Unmanned sweeper path planning method in open scene

Abstract

The invention relates to the field of unmanned motor sweepers, and provides an unmanned motor sweeper path planning method in an open scene, which comprises the following steps: acquiring a cleaning map of a cleaning area, and dividing the cleaning area into a plurality of sub-cleaning maps; constructing a sub-cleaning process corresponding to each sub-cleaning map; constructing a fixed circulation route based on the cleaning map and the cleaning process; determining the active time of the sub-cleaning map based on a historical monitoring picture; confirming a sub-periodicity based on the historical monitoring picture and the periodic time; and constructing a cleaning path of the cleaning map based on the cleaning task. During use, the sweeping times are distributed based on the sub-periodicity, the frequency of the low-garbage increment area is reduced, the frequency of the high-garbage increment area is increased according to needs, equipment idling and energy consumption waste caused by the same frequency of the whole area of a traditional fixed route are avoided, and the energy consumption and mechanical loss of the sweeper are reduced.

Description

Technical Field

[0001] This invention relates to the field of unmanned cleaning vehicles, and more specifically, to a path planning method for unmanned cleaning vehicles in open scenarios. Background Technology

[0002] As a daily recreational area for most residents, parks are generally divided into different zones such as children's play areas, elderly morning exercise areas, lawn areas, tree areas, and lake areas. These zones target different groups of people, so the activity times of each zone are different. At the same time, some activity areas are cyclical. For example, children's play areas may have higher traffic on weekends, and higher traffic will lead to a greater amount of garbage in this area. Existing autonomous vehicles typically follow fixed routes to clean parks in a cyclical manner. However, during peak hours, the high volume of pedestrian traffic increases the difficulty of obstacle avoidance for these vehicles. Additionally, during peak seasons, the amount of litter significantly increases, and cyclical cleaning cannot effectively address the litter generated in these areas. Conversely, some areas experience lower pedestrian traffic and less litter, requiring less frequent cleaning. This indicates that the existing autonomous cleaning vehicles' path allocation is not dynamically adjusted based on actual litter volume, leading to over-cleaning in low-litter areas, resulting in unreasonable resource allocation and waste.

[0003] To address these issues, we propose an open-scenario path planning method for unmanned sweepers. Summary of the Invention

[0004] Technical problems to be solved To address the problems existing in the prior art, the purpose of this invention is to provide a path planning method for unmanned sweepers in open scenarios, which solves the problem of uneven task allocation for sweepers.

[0005] Technical solution To solve the above problems, the present invention adopts the following technical solution.

[0006] A path planning method for an unmanned sweeper in an open scenario includes: Obtain the cleaning map of the cleaning area, obtain the building information of the cleaning map, and divide the cleaning map into multiple sub-cleaning maps based on the building information; Construct sub-cleaning processes corresponding to each of the sub-cleaning maps, wherein each sub-cleaning process includes a sub-cleaning path and a sub-cleaning time; A fixed loop route is constructed based on the cleaning map and the cleaning process; Using a daily cycle, the hourly garbage increment of the sub-cleaning map is confirmed based on historical monitoring images. The active time of the sub-cleaning map is determined based on the hourly garbage increment, and no-cleaning zones are set based on the active time. A periodic time is determined, and the unit garbage increment of the sub-cleaning map is confirmed based on the historical monitoring images and the periodic time. The sub-periodicity is determined based on the unit garbage increment. Based on the sub-periodicity corresponding to the sub-cleaning map and the no-cleaning time, cleaning tasks are allocated, and the cleaning path of the cleaning map is constructed based on the cleaning time, the cleaning task, and the fixed loop route of each sub-cleaning map.

[0007] Preferably, the step of dividing the cleaning map into multiple sub-cleaning maps includes: Obtain a cleaning map of the cleaning area and obtain building information from the cleaning map, wherein the building information includes building address and building function; Based on the building address, the cleaning map is divided into sub-cleaning maps centered on the building address.

[0008] Preferably, the step of constructing the sub-cleaning process corresponding to each of the sub-cleaning maps includes: Confirm the starting and ending positions of each of the sub-cleaning maps, wherein the starting and ending positions are the same; A sub-cleaning process is constructed based on the starting point location, the ending point location, and the sub-cleaning map. The sub-cleaning process includes a sub-cleaning path and a sub-cleaning time.

