Robot escape control method and device and cleaning robot

Abstract

The invention relates to a robot escape control method and device and a cleaning robot. The method comprises the steps that if the robot is in a motion limited state, a determination step is executed, specifically, according to the current position of the robot, a to-be-executed state change action of the robot is determined, and the state change action is used for controlling the robot to break away from the motion limited state; a detection execution step: detecting the position of a motion forbidden zone near the robot when the state change action is executed; if the position of the motion forbidden zone conflicts with the state change action, updating the state change action, and returning to execute the detection step; and if the movement forbidden zone position does not conflict with the state change action, the robot is controlled to execute the state change action. By adopting the method, the escape efficiency of the robot can be improved, the escape success rate of the robot is also improved, and the reliability of the escape method provided by the invention is higher.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robot control, in particular to a robot escape control method and device, a cleaning robot and a computer readable storage medium. BACKGROUND

[0002] A robot is a device integrating multiple functions such as environment perception, dynamic decision and planning, behavior control and execution, including industrial robots, exploration robots, service robots and military robots. Among them, a cleaning robot as a relatively common service robot is widely applied to people's daily life and brings great convenience to people's production and life.

[0003] Taking a cleaning robot as an example, in the process of executing a cleaning task, the cleaning robot may be trapped due to obstacles or changes in house topography; therefore, the cleaning robot should have certain autonomous escape ability. In the traditional technical solution, a method of executing corresponding escape actions according to preset rules is adopted. When the cleaning robot wanders in a small range, it is judged to be trapped and enters an escape mode; at this time, the cleaning robot will call escape actions in the rules one by one, such as executing forward, backward, rotation and other actions, until it escapes.

[0004] However, the escape strategy adopted by the traditional technical solution is constructed in advance by artificial means. When the pre-constructed escape actions are few, it may lead to a low escape success rate, that is, the reliability of the escape method adopted by the traditional technical solution is low. SUMMARY

[0005] Therefore, it is necessary to provide a robot escape control method and device with higher reliability, a cleaning robot, a computer readable storage medium and a computer program product in view of the above technical problems.

[0006] In a first aspect, the present application provides a robot escape control method. The method comprises:

[0007] If the robot is in a motion-restricted state, a determination step is executed: according to the current position of the robot, a state change action to be executed by the robot is determined, wherein the state change action is used to control the robot to escape from the motion-restricted state;

[0008] A detection step is executed: the motion prohibited area position near the robot when the state change action is executed is detected;

[0009] If the motion prohibited area position conflicts with the state change action, the state change action is updated, and the detection step is returned;

[0010] If the motion prohibited area position does not conflict with the state change action, the robot is controlled to execute the state change action.

[0011] In one of the embodiments, after the detecting step is performed, the method further comprises:

[0012] predicting a motion trajectory generated when the robot performs the state change action;

[0013] if the motion trajectory intersects with the motion restricted area position, determining that the motion restricted area position and the state change action conflict;

[0014] if the motion trajectory does not intersect with the motion restricted area position, determining that the motion restricted area position and the state change action do not conflict.

[0015] In one of the embodiments, after the motion trajectory generated when the robot performs the state change action is predicted, the method further comprises:

[0016] determining a plurality of first position coordinates of the motion trajectory and a plurality of second position coordinates of the motion restricted area position;

[0017] if there are overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, determining that the motion trajectory and the motion restricted area position intersect;

[0018] if there are no overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, determining that the motion trajectory and the motion restricted area position do not intersect.

[0019] In one of the embodiments, after the robot performs the state change action, the method further comprises:

[0020] if the robot does not reach the first escape target position within a preset time length, returning to perform the determining step.

[0021] In one of the embodiments, according to the current position of the robot, the state change action to be performed by the robot is determined, comprising:

[0022] generating a first action state parameter for escaping the motion restricted state according to the current position of the robot and a first escape target position of the robot for escaping the motion restricted state;

[0023] determining the state change action according to the current state parameter of the robot and the first action state parameter.

[0024] In one of the embodiments, the state change action is updated, comprising:

[0025] updating the state change action according to the relative position relationship between the current position of the robot and the motion restricted area position.

[0026] In one of the embodiments, the state change action is updated according to the relative position relationship between the current position of the robot and the motion forbidden zone position, including:

[0027] determining a second escape target position of the robot out of the motion restricted state and the relative position relationship between the current position of the robot and the motion forbidden zone position;

[0028] if the distance between the second escape target position and the current position of the robot is less than the preset distance threshold, generating a second action state parameter of the robot out of the motion restricted state according to the second escape target position and the relative position relationship;

[0029] updating the state change action according to the second action state parameter.

[0030] In one of the embodiments, if the distance between the second escape target position and the current position of the robot is greater than or equal to the preset distance threshold, determining a third escape target position between the second escape target position and the current position;

[0031] generating a third action state parameter of the robot out of the motion restricted state according to the third escape target position and the relative position relationship;

[0032] updating the state change action according to the third action state parameter.

[0033] In a second aspect, the application further provides a robot escape control device. The device includes:

[0034] a data acquisition module, configured to, if the robot is in a motion restricted state, execute a determination step of determining a first position of the robot and a state change action to be executed, wherein the state change action is used to control the robot to escape from the motion restricted state;

[0035] a state detection module, configured to execute a detection step of detecting a motion forbidden zone position near the robot when the state change action is executed;

[0036] an action updating module, configured to, if the motion forbidden zone position conflicts with the state change action, update the state change action and return to execute the detection step;

[0037] an action execution module, configured to, if the motion forbidden zone position does not conflict with the state change action, control the robot to execute the state change action.

[0038] In a third aspect, the present application provides a cleaning robot. The cleaning robot comprises a body, a driving assembly, a cleaning assembly, a detection sensor, a memory and a processor, the driving assembly, the cleaning assembly and the detection sensor are all mounted on the body, the driving assembly is used to drive the body to walk on a working surface, the cleaning assembly is used to clean the working surface, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0039] If the robot is in the motion-restricted state, a determination step is performed: according to the current position of the robot, a state change action to be performed by the robot is determined, wherein the state change action is used to control the robot to get rid of the motion-restricted state;

[0040] A detection step is performed: the motion-restricted area position near the robot when the state change action is performed is detected;

[0041] If the motion-restricted area position conflicts with the state change action, the state change action is updated, and the detection step is returned to be performed;

[0042] If the motion-restricted area position does not conflict with the state change action, the robot is controlled to perform the state change action.

