A robotic grooming method based on spatiotemporal data

Abstract

The present invention relates to the field of robot grooming technology, and in particular to a robot grooming method based on spatiotemporal data. The method comprises: obtaining a starting point and an end point of a grooming route of an area to be groomed, and determining an initial grooming path S according to the starting point and the end point of the grooming route and the distribution information of a target object to be groomed in the area to be groomed; discretizing and sampling the initial grooming path S to obtain a plurality of navigation points, and generating a discretized path S based on the navigation points. e ; For the discretized path S e The path is segmented by a rough combination method to obtain a plurality of segmented paths, and path tracking is completed for each segmented path. The present invention can reduce the error of the grooming robot when performing path tracking, improve the path tracking effect, and further improve the grooming efficiency.

Description

Technical Field

[0001] The present invention relates to the field of robot grooming technology, and in particular to a robot grooming method based on spatiotemporal data. Background Art

[0002] With the continuous development of robot technology, robots have been widely used in various fields, such as home, industry, medical treatment, etc. In the home environment, robots can undertake many household tasks, such as cleaning, cooking, caring for the elderly and children, etc.

[0003] In the prior art, for example, when a sweeping robot is cleaning a room or a lawnmower robot is taking care of vegetation, real-time path tracking is required to ensure that the lawnmower robot does not easily deviate from the care path while taking care of it. However, the prior art has poor versatility in path tracking applications, especially when tracking complex paths, the tracking error will continue to accumulate, resulting in poor path tracking effect, which in turn leads to low care efficiency and consumes the battery of the care robot.

[0004] Therefore, in response to the above technical problems, a robotic care method based on spatiotemporal data is urgently needed. Summary of the invention

[0005] The purpose of the present invention is to provide a robot grooming method based on spatiotemporal data: to solve the technical problem that the grooming robots in the existing solutions have poor versatility in path tracking, especially when tracking complex paths, the tracking errors will continue to accumulate, resulting in poor path tracking effects, which in turn leads to low grooming efficiency.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A robot care method based on spatiotemporal data, the method comprising:

[0008] Obtaining the starting point and the end point of the care route of the area to be cared for, and determining the initial care path S according to the starting point and the end point of the care route and the distribution information of the target object to be cared for in the area to be cared for;

[0009] Discretize the initial path S to obtain several navigation points, and generate a discretized path S based on the navigation points. e : Preset multiple collection points on the initial grooming path S, and record the collection points as navigation points, and obtain the position coordinates P of each navigation point n (X n , Y n ), where n is a positive integer, and two adjacent navigation points P are calculated respectively. i (X i , Y i ), Pi+1 (X i+1 , Y i+1 ), where i = {1, 2…n}, and the discretized path S is determined based on the curvature δ e The path sampling interval value d is used to generate the discretized path S according to the path sampling interval value d. e = {P1 P2 P3…P n};

[0010] For the discretized path S e The path is segmented by a rough combination method to obtain several segmented paths, and path tracking is completed for each segmented path: the starting point coordinates (X start , Y start ), the coordinates of the end point (X dst , Y dst ), substitute the coordinates of the starting point and the ending point into the calculation formula to obtain the target heading angle θ dis , according to the target heading angle θ dis Control the robot to drive from the starting point to the end point on the segmented path, and determine whether the robot reaches the end point. If so, the segmented path completes path tracking. If not, calculate the error e(k) of the tracking curve between the current position of the robot and the nearest navigation point in the segmented path, and calculate the angle mapping value that the robot needs to adjust when completing path tracking based on the error e(k), and control the robot to adjust the direction of travel based on the mapping value to complete path tracking;

[0011] The monitoring information of the environment in the area to be managed after the interval T and the discretized path S after completing the path tracking e Mark as historical spatiotemporal data and store it.

[0012] Furthermore, the process of determining the initial care route S according to the starting point of the care route, the end point of the care route and the distribution information of the target objects to be cared for in the care area includes:

[0013] The straight line between the starting point and the end point of the grooming route is taken as the influence line;

[0014] Based on the distribution information of the target objects to be taken care of, obtaining the geometric shape corresponding to the distribution area of ​​the target objects to be taken care of and the position information of the distribution area of ​​the target objects to be taken care of;

[0015] Determine the distribution area of ​​the target to be managed that has an intersection with the influence line or is within a predetermined threshold distance from the influence line according to the position information of the distribution area of ​​the influence line and the target to be managed, and record the distribution area of ​​the target to be managed as the target management area;

[0016] The two points farthest apart in the geometric shape corresponding to the target grooming area are selected as reference points, and the starting point and the end point of the grooming route are connected with the reference points, and the connection is used as the initial grooming path S.

