Path planning method in complex scene and cleaning robot

Abstract

A path planning method in a complex scene and a cleaning robot relate to the field of path planning, and the method comprises the following steps: acquiring current cleaning task information and recognizing terrain types in a cleaning area based on an image recognition algorithm, so that the cleaning area can be accurately divided, and the area complexity level can be calculated, thereby realizing accurate task allocation; the total energy demand is calculated according to the task type and the task precision requirement, then energy distribution is carried out according to the current electric quantity and the regional complexity level of the robot, and it is ensured that the cleaning task can be completed under the optimal energy consumption. Besides, by determining the task priority and the task execution parameter of each sub-cleaning task, the reasonability and the high efficiency of path planning are ensured, so that the cleaning robot can reasonably plan a path in a complex scene, the power consumption is reduced, and the cleaning efficiency is improved.

Description

Technical Field

[0001] The present application relates to the field of path planning, and in particular to a path planning method and a cleaning robot in complex scenarios. Background Art

[0002] With the development of science and technology, cleaning robots are increasingly widely used. Cleaning robots can automatically complete various cleaning tasks, improve work efficiency, and reduce people's labor intensity. However, path planning of cleaning robots in complex scenarios has always been a technical problem, especially in environments with complex terrain and diverse tasks. How to efficiently plan paths, allocate tasks, and manage energy has become a research focus.

[0003] In the prior art, cleaning robots usually use preset path planning algorithms to complete cleaning tasks. These algorithms are usually based on simple terrain models, such as plane maps or two-dimensional grid maps, and use fixed rules and strategies for path planning and task execution. For example, some cleaning robots use random roaming or rule-based path planning methods to complete cleaning tasks by continuously detecting and avoiding obstacles. These methods perform well in simple scenarios, but in environments with complex terrain, cleaning efficiency and effects are often difficult to guarantee.

[0004] Although existing technologies can solve the path planning problem of cleaning robots to a certain extent, when faced with complex terrain and diverse tasks, the robot may repeat the previously cleaned path when the battery is already low, resulting in excessive energy consumption and failure to complete the cleaning task. Summary of the invention

[0005] The present application provides a path planning method and a cleaning robot in complex scenarios, which are used to reduce the energy consumption of the cleaning robot.

[0006] In a first aspect, the present application provides a path planning method in a complex scenario, which is applied to a cleaning robot, and the method comprises: Get the current cleaning task information sent by the client, which includes the current cleaning area, task type and task accuracy requirements; Identify the terrain type in the current cleaning area based on an image recognition algorithm, and divide the current cleaning area according to the terrain type to obtain a plurality of current sub-areas; Obtain the terrain types of all current sub-areas, and calculate the complexity levels of all current sub-areas according to the terrain types; Calculate the total energy requirement of the current cleaning task information according to the area range of the current cleaning area, the terrain type of the current cleaning area, the task type and the task accuracy requirement; Allocate energy to the current sub-area based on the current power of the cleaning robot and the complexity level to obtain a current sub-cleaning task for each current sub-area, wherein the sum of the sub-energy requirements of all the current sub-cleaning tasks is equal to the total energy requirement; The task priority of each current sub-cleaning task is determined according to the sub-energy demand, and the task execution parameter of each current sub-cleaning task is determined according to the position information of the current sub-region.

[0007] In the above embodiment, by obtaining the current cleaning task information and identifying the terrain type in the cleaning area based on the image recognition algorithm, the cleaning area can be accurately divided and the regional complexity level can be calculated, thereby achieving accurate task allocation, calculating the total energy demand according to the task type and task accuracy requirements, and then allocating energy according to the current power of the robot and the regional complexity level, ensuring that the cleaning task can be completed with the most optimized energy consumption. In addition, by determining the task priority and task execution parameters of each sub-cleaning task, the rationality and efficiency of path planning are guaranteed, so that the cleaning robot can reasonably plan the path in complex scenes, reduce power consumption, and improve cleaning efficiency.

[0008] In combination with some embodiments of the first aspect, in some embodiments, the step of identifying the terrain type in the current cleaning area based on the image recognition algorithm and dividing the current cleaning area into a plurality of current sub-areas according to the terrain type specifically includes: Acquiring image information in the current cleaning area based on the camera device of the cleaning robot; Comparing the image information with preset images in a preset image database to determine all terrain types in the current cleaning area; According to the position information of all the terrain categories, a segmentation algorithm is used to divide the current cleaning area into a number of current sub-areas, so that the proportion of the main terrain type in all the terrain types in each current sub-area is greater than a preset proportion threshold.

[0009] In the above embodiment, the image information is obtained by the camera of the cleaning robot and compared with the image in the preset image database to accurately determine all the terrain types in the cleaning area. The segmentation algorithm is used to divide the cleaning area into several sub-areas, so that the proportion of the main terrain type in each sub-area is greater than the preset threshold, thereby achieving accurate area division, which not only improves the accuracy and efficiency of terrain recognition, but also makes the subsequent energy allocation and task priority determination more scientific and reasonable, further improving the path planning and task execution effect of the cleaning robot.