[0043] In a fourth aspect, the present application provides a computer readable storage medium. The computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the following steps:

[0044] If the robot is in the motion-restricted state, a determination step is performed: according to the current position of the robot, a state change action to be performed by the robot is determined, wherein the state change action is used to control the robot to get rid of the motion-restricted state;

[0045] A detection step is performed: the motion-restricted area position near the robot when the state change action is performed is detected;

[0046] If the motion-restricted area position conflicts with the state change action, the state change action is updated, and the detection step is returned to be performed;

[0047] If the motion-restricted area position does not conflict with the state change action, the robot is controlled to perform the state change action.

[0048] In a fifth aspect, the present application provides a computer program product. The computer program product comprises a computer program, and the computer program is executed by a processor to implement the following steps:

[0049] If the robot is in the motion-restricted state, a determination step is performed: according to the current position of the robot, a state change action to be performed by the robot is determined, wherein the state change action is used to control the robot to get rid of the motion-restricted state;

[0050] performing a detection step: detecting a motion forbidden zone position near the robot when performing the state change action;

[0051] if the motion forbidden zone position conflicts with the state change action, updating the state change action, and returning to the performing detection step;

[0052] if the motion forbidden zone position does not conflict with the state change action, controlling the robot to perform the state change action.

[0053] The robot escape control method, device, cleaning robot and computer readable storage medium provided by the present application; when it is detected that the robot is in a motion restricted state, the robot and the state change action to be performed will be determined; and before controlling the robot to perform the state change action, the motion forbidden zone position near the robot when performing the state change action will be detected; if the motion forbidden zone position conflicts with the state change action, the state change action needs to be updated and adjusted; on the contrary, if there is no conflict, the robot will be controlled to perform the state change action to escape from the current motion restricted state. Before controlling the robot to perform the corresponding action to complete the escape, the obstacles in the forbidden zone that can exist in the action process of the robot are identified, which can effectively avoid the waste caused by the robot performing invalid state change actions, not only improves the escape efficiency of the robot, but also improves the success rate of the escape of the robot, so that the reliability of the escape method provided by the present application is higher. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 It is an application environment diagram of the robot escape control method in one embodiment;

[0055] Figure 2 It is a flowchart of the robot escape control method in one embodiment;

[0056] Figure 3 It is a flowchart of the motion trajectory conflict detection sub-step of the step in one embodiment;

[0057] Figure 4 It is a flowchart of training the escape decision network in one embodiment;

[0058] Figure 5 It is a flowchart of the robot escape control method in another embodiment;

[0059] Figure 6 It is a structure block diagram of the robot escape control device in one embodiment;

[0060] Figure 7 It is an internal structure diagram of the cleaning robot in one embodiment. DETAILED DESCRIPTION

[0061] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and not to limit the present application.

[0062] When the robot is controlled to perform a corresponding task in the conventional technical solution, many uncertain factors in the task environment can affect the task process of the robot. For example, in the process of controlling the cleaning robot to perform the house cleaning task, the house floor may have a restricted area that hinders the cleaning robot from performing the cleaning task, such as large furniture placed on the house floor, unexpected obstacles, or steps that can cause a fall, etc. The above-mentioned restricted area can cause the cleaning robot to be trapped. The control strategy for assisting the robot to escape in the conventional technical solution is to determine that the cleaning robot is in a trapped state when it moves back and forth in a small range or performs repetitive actions. In the trapped state, the pre-set escape strategy is called one by one to perform forward, backward or rotation actions until the cleaning robot escapes or returns a signal of escape failure after calling all the escape strategies. Since the escape strategy is a rule based on experience and subjective formulation, it has certain limitations. Therefore, when the escape strategy is less, the escape success rate is usually not high, and when the escape strategy is too much, the calling one by one will consume a lot of time. At the same time, in the strategy optimization of the conventional technical solution, in order to avoid performance regression, new rules are often added on the basis of old rules, which makes the number of rules increase continuously with the optimization of the strategy, further increasing the time consumption of the cleaning robot to escape.

[0063] In order to overcome the above-mentioned defects in the conventional technical solution, the embodiments of the present application provide a robot control method which can be applied in the application environment as shown in Figure 1 Similarly, taking the cleaning robot performing the cleaning task as an example, Figure 1 As shown in the application environment, the cleaning robot 102 communicates with the server 104 in a wireless communication manner. The server 104 is built-in with a data storage system which can store data related to the cleaning task performed by the cleaning robot, such as the motion trajectory of the cleaning robot performing the cleaning task and various restricted area information in the cleaning task. The data storage system can be integrated on the server 104, or realized in a cloud manner and distributed storage architecture; and the server 104 can be realized by an independent server or a server cluster composed of multiple servers.

[0064] In the application environment, the cleaning robot 102 starts to perform a cleaning task to clean the floor of a house in response to a cleaning task instruction. The cleaning task instruction can be a control instruction issued by the server 104 to start the cleaning robot 102 and perform the cleaning task based on a work map corresponding to the cleaning task instruction. In the process of performing the cleaning task, the running state of the cleaning robot 102 is detected in real time. When the cleaning robot 102 appears to reciprocate in a small range or perform a repetitive action, it is determined that it is in a motion-limited state. In the motion-limited state, a determination step needs to be performed first to determine a state change action to be performed by the cleaning robot 102 according to the current position of the cleaning robot 102. The state change action is used to control the cleaning robot 102 to escape from the motion-limited state. After the state change action to be performed is determined, a detection step is further performed to detect whether the state change action of the cleaning robot 102 is reasonable, i.e., whether the cleaning robot 102 will conflict with a motion prohibited area near its motion area when performing the state change action. In the case where the position of the motion prohibited area is determined, if the position of the motion prohibited area conflicts with the state change action, the state change action of the cleaning robot 102 is updated, and the detection step is returned to perform. If the position of the motion prohibited area does not conflict with the state change action, the cleaning robot 102 is controlled to perform the state change action to escape from the motion-limited state.