[0017] Furthermore, the discretized path S is determined based on the curvature δ e The path sampling interval value d is used to generate the discretized path S according to the path sampling interval value d. e The process includes:

[0018] Calculate two adjacent navigation points P i (X i , Y i ), P i+1 (X i+1 , Y i+1 )’s curvature δ:

[0019]

[0020] The path sampling interval d is determined according to the curvature δ and the preset curvature threshold δ0: if δ≤δ0, then d=|X i+1 -X i |, if δ>δ0, then d=|Y i+1- Y i |;

[0021] Generate a discretized path S according to the path sampling interval value d e = {P1 P2 P3…P n}.

[0022] Furthermore, for the discretized path S e The process of segmenting the path by rough combination to obtain several segmented paths includes:

[0023] Step 1: Set t as the coarse combination parameter, and for each t discretized path S e The navigation points are roughly combined according to the sampling order of the sampling interval value to obtain several rough combination groups;

[0024] Step 2: The first and second navigation points of each rough combination group form a reference vector m, and the remaining j-th point and the j-1-th navigation point form a vector m j , where j>2, calculate m respectively j The angle θ between j , where the angle θ j The calculation formula is as follows:

[0025]

[0026] Step 3: Set θ j With the threshold angle θth For comparison, if the θ of the jth point j Greater than the threshold angle θ th , then the navigation points arranged in front of the j-th point are subdivided into a group, and the navigation points arranged in front of the j-th point are repeated with steps 2 and 3 to complete the subdivision of the rough combination, and each subdivided group is recorded as a segmented path.

[0027] Furthermore, the coordinates of the starting point and the ending point are substituted into the calculation formula to obtain the target heading angle θ dis The calculation formula is as follows:

[0028]

[0029] Where, disY=(Y dst -Y start ), disX=(X dst -X start ).

[0030] Further, the process of calculating the error e(k) of the tracking curve between the current position of the robot and the nearest navigation point in the segmented path, and calculating the angle mapping value that needs to be adjusted when the robot completes the path tracking according to the error e(k) includes:

[0031] Get the robot's current position (X r , Y r ), the end point of the segment path (X dst , Y dst ), the error e(k) of the tracking curve between the robot's current position and the nearest navigation point in the segmented path is obtained according to the calculation formula, which is as follows:

[0032]

[0033] Wherein, k is the sampling interval time sequence number corresponding to the path sampling interval value d, k>2, and k is a positive integer;

[0034] According to the error e(k), the angle mapping value u(k) that the robot needs to adjust when completing path tracking is calculated:

[0035] u(k)=K P [e(k)-e(k-1)]+d(e(k)-2e(k-1)+e(k-2))

[0036] Among them, K P is a constant set by the system, e(k-1) is the error corresponding to the sampling time sequence number k-1, and e(k-2) is the error corresponding to the sampling time sequence number k-2.

[0037] Further, the method further includes: substituting the historical spatio-temporal data corresponding to the area to be maintained into the correlation formula to calculate the maintenance frequency reference index, and planning the maintenance frequency of the area to be maintained according to the maintenance frequency reference index, including the following steps:

[0038] The monitoring information of the environment includes the monitoring information of the vegetation;

[0039] Obtain the first height H of the vegetation according to the monitoring information of the vegetation in the area to be maintained after the interval time T in the historical spatio-temporal data corresponding to the area to be maintained, where the first height H is the growth height of the vegetation in the area to be maintained after the interval time T;

[0040] Obtain the average working current A when the area to be maintained was last maintained according to the historical spatio-temporal data, where the historical spatio-temporal data stores the working current change data when the area to be maintained was last maintained;

[0041] Obtain the soil humidity S and nutrient content Y of the area to be maintained monitored by the near-earth sensor during the interval time;

[0042] Substitute the first height H, average working current A, soil humidity S, and nutrient content Y into the correlation calculation formula to calculate the maintenance frequency reference index LMO. The correlation formula is as follows:

[0043]

[0044] Among them, G is the height of the cutter from the ground when the robot maintains the vegetation, the value of z is 2.72, C is the weed type coefficient, and C is taken according to the main vegetation type coefficient in the marked area;

[0045] Plan the maintenance frequency of the area to be maintained according to the maintenance frequency reference index: set a preset threshold, compare the maintenance frequency reference index with the preset threshold, and plan the maintenance frequency of the target area according to the comparison result. If the maintenance frequency reference index is greater than the preset threshold, calculate the ratio B of the maintenance frequency reference index to the preset threshold. If U < B < U + 1, plan the maintenance frequency of the area to be maintained to be once every T / U + 1, where U is a positive integer. If the maintenance frequency reference index is less than the preset threshold, calculate the ratio B of the maintenance frequency reference index to the preset threshold. If B < 1 / 3, plan the maintenance frequency of the area to be maintained to be once every 3T. If 1 / 3 < B < 1 / 2, plan the maintenance frequency of the area to be maintained to be once every 2T. If 1 / 2 < B < 1, plan the maintenance frequency of the area to be maintained to be once every T.