[0010] In combination with some embodiments of the first aspect, in some embodiments, obtaining the terrain type of all the current sub-areas and calculating the complexity level of all the current sub-areas according to the terrain type specifically includes: When the number of terrain types in the first current sub-region is equal to one, determining the complexity level of the first current sub-region according to a preset terrain type-complexity level mapping table; When there is more than one terrain type in the second current sub-region, the proportion of different terrain types in the second current sub-region is calculated and the complexity levels corresponding to the terrain types are averaged according to the proportion to obtain the complexity level of the second current sub-region.

[0011] In the above embodiment, by obtaining the terrain type of each sub-area, its complexity level is determined according to the number and proportion of terrain types. When there is only one terrain type in the sub-area, the complexity level is directly determined according to the preset terrain type-complexity level mapping table. When there is more than one terrain type, the complexity level of the sub-area is obtained by calculating the proportion of different terrain types and averaging the complexity level according to the proportion. This method can accurately reflect the complexity of each sub-area, provide a scientific basis for subsequent energy allocation and task priority determination, help improve the working efficiency and cleaning effect of the cleaning robot in complex scenes, and ensure the execution stability and reliability of cleaning tasks in complex terrains.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the step of calculating the total energy requirement of the current cleaning task information according to the area range of the current cleaning area, the terrain type of the current cleaning area, the task type and the task accuracy requirement, the method further includes: Inputting the area range and the terrain type into an energy estimation model to obtain a first estimated energy of the current cleaning area; An energy calculation coefficient is calculated according to the task type and the task accuracy requirement, and the energy calculation coefficient is proportional to the task accuracy requirement; The product of the first estimated energy and the energy calculation coefficient is used as the total energy requirement.

[0013] In the above embodiment, the first estimated energy of the area is obtained by inputting the range and terrain type of the cleaning area into the energy estimation model, and the energy calculation coefficient is calculated according to the task type and task accuracy requirements. The energy calculation coefficient is proportional to the task accuracy. By taking the product of the first estimated energy and the energy calculation coefficient as the total energy demand, the accurate calculation of the total energy demand is achieved, which not only can flexibly adjust the energy demand according to different cleaning task types and accuracy requirements, but also improves the energy utilization rate of the cleaning robot, ensuring that the robot can complete efficient and accurate cleaning work in complex scenes.

[0014] In combination with some embodiments of the first aspect, in some embodiments, the step of determining the task priority of each current sub-cleaning task according to the sub-energy demand, and determining the task execution parameter of each current sub-cleaning task according to the position information of the current sub-area specifically includes: Calculate the geometric center point of all the position information; Based on the path search algorithm, an optimal path search is performed for all the geometric center points to obtain an optimal cleaning route, so that the power consumed by the cleaning robot by the optimal cleaning route is lower than a preset power threshold; Determine the task priority of each current sub-cleaning task according to the optimal cleaning route; Obtain the task execution parameters corresponding to the sub-energy demand in the preset cleaning intensity table.

[0015] In the above embodiment, the geometric center point of all position information is calculated, and the optimal path search is performed on the geometric center point based on the path search algorithm to obtain the optimal cleaning route that can make the robot's power consumption lower than the preset power threshold. Then, the task priority is determined according to the optimal cleaning route, and the task execution parameters corresponding to the sub-energy requirements are obtained in the preset cleaning intensity table. This can ensure that the cleaning robot can select the optimal path when performing the task, reduce unnecessary power consumption, and improve cleaning efficiency and the accuracy of task completion.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the task priority and task execution parameters of each current sub-cleaning task according to the sub-energy demand and the location information of the current sub-area, the method further includes: Get the real-time power of the cleaning robot; Determine whether the real-time power can cover all energy requirements of all remaining unfinished sub-cleaning tasks of the current task; If not, update the task execution parameters so that the energy consumption of the current sub-cleaning task is reduced.

[0017] In the above embodiment, the flexibility and adaptability of the cleaning robot are further improved by dynamically adjusting the task execution parameters during the task execution process. After determining the task priority and task execution parameters of each current sub-cleaning task, the power of the cleaning robot is obtained in real time, and it is determined whether it can cover all energy requirements of the remaining unfinished tasks. If the real-time power is not enough to complete all tasks, the task execution parameters are updated to reduce the energy consumption of the current sub-cleaning task to ensure that the robot can complete as many cleaning tasks as possible. This not only improves the energy management capability of the cleaning robot, but also enhances its task execution stability and reliability in complex scenarios. Through real-time power monitoring and parameter adjustment, task interruptions due to insufficient power can be effectively avoided, thereby ensuring the continuity and efficiency of cleaning tasks.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining whether the real-time power can cover all energy requirements of all remaining unfinished current sub-cleaning tasks, the method further includes: If possible, the real-time completion progress of the current sub-cleaning task is sent to the client.

[0019] In the above embodiment, by real-time monitoring of the power of the cleaning robot and judging whether it can cover all the energy requirements of the remaining unfinished tasks, the task management and execution efficiency of the cleaning robot is further enhanced. The specific steps include sending the real-time completion progress of the current sub-cleaning task to the client after determining that the real-time power covers all the remaining tasks, which not only improves the transparency of the cleaning task, but also provides users with real-time task progress information, which is convenient for users to monitor and manage.

[0020] In a second aspect, an embodiment of the present application provides a cleaning robot, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the cleaning robot to execute the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, an embodiment of the present application provides a computer program product comprising instructions, which, when executed on a cleaning robot, enables the cleaning robot to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, comprising instructions, which, when executed on a cleaning robot, enable the cleaning robot to execute the method described in the first aspect and any possible implementation of the first aspect.

