Sanitation vehicle road sweeping operation method and system based on big data

Abstract

The invention relates to the technical field of data processing, in particular to a sanitation vehicle road sweeping operation method and system based on big data, and the method comprises the steps: taking the starting point and the ending point of a sweeping task and the position of an obstacle point as input, and constructing a configuration space of an RRT algorithm; in the configuration space, a random tree is constructed by using an RRT algorithm based on a preset step length by taking the starting point as the starting point, and in the growth process of the random tree, if a connecting line of a randomly sampled target point and a node closest to the target point in the current random tree does not intersect with a current obstacle point, the target point is taken as a growth point; and if the candidate points intersect, selecting a point with the maximum optimization degree from a plurality of randomly generated candidate points as a growing point so as to continuously expand the current random tree until a preset termination condition is met, and obtaining a cleaning path. According to the invention, the accuracy of path planning can be improved, so that when the cleaning task is executed, the obstacle can be effectively avoided, and the cleaning work can be comprehensively completed.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a road cleaning operation method and system for sanitation vehicles based on big data. Background Art

[0002] As the demand for road cleaning increases dramatically, the application of big data and artificial intelligence provides technical support for the introduction of intelligent sanitation vehicles, effectively reducing the labor burden and improving cleaning efficiency. However, the complexity of actual roads makes path planning a major challenge, and it is difficult for cleaning operations to accurately match road conditions, thus affecting work efficiency. Therefore, combining actual road condition data for intelligent path planning to ensure that cleaning operations are highly consistent with road conditions has become the key to improving efficiency.

[0003] In the related art, a path planning method based on the triangle inner core guidance RRT algorithm is disclosed in a patent application document with publication number CN110705803A. The method includes: forming three vertices of a triangle by using a random point generated by a random function, a point in a random tree that is closest to the random point, and a target point, and then calculating the inner core coordinates of the triangle, and using the inner core coordinates as the growth direction of the random tree; in addition, the sampling method is adjusted by recording the number of times the inner core guidance is used and the growth of the random tree under a certain number of cycles.

[0004] However, the above scheme determines the new node of the random tree based on the incenter of the triangle formed by the random point generated by the random function, the point in the random tree that is closest to the random point, and the target point. However, when the difference between the starting point and the target point is too large or too small, introducing the incenter of the triangle may increase the complexity of the calculation, and the flexibility of the new node after the introduction of the incenter is poor, which may result in the determined new node being able to effectively avoid obstacles but being far away from the obstacles, thereby failing to effectively plan the path around the obstacle, reducing the coverage of the effective planning area, and ultimately affecting the accuracy of path planning. Summary of the invention

[0005] In order to solve the problem of low accuracy of path planning based on the RRT algorithm, the present invention provides a road cleaning operation method and system for sanitation vehicles based on big data.

[0006] According to a first aspect of the present invention, a road cleaning operation method of a sanitation vehicle based on big data is provided, comprising: The starting point, end point and positions of all obstacle points of the pre-acquired cleaning task are used as input to construct the configuration space of the RRT algorithm; In the configuration space, starting from the starting point, a random tree is constructed using the RRT algorithm based on a preset step size. During the growth of the random tree, if the line connecting the randomly sampled target point and the node closest to the target point in the current random tree does not intersect with the current obstacle point, the target point is used as the growth point; if they intersect, the point with the highest preference is selected from the multiple randomly generated candidate points as the growth point to continue expanding the current random tree until the preset termination condition is met and a cleaning path is obtained; The method for obtaining the priority includes: calculating the importance of any candidate point, where the importance represents the distance between the candidate point and the current obstacle point within the safety distance; The vector formed by the candidate point and the latest node grown in the current random tree is used as the first vector, and the vector formed by the end point and the latest node grown in the current random tree is used as the second vector; the preference of the candidate point is calculated, and the preference is positively correlated with the importance, as well as the normalized value of the cosine similarity between the first vector and the second vector.