[0009] Preferably, the step of constructing a fixed route based on the cleaning map and the cleaning process includes: Based on all the sub-cleaning processes, the corresponding starting point position of all the sub-cleaning maps is determined; Based on the cleaning map, all the starting points are connected to form a fixed loop route.

[0010] Preferably, the step of confirming the hourly garbage increment of the sub-cleaning map based on historical monitoring images on a daily basis, and determining the active time of the sub-cleaning map based on the hourly garbage increment, includes: The sweeper is connected to the same smart management platform or data interaction platform as the park's monitoring system. The sweeper can obtain historical monitoring images from the park's monitoring system through the data interaction platform. Obtain historical monitoring images of the sub-cleaning map over multiple days, and obtain the hourly garbage increment based on the historical monitoring images; A daily periodic time series is constructed based on the day as the periodic time, and an hourly garbage increment series is constructed based on the hourly garbage increment. The similarity values ​​of the daily periodic time series and the hourly garbage increment series are obtained using the cosine similarity formula. The active time is determined based on the similarity value and the time corresponding to the maximum garbage increment in the sub-cleaning map. Set the active time corresponding to the sub-cleaning map as the no-cleaning time.

[0011] Preferably, the step of determining a periodic time, confirming the unit garbage increment of the sub-cleaning map based on the historical monitoring images and the periodic time, and determining the sub-periodicity based on the unit garbage increment includes: Select one of the aforementioned sub-cleaning maps as the pre-cleaning map; Define a period of time, wherein the period of time consists of multiple units of time; Obtain the historical monitoring footage corresponding to the pre-sweeping map, and use the historical monitoring video to confirm the garbage increment per unit time as the unit garbage increment; Obtain all unit garbage increments within multiple periods; Construct a garbage increment sequence based on the unit garbage increment; Construct a periodic time series based on the periodic time; The waste increment sequence and the periodic time series are used to determine the sub-periodicity of the pre-sweeping map using a cosine similarity formula; Based on the unit waste increment, the maximum unit waste increment of the pre-sweeping map is determined, and the corresponding cycle date is determined based on the maximum unit waste increment; Repeat the above steps to confirm the cycle date for all the sub-cleaning maps.

[0012] Preferably, the step of allocating cleaning frequency based on the sub-periodicity corresponding to the sub-cleaning map and the no-cleaning time, and constructing the cleaning path of the cleaning map based on the cleaning time, the cleaning frequency, and the fixed loop route of each sub-cleaning map, includes: The working time of the sweeper is divided into multiple hourly units; Determine whether the sub-cleaning map is a periodic cleaning task or a daily cleaning task; If the sub-cleaning map is the periodic cleaning task, then a unit cleaning task is allocated once per hourly unit of time based on the cleaning frequency; Confirm all cleaning tasks within the hourly unit of time, including pre-cleaning tasks, post-cleaning tasks, and the unit cleaning task; Based on all the cleaning tasks, the corresponding sub-cleaning process is confirmed, and a unit cleaning path is constructed based on the fixed loop route and all the sub-cleaning processes. Estimate the unit cleaning time of the unit cleaning path. If the unit cleaning time is greater than the hourly unit time, assign the cleaning tasks other than the start and end points of the unit cleaning path to the new cleaning vehicle. The cleaning path is constructed by concatenating all the unit cleaning paths of all the sweeping vehicles based on the fixed loop path.

[0013] Preferably, the step of determining the sub-cleaning process corresponding to the sub-cleaning map based on the cleaning task, and constructing a unit cleaning path based on the fixed loop route and the sub-cleaning process, includes: The first hourly unit time is obtained starting from the start time of the sweeper, the parking point is confirmed as the starting point, and the unit cleaning path corresponding to the hourly unit time is constructed based on the fixed loop path. Select the next hourly unit time, confirm the end point of the unit cleaning path corresponding to the previous hourly unit time as the starting point, and construct the unit cleaning path for the hourly unit based on the fixed loop path; Repeat the above steps to construct the cleaning path corresponding to all the hourly units of time.

[0014] Preferably, the daily cleaning tasks include: The active time corresponding to the sub-cleaning map is set before the preceding cleaning task, and the active time corresponding to the sub-cleaning map is set before the following cleaning task.

