dynamic-3DA* path planning algorithm proposed in this paper, the present application selects an area of 1300x1300 square meters near the lunar south pole as the study area (ROI). First, the present application obtains the digital elevation model (DEM) generated by the American NASA lunar reconnaissance orbiter in the ROI area, and then generates the terrain slope map of the same area with a resolution of 20 meters / pixel based on the DEM of the area. At the same time, the present application also simulates the average light of the area from November 1, 2026 to December 31, 2026 with a time resolution of 1 hour. In order to simplify the calculation, the present application assumes that the lunar rover moves at a constant speed, and the motion speed v=20 m / h.

[0102] Finally, the simulation experiments of the proposed algorithm are conducted in the simulation environment of Python 3.11. In order to verify the effectiveness of the proposed algorithm, the three kinds of A* algorithms are compared, which only consider the terrain factor (Terrain-A*), the distance factor (Distance-A*) and the illumination factor (Illumination-A*), respectively. The parameter settings of the above-mentioned three control algorithms and the algorithm proposed in this paper are shown in Table 1. In addition, since the illumination conditions on the lunar surface are periodically dynamic, the influence of different starting times on the path planning results is also simulated. The settings of some parameters in this paper are shown in Table 1.

[0103] Table 1: Four groups of experimental parameters and four experimental scenarios

[0104]

[0105] Table 2: Other parameter settings involved in the experiment

[0106]

[0107] In order to evaluate the experimental results of the present application, the following evaluation indexes are set: 1) total path length Lenth. 2) path node number N. 3) average terrain undulation index IndexT. 4) cumulative illumination intensity Isum of the generated path.

[0108] Specifically, the path length Lenth and the path node number N are used to measure the length of the path generated by the path planning algorithm from the starting point to the target point; the average terrain undulation index IndexT is used to measure the terrain undulation degree experienced by the generated path, which is defined as: The meaning of the above formula is the standard deviation of the elevation, the standard deviation of the slope, and the average of the terrain roughness of the node list contained in the generated path. Isum is used to measure the cumulative illumination intensity on the generated path from the starting point to the target point, which is defined as

[0109] The simulation experiments are conducted in the region defined in the present application using the above-mentioned four path planning algorithms, and the experimental results are shown in Figure 7 Figure 7 (a) shows the path snapshots generated by the algorithm under different illumination intensity backgrounds, and it can be seen that the paths generated by the Terrain-A* algorithm and the Distance-A* algorithm will not be affected by the dynamic illumination intensity. The paths generated by the Illumination-A* algorithm and the Dynamic-3DA* algorithm will avoid the illumination shadow area, and successfully bypass when encountering an area that does not meet the illumination conditions, and select an area that meets the illumination intensity to pass through. Figure 7 ​(b) shows the path snapshots generated by the background of the terrain slope at different time. As can be seen from the figure, with the passage of time, the Terrain-A* algorithm and the Distance-A* algorithm will only select the node with the lowest terrain cost and the nearest distance to expand and generate the path, while the Illumination-A* algorithm and the Dynamic-3DA* algorithm will select the node with the best illumination condition to expand and generate the path.

[0110] The path generated by the Dynamic-3DA* algorithm is not completely affected by the terrain and slope.

[0111] Table 3: Experimental results of different path planning algorithms

[0112]

[0113] Table 3 shows the specific experimental results of different path planning algorithms. It can be seen that the Terrain-A* algorithm and the Distance-A* algorithm have the shortest path length and the number of path nodes, while the Illumination-A* algorithm and the Dynamic-3DA* algorithm have the longest path length and the largest number of path nodes.

[0114] The path generated by the Dynamic-3DA* algorithm is not completely affected by the terrain and slope.