[0007] When the present invention uses the RRT algorithm for path planning, after the randomly sampled target point is abandoned, the growth point is determined based on the preference of each candidate point, which avoids the growth point of the random tree from deviating from the obstacle, avoids the path from deviating from the target, and can maximize the cleaning operation area. At the same time, the growth point always develops in the direction of the end point, which improves the accuracy of path planning, and can efficiently and comprehensively complete the cleaning task based on the precise cleaning path.

[0008] Preferably, before calculating the importance of any candidate point, the method further comprises calculating a distance influence factor of any candidate point; The method for obtaining the distance influence factor includes: If the distance between any candidate point and the current obstacle point is greater than or equal to the preset safety distance, the reciprocal of the distance is used as the distance influence factor of the candidate point; If it is less than the safety distance, the distance influence factor is set to a preset value; wherein the preset value is used to reduce the importance of the candidate point.

[0009] The present invention takes into account the safe distance from obstacles, and the set distance influence factor can reduce the importance of candidate points that are not within the safe distance, thereby accurately measuring the importance of each candidate point.

[0010] Preferably, the importance satisfies the following relationship: ; In the formula, For the The importance of each candidate point; For the The distance influencing factor of each candidate point; For the The distance between the candidate point and the current obstacle point; To preset a safe distance; It is the maximum distance between all candidate points and the current obstacle point.

[0011] The present invention can set a greater degree of influence for candidate points within a safe distance and close to obstacles, so that when the growth points are determined based on the preference of each candidate point, the cleaning operation area can be maximized.

[0012] Preferably, calculating the preference of the candidate point includes: Calculate the normalized value of cosine similarity to satisfy the following relationship: ; In the formula, For the The normalized value of the cosine similarity between the first vector and the second vector corresponding to the candidate point; For the The first vector corresponding to the candidate point; For the The second vector corresponding to the candidate points; It is a representation of cosine similarity; The product of the normalized value and the importance is taken as the preference of the corresponding candidate point.

[0013] The present invention determines the optimization degree of candidate points by integrating various data, thereby ensuring the accuracy of the optimization degree.

[0014] Preferably, the method for obtaining the vector component of the first vector includes: Take the latest node grown in the current random tree as the target node, and calculate the horizontal coordinate difference or vertical coordinate difference between any candidate point and the target node; The abscissa difference and the ordinate difference are used as vector components of the first vector.

[0015] Preferably, the vector components of the second vector are determined in the same manner as the vector components of the first vector.

[0016] Preferably, the starting positions of the first vector and the second vector are both target nodes.

[0017] The present invention sets the first vector and the second vector to the same starting position, which can measure the deviation between the direction of the candidate point and the end point.

[0018] According to a second aspect of the present invention, a road cleaning operation system for sanitation vehicles based on big data is provided. The system includes a memory and a processor. A computer program is stored in the memory. The processor executes the computer program to implement the steps of the first aspect of the present invention.

[0019] The present invention has the following effects: 1. When the present invention uses the RRT algorithm to determine the cleaning path, the growth points are screened based on the preference of each candidate point, thereby avoiding the distance between the selected growth point and the obstacle being too far, maximizing the cleaning operation area, and preventing deviation from the end point of the cleaning task, thereby ensuring the accuracy of the cleaning path and ultimately achieving efficient and comprehensive completion of the cleaning task.

[0020] 2. The present invention determines the preference of any candidate point by integrating various data, thereby ensuring the accuracy of the determined preference. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1 It is a schematic diagram of the steps of a road cleaning method for sanitation vehicles based on big data in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] 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 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0023] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] Reference Figure 1 The road cleaning operation method of a sanitation vehicle based on big data includes steps S1 and S2, which are as follows: S1: Take the pre-acquired starting point, end point and positions of all obstacle points of the cleaning task as input to construct the configuration space of the RRT algorithm.

[0025] First, the sanitation vehicle manager can determine the target area for the cleaning task, select specific streets and roads, and download map slices of the corresponding area from the high-precision map system.

[0026] Then, by calling the real-time traffic API interface provided by the map service provider, the latest traffic conditions information of the target area is extracted, including the starting point and end point of the cleaning task and the location information of all marked areas in the area that may affect the passage of the cleaning vehicle (such as construction areas, location information of traffic accident sites, etc.).