[0065] In another implementation environment, the robot is built-in with a processor and a memory. The processor can control the robot to perform a corresponding action to escape from a running-limited state when the robot is in the running-limited state. The memory can store map data and the like of the robot in a task work process.

[0066] Similarly, taking the cleaning robot as an example, in the process of performing a cleaning task by the cleaning robot, the processor collects robot running state parameters and prohibited area information in a work environment by calling various sensors built-in in the robot. If the processor determines that the robot is in a motion-limited state according to the collected running state parameters, determines a state change action to escape from the motion-limited state according to the current position of the robot. Further, the rationality of the state change action is judged based on the position of the motion prohibited area described by the prohibited area information. If the position of the motion prohibited area conflicts with the state change action, the state change action is updated, and the rationality is re-judged. If the position of the motion prohibited area does not conflict with the state change action, the processor controls the robot to perform the state change action, so that the robot can smoothly escape from the motion-limited state.

[0067] In order to more specifically describe the implementation process of the method provided in the present application, as shown in Figure 2 In one embodiment, a robot control method is provided, which can be implemented by Figure 1The server 104 in the server 104 executes, or is independently executed by a robot device with data processing capability. The robot control method comprises the following steps:

[0068] Step 202, if the robot is in a motion restricted state, a determination step is performed: according to the current position of the robot, a state change action to be performed by the robot is determined. The state change action is used to control the robot to escape from the motion restricted state.

[0069] In the embodiment, the robot comprises a plurality of devices integrating environment perception, dynamic decision and planning, behavior control and execution, etc., such as express sorting robots and intelligent cleaning robots. The motion restricted state in the embodiment is a trapped state. From an objective factor, the robot is in a motion restricted state, which may be caused by environmental changes or the complexity of the task being performed. In the motion restricted state, the robot may reciprocate in a small operating range, or perform repetitive actions for a long time. The state change action in the embodiment is an escape action, which is used to make the robot escape from the current trapped state and restore to a normal task operating state.

[0070] Taking a cleaning robot performing a house floor cleaning task as an example, during the cleaning robot performing the cleaning task operation, various sensors or other acquisition devices in the cleaning robot are called in real time to monitor the motion state of the cleaning robot. Illustratively, when the sensor detects that the cleaning robot travels to the bottom of furniture and is blocked by a large amount of debris; or when cleaning the bottom of the sofa, the cleaning robot is restricted due to the collapse of the sofa. When the foregoing trapped situation occurs, according to the current position information of the cleaning robot, a corresponding escape action is triggered. In the embodiment, the triggering mode of the escape action can be to match the corresponding (previously set) escape action according to different scenes and position information according to the preset escape rule; or, in the embodiment, according to the trapped state information of the cleaning robot, such as the current position, pose, action and environment of the cleaning robot, an escape action corresponding to the current position is generated by calling an escape strategy based on reinforcement learning.

[0071] Step 204, a detection step is performed: detecting a motion restricted area position near the robot when the state change action is performed.

[0072] In embodiments, the state change action includes, but is not limited to, actions such as adjusting a pose, advancing, retreating, and rotating. The motion forbidden zone position near the robot refers to a position of a motion forbidden zone that may intersect with a motion area of the robot during the process of freeing the cleaning robot; the motion area refers to an area that the cleaning robot needs to pass through when performing a state change action; for example, when the state change action is an action content such as advancing or retreating, the motion area refers to a passing area of the cleaning robot when advancing or retreating. For another example, when the state change action is an action such as adjusting a pose or rotating, the motion area refers to a deflection area of the cleaning robot when the front face is adjusted at a certain deflection angle. In addition, the motion area in embodiments can also be a combination of the aforementioned passing area and deflection area. In embodiments, the motion forbidden zone position is a position that the cleaning robot is prohibited to reach in a complete plane map. The motion forbidden zone position represents the position of the motion forbidden zone, and the motion forbidden zone in embodiments can be determined according to various obstacles in the floor of the house, large furniture, pre-set virtual walls, and floor undulations such as steps. The various motion forbidden zones described above can all cause the cleaning robot to be trapped or damaged.

[0073] Further, in embodiments, various obstacle factors collected by sensors can be marked as weak obstacles, and various obstacle factors introduced in the rationality judgment process, such as virtual walls and steps, can be marked as strong obstacles. In the process of determining the motion forbidden zone in the operation map, the obstacle information corresponding to the strong obstacles can be directly selected to construct the motion forbidden zone, so that the process of delimiting the motion forbidden zone is more convenient and efficient, and at the same time, the motion or action of the cleaning robot is not too limited due to the setting of too many motion forbidden zones, which causes the task completion efficiency to be too low.

[0074] As an example, assuming that the initial interval distance between the selected freeing target position for the cleaning robot and the cleaning robot is d, the positions of the motion forbidden zones within the initial interval distance d from the cleaning robot can all be used as the motion forbidden zone position near the cleaning robot during this freeing.

[0075] Exemplarily, after the cleaning robot starts to perform the cleaning task, the environment information around the cleaning robot can be acquired by the sensor or the laser radar, whether in the motion-restricted state or not, and the hindering factors in the environment, such as obstacles, furniture and steps, etc., can be identified by the environment mapping and target detection methods, and the attribute information of the obstacles can be determined by target recognition. For example, the cleaning robot identifies by the ultrasonic sensor that there is a step at 50 centimeters in front of the current position. Further, the width information of the step is determined, and based on the distance from the step and the width of the step, a motion-restricted area and the position corresponding to the motion-restricted area are constructed in the cleaning operation map. In addition, in the embodiments, the motion-restricted area can also be constructed by pre-setting a virtual wall, for example, a virtual wall is set in the operation map of the cleaning task of the cleaning robot by the interactive operation of the user, so that it cannot pass directly; according to the thickness and position of the virtual wall set by the user, the motion-restricted area corresponding to the virtual wall is constructed in the operation map of the cleaning task, and the range and position of the motion-restricted area are determined according to the thickness and position of the virtual wall.

[0076] In step 206, if the motion-restricted area position conflicts with the state change action, the state change action is updated, and the detection step is returned to be executed.