[0015] Beneficial effects: Compared with the prior art, the advantages of this invention are: Based on the active times of different areas in the park, corresponding no-sweeping times are set to avoid sweeping vehicles entering active areas during peak hours to wait or repeatedly avoid them, thus reducing time consumption. Based on the sub-periodic allocation of sweeping frequency, the frequency is reduced in areas with low garbage increase and increased as needed in areas with high garbage increase, avoiding the equipment idling and energy waste caused by the traditional fixed route "same frequency in the whole area", thus reducing the energy consumption and mechanical wear of sweeping vehicles. Set up pre-cleaning before peak hours and post-cleaning afterwards to ensure the area remains clean during peak traffic periods and prevent trash buildup from affecting the user experience. Attached Figure Description

[0016] Figure 1 This is an overall flowchart of the present invention; Figure 2 This is a flowchart illustrating the construction of a sub-cleaning map according to the present invention; Figure 3 This is a flowchart illustrating the construction process of the sub-cleaning of the present invention; Figure 4This is a flowchart illustrating the construction of a fixed loop route according to the present invention; Figure 5 This is a flowchart illustrating the process of determining active time according to the present invention. Figure 6 This is a flowchart illustrating the determination of sub-periodicity in this invention; Figure 7 This is a schematic diagram of the sub-cleaning path construction of the present invention; Figure 8 This is a schematic diagram of the unit cleaning path construction of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 A flowchart of a path construction method for an open-scenario unmanned sweeper path planning method, including: S100. Obtain the cleaning map of the cleaning area. There are two ways to obtain the cleaning map: one is to build it independently after actual inspection, and the other is to obtain it by connecting with the APP through the API interface. Obtain the building information of the cleaning map. The building information in the park includes children's play area, adult exercise area, etc. Divide the area into multiple sub-cleaning maps based on the building information. During the division process, it is necessary to ensure that the boundaries of each sub-cleaning map are clear, the cleaning range does not overlap and there are no omissions, forming a sub-map division pattern of "one building as the core and multiple areas fully covered".

[0019] S200. Construct sub-cleaning processes corresponding to each sub-cleaning map. The sub-cleaning process is mainly monitored to understand the path to be taken and the time required for cleaning the sub-cleaning map. The sub-cleaning process includes the sub-cleaning path and the sub-cleaning time. The sub-cleaning time refers to the time used to clean the sub-cleaning map.

[0020] S300 constructs a fixed loop route based on the cleaning map and the cleaning process. The construction of the fixed loop route needs to follow the principles of prioritizing the connection of adjacent areas, closing the overall path, and minimizing the total travel distance. Since there are many tourists in the central path of the area, in order to avoid too many tourists and cause a large amount of obstacle avoidance work, a fixed loop route is constructed on the periphery of the area, mainly for the cleaning vehicle to move between different areas.

[0021] S400 uses a daily time cycle to confirm the hourly garbage increase of the sub-cleaning map based on historical monitoring images. It selects monitoring images from multiple consecutive days and uses image recognition technology to count the amount of garbage in the images. It calculates the increase in the amount of garbage in the sub-cleaning map every hour. Based on the hourly garbage increase, it determines the active time of the sub-cleaning map. Since an increase in the number of people will increase the garbage increase, the time with the largest hourly garbage increase value can be set as the active time. Based on the active time, no-sweeping zones are set. During the active time, all sweeping vehicles are prohibited from entering the corresponding sub-cleaning map to operate. This can avoid collisions between sweeping vehicles and pedestrians, interference with normal activities in the area, and reduce the invalid waiting time of sweeping vehicles during high-traffic periods.

[0022] S500: Determine a cycle time, and based on historical monitoring images and the cycle time, confirm the unit garbage increment of the sub-cleaning map. Determine the sub-periodicity based on the unit garbage increment. Since the cycle duration is different between different areas, some areas have more people on Sundays, some have more people at the beginning of the month, and some areas are seasonal, such as the amount of fallen leaves on trees increases significantly in autumn. Therefore, when determining the sub-periodicity, it is necessary to first confirm the cycle duration, then confirm the corresponding garbage increment based on the cycle duration, and confirm whether the garbage increment has periodicity based on the garbage increment and cycle duration. If it has periodicity, then confirm the cycle date.

[0023] S600: Based on the sub-periodicity of the sub-cleaning map and the no-cleaning time, cleaning tasks are assigned. The no-cleaning time is the active time corresponding to each sub-cleaning map. The cleaning path of the cleaning map is constructed based on the cleaning time, cleaning tasks and fixed loop routes of each sub-cleaning map. The maximum number of cleaning tasks that a cleaning truck can clean is limited. Therefore, if the park area is large, multiple cleaning trucks need to be dispatched to clean the park.