[0027] Afterwards, before the cleaning vehicle is officially put into operation, the starting point, end point and all obstacle points of the cleaning task are obtained by extracting the latest traffic condition information of the target area.

[0028] Finally, the starting position, the end position and the positions of all obstacle points of the cleaning task are used as the input of the Rapidly-exploring Random Tree (RRT) algorithm to construct the configuration space of the RRT algorithm. It should be noted that the process of constructing the configuration space based on the input information is a prior art and is not described in detail in this embodiment. Among them, the configuration space is a professional term in the RRT algorithm and is not described in detail in the present invention.

[0029] S2: In the configuration space, starting from the starting point, a random tree is constructed using the RRT algorithm based on a preset step size. During the growth of the random tree, if the line connecting the randomly sampled target point and the node closest to the target point in the current random tree does not intersect with the current obstacle point, the target point is used as the growth point; if they intersect, the point with the highest preference is selected from multiple randomly generated candidate points as the growth point to continue expanding the current random tree until the preset termination condition is met and a cleaning path is obtained.

[0030] It should be noted that the RRT algorithm is an algorithm for path planning, especially suitable for path planning problems of robots or autonomous systems in high-dimensional space. The basic principle is: by gradually building a random tree in the configuration space (C-space), continuously expanding from the existing nodes of the random tree to quickly explore the configuration space until a path from the starting point to the end point is found.

[0031] It should be further explained that when the traditional RRT algorithm expands the random tree, it will directly randomly select a new target point after discarding the target point, until the line connecting the target point and the node closest to the target point in the random tree no longer intersects with the obstacle point, and then expand the random tree. However, if the newly selected target point is far away from the obstacle point, it may cause the path to deviate from the predetermined target direction, thereby causing a deviation. Therefore, the present invention improves the traditional RRT algorithm. The specific improvement is: after discarding the randomly sampled target point, by calculating the preference of multiple candidate points, and selecting the candidate point with the largest preference as the growth point, the accuracy of the expanded nodes is ensured, and the accuracy of the path is further ensured.

[0032] In an exemplary embodiment of the present invention, the target point refers to a data point randomly sampled in the configuration space; the preset step size refers to the distance extended each time the random tree is expanded, and the preset step size set in this embodiment is 0.5 meters; the growth point refers to a data point that can be added to the random tree to become a node of the random tree during the expansion process of the random tree.

[0033] It should be noted that the distance between the target point and the node closest to the target point in the random tree is a preset step size, that is, 0.5 meters; the distance between each candidate point and the node closest to the corresponding candidate point in the random tree is a preset step size, that is, 0.5 meters.

[0034] The preference degree refers to the preference degree of any candidate point as a growth point, which is used as a data basis for selecting growth points from a plurality of randomly generated candidate points. For example, when the preference degree of any candidate point is large, the preference degree of the candidate point as a growth point is relatively large; on the contrary, when the preference degree of any candidate point is small, the preference degree of the candidate point as a growth point is relatively small.

[0035] Specifically, during the growth of the random tree, if the line connecting the randomly sampled target point and the node closest to the target point in the current random tree intersects with the obstacle point, the preference for screening the growth point can be determined by the following steps: Step 1: Randomly generate multiple candidate points in the configuration space, and calculate the importance of any candidate point. The importance represents the distance between the candidate point and the current obstacle point within the safety distance. It should be noted that cleaning operations require high efficiency. Even in an environment with obstacles, it is necessary to reasonably avoid obstacles when approaching them to ensure maximum coverage of the operation. However, in the processing of the traditional RRT algorithm, if a randomly selected target point crosses an obstacle, the point must be abandoned and a new target point must be reselected. This process may cause the cleaning path to avoid obstacles at a distance, thereby ignoring the area around the obstacle that should be cleaned, thereby affecting the efficiency of the operation. Therefore, when abandoning a randomly generated target point, the reselected point should be as close to the obstacle as possible while ensuring a safe distance, so as to ensure that the cleaning path can operate closely along the edge of the obstacle, thereby maximizing the efficiency of the operation and the cleaning coverage.