[0046] Furthermore, when the robot is performing maintenance according to the maintenance path, if it detects that an obstacle appears at a certain position on the maintenance path, the historical spatiotemporal data of the certain position is obtained. If the historical spatiotemporal data of the certain position does not have obstacle information, it is determined that the type of obstacle appearing at the position is a temporary obstacle. After the robot has completed maintenance of other target areas, it will try to maintain the target areas with temporary obstacles in the maintenance plan again at a certain time interval L. In particular, when the number of historical spatiotemporal data does not meet the preset number, the robot will directly go to the place with the temporary obstacle for targeted maintenance after completing maintenance of other areas for the first time. If the temporary obstacle continues to exist at this position, it will try to maintain it again at an interval of L, and repeat Q times. If the Qth attempt to maintain is successful, If the moment is after the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made; if the temporary obstacle continues to exist, the obstacle will be defined as a permanent obstacle and recorded in the historical spatiotemporal data until the next time there is no obstacle at this location, the obstacle type will be converted to a temporary obstacle; if an obstacle has appeared at this location in the historical spatiotemporal data, the number of times the obstacle blocks the grooming path at this location and the grooming frequency will be obtained through the historical spatiotemporal data, and the ratio E of the number of occlusions to the grooming frequency will be calculated, and an interval E*L will be set according to the occlusion ratio E, and a targeted grooming attempt will be made after the interval E*L, and this will be repeated Q times. If the moment of the Qth grooming attempt is after the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made.

[0047] Furthermore, when the number of historical spatiotemporal data meets the preset number, when the robot is performing grooming according to the grooming path, if a temporary obstacle that has appeared in the spatiotemporal data is detected at a certain position on the grooming path, the optimization algorithm will prioritize the time period closest to the temporary obstacle encountered this time, when there are fewer or even no temporary obstacles in the spatiotemporal data, to try grooming at that position in the first attempt based on the approximate time distribution of the temporary obstacles in a cycle; if there are still temporary obstacles in the first attempt, then the second attempt will be to try grooming at that position in a time period when there are fewer or even no temporary obstacles in the next spatiotemporal data, and this will be repeated Q times. If the time of the Qth grooming attempt is after the scheduled grooming plan time, only the Q-1th attempt plan will be executed and no further attempts will be made.

[0048] Compared with the existing solutions, the present invention achieves the following beneficial effects:

[0049] The present invention obtains the starting point and the end point of the care route of the area to be cared for, and determines the initial care path S according to the starting point and the end point of the care route and the distribution information of the target object to be cared for in the area to be cared for; discretizes and samples the initial care path S to obtain a plurality of navigation points, and generates a discretized path S based on the navigation points.e ; For the discretized path S e The path is segmented by a rough combination method to obtain a plurality of segmented paths, and path tracking is completed for each segmented path. The present invention can reduce the error of the grooming robot when performing path tracking, improve the path tracking effect, and further improve the grooming efficiency.

[0050] On the other hand, the present invention substitutes the historical spatiotemporal data corresponding to the area to be cared for into the correlation formula to calculate the care frequency reference index, and plans the care frequency of the area to be cared for according to the care frequency reference index. This enables more accurate and targeted cutting and care of the vegetation in the area to be cared for or precise cleaning of the rooms in the area to be cared for, thereby improving the care efficiency of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0052] Figure 1 It is a flow chart of a robot care method based on spatiotemporal data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] In addition, the described features, structures or characteristics may be combined in one or more example embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the example embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, steps, etc. may be adopted. In other cases, well-known structures, methods, implementations or operations are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0055] This embodiment provides a robot care method based on spatiotemporal data. Figure 1 is a flow chart of a robot care method based on spatiotemporal data according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0056] Step S101: obtaining the starting point and the end point of the care route of the area to be cared for, and determining the initial care path S according to the starting point and the end point of the care route and the distribution information of the target objects to be cared for in the area to be cared for.