[0023] It is understandable that the cleaning robot provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the method provided in the embodiment of the present application. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method, which will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. This application can accurately divide the cleaning area and calculate the regional complexity level by obtaining the current cleaning task information and identifying the terrain type in the cleaning area based on the image recognition algorithm, so as to achieve accurate task allocation, calculate the total energy demand according to the task type and task accuracy requirements, and then allocate energy according to the current power of the robot and the regional complexity level to ensure that the cleaning task can be completed with the most optimized energy consumption. In addition, by determining the task priority and task execution parameters of each sub-cleaning task, the rationality and efficiency of path planning are guaranteed, so that the cleaning robot can reasonably plan the path in complex scenes, reduce power consumption, and improve cleaning efficiency.

[0025] 2. This application obtains the terrain type of each sub-area and determines its complexity level according to the number and proportion of terrain types. When there is only one type of terrain in the sub-area, the complexity level is determined directly according to the preset terrain type-complexity level mapping table. When there is more than one type of terrain, the complexity level of the sub-area is obtained by calculating the proportion of different terrain types and averaging the complexity level according to the proportion. This method can accurately reflect the complexity of each sub-area, provide a scientific basis for subsequent energy allocation and task priority determination, help improve the work efficiency and cleaning effect of the cleaning robot in complex scenes, and ensure the execution stability and reliability of cleaning tasks in complex terrains.

[0026] 3. This application calculates the geometric center point of all position information, and searches for the optimal path for the geometric center point based on the path search algorithm to obtain the optimal cleaning route that can make the robot's power consumption lower than the preset power threshold, and then determines the task priority based on the optimal cleaning route, and obtains the task execution parameters corresponding to the sub-energy requirements in the preset cleaning intensity table, which can ensure that the cleaning robot can choose the optimal path when performing tasks, reduce unnecessary power consumption, and improve cleaning efficiency and task completion accuracy. At the same time, through reasonable task priority setting and determination of execution parameters, the orderly progress of cleaning tasks and the maximization of the overall cleaning effect are guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of a path planning method in a complex scenario in an embodiment of the present application; Figure 2 is another flowchart of the path planning method in a complex scenario in an embodiment of the present application; Figure 3 It is a schematic diagram of the physical device structure of the cleaning robot in the embodiment of the present application. DETAILED DESCRIPTION

[0028] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification of the present application, the singular expressions "one", "a kind of", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more of the listed items.

[0029] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "multiple" means two or more.

[0030] The following is a description of the process of the method provided by this implementation. Figure 1 , which is a flow chart of a path planning method in a complex scenario in an embodiment of the present application.

[0031] S101. Obtain current cleaning task information sent by the client, where the task information includes the current cleaning area, task type, and task accuracy requirements.

[0032] The client is a device that users use to send instructions to the cleaning robot through mobile terminals such as mobile phones or tablets. The current cleaning task information refers to the relevant parameters of this cleaning, including the area to be cleaned and the type of this cleaning task. For example, the current cleaning area may be a camping site, a tourist gathering area, etc., and the task types include cleaning garbage and removing dust.

[0033] Specifically, users can select a campsite in the mobile app for a garbage cleaning task and set medium precision requirements. The client will then send the current cleaning task information, including the campsite range information, garbage cleaning task type, and medium cleaning quality standard requirements, to the cleaning robot in the mountain environment. Different task precision requirements correspond to different task execution parameters set by the subsequent robot.

[0034] For example, if Camp No. 7 is selected for garbage cleaning and the cleaning quality is required to reach a medium standard, the client will send information including the corresponding area, task type, and quality requirements.

[0035] S102: Identify the terrain type in the current cleaning area based on an image recognition algorithm, and divide the current cleaning area into a plurality of current sub-areas according to the terrain type.

[0036] Among them, the image recognition algorithm is an algorithm that can automatically analyze images and identify targets; the terrain type represents the surface state within the current cleaning area, such as rocks, grass, etc.; and the area division is to divide a complete area into multiple small areas.

[0037] Specifically, the cleaning robot will first use its own camera equipment to collect image information in the area that needs to be cleaned. Then it will use the image recognition algorithm to analyze the content in the image and determine what types of terrain exist, such as grass, water, etc. After clarifying the terrain composition of the current area, the cleaning robot will use the corresponding area segmentation algorithm to divide the current complete cleaning area into multiple small areas with representative terrain according to the terrain distribution. The division can help the cleaning robot better plan and locate the cleaning path.

[0038] In some embodiments, the recognition and regional division of terrain can be achieved in the following two ways: Optionally, using a deep learning-based image classification algorithm to identify the terrain type and perform regional division according to the type distribution boundary; Optionally, matching the extracted image features with the preset terrain sample features, judging the terrain type according to the matching results, and performing regional division using a clustering algorithm. It is understandable that other image processing methods can also be used to achieve terrain recognition and regional division, which are not limited here.

[0039] S103: Obtain the terrain types of all the current sub-areas, and calculate the complexity levels of all the current sub-areas according to the terrain types.

[0040] Among them, the current sub-region represents each small area in the clean area divided in the previous step; the terrain type represents the surface state in each sub-region; and the complexity level is used to represent the complexity of the terrain of the sub-region.