[0077] In the embodiments, the specific way of judging whether the motion-restricted area position conflicts with the state change action is to judge whether all positions covered by the motion-restricted area contact or overlap with the motion area corresponding to the state change action. If there is contact or overlap in position, it is determined that the motion-restricted area position conflicts with the state change action; otherwise, it is determined that the motion-restricted area position does not conflict with the state change action. In the embodiments, the update of the state change action can be to reselect a state change action from a plurality of alternative state change actions to replace the state change action in conflict; or the action parameters of the state change action can be adjusted to realize the update of the state change action.

[0078] Specifically in the embodiment, when calling the escape strategy based on reinforcement learning to generate an escape action corresponding to the current position of the cleaning robot, the escape strategy may generate a plurality of alternative escape action parameters. Before controlling the cleaning robot to perform the corresponding escape action based on the action parameters, the rationality of the action should be judged. The process of rationality judgment in the embodiment is to judge whether the cleaning robot will touch or cross the running restricted area when performing the corresponding escape action. More specifically, the way to judge whether the running restricted area will be touched or crossed in the embodiment is to predict whether the cleaning robot will conflict with the position of the running restricted area during the execution of the escape action, that is, whether the escape action will cause the cleaning robot to touch or cross the running restricted area. When it is determined that there may be a conflict with the position of the running restricted area, it is necessary to update the escape action, that is, reselect a new set of escape action parameters from the multiple alternative escape action parameters, and make another rationality judgment on the escape action corresponding to the new escape action parameters.

[0079] Step 208: If the position of the restricted area does not conflict with the state change action, the robot is controlled to execute the state change action.

[0080] Specifically in the embodiment, if the escape strategy based on reinforcement learning is called to generate an escape action corresponding to the current position of the cleaning robot, and after rationality judgment it is determined that the escape action does not conflict with the position of the motion restricted area, that is, the cleaning robot will not have any contact with the motion restricted area during the process of executing the escape action; then the cleaning robot will be controlled to execute the escape action, so that it can successfully escape from the trapped state.

[0081] In the above-mentioned robot escape control method, when it is detected that the robot is in a restricted motion state, the robot and the state change action to be executed will be determined; and before controlling the robot to execute the state change action, the position of the motion restricted zone near the robot when executing the state change action will be detected; if the motion restricted zone position conflicts with the state change action, the state change action needs to be updated and adjusted; on the contrary, if no conflict occurs, the robot will be controlled to execute the state change action to escape from the current restricted motion state. Before controlling the robot to execute the corresponding action to complete the escape, by identifying the obstacles of the restricted zone that may exist during the robot's action, it can effectively avoid the robot from executing invalid state change actions and causing losses, which not only improves the robot's escape efficiency, but also improves the robot's escape success rate, making the escape method provided by the present application more reliable.

[0082] In one embodiment, Figure 3 As shown, the process of judging the rationality of the state change action may include the following steps:

[0083] Step 302, predict the motion trajectory generated when the robot performs the state change action.

[0084] In embodiments, various state change actions can be formed due to different motion restricted states. As before, the state change action can include actions such as adjusting the pose, advancing, retreating, and rotating; and correspondingly, different motion trajectories can be formed during the execution of different state change actions.

[0085] Illustratively, again taking the cleaning robot as an example, if the cleaning robot performs the action of advancing or retreating to escape the predicament, the motion trajectory formed is a plurality of straight line paths with overlapping portions; if the cleaning robot performs the action of adjusting the pose or rotating to escape the predicament, the initial orientation of the cleaning robot is taken as the initial edge, and the orientation after the action is completed is taken as the terminal edge, and the deflection trajectory formed between the initial edge and the terminal edge corresponding to the turning angle is taken as the motion trajectory of the escape action. In a more complex predicament, the cleaning robot needs to form a corresponding escape action by combining multiple actions, for example, not only does it need to adjust the pose of the cleaning robot multiple times, but it also needs to perform the actions of advancing and retreating after each adjustment of the pose. Therefore, in the case of multiple action combinations forming an escape action, the corresponding motion trajectory is also formed by combining the motion estimates corresponding to the multiple actions, i.e., it includes both the deflection trajectory and the straight line path.

[0086] Step 304, if the motion trajectory intersects with the motion restricted area position, it is determined that the motion restricted area position and the state change action are in conflict.

[0087] Illustratively, before the embodiment controls the cleaning robot to perform the escape action, a rationality judgment of the action needs to be performed. The process of the rationality judgment is to compare the range covered by the motion trajectory with the range covered by the motion restricted area, so as to calculate whether the motion trajectory of the cleaning robot performing the escape action is in contact or coincides with the motion restricted area position. More specifically, the way to judge whether there is contact or coincidence is to determine whether the position information (such as a plurality of coordinate points) of the motion trajectory and the position information (also a plurality of coordinate points) of the motion restricted area are the same. If there is a coincidence of positions, it is determined that the current escape action to be performed may be in conflict with the motion restricted area in the map.

[0088] Step 306, if the motion trajectory does not intersect with the motion restricted area position, it is determined that the motion restricted area position and the state change action are not in conflict.

[0089] For example, the embodiment compares the calculated motion trajectory coordinate points with the coordinate points in the motion restricted area and the coordinate points of the motion restricted area edge, and if none of them are the same, it is determined that the current to-be-executed escape action does not conflict with the motion restricted area position. The embodiment makes a reasonable judgment on the to-be-executed escape action through accurate comparison of the position information, so that the judgment result is more accurate and reliable.

[0090] In one embodiment, after predicting the motion trajectory generated when the robot executes the state change action, the above method further comprises:

[0091] determining a plurality of first position coordinates of the motion trajectory and a plurality of second position coordinates of the motion restricted area position; if there are overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is determined that the motion trajectory and the motion restricted area position have an intersection; if there are no overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is determined that the motion trajectory and the motion restricted area position have no intersection.

[0092] Specifically, all trajectory point coordinates of the motion trajectory are obtained as the plurality of first position coordinates, and all region point coordinates of the motion restricted area position near the cleaning robot are obtained as the plurality of second position coordinates; if there are any same first position coordinates and second position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is considered that the motion trajectory and the motion restricted area position have an intersection; if there are no same first position coordinates and second position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is considered that the motion trajectory and the motion restricted area position have no intersection.