[0115] In terms of average terrain undulation index, the Terrain-A* algorithm and the Distance-A* algorithm have smaller terrain undulation, and at the same time, since the Terrain-A* algorithm only considers the influence of terrain in the process of generating the path, the path generated by the algorithm has the smallest average terrain undulation. In addition, the Illumination-A* algorithm and the Dynamic-3DA* algorithm need to go through higher terrain undulation in order to avoid the influence of dynamic light. In terms of cumulative light intensity of the generated path, since the Illumination-A* algorithm needs to dynamically find the best light area, the path generated by the algorithm has the best cumulative light value. At the same time, since the Dynamic-3DA* algorithm also considers the factor of dynamic light in the process of generating the path, the cumulative light value on the path generated by the algorithm is also good. In addition, the Terrain-A* algorithm and the Distance-A* algorithm have poor cumulative light value, and at the same time, they will inevitably fall into the area with poor light conditions.

[0116] Figure 8The dynamic changes of elevation, slope, terrain roughness and illumination intensity during the path generation process of the four path planning algorithms are shown. It can be seen that the Terrain-A* algorithm and the Distance-A* algorithm have smaller fluctuation values in elevation, slope and terrain roughness during the path generation process. The Illumination-A* algorithm and the Dynamic-3DA* algorithm have relatively large fluctuations in elevation, slope, terrain roughness and other terrains due to the need to consider the influence of dynamic illumination. From the illumination intensity trend chart of the path points generated by the four algorithms, it can be seen that the Terrain-A* algorithm and the Distance-A* algorithm will pass through points without illumination during the path generation process, which cannot meet the illumination constraints required by path generation, while the paths generated by the Illumination-A* algorithm and the Dynamic-3DA* algorithm meet the dynamic illumination constraints.

[0117] Since the illumination conditions of the lunar south pole will experience a complete change in a lunar month, in order to explore the influence of different starting times of the algorithm on path generation in a dynamic illumination environment, the present application takes November 2026 as the research target, and counts the time of this month in units of hours, a total of 720 hours. At the same time, 4 time points are selected as the starting time of path planning for simulation experiment in this period, that is, 0 o'clock, 180 o'clock, 360 o'clock and 540 o'clock are selected respectively, and the experimental results are as follows.

[0118] From the above experimental results, it can be seen that the path planning algorithms can successfully complete the path planning task in the following three scenarios: Figure 9 It can be seen that in (a), (b), (d) figures, that is, starting from the 0th, 180th, 540th hour of November 1, 2026, the algorithm can successfully complete the path planning task. The above three scenarios have a reachable path from the starting point to the target point under the influence of dynamic illumination. However, in (c) figure, since the entire region is basically covered by shadows, an effective path from the starting point to the target point cannot be planned under the constraints of the present application.

[0119] Although the algorithms in (a), (b), (d) figures have found effective motion paths for the rover, it is very important to select a suitable path planning starting time for the rover in the lunar south pole illumination condition dynamic change environment to generate a high-quality and robust rover path. Table 4 and Figure 10The time for the algorithm to generate the path and the length of the generated path in the above scenarios are counted. It can be seen that among the four time points selected in the present application, the path generated by path planning starting from the 180th time point is the shortest in length, and the algorithm has the least execution time when path planning starts from the 540th time point. Compared with the above two scenarios, if path planning starts at the 0th time point, the generated path is relatively tortuous due to the sparse and scattered shadows caused by the light, and the length of the generated path and the required algorithm execution time are the longest.

[0120] Table 4: Experimental results of different path planning start times

[0121]

[0122] Embodiment Two: In order to achieve the above purpose, as shown in Figure 11 based on embodiment one, the present application discloses a path planning system based on dynamic lighting, comprising:

[0123] The data processing module 11 is used for receiving DEM map data, dynamic lighting map data, path planning starting point and path planning target point, generating slope map data based on the DEM map data; wherein the DEM map data is data generated by a pre-established digital elevation model, containing elevation values distributed according to grid size; the slope map data and the dynamic lighting map data are processed to generate a dynamic passability map, wherein the dynamic passability map is a time-varying passability constraint of whether each position is passable in each time step;