[0041] Specifically, after dividing the complete cleaning area, the cleaning robot needs to clarify what types of terrain exist in each current sub-area, such as rocks, trees, etc. After figuring out the terrain composition of each sub-area, a comprehensive complexity level of each sub-area is calculated based on the complexity of different terrains. This indicates the complexity and difficulty of the terrain composition of the area for cleaning work.

[0042] In some embodiments, the sub-region complexity calculation can be implemented in the following two ways: optionally, query the terrain image in the sub-region, use the semantic segmentation algorithm to identify various types of terrain, and calculate the complexity level of the sub-region according to a predetermined rule; optionally, retrieve the terrain data of multiple points in the sub-region, count the proportions of different terrains, and calculate the weighted complexity level of the region according to the proportion. It is understandable that other methods can also be used to calculate the complexity according to the terrain type, which is not limited here.

[0043] S104: Calculate the total energy requirement of the current cleaning task information according to the area range of the current cleaning area, the terrain type of the current cleaning area, the task type, and the task accuracy requirement.

[0044] Among them, the area range represents the area size of the entire current cleaning area; the terrain type refers to the main terrain composition of the entire area; the task type represents the specific category of cleaning work that needs to be performed; the task accuracy requirement represents the quality requirement for the final cleaning effect; the total energy demand is used to represent the estimated energy value required to complete the entire current cleaning task.

[0045] Specifically, the cleaning robot will consider the size of the entire area to be processed, the main terrain type of the area, the specific type of cleaning operation to be performed, and the user's requirements for the quality of the cleaning results. All of these factors will affect the total amount of energy required to complete the task. The cleaning robot will input this key information into the preset energy demand calculation model, which will calculate the total energy demand value to meet the current task according to the established algorithm.

[0046] In some embodiments, the total energy demand can be obtained by: optionally, using a multivariate regression model, training an energy prediction model based on data such as terrain type and area size, and calculating the total demand; optionally, querying a preset cleaning operation database, obtaining historical energy consumption data for similar tasks, and calculating the total demand. It is understandable that other calculation methods can also be used to obtain an estimated value of the overall energy demand, which is not limited here.

[0047] S105. Allocate energy to the current sub-area based on the current power of the cleaning robot and the complexity level to obtain a current sub-cleaning task for each current sub-area, and the sum of the sub-energy requirements of all the current sub-cleaning tasks is equal to the total energy requirement.

[0048] The power level refers to the remaining available power of the cleaning robot. The complexity level is the complexity calculated in the previous step based on the terrain conditions of each sub-area. The energy allocation to the sub-areas is the estimation and allocation of the energy required for cleaning each sub-area based on the power status and regional complexity. The current sub-cleaning task is the specific cleaning task divided for each sub-area. The sub-energy demand is the energy value required for each sub-cleaning task. The total energy demand is the total energy required for the overall cleaning task of the entire area.

[0049] Specifically, the cleaning robot first detects its current remaining power, for example, 40%. Then, according to the complexity level of each sub-area, for example, area A is level 1 and area B is level 3. The cleaning robot will allocate energy requirements to areas of different complexity according to the power status and a certain proportion, setting the energy requirements of high-complexity areas higher and low-complexity areas lower, and at the same time making the sum of the energy requirements of each area equal to the estimated total energy requirement, so that the current sub-cleaning task of each sub-area and the sub-energy requirements corresponding to each task can be obtained. Finally, the cleaning robot checks whether the sum of the energy requirements of all sub-tasks is equal to the total energy requirement. If they are not equal, further adjustments are required until the two are equal.

[0050] S106 , determining a task priority of each current sub-cleaning task according to the sub-energy demand, and determining a task execution parameter of each current sub-cleaning task according to the position information of the current sub-region.

[0051] Among them, the sub-energy demand represents the energy value required for each subtask obtained previously; the task priority is used to represent the execution order of multiple subtasks; the location information includes the coordinates of each sub-area, etc.; the task execution parameters represent the relevant settings for completing the cleaning.

[0052] Specifically, the cleaning robot will determine the task priority according to the energy demand of each sub-area cleaning task, and the task that requires more energy will have a higher priority. At the same time, the specific location coordinates and other information of all sub-areas will be retrieved to determine the best cleaning route, cleaning mode and other execution parameters for each task. In this way, the order of tasks can be reasonably arranged according to energy consumption, and each subtask can be completed with the optimal parameters to achieve energy saving and maximum efficiency. In some embodiments, the task execution plan can be planned in the following ways: optionally, a corresponding table of energy demand and priority, a matching table of task location and execution parameters are constructed, and query planning is performed; optionally, a rule-based planning program is designed, demand and location data are input, and the program outputs an execution plan. It can be understood that task priority and execution parameters can also be determined in other ways, which are not limited here.

[0053] The following is a more detailed description of the process of the method provided by this implementation. Figure 2, which is another flow chart of the path planning method in a complex scenario in an embodiment of the present application.

[0054] S201. Obtain current cleaning task information sent by the client, where the task information includes the current cleaning area, task type, and task accuracy requirements.

[0055] It can be understood that this step is similar to step S101 and will not be described in detail here.

[0056] S202: Acquire image information in the current cleaning area based on the camera device of the cleaning robot.

[0057] The camera device refers to a camera installed on the cleaning robot, which can capture environmental images. The image information in the current cleaning area refers to the image data of the interior of the area to be cleaned captured by the cleaning robot.