[0024] In steps S100 to S600, the number of cleaning times for sub-cleaning maps is planned based on the relationship between the increase in garbage volume and time. This allows for multiple cleaning trips of cleaning vehicles if a sub-cleaning map has a large amount of garbage, while reducing the number of cleaning trips for sub-cleaning maps with a small amount of garbage, thus preventing waste of resources.

[0025] Please see Figure 2 A flowchart of a sub-cleaning map construction method for a path planning method for an unmanned sweeping vehicle in an open scenario, including: S101. Obtain the cleaning map of the cleaning area and the building information of the cleaning map. The building information is mainly named based on the functional requirements of the location. The building information includes the building address and building function.

[0026] S102. Based on the building address, divide the cleaning map into sub-cleaning maps centered on the building address.

[0027] In steps S101 to S102, the cleaning map is divided into multiple sub-cleaning maps based on function and building.

[0028] Please see Figure 3 A flowchart of a sub-cleaning process construction method for a path planning method for an unmanned sweeping vehicle in an open scenario, including: S201. Confirm the starting and ending positions of each sub-cleaning map. The starting and ending positions are the same.

[0029] S202. Construct a sub-cleaning process based on the starting point, ending point, and sub-cleaning map. The sub-cleaning process includes a sub-cleaning path and a sub-cleaning time. For each cleaning map, use a path planning algorithm (such as A* algorithm or Dijkstra algorithm) to plan the optimal sub-cleaning path covering the entire sub-cleaning map. The optimal path must meet the principles of "shortest path, fewest turns, and highest obstacle avoidance efficiency", and the starting point and ending point of the path must be the same. Estimate the sub-cleaning time for the constructed sub-cleaning path.

[0030] In steps S201 to S202, the sub-cleaning path of the sub-cleaning map is planned based on the terrain of the sub-cleaning map.

[0031] Please see Figure 4 A flowchart of a fixed-loop path construction method for an open-scenario unmanned sweeper path planning method, including: S301. Based on all sub-cleaning processes, confirm the corresponding starting point position of all sub-cleaning maps.

[0032] S302. Based on the cleaning map, all starting positions are connected to form a fixed loop route. The fixed loop route is used to connect various sub-cleaning maps, and is not located at the activity center of each sub-cleaning map to reduce obstacle avoidance work during cleaning. When connecting, the principle of "prioritizing the connection of adjacent areas, closed loop of the overall path, and shortest total travel distance" must be followed.

[0033] In steps S301 to S302, a fixed loop route is set based on the cleaning map, and the various sub-cleaning maps are connected in series. The cleaning vehicle transfers between the various sub-cleaning maps through the fixed loop route.

[0034] Please see Figure 5 A flowchart of an active time confirmation method for path planning of an unmanned sweeper in an open scenario, including: S401. Connect the sweeper and the park monitoring system to the same smart management platform or data interaction platform. The sweeper can obtain historical monitoring images from the park monitoring system through the data interaction platform. S402. Obtain multi-day historical monitoring images of the sub-cleaning map, obtain hourly garbage increments based on the historical monitoring images, construct a daily periodic time series based on the day as the periodic time, and construct an hourly garbage increment series based on the hourly garbage increments.

[0035] S403. The similarity between the daily cycle time series and the hourly garbage increment series is obtained by using the cosine similarity formula. Cosine similarity: Assuming the period is (like =7 days), count m complete cycles (e.g., m=4 weeks, total sample size ( ),definition: · (Garbage increment sequence): · (A periodic time series, where time values ​​repeat within each period).

[0036] Cosine similarity formula: Result judgment: · The range is [0,1] (because both time and increment are non-negative). • The closer to 1: the stronger the periodic correlation (e.g. =0.9, indicating that the pattern of garbage increase changing with the "weekly cycle" is very obvious. • Close to 0: No obvious periodic correlation.

[0037] S404. Based on the similarity and the time corresponding to the maximum garbage increment of the sub-cleaning map as the active time, the active time corresponding to the sub-cleaning map is set as the no-sweeping time. During the no-sweeping time, all cleaning vehicles are prohibited from entering the corresponding sub-cleaning map for cleaning.