[0036] It should be further explained that, when calculating the importance of a corresponding candidate point based on the distance between any candidate point and the current obstacle point, the present invention takes into account that when a randomly sampled target point is discarded, the reselected point should be as close to the obstacle as possible on the basis of ensuring a safe distance, so as to ensure the accuracy of the importance calculation. In this embodiment, the number of candidate points is 10, and this embodiment does not specifically limit the number of candidate points.

[0037] In an exemplary embodiment of the present invention, before calculating the importance of any candidate point, it is also necessary to calculate the distance influence factor of any candidate point; specifically, the distance influence factor of any candidate point can be determined by the following steps: If the distance between any candidate point and the current obstacle point is greater than or equal to the preset safety distance, the reciprocal of the distance is used as the distance influence factor of the candidate point; if it is less than the safety distance, the distance influence factor is set to a preset value; wherein the preset value is used to reduce the importance of the candidate point.

[0038] Specifically, the distance influence factor of any candidate point satisfies the following relationship: ; In the formula, For the The distance influencing factor of each candidate point; For the The distance between the candidate point and the current obstacle point; To preset a safe distance, in this embodiment =1m (meter).

[0039] Among them, when , then it indicates If the distance between a candidate point and the current obstacle point is not within the safe distance, and the candidate point is used as the growth point of the random tree, the cleaning vehicle will collide with the current obstacle at this position. Therefore, in order to ensure the accuracy of the cleaning path, the distance impact factor of the candidate point that is not within the safe distance needs to be set to a smaller value, so as to avoid selecting the candidate point that is not within the safe distance as the growth point.

[0040] when , explain If the distance between a candidate point and the current obstacle point is within the safe distance, the inverse of the distance between the candidate point and the current obstacle point is used as the distance influence factor of the candidate point, so that the candidate point that remains within the safe distance and close to the current obstacle has a greater influence.

[0041] Furthermore, after determining the distance influence factor of each candidate point, the importance of any candidate point can be calculated. Specifically, the importance of any candidate point satisfies the following relationship: ; In the formula, For the The importance of each candidate point; For the The distance influencing factor of each candidate point; For the The distance between the candidate point and the current obstacle point; To preset a safe distance; It is the maximum distance between all candidate points and the current obstacle point.

[0042] in, ; In the formula, For the The horizontal coordinates of the candidate points; For the The vertical coordinate of the candidate point; is the horizontal coordinate of the current obstacle point; is the ordinate of the current obstacle point; the smaller the value, the closer the candidate point is to the current obstacle point.

[0043] when ,and If it is larger, the The distance between the candidate point and the current obstacle point is within the safe distance, and the candidate point is relatively close to the safety boundary, which means that the candidate point is more inclined to operate at the safety edge, thereby maximizing the cleaning area, and the corresponding candidate point is more important.

[0044] when = , then it indicates The distance between the candidate point and the current obstacle point is not within the safe distance. The value of is larger or smaller due to The existence of will make the importance of the candidate point at a lower level, thereby ensuring the accuracy of the determined importance.

[0045] Step 2: The vector formed by the candidate point and the latest node grown in the current random tree is used as the first vector, and the vector formed by the end point and the latest node grown in the current random tree is used as the second vector; determine the cosine similarity between the first vector and the second vector; In an exemplary embodiment of the present invention, the vector component of the first vector corresponding to any candidate point may be determined by the following steps: The latest grown node in the current random tree is taken as the target node, and the abscissa difference or ordinate difference between any candidate point and the target node is calculated; the abscissa difference and the ordinate difference are taken as the vector components of the first vector.

[0046] Specifically, The first vector corresponding to the candidate point The vector components of ,as well as ;in, For the The horizontal coordinates of the candidate points; is the horizontal coordinate of the target node; For the The vertical coordinate of the candidate point; is the ordinate of the target node; The first vector corresponding to the candidate point It can be expressed as .

[0047] In an exemplary embodiment of the invention, the vector components of the second vector are determined in the same manner as the vector components of the first vector.

[0048] Specifically, The second vector corresponding to the candidate point The vector components of ,as well as ;in, is the horizontal coordinate of the end point; is the horizontal coordinate of the target node; is the ordinate of the end point; is the ordinate of the target node; The second vector corresponding to the candidate point It can be expressed as It should be noted that the second vectors corresponding to multiple candidate points randomly generated at the same time are the same.