[0058] Specifically, after receiving the task of cleaning a campsite in the previous step, the cleaning robot will drive into the campsite and start the 360-degree camera to take a panoramic photo of the area, obtaining image information from various angles in order to fully understand the specific environmental conditions of the area and provide a basis for the subsequent formulation of cleaning strategies.

[0059] S203: Compare the image information with preset images in a preset image database based on an image recognition algorithm to determine all terrain types in the current cleaning area.

[0060] The cleaning robot will pre-build an image database, which contains preset images of various standard terrain styles, such as grass, mud, stone, water, etc. After collecting the image information of the current area, the cleaning robot will use the image recognition algorithm to compare and analyze the acquired image information with the preset image database, and determine the terrain types in the current area through similarity matching, such as whether there is grass, stone, water, etc. in the area. Ultimately, all terrain types in the current cleaning area can be determined.

[0061] Specifically, by comparing with the standard image, the cleaning robot determines that the collected camping site image contains three types of terrain: grass, mud and rock. Therefore, it can be determined that the terrain types in the camping site area include grass, mud and rock. This provides a basis for the subsequent formulation of cleaning strategies for different terrains.

[0062] S204: Divide the current cleaning area into a number of current sub-areas using a segmentation algorithm according to the location information of all the terrain types, so that the proportion of the main terrain type in all the terrain types in each current sub-area is greater than a preset proportion threshold.

[0063] The location information of the terrain category refers to the specific distribution locations of different terrain types in the current cleaning area obtained through the previous step. The segmentation algorithm is an algorithm that can automatically divide an area into multiple sub-areas. The proportion threshold refers to a set threshold used to determine whether the proportion of a certain terrain in the sub-area is greater than the threshold.

[0064] Specifically, after determining that there are multiple terrain categories in the current cleaning area, the cleaning robot will count the specific location distribution of each terrain category, and then the cleaning robot will enable the segmentation algorithm to automatically divide the entire current cleaning area into multiple small current sub-areas according to the location information of various terrains. During the segmentation process, the cleaning robot will require that in each current sub-area generated by the segmentation, there must be a terrain type with a proportion greater than a preset proportion threshold, for example, more than 70%, that is, each sub-area is required to have a main terrain type, so that each sub-area has a representative main terrain.

[0065] For example, an area with three types of terrain, namely grass, mud, and water, can be divided into sub-area 1 dominated by grass, sub-area 2 dominated by mud, and sub-area 3 dominated by water. The proportion of the main terrain type in each sub-area exceeds 70%.

[0066] S205: When the number of terrain types in the first current sub-region is equal to one, determine the complexity level of the first current sub-region according to a preset terrain type-complexity level mapping table.

[0067] The preset terrain type-complexity level mapping table refers to a pre-established table that clearly specifies the complexity levels corresponding to different terrain types. When there is only one type of terrain in a sub-area, the complexity level corresponding to the terrain can be directly found out based on this mapping table, which is the complexity level of the sub-area.

[0068] Specifically, the cleaning robot will record the complexity of different terrains in advance. For example, grass is level 1, mud is level 2, and rock is level 3. If the current sub-area obtained by segmentation contains only grass, then the complexity level corresponding to grass can be found in the mapping table as level 1, so the complexity level of the current sub-area is determined to be level 1. This correspondence can help quickly determine the complexity of a single terrain sub-area.

[0069] S206. When there is more than one terrain type in the second current sub-region, calculate the proportion of different terrain types in the second current sub-region and average the complexity levels corresponding to the terrain types according to the proportion to obtain the complexity level of the second current sub-region.

[0070] When there are multiple terrains in a sub-region, its complexity needs to take into account the influence of various terrains. At this time, we can first calculate the area ratio of various terrains in the sub-region, and then perform weighted average according to the complexity level corresponding to each terrain itself, and then we can get the overall complexity level of this sub-region that takes into account the influence of various terrains.

[0071] Specifically, if a sub-area has two types of terrain, grass and rock, with grass accounting for 80% and rock accounting for 20%, and the grass complexity is level 1 and the rock complexity is level 3, then you can multiply the 80% proportion of the grass by its complexity level 1, and add the 20% proportion of the rock multiplied by the complexity level 3, and then take the average, to get the complexity level of the sub-area to be approximately 1.4.

[0072] S207: Input the area range and the terrain type into an energy estimation model to obtain a first estimated energy of the current cleaning area.

[0073] Specifically, the cleaning robot first uses its own ranging device to scan and calculate the detailed area size of the target area that needs to be cleaned, and determines it to be 1,000 square meters. Then, the cleaning robot collects image information and analyzes and judges the area in the early stage, and clarifies that the area contains two main types of terrain, namely rocky land and grassland. Next, the cleaning robot inputs the measured precise area data of 1,000 square meters, as well as the two terrain types of rocky land and grassland obtained by analysis, into the pre-established energy demand prediction model. This prediction model comprehensively considers the impact of area size and terrain factors on energy, and after multi-dimensional calculation and analysis, outputs a rough energy prediction value, that is, the energy required to complete all cleaning tasks in the current area is predicted to be 50,000 joules. This is the first estimated energy result for the area. After obtaining this data, the cleaning robot can preliminarily determine the approximate energy level required for cleaning the area.