[0038] In steps S400 to S404, the sub-sweeping area is observed. Figure 1 If the garbage increase within a day is found to suddenly increase at a certain time, the garbage increase over multiple days is retrieved. Based on the cosine formula, a garbage increase sequence and a periodic time sequence are set up to determine the similarity between the garbage increase and the periodic time. If the similarity is high, it means that a large amount of garbage appears in the sub-cleaning map at a specific time every day. Since the generation of garbage is usually closely related to the flow of people, the time when the garbage increase in the sub-cleaning map is large is taken as the active time.

[0039] Please see Figure 6 A flowchart of a cycle time confirmation method for path planning of an unmanned sweeper in an open scenario, including: S501. Select a sub-cleaning map as the pre-cleaning map; confirm a cycle time, wherein the cycle time consists of multiple unit times.

[0040] S502. Obtain the historical monitoring screen corresponding to the pre-sweeping map. The historical monitoring video is divided into unit historical monitoring video based on unit time. The garbage increment within a unit time is confirmed as the unit garbage increment based on the unit historical monitoring video.

[0041] S503. Obtain all unit garbage increments within multiple periods; construct a garbage increment sequence based on the unit garbage increments; construct a periodic time series based on the periodic time; determine the sub-periodicity of the pre-sweeping map using the garbage increment sequence and the periodic time series through the cosine similarity formula.

[0042] S504. Confirm the maximum unit garbage increment of the pre-sweeping map based on the unit garbage increment, and confirm the corresponding cycle date based on the maximum unit garbage increment. Repeat the above steps to confirm the cycle dates for all sub-cleaning maps.

[0043] In steps S500 to S504, by observing the garbage increase of the sub-cleaning map over multiple days, if it is found that the garbage increase suddenly increases on a certain day, the garbage increase of each cycle is retrieved, and the garbage increase sequence and cycle time sequence are set based on the cosine formula. The similarity between the garbage increase and the cycle time is judged. If the similarity is high, it means that a large amount of garbage appears in the sub-cleaning map at a specific time in each cycle (such as Sunday of each week, the 15th of each month, or autumn of each year). Therefore, the time when the garbage increase of the sub-cleaning map is large is taken as the cycle date.

[0044] Please see Figure 7 A structural diagram of a path planning method for an unmanned sweeper in an open scenario, comprising: dividing the sweeper's working time into multiple hourly units; S601. Compare the cycle date corresponding to the sub-cleaning map with the real-time date. If the comparison is successful, the sub-cleaning map is a cycle cleaning task; otherwise, it is a daily cleaning task. S602. If it is a daily cleaning task, the sweeper needs to complete the preparatory cleaning task before the start time of work and the sweeper needs to complete the final cleaning task after the end time of work. Both the preparatory cleaning task and the final cleaning task are to clean the entire cleaning map. The daily cleaning task includes the pre-cleaning task and post-cleaning task, preparatory cleaning task and final cleaning task corresponding to the sub-cleaning map.

[0045] S603. If the sub-cleaning map is a periodic cleaning task, then based on the cleaning frequency, one unit cleaning task is allocated per hourly unit of time. The cleaning frequency is set based on... The value, if A value ≥0.9 indicates a high frequency, and the cleaning frequency can be increased to once every hour on certain dates; if 0.7 ≤ If the similarity is less than 0.9, it is considered medium frequency, occurring once every 2 hours; if the similarity is less than or equal to 0.5, it is considered medium frequency. If the value is <0.7, it is considered a low-frequency word, so it should be read every three hours; Next, determine if there is an active period in the sub-sweep map; If there is no active time, the start time of the sweeper's work will be the starting point and the end time will be the ending point, based on... The value is used to assign one cleaning task to the corresponding hourly unit of time. If there is an active time, first assign the pre-cleaning task within the hourly unit time before the active time, and then assign the post-cleaning task within the hourly unit time after the active time. Based on the sub-cleaning map, the preceding and subsequent cleaning tasks are determined. The corresponding hourly time unit is determined based on the preceding cleaning tasks, with the start time of each hourly time unit as the starting point and the start time of the sweeper as the ending point. The value is used to assign a cleaning task to the corresponding hourly unit time, thus associating the cleaning task with the corresponding hourly unit time. Based on the hourly unit time corresponding to the subsequent cleaning task, the end time of the hourly unit time is used as the start point, and the end time of the sweeper's work is used as the end point, and based on... The value assigns a cleaning task to the corresponding hourly unit time, thus associating the cleaning task with the corresponding hourly unit time.

