Snow sweeping robot path planning system and method based on optimized Astar algorithm

Abstract

The invention relates to the technical field of robot path planning, and discloses a snow sweeping robot path planning system and method based on an optimized Astar algorithm, the snow sweeping robot path planning system comprises an accumulated snow thickness measurement module, a dynamic weight adjustment module, a path planning module and a path smoothing module, the thickness measurement module comprises a depth data acquisition unit and an accumulated snow depth calculation unit, the dynamic weight of the accumulated snow thickness is introduced into the cost function, and the self-adaptive adjustment mechanism of the dynamic weight is particularly important in extreme snowy weather, so that the robot can flexibly adjust path planning in a complex environment to better adapt to continuously changing snow depth, in addition, the accumulated snow thickness information is acquired in real time through a sensor, and the accuracy of the accumulated snow thickness information is improved. And denoising processing is carried out through median filtering, so that the influence of violent snow on the sensor can be well eliminated, the robot can preferentially select a path with less accumulated snow or a cleaned path, and thus the driving stability of the snow sweeping robot is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot path planning, and in particular to a snow-clearing robot path planning system and method based on an optimized Astar algorithm. Background Art

[0002] With the rapid development of robotics technology, the application scenarios of snow-clearing robots are becoming more and more extensive. How to achieve automated and efficient snow-clearing has become an important research topic at this stage. On the one hand, most snow-clearing robots can operate smoothly under normal weather conditions, but their ability to operate in extreme weather (such as blizzards) has not been fully verified. The low visibility and large white background under heavy snow conditions pose severe challenges to the performance of visual sensors, resulting in reduced map construction accuracy. On the other hand, the uneven distribution of snow thickness also poses difficulties for the path planning of snow-clearing robots. On sections with large differences in thickness, traditional path planning algorithms often find it difficult to provide an effective driving path, thereby reducing the efficiency and accuracy of snow-clearing operations. In the navigation process of snow-clearing robots, the Astar algorithm is widely used for path planning in static environments. The Astar algorithm finds the optimal path by evaluating the cost function of the path.

[0003] At present, most of the research on optimizing the Astar algorithm focuses on improving the algorithm's search efficiency, such as reducing computational complexity by improving heuristic functions, reducing the number of node expansions, optimizing data structures, etc. However, these optimization methods mainly target the algorithm's operating performance and do not fully consider the impact of the dynamic characteristics of complex environments (such as changes in snow thickness in the snow) on path planning. They cannot dynamically adjust the path cost under extreme snow conditions while ensuring navigation stability. At the same time, the algorithm does not have adaptive capabilities and cannot flexibly adjust the planning strategy under different snow conditions, which reduces the system's environmental adaptability.

[0004] To this end, we propose a snow-clearing robot path planning system and method based on the optimized Astar algorithm. Summary of the invention

[0005] The present invention mainly solves the technical problems existing in the above-mentioned prior art and provides a snow-clearing robot path planning system and method based on an optimized Astar algorithm.

[0006] In order to achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a snow-clearing robot path planning system based on an optimized Astar algorithm includes a snow thickness measurement module, a dynamic weight adjustment module, a path planning module and a path smoothing module. The snow thickness measurement module is used to obtain the snow depth data of the front area in real time through a depth camera. The snow thickness measurement module includes a depth data acquisition unit and a snow depth calculation unit. The depth data acquisition unit is used to collect the depth information of the front area of ​​the robot, and remove the influence of snowflakes on the depth camera during heavy snow by median filtering. The median filtering formula is as follows:

[0007] y[n]=median(x[nk],…,x[n],…,x[n+k])

[0008] Where x is the original signal, y is the filtered signal, n is the current position, k is the window size, and the snow depth calculation unit calculates the snow thickness based on the installation height and tilt angle of the depth camera. The calculation formula of L(n) is as follows:

[0009]

[0010] Among them, the installation height of the depth camera is H camera , the tilt angle is θ, and the depth value (distance to the snow surface) measured by the depth camera is H distance .

[0011] Preferably, the dynamic weight adjustment module is used to generate a dynamic cost weight based on snow depth data. The dynamic weight adjustment module includes a basic weight setting unit and an adaptive weight adjustment unit. The basic weight setting unit is used to set a basic weight value of snow thickness. The adaptive weight adjustment unit dynamically adjusts the weight based on snow depth data. The adjustment formula is:

[0012]

[0013] Among them, Wo is the basic weight, which represents the cost weight of the normal path and is set to 1.0. α is the dynamic adjustment coefficient of the weight. The larger the value, the more sensitive it is to the change of snow depth. In the initial setting, α is set to 0.5 to obtain a medium dynamic response effect, and then gradually adjusted according to the experimental data. Dmin and Dmax are the minimum and maximum values ​​of the snow depth, respectively. When the snow depth is small, the weight factor is close to the basic weight Wo, and the path planning is mainly based on the distance. When the snow depth is large, the weight W(n) will increase significantly, prompting the robot to avoid areas with thicker snow. This adaptive adjustment mechanism of dynamic weights ensures that the robot can respond flexibly to different snow thicknesses.