[0093] In one embodiment, after controlling the robot to execute the state change action, the method in the embodiment can further comprise the following steps: if the robot does not reach the first escape target position within a preset time length, returning to execute the determining step.

[0094] In the embodiment, the preset time length refers to the maximum time length for the robot to execute the state change action, and if the robot escapes from the motion restricted state by executing the state change action within this maximum time length, it is determined that the robot escapes successfully; if the robot fails to escape from the motion restricted state by executing the state change action within this maximum time length, it is determined that the robot escapes overtime. The first escape target position in the embodiment is any position that can make the robot escape from the motion restricted state. Moreover, the first escape target position in the embodiment is the target point of the trajectory.

[0095] For example, in the embodiment, one coordinate position in the work trajectory of the cleaning robot selected by the cleaning robot to perform the cleaning task is selected as the first escape target position. That is, the cleaning robot can successfully escape from the trapped state if it successfully reaches the coordinate position within a set maximum time length. After determining the reasonable escape action of the cleaning robot in the escape state through rationality judgment, the cleaning robot is controlled to start performing the escape action, and a built-in timer of the cleaning robot is started. When the timer reaches the maximum time length set in advance, the current position of the cleaning robot is compared with the aforementioned determined coordinate position. If they are the same, it is determined that the cleaning robot has successfully escaped from the trapped state, and the cleaning task is resumed according to the aforementioned work estimation; if they are not the same, it is determined that the cleaning robot has escaped from the trapped state for too long, and the step of generating the escape action in the embodiment method needs to be fed back. For example, in the embodiment, the escape action of the cleaning robot is generated based on the reinforcement learning escape strategy, so in the case where the cleaning robot escapes from the trapped state for too long, a new escape action needs to be generated again according to the new position reached by the cleaning robot after performing the escape action and the current state information of the robot by using the escape strategy. In the embodiment, the state change action is updated and adjusted in the case where the robot escapes from the trapped state for too long, so that the robot can escape from the trapped state more efficiently.

[0096] In one embodiment, the process of determining the state change action to be performed by the robot according to the current position of the robot in the method can include the following steps:

[0097] Step one, generating a first action state parameter for escaping from the motion restricted state according to the current position of the robot and the first escape target position of the robot escaping from the motion restricted state.

[0098] Step two, determining the state change action according to the current state parameter of the robot and the first action state parameter.

[0099] In embodiments, the first action parameter and the current state parameter are both physical parameters describing the motion state of the robot, including but not limited to the travel speed, acceleration, rotation speed, and task trajectory of the cleaning robot. More specifically, the current state parameter in embodiments is obtained by real-time collection of the action parameters of the cleaning robot through various sensors of the cleaning robot; for example, if the cleaning robot is currently in a trapped state, the current state parameter collected is the state parameter of the cleaning robot in the trapped state. The first action parameter refers to the action state parameter preset in the cleaning task instruction issued to the cleaning robot, for example, the action state parameter required to be executed when the cleaning robot reaches a specific location in the house or to change the state. In embodiments, the first action parameter needs to be determined, and the relative positional relationship between the current position of the cleaning robot and the specified trapped target position needs to be determined first, for example, the distance between the two positions; which serves as the basis for determining the action state parameter.

[0100] Exemplarily, when it is detected that the cleaning robot is in a trapped state, a specific position is first selected as the trapped target position of the cleaning robot from the pre-acquired task trajectory or task map corresponding to the cleaning task. After the trapped target position is determined, the action state parameter of the cleaning robot when reaching the trapped target position in the task trajectory or task map is determined according to the task planning corresponding to the cleaning task, including the travel speed, travel direction, and rotation speed of the cleaning robot. The above parameters are taken as the target action state parameter, and the action state parameter of the cleaning robot in the current trapped state is collected in real time. Further, the first action state parameter in embodiments is formed according to the distance between the current trapped position of the cleaning robot and the trapped target position, and the difference between the action state parameters before and after the trapped state is changed, so that the cleaning robot performs the corresponding state change action to get rid of the trapped state. In embodiments, the state change action of the robot is determined through careful comparison of the position and motion parameters, so that the robot can more smoothly and efficiently complete the release of the trapped state.

[0101] In one embodiment, the process of updating the state change action in the method can include the following steps: updating the state change action according to the relative positional relationship between the current position of the robot and the motion prohibited zone position.

[0102] In embodiments, the relative positional relationship represents the positional relationship between the current position of the robot and the motion prohibited zone (position). In embodiments, the relative positional relationship can be used to more accurately determine the precise position of the motion prohibited zone.

[0103] Exemplarily, in the trapped state, firstly, the current trapped position of the cleaning robot is determined; and based on the state parameters (such as the current pose, orientation and action parameters) of the cleaning robot in the trapped state, the corresponding state change action is determined, and then in the detection step, the motion forbidden area of the cleaning robot is detected, for example, according to the pre-loaded cleaning operation map, at the current position of the cleaning robot, 10 cm away from the current orientation, a virtual wall is set, the direction of the virtual wall is north-south, the wall length is 1 m and the wall thickness is 20 cm. After the parameter values of the virtual wall are determined, it is determined that the virtual wall is mapped to the forbidden area position in the plane map. After the forbidden area formed by the virtual wall and the distance between the cleaning robot and the virtual wall are determined, the state change action determined in the foregoing is adjusted and optimized, so that the updated state change action can perfectly avoid the virtual wall; for example, the updated state change action needs to first turn the cleaning robot by 90 degrees counterclockwise, and then control it to move 5 cm away from the wall in the direction parallel to the wall surface, until the end of the virtual wall. In the embodiment, based on the relative position relationship of the motion forbidden area position, a more accurate state change action can be formed to enable the robot to successfully escape from the motion trapped state.

[0104] More specifically, in one embodiment, the process of updating the state change action according to the relative position relationship between the current position of the robot and the motion forbidden area position in the method can include the following steps:

[0105] Step one, determine the second trapped target position of the robot to escape from the motion restricted state, and the relative position relationship between the current position of the robot and the motion forbidden area position.

[0106] Step two, if the distance between the second trapped target position and the current position of the robot is less than a preset distance threshold, then according to the second trapped target position and the relative position relationship, generate the second action state parameter of the robot to escape from the motion restricted state.

[0107] Step three, update the state change action according to the second action state parameter.