[0124] The cost acquisition module 12 is used for taking the path planning starting point as the starting position of path search, taking the path planning target point as the search target, and using a three-dimensional path planning algorithm based on dynamic lighting to search for a path based on the constraint of the dynamic passability map, and calculating the total cost of the planned path node based on the DEM map data, the slope map data and the dynamic lighting map data, wherein the total cost includes terrain cost, distance cost and lighting cost;

[0125] The path planning module 13 is used for determining the planned path node with the minimum total cost through multiple iterations, generating a path node sequence from the path planning starting point to the path planning target point based on the planned path node with the minimum total cost, taking the path node sequence as the optimal global path, thereby realizing the planning of the optimal path.

[0126] Based on the same inventive concept, the present application further provides a computer device, comprising: one or more processors, and a memory for storing one or more computer programs; the program comprises program instructions, and the processor is configured to execute the program instructions stored in the memory. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, and are configured to implement one or more instructions, and are specifically configured to load and execute one or more instructions in the computer storage medium to implement the above method.

[0127] It needs to be further explained that, based on the same inventive concept, the present application further provides a computer storage medium, which stores a computer program, and the computer program is executed by the processor to perform the above method. The storage medium can adopt any combination of one or more computer readable media. The computer readable medium can be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may, for example, be but is not limited to an electrical, magnetic, optical, electrical, magnetic, infrared, or semiconductor system, device or component, or any combination of the above. More specific examples (non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or component.

[0128] In the description of the present application, the description of the terms "one embodiment", "an example", "a specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present disclosure. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in one or more embodiments or examples.

[0129] The foregoing presents and describes the basic principles, main features and advantages of the present disclosure. It should be understood by those skilled in the art that the present disclosure is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only to illustrate the principles of the present disclosure. Without departing from the spirit and scope of the present disclosure, various changes and improvements can be made to the present disclosure, and all these changes and improvements fall within the scope of the present disclosure.

Claims

1. A path planning method based on dynamic illumination, characterized in that, The method includes the following steps: The system receives DEM map data, dynamic lighting map data, path planning start point, and path planning target point. It then generates slope map data based on the DEM map data. The DEM map data is generated from a pre-established digital elevation model and contains elevation values ​​distributed according to grid size. The system processes the slope map data and dynamic lighting map data to generate a dynamic accessibility map, which represents a time-varying access constraint for each location at each time step, indicating whether it is passable. The process of generating a dynamic accessibility map by processing slope map data and dynamic illumination map data includes: The slope map data and dynamic lighting map data are binarized to obtain the slope accessibility map and the lighting accessibility map for each hour. The slope accessibility map and the lighting accessibility map for each hour are then multiplied by matrix to obtain the dynamic accessibility map. The slope passability map represents a passable area where the slope does not exceed a preset angle; The hourly light accessibility map represents a passable area where the illuminance is not lower than a preset light threshold. Using the starting point of path planning as the starting position of path search and the target point of path planning as the search target, a three-dimensional path planning algorithm based on dynamic lighting is used to search the path based on the constraints of the dynamic accessibility map. The total cost of the planned path nodes is calculated based on DEM map data, slope map data and dynamic lighting map data, wherein the total cost includes terrain cost, distance cost and lighting cost. Through multiple iterations, the path nodes with the minimum total cost are determined. Based on the path nodes with the minimum total cost, a sequence of path nodes from the starting point to the target point of the path planning is generated. The sequence of path nodes is used as the optimal global path, thereby realizing the planning of the optimal path.

2. The path planning method based on dynamic illumination according to claim 1, characterized in that, The process of generating slope map data based on DEM map data includes: The slope map data is obtained by calculating the rate of change of elevation values ​​distributed by grid size in both horizontal and numerical directions.