[0074] The process of building the energy estimation model is as follows: First, collect a large amount of historical data related to cleaning tasks. The historical data includes area range: the area size of different cleaning areas, actual energy consumption: the actual energy consumed to complete each task, feature data includes area range and terrain type after one-hot encoding, label data is actual energy consumption, input data is area range and terrain type, and output data is the first estimated energy.

[0075] First, data preprocessing is performed by cleaning, feature engineering and normalizing the data. Then, the random forest regression model is selected for energy estimation. The data set is divided into a training set and a test set in a ratio of 8:2. The training set data is used to train the random forest regression model. The performance of the model is evaluated using indicators such as mean square error (MSE). Finally, the model is optimized by adjusting hyperparameters (such as the number of trees, the depth of the trees, etc.), and the best parameter combination is found using methods such as grid search or random search.

[0076] S208. Calculate an energy calculation coefficient according to the task type and the task accuracy requirement, where the energy calculation coefficient is proportional to the task accuracy requirement.

[0077] Specifically, on the basis of the first estimated energy, the cleaning robot also needs to consider the specific type of this cleaning task and the accuracy requirements to be achieved, because different task types and accuracy requirements will affect the final energy demand. Confirming that this task is sweeping the floor, and the user has set high-precision cleaning quality requirements, the cleaning robot inputs these two pieces of information into the energy calculation model. The model will calculate a reasonable energy calculation coefficient based on the high-precision sweeping task to adjust the final energy demand. Among them, the high-precision requirement is proportional to the required energy. The higher the accuracy required, the greater the corresponding energy demand and the higher the calculation coefficient. After comprehensive analysis, the model finally generated an energy calculation coefficient of 1.5 to adapt to high-precision sweeping tasks.

[0078] S209: taking the product of the first estimated energy and the energy calculation coefficient as the total energy demand.

[0079] Specifically, after obtaining the first estimated energy of 50,000 joules and the corresponding calculated energy calculation coefficient of 1.5, the cleaning robot inputs the two into the calculation program for multiplication, and the value obtained is 50,000*1.5=75,000 joules. This means that in order to complete the high-precision sweeping task in the current area, a total energy demand of 75,000 joules is expected. Through this precise calculation, the cleaning robot can clearly know the energy consumed throughout the entire cleaning task, so as to prepare in advance and ensure that subsequent cleaning work can be completed smoothly and continuously, avoiding interruption of the task due to insufficient energy.

[0080] S210. Allocate energy to the current sub-area based on the current power of the cleaning robot and the complexity level to obtain a current sub-cleaning task for each current sub-area, and the sum of the sub-energy requirements of all the current sub-cleaning tasks is equal to the total energy requirement.

[0081] It is understandable that this step is similar to step S105 and will not be described in detail here. S211, calculating the geometric center point of all the position information.

[0082] The location information refers to the specific coordinate position of each sub-region. The geometric center point refers to the coordinate of the center point for calculating the sub-region location information. By calculating the geometric center points of all sub-regions, the central representative point of each sub-region position can be obtained.

[0083] Specifically, the cleaning robot will retrieve the vertex coordinates of each sub-area to determine the contour shape and range of the area. Then, it will extract all vertex coordinate information and use the geometric algorithm in trigonometry to calculate the coordinates of the center point of this polygonal area. For example, if the vertex coordinates of a sub-area are (1, 1), (3, 1), (2, 3), then the coordinates of the geometric center point of this triangular area are (2, 1.67). The cleaning robot will calculate the geometric center points of all sub-areas one by one.

[0084] S212: performing an optimal path search for all the geometric center points based on a path search algorithm to obtain an optimal cleaning route, so that the power consumed by the cleaning robot by the optimal cleaning route is lower than a preset power threshold.

[0085] The path search algorithm is an algorithm that can search for the optimal path. The geometric center point has been calculated in the previous step. The optimal cleaning route is the best cleaning sequence determined after comprehensively considering time and energy factors. The power threshold is the preset lower limit requirement.

[0086] In this step, the cleaning robot needs to search for the optimal path for the geometric center points of all sub-areas based on the path search algorithm to obtain an optimal cleaning route. The geometric center point here refers to the center point coordinates of the position of each sub-area calculated in the previous step. The path search algorithm can adopt algorithms such as A* algorithm and Dijkstra algorithm. The goal is to search for the shortest path that can cover all geometric center points, that is, the optimal cleaning route, while meeting the preset power threshold requirements. This can minimize the energy consumption and repeated paths of the cleaning robot.

[0087] In some embodiments, specifically, the cleaning robot can use the A algorithm to search for the optimal path for the geometric center point of the sub-area. Assuming there are 5 sub-areas, the corresponding center point coordinates are (1, 2), (4, 6), (5, 3), (7, 4), and (9, 1), respectively. Then, these 5 center points can be used as the search starting point and end point of the A algorithm, and the open list and the closed list are initialized. The starting point (1, 2) is added to the open list, and the center point (4, 6) with the smallest f(n) value is selected from the open list and moved to the closed list. The g(n) and f(n) of the adjacent center point (5, 3) are added to the open list, and steps 2-3 are repeated until the end point (9, 1) is added to the open list. The optimal path search is completed, and the optimal cleaning route that passes through all center points is obtained by backtracking the path.

[0088] S213: Determine the task priority of each current cleaning sub-task according to the optimal cleaning route.