[0057] Specifically, the straight line where the starting point and the end point of the care route are located is used as the influence line; based on the distribution information of the target to be cared for, the geometric shape corresponding to the distribution area of ​​the target to be cared for and the position information of the distribution area of ​​the target to be cared for are obtained; according to the influence line and the position information of the distribution area of ​​the target to be cared for, the distribution area of ​​the target to be cared for that has an intersection with the influence line or is within a predetermined threshold distance from the influence line is determined, and the distribution area of ​​the target to be cared for is recorded as the target care area; the two farthest points in the geometric shape corresponding to the target care area are selected as reference points, and the starting point and the end point of the care route are connected to the reference point, and the connection line is used as the initial care path S. It is worth noting that the geometric shape corresponding to the target area to be cared for can be regular or irregular, and can be obtained according to GIS (Geographic Information System: Geographic Information System).

[0058] Step S102: Discretize and sample the initial grooming path S to obtain a number of navigation points, and generate a discretized path S based on the navigation points. e :

[0059] Preset multiple collection points on the initial grooming path S, and record the collection points as navigation points, and obtain the position coordinates P of each navigation point n (X n , Y n ), where n is a positive integer, and two adjacent navigation points P are calculated respectively. i (X i , Y i ), P i+1 (X i+1 , Y i+1 ), where i = {1, 2…n}, and the discretized path S is determined based on the curvature δ e The path sampling interval value d is used to generate the discretized path S according to the path sampling interval value d. e = {P1 P2 P3…P n};

[0060] Specifically, the discretized path S is determined based on the curvature δ e The path sampling interval value d is used to generate the discretized path S according to the path sampling interval value d. e The process includes:

[0061] Calculate two adjacent navigation points Pi (X i , Y i ), P i+1 (X i+1 , Y i+1 )’s curvature δ:

[0062]

[0063] The path sampling interval d is determined according to the curvature δ and the preset curvature threshold δ0: if δ≤δ0, then d=|X i+1 -X i |, if δ>δ0, then d=|Y i+1 -Y i |;

[0064] Generate a discretized path S according to the path sampling interval value d e = {P1 P2 P3…P n}.

[0065] Step S103: Discretize the path S e The path is segmented by a coarse combination method to obtain several segmented paths, and path tracking is completed for each segmented path:

[0066] Get the starting point coordinates (X start , Y start ), the coordinates of the end point (X dst , Y dst ), substitute the coordinates of the starting point and the ending point into the calculation formula to obtain the target heading angle θ dis , according to the target heading angle θ dis Control the robot to drive from the starting point to the end point on the segmented path, and determine whether the robot has reached the end point. Specifically, determine whether the coordinates of the robot's position and the coordinates of the end point are consistent. If so, the robot has reached the end point, and the segmented path completes path tracking. If not, calculate the error e(k) of the tracking curve between the robot's current position and the nearest navigation point in the segmented path, and calculate the angle mapping value that the robot needs to adjust when completing path tracking based on the error e(k), and control the robot to adjust the travel direction based on the mapping value to complete path tracking.

[0067] Furthermore, for the discretized path S e The process of segmenting the path by rough combination to obtain several segmented paths includes:

[0068] Step 1: Set t as the coarse combination parameter, and for each t discretized path S e The navigation points are roughly combined according to the sampling order of the sampling interval value to obtain several rough combination groups;

[0069] Step 2: The first and second navigation points of each rough combination group form a reference vector m, and the remaining j-th point and the j-1-th navigation point form a vector m j , where j>2, calculate m respectively j The angle θ between j , where the angle θ j The calculation formula is as follows:

[0070]

[0071] Step 3: Set θ j With the threshold angle θ th For comparison, if the θ of the jth point j Greater than the threshold angle θ th , then the navigation points arranged in front of the j-th point are subdivided into a group, and the navigation points arranged in front of the j-th point are repeated with steps 2 and 3 to complete the subdivision of the rough combination, and each subdivided group is recorded as a segmented path.

[0072] Substitute the coordinates of the starting point and the ending point into the calculation formula to obtain the target heading angle θ dis The calculation formula is as follows:

[0073]

[0074] Where, disY=(Y dst -Y start ), disX=(X dst -X start ).

[0075] The process of calculating the error e(k) of the tracking curve between the current position of the robot and the nearest navigation point in the segmented path, and calculating the angle mapping value that needs to be adjusted when the robot completes the path tracking according to the error e(k) includes:

[0076] Get the robot's current position (X r , Y r ), the end point of the segment path (X dst , Y dst ), the error e(k) of the tracking curve between the robot's current position and the nearest navigation point in the segmented path is obtained according to the calculation formula, which is as follows:

[0077]

[0078] Wherein, k is the sampling interval time sequence number corresponding to the path sampling interval value d, k>2, and k is a positive integer;

[0079] According to the error e(k), the angle mapping value u(k) that the robot needs to adjust when completing path tracking is calculated:

[0080] u(k)=K P [e(k)-e(k-1)]+d(e(k)-2e(k-1)+e(k-2))

[0081] Among them, K P is a constant set by the system, e(k-1) is the error corresponding to the sampling time sequence number k-1, and e(k-2) is the error corresponding to the sampling time sequence number k-2.