3. The path planning method based on dynamic illumination according to claim 1, characterized in that, The dynamic lighting-based 3D path planning algorithm is a heuristic path planning algorithm improved from Dijkstra's algorithm. The heuristic function design of the dynamic lighting-based 3D path planning algorithm includes: For the current node The total cost function f(n) is defined as: (1) In the formula: This represents the distance from the starting node to the current node. The real cost, Indicates starting from the current node The cost estimated by the heuristic function to the destination node.

4. The path planning method based on dynamic illumination according to claim 1, characterized in that, The calculation process of the terrain cost: Assume the inspection device is located at position When, the corresponding slope is The details are as follows: (2) and These represent the rates of change of elevation in the horizontal and vertical directions, respectively, and are calculated as follows: (3) (4) In the above formula, They represent respectively with The elevation value at the corresponding location in the 3×3 grid centered on the center. and These represent the horizontal and vertical resolutions of the elevation map, respectively. The calculation of terrain cost mainly considers three factors: changes in terrain elevation, changes in slope, and terrain roughness, which are respectively represented by... , , The total terrain cost is expressed as: Mark the passable areas in the dynamic passability map as The area that cannot be passed is Let the current position of the patrol device be... The next position is The terrain-related costs are calculated as follows: (5) (6) (7) (8) In equation (5), and Representing positions respectively and Elevation value, The influence coefficient of elevation change is expressed in equation (6). and Representing positions respectively and The slope value, The coefficient representing the influence of slope variation is given in equation (7). and Representing positions respectively and The terrain roughness value, The coefficient representing the influence of roughness variation. In equation (8), the parameter , , satisfy conditions.

5. The path planning method based on dynamic illumination according to claim 1, characterized in that, The calculation process of the distance cost: The distance cost is calculated using Euclidean distance or diagonal distance, assuming the current location of the rovers is at position . Point, moving to in the next moment Point, then distance cost The following can be calculated: (9) In the above formula, For position and The distance between them This represents the influence coefficient of the distance factor.

6. The path planning method based on dynamic illumination according to claim 1, characterized in that, The calculation process of the illumination cost: Let time t be... The light intensity at the location is time t+1 The light intensity at the location is Then the cost of light It can be defined as: (10) In the above formula, The coefficient represents the influence of illumination factors on the illumination cost function. As the illumination intensity increases at time t+1, the overall illumination cost decreases.

7. The path planning method based on dynamic illumination according to claim 1, characterized in that, The heuristic function of the dynamic lighting-based 3D path planning algorithm, when simultaneously considering terrain cost, distance cost, and lighting cost, is defined as follows: (11) In the above formula, the parameter , , satisfy conditions.

8. A path planning system based on dynamic illumination, employing the path planning method based on dynamic illumination as described in any one of claims 1 to 7, characterized in that, include: The data processing module receives DEM map data, dynamic lighting map data, path planning start point, and path planning target point, and generates slope map data based on the DEM map data. The DEM map data is generated from a pre-established digital elevation model and contains elevation values ​​distributed according to grid size. The module processes the slope map data and dynamic lighting map data to generate a dynamic accessibility map, where the dynamic accessibility map represents the time-varying access constraints for each location at each time step. The cost acquisition module is used to search the path starting point as the starting position of the path planning and the path planning target point as the search target. Based on the constraints of the dynamic accessibility map, it uses a three-dimensional path planning algorithm based on dynamic lighting to search the path. Based on DEM map data, slope map data and dynamic lighting map data, it calculates the total cost of the planned path nodes. The total cost includes terrain cost, distance cost and lighting cost. The path planning module is used to determine the path nodes with the minimum total cost through multiple iterations. Based on the path nodes with the minimum total cost, a sequence of path nodes from the starting point to the target point of the path planning is generated. The sequence of path nodes is used as the optimal global path, thereby realizing the planning of the optimal path.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The memory stores a computer program that can run on the processor. When the processor loads and executes the computer program, it employs a path planning method based on dynamic lighting as described in any one of claims 1 to 7.

Images