[0046] S604, according to Figure 8 It can be seen that the steps to construct a unit cleaning path include: Based on the hourly unit time, all cleaning tasks within the unit are confirmed. The cleaning tasks include pre-cleaning tasks, post-cleaning tasks, and unit cleaning tasks. Based on the cleaning tasks, the corresponding sub-cleaning map is confirmed. The first hourly unit time is obtained with the start time of the cleaning vehicle as the starting point. The stopping point is confirmed as the starting point. Based on the fixed loop path, all the sub-cleaning paths corresponding to the hourly unit time are connected to construct the unit cleaning path. Select the next hourly unit time, confirm the end point of the unit cleaning path corresponding to the previous hourly unit time as the starting point, and construct the unit cleaning path by connecting all the sub-cleaning paths corresponding to the hourly unit time based on the fixed loop path. Repeat the above steps to construct the cleaning path corresponding to all the hourly units of time, and construct a return path based on the cleaning map for the end point and the stopping point of the last cleaning path.

[0047] S607. Estimate the unit cleaning time of the unit cleaning path based on the sub-cleaning time and the fixed loop route. If the unit cleaning time is greater than the hourly unit time, assign the cleaning tasks other than the start and end points of the unit cleaning path to a new cleaning vehicle until the unit cleaning time is less than the hourly unit time.

[0048] S608. Repeat the above steps to estimate the unit cleaning time of the new sweeper. If the unit cleaning time is greater than the hourly unit time, allocate a new sweeper until all unit cleaning times are less than the hourly unit time. Construct a cleaning path by connecting all unit cleaning paths of all sweepers based on a fixed loop path.

[0049] In steps S600 to S608, cleaning tasks are assigned to sub-cleaning maps based on previous cycle time and active time, the completion time of cleaning tasks is calculated, and the corresponding number of cleaning vehicles are dispatched to sub-cleaning maps based on the completion time, so that each cleaning can achieve perfect coordination.

[0050] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A path planning method for an unmanned sweeper in an open scenario, characterized in that, include: Obtain the cleaning map of the cleaning area, obtain the building information of the cleaning map, and divide the cleaning map into multiple sub-cleaning maps based on the building information; Construct sub-cleaning processes corresponding to each of the sub-cleaning maps, wherein each sub-cleaning process includes a sub-cleaning path and a sub-cleaning time; A fixed loop route is constructed based on the cleaning map and the cleaning process; Using a daily cycle, the hourly garbage increment of the sub-cleaning map is confirmed based on historical monitoring images. The active time of the sub-cleaning map is determined based on the hourly garbage increment, and no-cleaning zones are set based on the active time. A periodic time is determined, and the unit garbage increment of the sub-cleaning map is confirmed based on the historical monitoring images and the periodic time. The sub-periodicity is determined based on the unit garbage increment. Based on the sub-periodicity corresponding to the sub-cleaning map and the no-cleaning time, cleaning tasks are allocated, and the cleaning path of the cleaning map is constructed based on the cleaning time, the cleaning task, and the fixed loop route of each sub-cleaning map.

2. The path planning method for an unmanned sweeper in an open scenario as described in claim 1, characterized in that, The step of dividing the cleaning map into multiple sub-cleaning maps includes: Obtain a cleaning map of the cleaning area and obtain building information from the cleaning map, wherein the building information includes building address and building function; Based on the building address, the cleaning map is divided into sub-cleaning maps centered on the building address.

3. The path planning method for an unmanned sweeper in an open scenario as described in claim 1, characterized in that, The steps for constructing the sub-cleaning process corresponding to each of the sub-cleaning maps include: Confirm the starting and ending positions of each of the sub-cleaning maps, wherein the starting and ending positions are the same; A sub-cleaning process is constructed based on the starting point location, the ending point location, and the sub-cleaning map. The sub-cleaning process includes a sub-cleaning path and a sub-cleaning time.

4. The path planning method for an unmanned sweeper in an open scenario as described in claim 3, characterized in that, The step of constructing a fixed route based on the cleaning map and the cleaning process includes: Based on all the sub-cleaning processes, the corresponding starting point position of all the sub-cleaning maps is determined; Based on the cleaning map, all the starting points are connected to form a fixed loop route.