[0082] Step S104: The monitoring information of the environment in the area to be managed after the interval T and the discretized path S after the path tracking is completed are recorded. e Mark as historical spatiotemporal data and store it.

[0083] In summary, the present invention obtains the starting point and the end point of the care route of the area to be cared for, and determines the initial care path S according to the starting point and the end point of the care route and the distribution information of the target object to be cared for in the area to be cared for; discretizes and samples the initial care path S to obtain a number of navigation points, and generates a discretized path S based on the navigation points. e ; For the discretized path S e The path is segmented by a rough combination method to obtain a plurality of segmented paths, and path tracking is completed for each segmented path. The present invention can reduce the error of the grooming robot when performing path tracking, improve the path tracking effect, and further improve the grooming efficiency.

[0084] In some embodiments, the method further includes: substituting the historical spatiotemporal data corresponding to the area to be maintained into the correlation formula to calculate a maintenance frequency reference index, and planning the maintenance frequency of the area to be maintained according to the maintenance frequency reference index, including the following steps:

[0085] Environmental monitoring information includes vegetation monitoring information;

[0086] Obtaining a primary height H of vegetation in the area to be managed after an interval T in the historical spatiotemporal data corresponding to the area to be managed, wherein the primary height H is the growth height of vegetation in the area to be managed after the interval T;

[0087] Obtaining the average working current A when the area to be treated was treated last time according to the historical spatiotemporal data, wherein the working current change data when the area to be treated was treated last time is stored in the historical spatiotemporal data;

[0088] The soil moisture S and nutrient content Y of the area to be managed are monitored by the near-ground sensor during the acquisition interval;

[0089] Substitute the primary height H, average working current A, soil humidity S, and nutrient content Y into the associated calculation formula to obtain the maintenance frequency reference index LMO. The associated formula is as follows:

[0090]

[0091] Among them, G is the height of the cutter of the robot from the ground when maintaining the vegetation, the value of z is 2.72, C is the weed type coefficient, and C is taken according to the main vegetation type coefficient in the marked area;

[0092] Maintain the area to be maintained according to the discretized path after completing path tracking.

[0093] It should be noted that when the maintenance robot is performing maintenance, in the area where the vegetation grows vigorously, the corresponding motor current is large, and the historical spatio-temporal data records that the robot consumes a large amount of electricity in this area. Combining with lidar, cameras, and infrared sensors to detect that the grass height is high, it is recorded in the spatio-temporal database that there is more grass in this area, that is, the frequency of maintenance required is higher, and the maintenance frequency needs to be higher than other places. After the robot completes autonomous operation, the user manually controls the robot to maintain the area more times, and the corresponding area naturally requires the robot to perform maintenance more times.

[0094] Plan the maintenance frequency of the area to be maintained according to the maintenance frequency reference index: set a preset threshold, compare the maintenance frequency reference index with the preset threshold, and plan the maintenance frequency of the target area according to the comparison result. If the maintenance frequency reference index is greater than the preset threshold, calculate the ratio B of the maintenance frequency reference index to the preset threshold. If U < B < U + 1, plan the maintenance frequency of the area to be maintained to be once every T / U + 1 intervals, where U is a positive integer. If the maintenance frequency reference index is less than the preset threshold, calculate the ratio B of the maintenance frequency reference index to the preset threshold. If B < 1 / 3, plan the maintenance frequency of the area to be maintained to be once every 3T intervals. If 1 / 3 < B < 1 / 2, plan the maintenance frequency of the area to be maintained to be once every 2T intervals. If 1 / 2 < B < 1, plan the maintenance frequency of the area to be maintained to be once every T intervals.

[0095] In summary, the present invention substitutes the historical spatio-temporal data corresponding to the area to be maintained into the associated formula to obtain the maintenance frequency reference index, and plans the maintenance frequency of the area to be maintained according to the maintenance frequency reference index, which can perform more precise and targeted cutting and maintenance on the vegetation in the area to be maintained, and improve the maintenance efficiency of the robot.