[0014] Preferably, the path planning module is used to execute the optimized Astar algorithm in combination with the dynamic cost weight to generate a navigation path that avoids areas with thick snow. The path planning module includes a path cost calculation unit and an optimal path generation unit. The path cost calculation unit calculates the path cost by using the optimized A* algorithm. The improved cost function is defined as:

[0015] f′(n)=g(n)+h(n)

[0016] g′(n)=g(n)+W(n)·L(n)

[0017] Among them, g(n) is the actual cost from the starting point to the current node, h(n) is the heuristic estimate, L(n) is the snow thickness of the current node, and www is the dynamic weight value. The optimal path generation unit generates the optimal path to avoid the thick snow area based on the calculated cost. In order to further improve the adaptability of path planning, the threshold setting of snow thickness is introduced in the optimized A* algorithm: when the snow thickness is less than or equal to 3cm, it is considered that the impact of this area on robot navigation can be ignored, and path planning is only calculated based on the distance cost of the traditional A* algorithm. When the snow thickness is in the range of 3cm-15cm, the snow thickness is included in the path cost calculation. The path priority is adjusted by the dynamic weight W(n). The robot tends to choose the area with shallow snow. When the snow thickness exceeds 15cm, it is regarded as an inaccessible area, which is equivalent to a static obstacle. After the threshold is introduced, the path planning cost function is adjusted as follows:

[0018]

[0019] Preferably, the path smoothing module is used to perform curve smoothing on the planned path to reduce path turning points and improve driving stability. The path smoothing module includes a smooth curve generation unit and a path optimization unit. The smooth curve generation unit uses a cubic B-spline curve to smooth the generated path. Suppose the control point sequence is P = {P0, P1, ..., P n n}, the defined cubic B-spline curve C(u) can be expressed as:

[0020]

[0021] Among them, N i,k is the B-spline basis function, k is the order of the curve, and the basis function is recursively defined as follows:

[0022]

[0023] The method for snow-clearing robot path planning based on the optimized Astar algorithm includes the above-mentioned snow-clearing robot path planning system based on the optimized Astar algorithm, and specifically includes the following steps:

[0024] S1: collect ground information through the depth camera and perform denoising through median filtering;

[0025] S2: Calculate the snow thickness in the area in front of the robot;

[0026] S3: Generate a dynamic weight of snow thickness through a dynamic weight adjustment module according to the snow thickness data;

[0027] S4: In the path planning module, the optimized Astar algorithm is executed in combination with the dynamic weight value to generate the optimal path avoiding the thick snow area;

[0028] S5: using a path smoothing module to perform curve smoothing on the optimal path to generate a smooth path;

[0029] S6: The navigation path after path smoothing is used to guide the driving of the snow-clearing robot.

[0030] Preferably, in S1, a depth camera is used to obtain the snow depth in real time and a median filter is performed.

[0031] Preferably, in S2, the depth value is converted into snow thickness by calculation.

[0032] Preferably, in S3, the path cost weight is dynamically adjusted according to the snow thickness.

[0033] Preferably, in S4, the optimal path is generated by combining the optimized Astar algorithm with the cost function.

[0034] Preferably, in S5, the path is smoothed using a cubic B-spline curve to ensure the execution effect of the path planning.

[0035] The present invention provides a snow-clearing robot path planning system and method based on an optimized Astar algorithm.

[0036] It has the following beneficial effects:

[0037] 1. The snow-clearing robot path planning system and method based on the optimized Astar algorithm introduces the dynamic weight of snow thickness into the cost function. In extreme blizzard weather, the adaptive adjustment mechanism of the dynamic weight is particularly important, so that the robot can flexibly adjust the path planning in a complex environment to better adapt to the changing snow depth. In addition, the snow thickness information is obtained in real time by the sensor, and denoising is performed by median filtering, which can better eliminate the impact of blizzards on the sensor, so that the robot can give priority to paths with less snow or cleared paths, thereby improving the driving stability of the snow-clearing robot.