[0108] In an embodiment, the second escape target position is also a coordinate position selected in the operation trajectory or operation map in which the cleaning robot performs the cleaning task. Different from the first escape target position in the foregoing embodiment, the second escape target position in the embodiment is a coordinate position selected again in the operation trajectory or operation map as a new escape target position after the first escape target position is determined, in a case where a reasonable state change action cannot be formed due to the restriction of the motion forbidden area. In addition, before the escape target position is reselected and the corresponding state change action is performed or in the process of performing the state change action, it is also necessary to judge the rationality of the selection of the new escape target position. In the embodiment, in order to improve the success rate of robot escape, when judging the rationality of the reselected escape target position, it is necessary to judge whether the distance between the escape target position and the current position of the robot is less than a preset distance threshold. The preset distance threshold can be a specific numerical value as a hyperparameter in the method, that is, a fixed specific numerical value is set as the preset distance threshold; or the preset distance threshold can be determined according to the distance between the previous escape target position and the current position, and specifically the preset distance threshold can be set to be equal to or less than the distance between the previous escape target position and the current position, so that the distance between the newly determined escape target position and the current position should be closer than the distance between the previous escape target position, thereby enabling the robot to escape faster. The second action state parameter in the embodiment is an action state parameter formed according to the distance between the current trapped position of the cleaning robot and the reselected escape target position, and the difference between the action state parameters before and after escape.

[0109] Exemplarily, in the embodiment, in the trapped state of the cleaning robot, the first escape target position is determined, and based on the escape target position and the motion state parameter of the cleaning robot, the corresponding escape action parameter is generated through the escape strategy of reinforcement learning. However, the rationality judgment of the escape action of the escape action parameter is that the escape action is unreasonable, that is, the escape action may conflict with the motion forbidden area, or in the detection of escape timeout, it is determined that the cleaning robot escapes timeout. Therefore, it is necessary to adjust the escape target position to obtain the second escape position, and at the same time obtain the corresponding escape action parameter through the same way. Before the rationality judgment of the new escape action, it is necessary to judge the rationality of the selection of the second escape position in advance, for example, the distance between the first escape position and the current position is 70 cm, which is taken as the distance threshold; and the distance between the newly determined second escape position and the current position is 55 cm, which is less than the foregoing determined distance threshold. Therefore, the selection of the second escape position is reasonable, so that the corresponding action state parameter is formed according to the foregoing escape strategy, and the current escape action of the cleaning robot is updated according to the action state parameter.

[0110] In an embodiment, if the re-determined escape target position does not satisfy the distance threshold condition, a new escape target position needs to be determined by a more optimal position determination strategy to avoid the escape efficiency being relatively low due to the calculation process of the escape strategy being too complicated. The method specifically includes the following steps:

[0111] Step one, if the distance between the second escape target position and the current position of the robot is greater than or equal to the preset distance threshold, a third escape target position is determined between the second escape target position and the current position.

[0112] Step two, according to the third escape target position and the relative position relationship, a third action state parameter of the robot escaping from the motion restricted state is generated;

[0113] Step three, according to the third action state parameter, the state change action is updated.

[0114] In an embodiment, the third escape target position is a new escape target position determined again between the current position of the robot and the determined second escape target position. Further, in order to improve the escape efficiency, the foregoing position determination strategy can also be more specifically constrained and limited in the embodiment, for example, the midpoint position between the current position and the second escape target position is taken as the new escape target position until the new escape target position satisfies the distance threshold condition. Similarly, the third action state parameter in the embodiment is an action state parameter formed according to the distance between the current trapped position of the cleaning robot and the re-determined escape target position, and the difference between the action state parameters before and after the escape.

[0115] Exemplarily, in the embodiment, the distance between the escape target position determined after the update of the escape target position of the cleaning robot in the trapped state and the current position of the cleaning robot is 80 centimeters, which does not satisfy the threshold distance of 70 centimeters between the first determined escape target position and the current position. Further, the escape target position still needs to be adjusted; the embodiment will select the midpoint between the updated determined escape target position and the current position as the new escape target position, and the distance between the escape target position and the current position is 40 centimeters, which also satisfies the threshold distance condition of 70 centimeters. Finally, the selection of the third escape position obtained after the update is reasonable, so that the corresponding action state parameter is formed according to the foregoing escape strategy, and the current escape action of the cleaning robot is updated according to the action state parameter.

[0116] It needs to be pointed out that the escape strategy in the embodiment is obtained by training in a reinforcement learning manner. Specifically, the escape strategy in the embodiment is implemented based on a neural network model of reinforcement learning. After the training of the network model is completed, the network model is transplanted into a robot or a server in communication with the robot.

[0117] More specifically, as shown in Figure 4 the training process of the escape strategy network model in the embodiment includes the following steps:

[0118] Step one, the escape strategy network model is trained in a simulator, wherein the simulator includes but is not limited to tools such as mujoco, pybullet, issac sim, etc. Different trapped scenes for training are built in the simulator, such as: table and chair group, narrow mouth, narrow channel, groove, door sill and obstacle blocking, etc. At the same time, the scene layout is randomized to increase the diversity of the scene.

[0119] Step two, the domain randomization is used to improve the feasibility of the output result of the escape strategy, wherein the objects of the domain randomization include but are not limited to the mass, inertia tensor, friction coefficient, state input, action output, etc. of the robot.

[0120] Step three, the escape target position, the current state parameter of the robot and the current position of the robot are input into the escape strategy network model, and feature extraction and dimension reduction processing are performed through the model. In addition, in order to improve the accuracy of the model output, the input of the model can also include strong obstacle information, weak obstacle information and historical escape action parameters in the map, etc.

[0121] Step four, the escape action output by the model is reasonably judged, and the weak obstacles which are not labeled as strong obstacles are labeled as strong obstacles.

[0122] Step five, the reward value or the penalty value of the escape strategy network model is determined. Specifically, in the embodiment, the application of the reward value or the penalty value includes but is not limited to the following cases: 1, repulsive potential field: when the distance to the obstacle is less than a threshold, the penalty value starts from 0 and increases, the closer to the obstacle, the larger the penalty value, wherein the penalty value of the strong obstacle is larger than that of the weak obstacle. 2, attractive potential field: the closer to the target point, the smaller the penalty value. 3, collision penalty: after collision, a penalty value is given; when continuous collision occurs, the penalty value of collision is lowered. 4, direction penalty: if there is an obstacle within a certain distance in front or behind, a penalty value is given. 5, arrival reward: when reaching the target point, a reward value is given.