5. The path planning method for an unmanned sweeper in an open scenario as described in claim 1, characterized in that, The process of determining the hourly garbage increment of the sub-cleaning map based on historical monitoring footage, using a daily time period, and determining the active time of the sub-cleaning map based on the hourly garbage increment, includes: The sweeper is connected to the same smart management platform or data interaction platform as the park's monitoring system. The sweeper can obtain historical monitoring images from the park's monitoring system through the data interaction platform. Obtain historical monitoring images of the sub-cleaning map over multiple days, and obtain the hourly garbage increment based on the historical monitoring images; A daily periodic time series is constructed based on the day as the periodic time, and an hourly garbage increment series is constructed based on the hourly garbage increment. The similarity values ​​of the daily periodic time series and the hourly garbage increment series are obtained using the cosine similarity formula. The active time is determined based on the similarity value and the time corresponding to the maximum garbage increment in the sub-cleaning map. Set the active time corresponding to the sub-cleaning map as the no-cleaning time.

6. The path planning method for an unmanned sweeper in an open scenario as described in claim 1, characterized in that, The steps of determining a periodic time, confirming the unit garbage increment of the sub-cleaning map based on the historical monitoring images and the periodic time, and determining the sub-periodicity based on the unit garbage increment include: Select one of the aforementioned sub-cleaning maps as the pre-cleaning map; Define a period of time, wherein the period of time consists of multiple units of time; Obtain the historical monitoring footage corresponding to the pre-sweeping map, and use the historical monitoring video to confirm the garbage increment per unit time as the unit garbage increment; Obtain all unit garbage increments within multiple periods; Construct a garbage increment sequence based on the unit garbage increment; Construct a periodic time series based on the periodic time; The waste increment sequence and the periodic time series are used to determine the sub-periodicity of the pre-sweeping map using a cosine similarity formula; Based on the unit waste increment, the maximum unit waste increment of the pre-sweeping map is determined, and the corresponding cycle date is determined based on the maximum unit waste increment; Repeat the above steps to confirm the cycle date for all the sub-cleaning maps.

7. The path planning method for an unmanned sweeper in an open scenario as described in claim 3, characterized in that, The step of allocating cleaning frequency based on the sub-periodicity corresponding to the sub-cleaning map and the no-cleaning time, and constructing the cleaning path of the cleaning map based on the cleaning time, the number of cleanings, and the fixed loop route of each sub-cleaning map, includes: The working time of the sweeper is divided into multiple hourly units; Determine whether the sub-cleaning map is a periodic cleaning task or a daily cleaning task; If the sub-cleaning map is the periodic cleaning task, then a unit cleaning task is allocated once per hourly unit of time based on the cleaning frequency; Confirm all cleaning tasks within the hourly unit of time, including pre-cleaning tasks, post-cleaning tasks, and the unit cleaning task; Based on all the cleaning tasks, the corresponding sub-cleaning process is confirmed, and a unit cleaning path is constructed based on the fixed loop route and all the sub-cleaning processes. Estimate the unit cleaning time of the unit cleaning path. If the unit cleaning time is greater than the hourly unit time, assign the cleaning tasks other than the start and end points of the unit cleaning path to the new cleaning vehicle. The cleaning path is constructed by concatenating all the unit cleaning paths of all the sweeping vehicles based on the fixed loop path.

8. The path planning method for an unmanned sweeper in an open scenario as described in claim 7, characterized in that, The steps of determining the sub-cleaning process corresponding to the sub-cleaning map based on the cleaning task, and constructing a unit cleaning path based on the fixed loop route and the sub-cleaning process, include: The first hourly unit time is obtained starting from the start time of the sweeper, the parking point is confirmed as the starting point, and the unit cleaning path corresponding to the hourly unit time is constructed based on the fixed loop path. Select the next hourly unit time, confirm the end point of the unit cleaning path corresponding to the previous hourly unit time as the starting point, and construct the unit cleaning path for the hourly unit based on the fixed loop path; Repeat the above steps to construct the cleaning path corresponding to all the hourly units of time.

9. The path planning method for an unmanned sweeper in an open scenario as described in claim 7, characterized in that, The routine cleaning tasks include: The active time corresponding to the sub-cleaning map is set before the preceding cleaning task, and the active time corresponding to the sub-cleaning map is set before the following cleaning task.

Images