[0089] Specifically, the cleaning robot will search the order of the optimal path and set the priority of the corresponding subtasks according to the order of appearance of each sub-area on the path. For example, if the optimal path order is area A->area B->area C->area D, the cleaning robot will set the cleaning task of area A to have the highest priority, followed by area B, then area C, and finally area D. Because of the power limit, the area near the starting point of the path must be cleaned first. This ensures that the area with less power consumption is cleaned first, and then the area with more power consumption is cleaned, maximizing the cleaning area completed under limited power.

[0090] S214. Obtain task execution parameters corresponding to the sub-energy demand from a preset cleaning intensity table.

[0091] The preset cleaning intensity table refers to a lookup table that stores the corresponding task execution parameters under different cleaning intensities. The sub-energy demand refers to the cleaning energy required for each sub-area calculated based on the regional complexity. The task execution parameters refer to some adjustable execution variables required to complete the cleaning. Obtaining the corresponding parameters means obtaining a matching execution parameter combination in the lookup table according to the energy demand of the sub-area. The task execution parameters include the following parameters: cleaning speed refers to the moving speed of the cleaning robot. Different speed levels can be set, such as slow, medium, and high speed. The faster the moving speed, the larger the cleaning area that can be covered per unit time, but the energy consumption is also higher; the number of cleaning times refers to the number of rounds required to complete a cleaning task. It can be set to multiple cleanings to increase the cleaning intensity. The more cleaning times, the better the cleaning effect, but the higher the energy consumption; cleaning mode refers to different cleaning methods, such as standard mode, strong mode, and light sweeping mode. The cleaning intensity and energy consumption are different in different modes; cleaning route refers to the planning method of the cleaning path, such as one-way cleaning, S-shaped cleaning, random cleaning and other route modes. The coverage efficiency and energy efficiency of different route modes are different; cleaning time refers to the time length setting to complete a cleaning. The longer the time, the more complete the cleaning area can be covered.

[0092] Specifically, the cleaning robot has a preset cleaning intensity parameter table, which lists the execution parameters such as speed, number of cleanings, cleaning mode, etc. at different intensities. For example, intensity level 1 corresponds to slow speed, 1 cleaning, and standard mode, and intensity level 2 corresponds to medium speed, 2 cleanings, and strong mode. The cleaning robot will find the energy demand value corresponding to each sub-area, and find a set of execution parameters corresponding to the matching level in the table. For example, if the energy demand of an area belongs to level 2, the table will be looked up to obtain the combination parameters of medium speed, 2 times, and strong mode as the optimal execution parameters for the area. In this way, reasonable cleaning parameters are matched for each sub-area, so that the cleaning intensity and energy consumption are optimally matched.

[0093] After step S214, the following steps are also included: Get the real-time power level of the cleaning robot.

[0094] This step is the real-time power refers to the current remaining power value of the cleaning robot during the cleaning task. The real-time power is obtained to monitor the power usage so as to adjust the cleaning strategy in time.

[0095] Specifically, during the cleaning process, the cleaning robot will continuously monitor the remaining power of the battery pack through its own power sensor. It will also obtain its current remaining power value, that is, the real-time power, at regular intervals. For example, it will detect and record the current power every 5 minutes. This allows the cleaning robot to obtain its own remaining power status in a timely manner.

[0096] Determine whether the real-time power can cover all energy requirements of all remaining unfinished sub-cleaning tasks of the current cleaning task.

[0097] Judging coverage means comparing the real-time power and the power required for the unfinished subtasks to see whether the remaining power is sufficient to complete all the remaining subtasks.

[0098] Specifically, the cleaning robot will compare the real-time power obtained in this step with the estimated total energy demand of the unfinished subtasks in the previous step. If the real-time power value is greater than or equal to the sum of the energy demands of all remaining subtasks, it is determined that there is enough power to cover; if the real-time power is less than the sum of the remaining demands, it is determined that the power cannot cover.

[0099] For example, if there are 30,000 joules of power left, but the uncompleted subtasks require a total of 40,000 joules, the cleaning robot will determine that the remaining power cannot cover the remaining task requirements.

[0100] If not, update the task execution parameters so that the energy consumption of the current sub-cleaning task is reduced.

[0101] Specifically, if the cleaning robot determines that the real-time power consumption cannot cover the needs of all remaining subtasks, it will start the energy consumption optimization program, reset the execution parameters of the current sub-cleaning task, lower the cleaning speed, reduce the number of cleaning times, and other measures to reduce the energy consumption of the current task. When the new parameters are executed, the energy required will be lower than the previously estimated sub-task energy consumption. This allows limited power to cover more tasks.

[0102] If possible, the real-time completion progress of the current sub-cleaning task is sent to the client.

[0103] If the battery is sufficient, the user will be provided with real-time feedback on the execution progress of the current sub-cleaning task.

[0104] Specifically, if the cleaning robot determines that the battery is sufficient, it will calculate the specific completion progress of the current task after completing the current sub-cleaning task, for example, 80% completion. This real-time progress data will be sent to the remote control client through network transmission. After receiving the data, the client can display it to the user, allowing the user to understand the progress of the task at any time. This real-time progress feedback can improve task transparency and allow users to understand the cleaning status at any time.