[0096] In some embodiments, the historical spatiotemporal data is updated at time intervals T, and the historical spatiotemporal data with the longest storage time is eliminated. Specifically, the historical spatiotemporal data collected at F time intervals T form a historical spatiotemporal data group, and the historical spatiotemporal data with the longest storage time in the historical spatiotemporal data group is deleted. Here, an example can be given, such as starting to take care of the first day at 10 o'clock in the morning, starting to take care of the second day at 11 o'clock in the morning, starting to take care of the third day at 12 o'clock in the morning, and so on to 10 o'clock on the evening of the 13th day. 13 days cover all working time periods in order to update the complete historical spatiotemporal data, so that when taking care of the target area, the decision-making is more accurate.

[0097] In some embodiments, based on the robot's cleaning path information in historical spatiotemporal data, when the robot detects an obstacle ahead on the cleaning path, if the obstacle has appeared in the cleaning path information, the occlusion ratio of the obstacle blocking the cleaning path is calculated, and it is determined whether the occlusion ratio exceeds 20%. If so, the obstacle is bypassed to clean other target areas.

[0098] In some embodiments, if the obstacle has never appeared in the historical spatiotemporal data, the obstacle type will be determined to be a temporary obstacle. After taking care of other target areas, the robot will try to take care of the target area with temporary obstacles in the current care plan again at a certain time interval. For example, if the historical spatiotemporal data is insufficient, the mowing robot will try 3 times. The first time, after taking care of other areas, it will directly go to the place with temporary obstacles to try to take care of them. If there are still obstacles, it will try to take care of them again after 1 day. If there are still obstacles, it will try to take care of them again after 3 days. If there are still obstacles after 3 attempts, it will not try again, and the obstacle will be determined as a permanent obstacle and recorded in the historical spatiotemporal data.

[0099] Furthermore, when the robot is performing maintenance according to the maintenance path, if it detects that an obstacle appears at a certain position on the maintenance path, the historical spatiotemporal data of the certain position is obtained. If the historical spatiotemporal data of the certain position does not have obstacle information, it is determined that the type of obstacle appearing at the position is a temporary obstacle. After the robot has completed maintenance of other target areas, it will try to maintain the target areas with temporary obstacles in the maintenance plan again at a certain time interval L. In particular, when the number of historical spatiotemporal data does not meet the preset number, the robot will directly go to the place with the temporary obstacle for targeted maintenance after completing maintenance of other areas for the first time. If the temporary obstacle continues to exist at this position, it will try to maintain it again at an interval of L, and repeat Q times. If the Qth attempt to maintain is successful, If the moment is after the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made; if the temporary obstacle continues to exist, the obstacle will be defined as a permanent obstacle and recorded in the historical spatiotemporal data until the next time there is no obstacle at this location, the obstacle type will be converted to a temporary obstacle; if an obstacle has appeared at this location in the historical spatiotemporal data, the number of times the obstacle blocks the grooming path at this location and the grooming frequency will be obtained through the historical spatiotemporal data, and the ratio E of the number of occlusions to the grooming frequency will be calculated, and an interval E*L will be set according to the occlusion ratio E, and a targeted grooming attempt will be made after the interval E*L, and this will be repeated Q times. If the moment of the Qth grooming attempt is after the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made.

[0100] Furthermore, when the number of historical spatiotemporal data meets the preset number, when the robot is performing grooming according to the grooming path, if a temporary obstacle that has appeared in the spatiotemporal data is detected at a certain position on the grooming path, the optimization algorithm will prioritize the time period closest to the temporary obstacle encountered this time, when there are fewer or even no temporary obstacles in the spatiotemporal data, to try grooming at that position in the first attempt based on the approximate time distribution of the temporary obstacles in a cycle; if there are still temporary obstacles in the first attempt, then the second attempt will be to try grooming at that position in a time period when there are fewer or even no temporary obstacles in the next spatiotemporal data, and this will be repeated Q times. If the time of the Qth grooming attempt is after the scheduled grooming plan time, only the Q-1th attempt plan will be executed and no further attempts will be made.

[0101] Those skilled in the art should understand that the technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0102] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.

[0103] Therefore, the scope of protection of this patent application should be based on the attached claims.