[0123] In the embodiment, the escape strategy is automatically generated by the reinforcement learning manner, which not only effectively improves the response efficiency of the escape action, but also improves the accuracy of the escape action, so that the robot can successfully escape.

[0124] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description, serve to explain the principles of the application. Figure 5 The complete process of the robot control method provided by the embodiments in the present application is described as follows:

[0125] Step one, when the cleaning robot is found to be trapped, a target point in the pre-generated cleaning task trajectory is selected as the target position input, combined with the current state information of the cleaning robot, a reinforcement learning-based escape strategy is called to output the escape action. The state information includes but is not limited to the current position, pose, action, strong obstacle markers, weak obstacle markers and historical action execution results of the cleaning robot in the vicinity of the motion range.

[0126] Step two, it is judged whether the current to-be-executed escape action is reasonable based on rules. For example, it is judged whether the cleaning robot will collide with obstacles, whether it will cross a virtual wall or whether it will reach a step area when executing the escape action.

[0127] Step three, after passing the reasonable detection step of step two, the escape timeout detection will also be performed in the process of the cleaning robot executing the escape action. If the escape timeout, the current position of the cleaning robot and the key position in the trajectory of the current target point are selected as the new target position input, the reinforcement learning-based escape strategy is called again to output the escape action, and then the rationality judgment is performed.

[0128] Step four, steps two and three are repeated until the cleaning robot successfully escapes. Or when the number of cycles of steps two and three is greater than the preset number of rounds, the cleaning robot still fails to escape successfully, the corresponding alarm information will be triggered, and the manual intervention will be extracted to assist the cleaning robot to escape.

[0129] It should be understood that although each step in the flowchart involved in the above embodiments is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise stated herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other sequences. Moreover, at least part of the steps in the flowchart involved in the above embodiments can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence of these steps or stages is not necessarily sequential, but can be alternately executed with other steps or steps or stages in other steps.

[0130] Based on the same inventive concept, the embodiments of the present application also provide a robot escape control device for implementing the robot escape control method described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more robot escape control device embodiments provided below can refer to the limitations of the robot escape control method described above, which will not be repeated here.

[0131] In one embodiment, as shown in Figure 6 A robot escape control device 600 is provided, comprising a data acquisition module 601, a state detection module 602, an action update module 603 and an action execution module 604, wherein:

[0132] The data acquisition module 601 is configured to, if the robot is in a motion-restricted state, perform a determination step of determining a first position of the robot and a state change action to be executed, wherein the state change action is used to control the robot to escape from the motion-restricted state.

[0133] The state detection module 602 is configured to perform a detection step of detecting a motion prohibited area position near the robot when the state change action is executed.

[0134] The action update module 603 is configured to, if the motion prohibited area position conflicts with the state change action, update the state change action and return to perform the detection step.

[0135] The action execution module 604 is configured to, if the motion prohibited area position does not conflict with the state change action, control the robot to execute the state change action.

[0136] In one embodiment, the device 600 further comprises a trajectory detection module, which is further configured to predict a motion trajectory generated when the robot executes the state change action; if the motion trajectory and the motion prohibited area position have an intersection, it is determined that the motion prohibited area position and the state change action conflict; if the motion trajectory and the motion prohibited area position do not have an intersection, it is determined that the motion prohibited area position and the state change action do not conflict.

[0137] In one embodiment, the trajectory detection module is further configured to determine a plurality of first position coordinates of the motion trajectory and a plurality of second position coordinates of the motion prohibited area position; if there are overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is determined that the motion trajectory and the motion prohibited area position have an intersection; if there are no overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is determined that the motion trajectory and the motion prohibited area position do not have an intersection.

[0138] In one embodiment, the action update module 603 is further configured to, if the robot does not reach the first escape target position within a preset time length, return to perform the determination step.

[0139] In an embodiment, the data collection module 601 is further configured to generate a first action state parameter for the robot to escape from the motion restricted state according to a current position of the robot and a first escape target position of the robot to escape from the motion restricted state, and determine the state change action according to a current state parameter of the robot and the first action state parameter.

[0140] In an embodiment, the action update module 603 is further configured to update the state change action according to a relative position relationship between the current position of the robot and the motion forbidden zone position.

[0141] In an embodiment, the action update module 603 is further configured to determine a second escape target position of the robot to escape from the motion restricted state and a relative position relationship between the current position of the robot and the motion forbidden zone position, generate a second action state parameter for the robot to escape from the motion restricted state according to the second escape target position and the relative position relationship if a distance between the second escape target position and the current position of the robot is less than a preset distance threshold, and update the state change action according to the second action state parameter.

[0142] In an embodiment, the action update module 603 is further configured to determine a third escape target position between the second escape target position and the current position of the robot if the distance between the second escape target position and the current position of the robot is greater than or equal to the preset distance threshold, generate a third action state parameter for the robot to escape from the motion restricted state according to the third escape target position and the relative position relationship, and update the state change action according to the third action state parameter.

[0143] The above-mentioned modules in the robot escape control device can be realized by software, hardware, or a combination thereof. The above-mentioned modules can be embedded in or independent of a processor in the cleaning robot in hardware form, or stored in a memory in the cleaning robot in software form, so as to be called and executed by the processor to perform the operations corresponding to the above-mentioned modules.

[0144] In an embodiment, a cleaning robot is provided, and an internal structure diagram of the cleaning robot can be as shown in Figure 7The cleaning robot includes a body, a processor, a memory, a communication interface, an input device, a driving assembly, a cleaning assembly, and a detection sensor, the driving assembly, the cleaning assembly, and the detection sensor are all installed on the body, the driving assembly is used to drive the body to walk on a working surface, and the cleaning assembly is used to clean the working surface. The processor of the cleaning robot is used to provide computing and control capabilities. The memory of the cleaning robot includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the cleaning robot is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, a mobile cellular network, NFC (near field communication), or other technologies. The computer program is executed by the processor to implement a control method of a cleaning robot.