[0105] The cleaning robot in the embodiment of the present invention is described below from the perspective of hardware processing. Figure 3 , is a schematic diagram of a physical device structure of a cleaning robot in an embodiment of the present application.

[0106] It should be noted that Figure 3The structure of the cleaning robot shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0107] like Figure 3 As shown, the cleaning robot includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In RAM 303, various programs and data required for system operation are also stored. CPU 301, ROM 302 and RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0108] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0109] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0110] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0111] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0112] Specifically, the cleaning robot of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the path planning method for complex scenarios provided in the above embodiment is implemented.

[0113] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the cleaning robot described in the above embodiment; or may exist independently without being assembled into the cleaning robot. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the cleaning robot, the cleaning robot implements the path planning method in complex scenarios provided in the above embodiment.

[0114] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0115] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0116] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A path planning method in complex scenarios, It is characterized in that Applied to a cleaning robot, the method comprises: Obtain the current cleaning task information sent by the client, the task information including the current cleaning area, task type and task accuracy requirements; Identifying the type of terrain in the current cleaning area based on an image recognition algorithm, and dividing the current cleaning area according to the type of terrain to obtain a plurality of current sub-areas; Acquire the terrain types of all the current sub-areas, and calculate the complexity levels of all the current sub-areas according to the terrain types; Calculating the total energy requirement of the current cleaning task information according to the area range of the current cleaning area, the terrain type of the current cleaning area, the task type and the task accuracy requirement; Allocating energy to the current sub-area based on the current power of the cleaning robot and the complexity level to obtain a current sub-cleaning task for each of the current sub-areas, wherein the sum of the sub-energy requirements of all the current sub-cleaning tasks is equal to the total energy requirement; The task priority of each of the current sub-cleaning tasks is determined according to the sub-energy demand, and the task execution parameter of each of the current sub-cleaning tasks is determined according to the position information of the current sub-region.

2. The method according to claim 1, It is characterized in that The step of identifying the terrain type in the current cleaning area based on an image recognition algorithm and dividing the current cleaning area according to the terrain type to obtain a plurality of current sub-areas specifically includes: Acquiring image information within the current cleaning area based on a camera device of the cleaning robot; Comparing the image information with preset images in a preset image database based on an image recognition algorithm to determine all terrain types in the current cleaning area; According to the position information of all the terrain categories, a segmentation algorithm is used to divide the current cleaning area into a number of current sub-areas, so that the proportion of the main terrain type in all the terrain types in each current sub-area is greater than a preset proportion threshold.

3. The method according to claim 1, It is characterized in that The obtaining of the terrain types of all the current sub-areas and calculating the complexity levels of all the current sub-areas according to the terrain types specifically includes: When the number of terrain types in the first current sub-region is equal to one, determining the complexity level of the first current sub-region according to a preset terrain type-complexity level mapping table; When there are more than one terrain types in the second current sub-region, the proportions of the different terrain types in the second current sub-region are calculated and the complexity levels corresponding to the terrain types are averaged according to the proportions to obtain the complexity level of the second current sub-region.

4. The method according to claim 1, It is characterized in that The step of calculating the total energy requirement of the current cleaning task information according to the area range of the current cleaning area, the terrain type of the current cleaning area, the task type and the task accuracy requirement, the method further includes: Inputting the area range and the terrain type into an energy estimation model to obtain a first estimated energy of the current cleaning area; Calculating an energy calculation coefficient according to the task type and the task accuracy requirement, wherein the energy calculation coefficient is proportional to the task accuracy requirement; The product of the first estimated energy and the energy calculation coefficient is used as the total energy requirement.

5. The method according to claim 1, It is characterized in that The step of determining the task priority of each of the current sub-cleaning tasks according to the sub-energy demand, and determining the task execution parameters of each of the current sub-cleaning tasks according to the position information of the current sub-area, specifically includes: Calculate the geometric center point of all the position information; Performing an optimal path search for all the geometric center points based on a path search algorithm to obtain an optimal cleaning route, so that the power consumed by the cleaning robot by the optimal cleaning route is lower than a preset power threshold; Determine the task priority of each of the current sub-cleaning tasks according to the optimal cleaning route; The task execution parameters corresponding to the sub-energy demand are obtained from the preset cleaning intensity table.

6. The method according to claim 1, It is characterized in that After the step of determining the task priority and task execution parameters of each current sub-cleaning task according to the sub-energy demand and the location information of the current sub-area, the method further includes: Obtaining the real-time power of the cleaning robot; Determine whether the real-time power can cover all energy requirements of all the remaining unfinished current sub-cleaning tasks; If not, the task execution parameters are updated so that the energy consumption of the current sub-cleaning task is reduced.

7. The method according to claim 6, It is characterized in that After the step of determining whether the real-time power can cover all energy requirements of all the remaining unfinished current sub-cleaning tasks, the method further includes: If possible, the real-time completion progress of the current sub-cleaning task is sent to the client.

8. A cleaning robot, It is characterized in that The cleaning robot comprises: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code comprises computer instructions, and the one or more processors call the computer instructions so that the cleaning robot executes the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, It is characterized in that When the instructions are executed on the cleaning robot, the cleaning robot is caused to perform the method according to any one of claims 1 to 7.

10. A computer program product, It is characterized in that When the computer program product runs on a cleaning robot, the cleaning robot is enabled to perform the method according to any one of claims 1 to 7.

Images