[0104] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A robot grooming method based on spatiotemporal data, characterized in that: The method includes: Obtaining the starting point of the grooming route and the ending point of the grooming route for the area to be groomed, and determining the initial grooming path S according to the starting point of the grooming route, the ending point of the grooming route, and the distribution information of the objects to be groomed in the area to be groomed; Discretize the initial path S to obtain several navigation points, and generate a discretized path S based on the navigation points. e : Preset multiple collection points on the initial grooming path S, and record the collection points as navigation points, and obtain the position coordinates P of each navigation point n (X n , Y n ), where n is a positive integer, and two adjacent navigation points P are calculated respectively. i (X i , Y i ), P i+1 (X i+1 , Y i+1 ), where i = {1, 2…n}, and the discretized path S is determined based on the curvature δ e The path sampling interval value d is used to generate the discretized path S according to the path sampling interval value d. e = {P1 P2 P3…P n }: Calculate two adjacent navigation points P i (X i , Y i ), P i+1 (X i+1 , Y i+1 )’s curvature δ: The path sampling interval d is determined according to the curvature δ and the preset curvature threshold δ0: if δ≤δ0, then d=|X i+1 -X i |, if δ>δ0, then d=|Y i+1 -Y i |; Generate a discretized path S according to the path sampling interval value d e = {P1 P2 P3…P n }; For the discretized path S e The path is segmented by a rough combination method to obtain several segmented paths, and path tracking is completed for each segmented path: the starting point coordinates (X start , Y start ), the coordinates of the end point (X dst , Y dst ), substitute the coordinates of the starting point and the ending point into the calculation formula to obtain the target heading angle θ dis , according to the target heading angle θ dis Control the robot to drive from the starting point to the end point on the segmented path, and determine whether the robot reaches the end point. If so, the segmented path completes path tracking. If not, calculate the error e(k) of the tracking curve between the current position of the robot and the nearest navigation point in the segmented path, and calculate the angle mapping value that the robot needs to adjust when completing path tracking based on the error e(k), and control the robot to adjust the direction of travel based on the mapping value to complete path tracking; The monitoring information of the environment in the area to be managed after the interval T and the discretized path S after completing the path tracking e Mark as historical spatiotemporal data and store it; Substituting the historical spatio-temporal data corresponding to the area to be groomed into the correlation formula to calculate the grooming frequency reference index, and planning the grooming frequency of the area to be groomed based on the grooming frequency reference index, including the following steps: The monitoring information of the environment includes the monitoring information of vegetation; Obtaining the primary height H of the vegetation based on the monitoring information of the vegetation in the area to be groomed after an interval time T in the historical spatio-temporal data corresponding to the area to be groomed, where the primary height H is the growth height of the vegetation in the area to be groomed after the interval time T; Obtaining the average working current A when grooming the area to be groomed last time according to the historical spatio-temporal data, where the historical spatio-temporal data stores the working current change data when grooming the area to be groomed last time; Obtaining the soil humidity S and nutrient content Y of the area to be groomed monitored by the near-earth sensor during the interval time; Substituting the primary height H, the average working current A, the soil humidity S, and the nutrient content Y into the correlation calculation formula to calculate the grooming frequency reference index LMO, and the correlation formula is as follows: Wherein, G is the height of the cutter of the robot when grooming the vegetation from the ground, the value of z is 2.72, C is the weed type coefficient, and C is valued according to the main vegetation type coefficient in the marked area; Planning the grooming frequency of the area to be groomed based on the grooming frequency reference index: setting a preset threshold, comparing the grooming frequency reference index with the preset threshold, and planning the grooming frequency of the target area according to the comparison result. If the grooming frequency reference index is greater than the preset threshold, calculating the ratio B of the grooming frequency reference index to the preset threshold. If U < B < U + 1, planning the grooming frequency of the area to be groomed to be once every T / U + 1, where U is a positive integer. If the grooming frequency reference index is less than the preset threshold, calculating the ratio B of the grooming frequency reference index to the preset threshold. If B < 1 / 3, planning the grooming frequency of the area to be groomed to be once every 3T. If 1 / 3 < B < 1 / 2, planning the grooming frequency of the area to be groomed to be once every 2T. If 1 / 2 < B < 1, planning the grooming frequency of the area to be groomed to be once every T.

2. The method according to claim 1, characterized in that The process of determining the initial grooming path S according to the starting point of the grooming route, the ending point of the grooming route, and the distribution information of the objects to be groomed in the area to be groomed includes: Taking the straight line where the starting point of the grooming route and the ending point of the grooming route are located as the influencing line; Obtaining the geometric shape of the distribution area of the objects to be groomed and the position information of the distribution area of the objects to be groomed based on the distribution information of the objects to be groomed; Determining the distribution area of the objects to be groomed that has an intersection with the influencing line or is within a predetermined threshold distance from the influencing line according to the influencing line and the position information of the distribution area of the objects to be groomed, and recording the distribution area of the objects to be groomed as the target grooming area; The two points farthest apart in the geometric shape corresponding to the target grooming area are selected as reference points, and the starting point and the end point of the grooming route are connected with the reference points, and the connection is used as the initial grooming path S.