[0049] In an exemplary embodiment of the present invention, the starting positions of the first vector and the second vector are both target nodes.

[0050] The target node is used as the starting position of the first vector and the second vector corresponding to any candidate point, and the direction of the first vector and the second vector corresponding to the candidate point can be determined. After the starting position of the first vector and the second vector corresponding to the candidate point, the The first vector corresponding to the candidate point The direction is from the target node to the The direction of the candidate point; The second vector corresponding to the candidate point The direction is the direction from the target node to the end point.

[0051] Further, after determining the vector components and directions of the first vector and the second vector corresponding to any candidate point, the first vector and the second vector corresponding to the candidate point can be obtained, and then the cosine similarity calculation formula can be used to calculate the cosine similarity of the first vector and the second vector corresponding to the candidate point. It should be noted that the calculation method of the cosine similarity between vectors is a prior art and is not described in detail in this embodiment.

[0052] Optionally, if the cosine similarity between the first vector and the second vector corresponding to any candidate point is large, it means that the angle between the corresponding two vectors is small, which further indicates that the direction of the candidate point is closer to the direction toward the end point.

[0053] Step 3: Calculate the preference of the candidate point. The preference is positively correlated with the importance and the normalized value of the cosine similarity between the first vector and the second vector.

[0054] In an exemplary embodiment of the present invention, the preference of any candidate point can be determined by the following steps: The normalized value of the cosine similarity between the first vector and the second vector corresponding to any candidate point is calculated, and the normalized value and the importance of the candidate point are multiplied to obtain the preference of the candidate point.

[0055] Specifically, the normalized value of the cosine similarity between the first vector and the second vector corresponding to any candidate point satisfies the following relationship: ; In the formula, For the The normalized value of the cosine similarity between the first vector and the second vector corresponding to the candidate point; For the The first vector corresponding to the candidate point; For the The second vector corresponding to the candidate points; It is a representation of cosine similarity; It should be noted that, since the value range of the cosine similarity between vectors is (-1, 1), the present invention limits the range of the cosine similarity between vectors to 0-1, which can reduce the amount of calculation in the subsequent calculation process, thereby reducing the difficulty of calculation.

[0056] Optionally, after determining the normalized value of the cosine similarity between the first vector and the second vector corresponding to any candidate point, the preference of the candidate point can be calculated. Specifically, the preference of any candidate point satisfies the following relationship: ; In the formula, For the The preference of candidate points; For the The normalized value of the cosine similarity between the first vector and the second vector corresponding to the candidate point; For the The importance of a candidate point.

[0057] in, The larger the The angle between the first vector and the second vector corresponding to the candidate point is small, which means that the direction of the candidate point is closer to the direction toward the end point, and the corresponding candidate point has a higher preference degree; The larger the value is, the higher the preference of the candidate point is.

[0058] In another embodiment, the preference of the candidate point may also be calculated by other calculation formulas, such as by a summation formula instead of a multiplication formula.

[0059] Furthermore, the preference of all preferred points can be calculated, and the candidate point with the largest preference can be selected as the growth point, so as to get as close to the obstacle point as possible while ensuring a safe distance between the growth point and the obstacle point, while avoiding the direction of the growth point deviating from the end point.

[0060] Next, the growth process of the random tree is described in detail: First, the starting point in the configuration space is used as the root node of the random tree, and then a target point is randomly sampled in the configuration space. If the line connecting the target point and the root node does not intersect with the current obstacle point, the target point is used as the growth point; if it intersects, the point with the highest preference is selected from the 10 randomly generated candidate points as the growth point, and the determined growth point is added to the root node to update the random tree, and the above steps are repeated until the distance between the newly generated node in the random tree and the end point is less than the preset step size, that is, less than 0.5 meters and there is no obstacle, then it is determined that the expansion process of the random tree meets the preset termination condition, and the random tree after expansion is obtained; then a path from the starting point to the end point is traced back in the random tree after expansion to obtain a cleaning path.