[0145] Those skilled in the art can understand that, Figure 7 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the cleaning robot to which the scheme of the present application is applied. A specific cleaning robot can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0146] In one embodiment, a cleaning robot is provided, including a body, a driving assembly, a cleaning assembly, a detection sensor, a memory, and a processor, the driving assembly, the cleaning assembly, and the detection sensor are all installed on the body, the driving assembly is used to drive the body to walk on a working surface, the cleaning assembly is used to clean the working surface, the memory stores a computer program, and the processor executes the computer program to implement the following steps:

[0147] If the robot is in a motion-restricted state, a determination step is performed: according to the current position of the robot, a state change action to be performed by the robot is determined, wherein the state change action is used to control the robot to escape from the motion-restricted state;

[0148] A detection step is performed: a motion-restricted area position near the robot when the state change action is performed is detected;

[0149] If the motion-restricted area position conflicts with the state change action, the state change action is updated, and the detection step is returned to be performed;

[0150] If the motion-restricted area position does not conflict with the state change action, the robot is controlled to perform the state change action.

[0151] In one embodiment, a computer readable storage medium is provided, having stored thereon a computer program, the computer program being executable by a processor to implement the following steps:

[0152] If the robot is in the motion restricted state, a determining step is performed to determine, according to the current position of the robot, a state change action to be performed by the robot, wherein the state change action is used to control the robot to get rid of the motion restricted state;

[0153] A detecting step is performed to detect a motion forbidden zone position in the vicinity of the robot when performing the state change action;

[0154] If the motion forbidden zone position conflicts with the state change action, the state change action is updated, and the detecting step is returned to perform;

[0155] If the motion forbidden zone position does not conflict with the state change action, the robot is controlled to perform the state change action.

[0156] In one embodiment, a computer program product is provided, comprising a computer program, the computer program being executable by a processor to implement the following steps:

[0157] If the robot is in the motion restricted state, a determining step is performed to determine, according to the current position of the robot, a state change action to be performed by the robot, wherein the state change action is used to control the robot to get rid of the motion restricted state;

[0158] A detecting step is performed to detect a motion forbidden zone position in the vicinity of the robot when performing the state change action;

[0159] If the motion forbidden zone position conflicts with the state change action, the state change action is updated, and the detecting step is returned to perform;

[0160] If the motion forbidden zone position does not conflict with the state change action, the robot is controlled to perform the state change action.

[0161] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0162] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0163] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent of the present application. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of protection of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A robot control method, characterized in that: The method comprises: If the robot is in a restricted motion state, performing a determining step: determining a state change action to be performed by the robot according to the current position of the robot, wherein the state change action is used to control the robot to exit the restricted motion state; Performing a detection step: detecting a motion restricted zone position near the robot when performing the state change action; If the position of the restricted area conflicts with the state change action, the state change action is updated and the detection step is returned to be executed; If the position of the movement restricted zone does not conflict with the state change action, the robot is controlled to execute the state change action.

2. The method according to claim 1, characterized in that After performing the detecting step, the method further comprises: Predicting a motion trajectory generated when the robot performs the state change action; If the motion trajectory intersects with the position of the motion forbidden zone, it is determined that there is a conflict between the position of the motion forbidden zone and the state change action; If the motion trajectory does not intersect with the position of the motion forbidden zone, it is determined that there is no conflict between the position of the motion forbidden zone and the state change action.

3. The method according to claim 2, characterized in that After predicting the motion trajectory generated by the robot when performing the state change action, the method further includes: Determining a plurality of first position coordinates of the motion trajectory and a plurality of second position coordinates of the motion restricted area; If there are overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, determining that there is an intersection between the motion trajectory and the position of the motion restricted zone; If there are no overlapping position coordinates between the plurality of first position coordinates and the plurality of second position coordinates, it is determined that there is no intersection between the motion trajectory and the position of the motion restricted zone.

4. The method according to claim 1, wherein After controlling the robot to perform the state change action, the method further includes: If the robot does not reach the first escape target position within the preset time, return to execute the determining step.

5. The method according to claim 1, wherein The step of determining a state change action to be performed by the robot according to the current position of the robot includes: generating a first action state parameter for escaping from the motion-restricted state according to the current position of the robot and a first escape target position of the robot for escaping from the motion-restricted state; A state change action is determined according to the current state parameter of the robot and the first action state parameter.

6. The method according to claim 1, characterized in that The updating of the state change action includes: The state change action is updated according to the relative position relationship between the current position of the robot and the position of the movement restricted area.

7. The method according to claim 6, characterized in that The updating of the state change action according to the relative positional relationship between the current position of the robot and the position of the movement restricted area includes: Determining a second escape target position for the robot to escape from the restricted motion state, and a relative positional relationship between the current position of the robot and the position of the restricted motion zone; If the distance between the second escape target position and the current position of the robot is less than a preset distance threshold, generating a second action state parameter for the robot to escape from the motion restricted state according to the second escape target position and the relative position relationship; The state change action is updated according to the second action state parameter.

8. The method according to claim 7, characterized in that The method further comprises: If the distance between the second escape target position and the current position of the robot is greater than or equal to a preset distance threshold, determining a third escape target position between the second escape target position and the current position; generating a third action state parameter for the robot to escape from the motion-restricted state according to the third escape target position and the relative position relationship; The state change action is updated according to the third action state parameter.

9. A robot escape control device, characterized in that: The device comprises: a data acquisition module, configured to, if the robot is in a restricted motion state, execute a determining step: determining a first position of the robot and a state change action to be performed, wherein the state change action is used to control the robot to exit the restricted motion state; A state detection module is used to perform a detection step: detecting a motion restricted zone position near the robot when performing the state change action; an action updating module, configured to update the state changing action if the position of the motion restricted area conflicts with the state changing action, and return to the detection step; An action execution module is used to control the robot to execute the state change action if the position of the movement restricted area does not conflict with the state change action.

10. A cleaning robot comprising a body, a driving component, a cleaning component, a detection sensor, a memory, and a processor, wherein the driving component, the cleaning component, and the detection sensor are all mounted on the body, the driving component is used to drive the body to move on a working surface, the cleaning component is used to clean the working surface, and the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

Images