[0038] 2. The snow-clearing robot path planning system and method based on the optimized Astar algorithm optimizes the application of the Astar algorithm in the snowy environment by introducing a dynamic weight mechanism, so that the snow-clearing robot can adapt to different snow depths and select the optimal path. In addition, the path smoothing process effectively improves the driving stability of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a system module diagram of the present invention;

[0040] Figure 2 It is the optimized Astar algorithm flow chart of the present invention;

[0041] Figure 3 It is a comparison diagram of the traditional Astar algorithm of the present invention and the optimized Astar algorithm in a simulated snow two-dimensional grid. DETAILED DESCRIPTION

[0042] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings described below are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0043] The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with the technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantial technical significance. Any structural modification, change in proportion or adjustment of size shall still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.

[0044] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0045] In the description of the embodiments of the present invention, it should be noted that the terms "center", "upper", "lower", "inner", "outer", "side", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the invented product is usually placed when in use, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0046] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific circumstances.

[0047] 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.

[0048] Example 1: Snow-clearing robot path planning system based on optimized Astar algorithm, such as Figure 1 As shown, it includes a snow thickness measurement module, a dynamic weight adjustment module, a path planning module and a path smoothing module. The snow thickness measurement module is used to obtain the snow depth data of the front area in real time through the depth camera. The snow thickness measurement module includes a depth data acquisition unit and a snow depth calculation unit. The depth data acquisition unit is used to collect the depth information of the front area of ​​the robot and remove the influence of snowflakes on the depth camera during heavy snow by median filtering. The median filtering formula is as follows:

[0049] y[n]=median(x[nk],…,x[n],…,x[n+k])

[0050] Where x is the original signal, y is the filtered signal, n is the current position, k is the window size, and the snow depth calculation unit calculates the snow thickness based on the installation height and tilt angle of the depth camera. The calculation formula of L(n) is as follows:

[0051]

[0052] Among them, the installation height of the depth camera is H camera , the tilt angle is θ, and the depth value (distance to the snow surface) measured by the depth camera is H distance .

[0053] The dynamic weight adjustment module is used to generate dynamic cost weights based on snow depth data. The dynamic weight adjustment module includes a basic weight setting unit and an adaptive weight adjustment unit. The basic weight setting unit is used to set the basic weight value of snow thickness. The adaptive weight adjustment unit dynamically adjusts the weight based on snow depth data. The adjustment formula is:

[0054]

[0055] Among them, Wo is the basic weight, which represents the cost weight of the normal path and is set to 1.0. α is the dynamic adjustment coefficient of the weight. The larger the value, the more sensitive it is to the change of snow depth. In the initial setting, α is set to 0.5 to obtain a medium dynamic response effect, and then gradually adjusted according to the experimental data. Dmin and Dmax are the minimum and maximum values ​​of the snow depth, respectively. When the snow depth is small, the weight factor is close to the basic weight Wo, and the path planning is mainly based on the distance. When the snow depth is large, the weight W(n) will increase significantly, prompting the robot to avoid areas with thicker snow. This adaptive adjustment mechanism of dynamic weights ensures that the robot can respond flexibly to different snow thicknesses.

[0056] The path planning module is used to execute the optimized Astar algorithm in combination with the dynamic cost weight to generate a navigation path that avoids areas with thick snow. The path planning module includes a path cost calculation unit and an optimal path generation unit. The path cost calculation unit calculates the path cost through the optimized A* algorithm. The improved cost function is defined as:

[0057] f′(n)=g′(n)+h(n)

[0058] g′(n)=g(n)+W(n)·L(n)

[0059] Among them, g(n) is the actual cost from the starting point to the current node, h(n) is the heuristic estimate, L(n) is the snow thickness of the current node, and www is the dynamic weight value. The optimal path generation unit generates the optimal path to avoid the thick snow area based on the calculated cost. In order to further improve the adaptability of path planning, the threshold setting of snow thickness is introduced in the optimized A* algorithm: when the snow thickness is less than or equal to 3cm, it is considered that the impact of this area on robot navigation can be ignored, and path planning is only calculated based on the distance cost of the traditional A* algorithm. When the snow thickness is in the range of 3cm-15cm, the snow thickness is included in the path cost calculation. The path priority is adjusted by the dynamic weight W(n). The robot tends to choose the area with shallow snow. When the snow thickness exceeds 15cm, it is regarded as an inaccessible area, which is equivalent to a static obstacle. After the threshold is introduced, the path planning cost function is adjusted as follows:

[0060]

[0061] The path smoothing module is used to perform curve smoothing on the planned path to reduce the turning points of the path and improve the driving stability. The path smoothing module includes a smoothing curve generation unit and a path optimization unit. The smoothing curve generation unit uses a cubic B-spline curve to smooth the generated path. Let the control point sequence be P = {P0, P1, …, P n n}, and the defined cubic B-spline curve C(u) can be expressed as:

[0062]

[0063] where N i,k is the B-spline basis function, k is the order of the curve, and the basis function is recursively defined as follows:

[0064]

[0065] By introducing the dynamic weight of the snow depth into the cost function, in extreme blizzard weather, the adaptive adjustment mechanism of the dynamic weight is particularly important, enabling the robot to flexibly adjust the path planning in a complex environment to better adapt to the changing snow depth. In addition, by obtaining the snow depth information in real time through the sensor and performing denoising processing through median filtering, the influence of the blizzard on the sensor can be better eliminated, and the robot can be enabled to preferentially select paths with less snow accumulation or cleared paths, thereby improving the driving stability of the snow-sweeping robot.

[0066] Embodiment 2: On the basis of Embodiment 1, Embodiment 2, and Embodiment 3, as Figure 1 shown, in order to facilitate the visualization of the data results, the real snow map is converted into a two-dimensional grid map with a grid side length of 0.5 m, a total of 256 grid cells, and the snow depth range is 3 cm to 20 cm. Static obstacles are added to simulate the actual complex environment. To improve the repeatability and intuitiveness of the experiment, where: the gray value R = 0 indicates the presence of obstacles or a snow depth greater than 15 cm (black), the gray value R = 255 indicates no obstacles or a snow depth less than 3 cm (white), and 0 < R < 255 indicates different snow depths. The larger the gray value, the shallower the snow. Two sets of starting points are randomly set to verify the effectiveness of the optimized Astar algorithm. The index of "the proportion of avoiding high-snow-depth areas" is introduced. When the snow depth is greater than 12 cm, it is defined as a thick snow area. Its calculation formula is as follows:

[0067]

[0068] where N total is the total number of coordinate points on the path; N high is the number of coordinate points on the path that fall in the high-snow-depth area. The experimental results are shown in Table 1:

[0069] Table 1 Experimental data table

[0070]

[0071] Although the optimized A* algorithm increases the total length of the path, it successfully achieves the goal of 100% avoiding high snow depth areas through the dynamic weight mechanism, and the path cost is significantly reduced, indicating that the optimized algorithm is more suitable for complex snow environments and effectively avoids the risk of high resistance driving in areas with thick snow. By introducing a dynamic weight mechanism, the application of the Astar algorithm in snowy environments is optimized, allowing the snow-clearing robot to adapt to different snow depths and select the optimal path. In addition, path smoothing effectively improves the robot's driving stability.

[0072] Embodiment 3: A method for snow-clearing robot path planning based on an optimized Astar algorithm, including the above-mentioned snow-clearing robot path planning system based on an optimized Astar algorithm, specifically includes the following steps: S1: collecting ground information through a depth camera and performing denoising through a median filter; S2: calculating the thickness of snow in the area in front of the robot; S3: generating a dynamic weight of snow thickness through a dynamic weight adjustment module based on snow thickness data; S4: in the path planning module, combining the dynamic weight value, executing the optimized Astar algorithm to generate an optimal path that avoids areas with thick snow; S5: using a path smoothing module to perform curve smoothing on the optimal path to generate a smooth path; S6: guiding the driving of the snow-clearing robot through the navigation path after path smoothing. In S1, the depth of snow is obtained in real time by using a depth camera and a median filter is performed. In S2, the depth value is converted into snow thickness by calculation. In S3, the path cost weight is dynamically adjusted according to the snow thickness. In S4, the optimal path is generated by combining the optimized Astar algorithm with the cost function. In S5, the path is smoothed using a cubic B-spline curve to ensure the execution effect of path planning.

[0073] Embodiment 4: A method for snow-clearing robot path planning based on an optimized Astar algorithm, based on Embodiment 4, specifically includes the following steps: Step 1: Acquire ground information through a depth camera and perform denoising through a median filter; Step 2: Calculate the thickness of the snow to provide data support for subsequent path planning. This step ensures that the robot can understand the environment ahead in real time; Step 3: Calculate the dynamic cost weight by processing the snow depth data. The dynamic adjustment of the weight can reflect the impact of the current snow thickness on the path cost, thereby providing a more environmentally adaptable weight for path selection; Step 4: First set the current position and target point of the robot, then calculate the path cost through the Astar algorithm adjusted based on the dynamic weight, and select the optimal path based on the total cost. This step optimizes the path selection of the traditional Astar algorithm, so that the snow-clearing robot can give priority to paths with thinner or cleared snow, and avoid areas with thicker snow; Step 5: Optimize the path through B-spline curve smoothing, reduce unnecessary turning points, ensure a smoother path, and improve the driving stability of the robot.