[0061] Furthermore, after the cleaning path is determined, the cleaning path can be input into the navigation system of the sanitation vehicle, so that cleaning operations can be performed based on the cleaning path to minimize deviation from the predetermined cleaning area while avoiding obstacles and maximize the cleaning area, thereby fully completing the cleaning task.

[0062] The present invention also provides a road cleaning operation system for sanitation vehicles based on big data. The system includes a memory and a processor, and a computer program is stored in the memory. The computer program integrates the functions of the road cleaning operation method for sanitation vehicles based on big data. When the computer program is executed, the accuracy of path planning can be improved through the road cleaning operation method for sanitation vehicles based on big data, so that when performing cleaning tasks, obstacles can be effectively avoided and the cleaning work can be fully completed.

[0063] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0064] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A road cleaning method for sanitation vehicles based on big data, characterized in that: include: The starting point, end point and positions of all obstacle points of the pre-acquired cleaning task are used as input to construct the configuration space of the RRT algorithm; In the configuration space, starting from the starting point, a random tree is constructed using the RRT algorithm based on a preset step size. During the growth of the random tree, if the line connecting the randomly sampled target point and the node closest to the target point in the current random tree does not intersect with the current obstacle point, the target point is used as a growth point; If they intersect, the point with the highest preference is selected from the randomly generated multiple candidate points as the growth point to continue expanding the current random tree until the preset termination condition is met and a cleaning path is obtained; The method for obtaining the priority includes: calculating the importance of any candidate point, wherein the importance represents the distance between the candidate point and the current obstacle point within the safety distance; A vector formed by the candidate point and the latest node grown in the current random tree is used as a first vector, and a vector formed by the end point and the latest node grown in the current random tree is used as a second vector; the preference of the candidate point is calculated, and the preference is positively correlated with the importance and the normalized value of the cosine similarity between the first vector and the second vector.

2. The road cleaning method of sanitation vehicles based on big data according to claim 1 is characterized in that: Before calculating the importance of any candidate point, the method further includes calculating a distance influence factor of any candidate point; The method for obtaining the distance influence factor includes: If the distance between any candidate point and the current obstacle point is greater than or equal to the preset safety distance, the reciprocal of the distance is used as the distance influence factor of the candidate point; If it is less than the safety distance, the distance influence factor is set to a preset value; wherein the preset value is used to reduce the importance of the candidate point.

3. The road cleaning method of sanitation vehicles based on big data according to claim 2 is characterized in that: The importance satisfies the following relationship: ； In the formula, For the The importance of each candidate point; For the The distance influencing factor of each candidate point; For the The distance between the candidate point and the current obstacle point; To preset a safe distance; It is the maximum distance between all candidate points and the current obstacle point.

4. The road cleaning method of sanitation vehicles based on big data according to claim 1, characterized in that: The calculating the preference of the candidate point includes: The normalized value of the cosine similarity is calculated to satisfy the following relationship: ； In the formula, For the The normalized value of the cosine similarity between the first vector and the second vector corresponding to the candidate point; For the The first vector corresponding to the candidate point; For the The second vector corresponding to the candidate points; It is a representation of cosine similarity; The product of the normalized value and the importance degree is used as the preference degree of the corresponding candidate point.

5. The road cleaning method of sanitation vehicles based on big data according to claim 4 is characterized in that: The method for obtaining the vector component of the first vector includes: The most recently grown node in the current random tree is taken as the target node, and the horizontal coordinate difference or vertical coordinate difference between any candidate point and the target node is calculated; The horizontal coordinate difference and the vertical coordinate difference are used as vector components of the first vector.

6. The road cleaning method of sanitation vehicles based on big data according to claim 5, characterized in that: The vector components of the second vector are determined in the same manner as the vector components of the first vector.

7. The road cleaning method of sanitation vehicles based on big data according to claim 6 is characterized in that: The starting positions of the first vector and the second vector are both target nodes.

8. A road cleaning system for sanitation vehicles based on big data, characterized in that: The big data-based sanitation vehicle road cleaning operation system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the big data-based sanitation vehicle road cleaning operation method as described in any one of claims 1-7.

Images