3. The method according to claim 1, characterized in that For the discretized path S e The process of segmenting the path by rough combination to obtain several segmented paths includes: Step 1: Set t as the coarse combination parameter, and for each t discretized path S e The navigation points are roughly combined according to the sampling order of the sampling interval value to obtain several rough combination groups; Step 2: The first and second navigation points of each rough combination group form a reference vector m, and the remaining j-th point and the j-1-th navigation point form a vector m j , where j>2, calculate m respectively j The angle θ between j , where the angle θ j The calculation formula is as follows: Step 3: Set θ j With the threshold angle θ th For comparison, if the θ of the jth point j Greater than the threshold angle θ th , then the navigation points arranged in front of the j-th point are subdivided into a group, and the navigation points arranged in front of the j-th point are repeated with steps 2 and 3 to complete the subdivision of the rough combination, and each subdivided group is recorded as a segmented path.

4. The method according to claim 1, characterized in that: Substitute the coordinates of the starting point and the ending point into the calculation formula to obtain the target heading angle θ dis The calculation formula is as follows: Where, disY=(Y dst -Y start ), disX=(X dst -X start ).

5. The method according to claim 1, characterized in that The process of calculating the error e(k) of the tracking curve between the current position of the robot and the nearest navigation point in the segmented path, and calculating the angle mapping value that needs to be adjusted when the robot completes the path tracking according to the error e(k) includes: Get the robot's current position (X r , Y r ), the end point of the segment path (X dst , Y dst ), the error e(k) of the tracking curve between the robot's current position and the nearest navigation point in the segmented path is obtained according to the calculation formula, which is as follows: Wherein, k is the sampling interval time sequence number corresponding to the path sampling interval value d, k>2, and k is a positive integer; According to the error e(k), the angle mapping value u(k) that the robot needs to adjust when completing path tracking is calculated: u(k)=K P [e(k)-e(k-1)]+d(e(k)-2e(k-1)+e(k-2)) Among them, K P is a constant set by the system, e(k-1) is the error corresponding to the sampling time sequence number k-1, and e(k-2) is the error corresponding to the sampling time sequence number k-2.

6. The method according to claim 1, characterized in that The method further includes: When the robot is cleaning according to the cleaning path, if it detects an obstacle at a certain position on the cleaning path, it obtains the historical spatiotemporal data of the certain position. If there is no obstacle information in the historical spatiotemporal data of the certain position, it determines that the obstacle type at the position is a temporary obstacle. After cleaning other target areas, the robot tries to clean the target areas with temporary obstacles in the cleaning plan again at a certain time interval L. Among them, when the number of historical spatiotemporal data does not meet the preset number, the robot directly goes to the place with temporary obstacles for targeted cleaning after cleaning other areas for the first time. If the temporary obstacle continues to exist at this position, it tries to clean it again at a time interval L, and repeats Q times. If the time of the Qth cleaning attempt is After the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made; if the temporary obstacle continues to exist, the obstacle will be defined as a permanent obstacle and recorded in the historical space-time data until the next time there is no obstacle at this location, the obstacle type will be converted to a temporary obstacle; if an obstacle has appeared at this location in the historical space-time data, the number of times the obstacle blocks the grooming path at this location and the grooming frequency will be obtained through the historical space-time data, and the ratio E of the number of occlusions to the grooming frequency will be calculated, and an interval E*L will be set according to the occlusion ratio E, and a targeted grooming attempt will be made at the interval E*L, and this will be repeated Q times. If the moment of the Qth grooming attempt is after the scheduled grooming time, only the Q-1th attempt plan will be executed and no further attempts will be made.

7. The method according to claim 6, characterized in that The method also includes: when the number of historical spatiotemporal data meets a preset number, when the robot performs grooming according to the grooming path, if a temporary obstacle that has appeared in the spatiotemporal data is detected at a certain position on the grooming path, a preferential algorithm is used to select, based on the approximate time distribution of the appearance of temporary obstacles in a cycle time, a time period closest to the temporary obstacle encountered this time, where temporary obstacles appear less frequently or even no temporary obstacles appear in the spatiotemporal data, to try grooming at the position for the first attempt; if there is still a temporary obstacle in the first attempt, then a second attempt is made to go to the position in a time period where temporary obstacles appear less frequently or even no temporary obstacles appear in the next spatiotemporal data to try grooming, and this is repeated Q times. If the time of the Qth grooming attempt is after the predetermined grooming planning time, only the Q-1th attempt plan is executed and no further attempts are made.

Images