[0074] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A snow-clearing robot path planning system based on the optimized Astar algorithm, characterized in that: It includes a snow thickness measurement module, a dynamic weight adjustment module, a path planning module and a path smoothing module. The snow thickness measurement module includes a depth data acquisition unit and a snow depth calculation unit. The depth data acquisition unit is used to collect the depth information of the area in front of the robot and remove the influence of snowflakes on the depth camera during heavy snow by median filtering. The median filtering formula is as follows: y[n]=median(x[nk],…,x[n],…,x[n+k]) Where x is the original signal, y is the filtered signal, n is the current position, k is the window size, and the snow depth calculation unit calculates the snow thickness based on the installation height and tilt angle of the depth camera. The calculation formula of L(n) is as follows: Among them, the installation height of the depth camera is H camera , the tilt angle is θ, and the depth value (distance to the snow surface) measured by the depth camera is H distance .

2. The snow-clearing robot path planning system based on the optimized Astar algorithm according to claim 1 is characterized in that: The dynamic weight adjustment module includes a basic weight setting unit and an adaptive weight adjustment unit. The basic weight setting unit is used to set the basic weight value of snow thickness. The adaptive weight adjustment unit dynamically adjusts the weight based on snow depth data. The adjustment formula is: Among them, Wo is the basic weight, which represents the cost weight of the normal path and is set to 1.

0. α is the dynamic adjustment coefficient of the weight. The larger the value, the more sensitive it is to the change of snow depth. In the initial setting, α is set to 0.5 to obtain a medium dynamic response effect, and then gradually adjusted according to the experimental data. Dmin and Dmax are the minimum and maximum values ​​of the snow depth, respectively. When the snow depth is small, the weight factor is close to the basic weight Wo, and the path planning is mainly based on the distance. When the snow depth is large, the weight W(n) will increase significantly, prompting the robot to avoid areas with thicker snow. This adaptive adjustment mechanism of dynamic weights ensures that the robot can respond flexibly to different snow thicknesses.

3. The snow-clearing robot path planning system based on the optimized Astar algorithm according to claim 1 is characterized in that: The path planning module includes a path cost calculation unit and an optimal path generation unit. The path cost calculation unit calculates the path cost by using the optimized A* algorithm. The improved cost function is defined as: f′(n)=g′(n)+h(n) g′(n)=g(n)+W(n)·L(n) Among them, g(n) is the actual cost from the starting point to the current node, h(n) is the heuristic estimate, L(n) is the snow thickness of the current node, and www is the dynamic weight value. The optimal path generation unit generates the optimal path avoiding the thick snow area based on the calculated cost. In order to further improve the adaptability of path planning, the threshold setting of snow thickness is introduced in the optimized A* algorithm: when the snow thickness is less than or equal to 3cm, it is considered that the impact of this area on robot navigation can be ignored, and path planning is only calculated based on the distance cost of the traditional A* algorithm. When the snow thickness is in the range of 3cm-15cm, the snow thickness is included in the path cost calculation. The path priority is adjusted by the dynamic weight W(n). The robot tends to choose the area with shallow snow to travel. When the snow thickness exceeds 15cm, it is regarded as an inaccessible area, which is equivalent to a static obstacle. After the threshold is introduced, the path planning cost function is adjusted as follows:

4. The snow-clearing robot path planning system based on the optimized Astar algorithm according to claim 1 is characterized in that: The path smoothing module includes a smooth curve generating unit and a path optimization unit. The smooth curve generating unit uses a cubic B-spline curve to smooth the generated path. Assume that the control point sequence is P = {P0, P1, ..., P n n}, the defined cubic B-spline curve C(u) can be expressed as: Among them, N i,k is the B-spline basis function, k is the order of the curve, and the basis function is recursively defined as follows:

5. A method for snow-clearing robot path planning based on optimized Astar algorithm, characterized in that: The snow-clearing robot path planning system based on the optimized Astar algorithm according to any one of claims 1 to 4 specifically comprises the following steps: S1: collect ground information through the depth camera and perform denoising through median filtering; S2: Calculate the snow thickness in the area in front of the robot; S3: Generate a dynamic weight of snow thickness through a dynamic weight adjustment module according to the snow thickness data; S4: In the path planning module, the optimized Astar algorithm is executed in combination with the dynamic weight value to generate the optimal path avoiding the thick snow area; S5: using a path smoothing module to perform curve smoothing on the optimal path to generate a smooth path; S6: The navigation path after path smoothing is used to guide the driving of the snow-clearing robot.

Images