A multi-inspection robot route collaborative control system

Abstract

The present invention belongs to the technical field of patrol robots, and specifically is a multi-patrol robot route collaborative control system, comprising a state monitoring module, a route planning module, a route optimization module, a route adjustment module and a battery life verification module; the state monitoring module is used to record the state information of each patrol robot in the multi-patrol robot group, and by deleting repeated patrol points, it is possible to avoid repeated patrols and collisions that may exist in the patrol process during the route construction stage, and then obtain a third patrol route. At this point, the patrol robot performs a battery life verification according to the third patrol route. If the verification passes, the patrol robot performs the patrol task according to the third patrol route, and can patrol all points to be inspected to avoid missed inspections. At the same time, based on the route adjustment module, repeated patrol points are deleted, which can avoid collision problems in the patrol process, and is also beneficial to reducing the patrol route length of the patrol robot, thereby optimizing the patrol efficiency of the patrol robot.

Description

Technical Field

[0001] The invention belongs to the technical field of inspection robots, and in particular is a multi-inspection robot route collaborative control system. Background Art

[0002] With the advancement of technology, multi-robot patrol systems have been widely used in many industries such as security monitoring and environmental monitoring. Such systems are usually composed of multiple autonomous mobile robots that work together to complete patrol tasks in a specific area.

[0003] At present, when multiple inspection robots perform inspection tasks, due to uneven distribution of system tasks and unreasonable inspection route design, it may lead to waste of resources or low inspection efficiency. In response to the above problems, the existing patent with publication number CN118376251A discloses a route collaborative control method and system for multiple inspection robots. The method includes: constructing a state information set of a multi-inspection robot group, obtaining M inspection points in a preset inspection area, combining the initial position of the robot for route planning, identifying repeated route combinations, extracting the first route and optimizing the route connection with the remaining routes, and then performing a robot endurance qualification check. If it passes, start the matching robot for dynamic obstacle detection, and finally, perform route collaborative optimization control. This application can solve the problem that the existing technology cannot accurately plan routes due to insufficient collaborative control capabilities, resulting in low inspection efficiency, further affecting the comprehensiveness and timeliness of the inspection, and realize the collaborative optimization control of the routes of multiple matching robots, thereby improving the inspection efficiency and safety.

[0004] In the above-mentioned prior art, the method of controlling multiple inspection robots can be summarized as follows: first, multiple first routes are constructed based on all the inspection points to be inspected, and then the same inspection robots assigned to different first routes are counted, and the first routes belonging to the same inspection robot are merged and connected to form a first optimized route. Finally, the endurance check is performed on the first optimized route. If it passes, the inspection task is performed according to the first optimized route. However, since the positions of the inspection robots in the factory are fixed, there may be a same inspection point at the same distance from two inspection robots. When constructing the first route, multiple inspection points are assigned to one inspection robot in a range manner. This may result in an inspection point existing on two first optimized routes at the same time, which will increase the collision risk of the inspection robots. At the same time, it will also cause the inspection routes of multiple inspection robots to become longer, resulting in problems such as inability to return and low efficiency.

[0005] To this end, the present invention provides a multi-inspection robot route collaborative control system. Summary of the invention

[0006] In order to make up for the deficiencies of the prior art, at least one technical problem raised in the background technology is solved.

[0007] The technical solution adopted by the present invention to solve the technical problem is: a multi-inspection robot route collaborative control system described in the present invention comprises:

[0008] A status monitoring module is used to record the status information of each inspection robot in the multi-inspection robot group, wherein the status information includes the inspection robot number, remaining power and initial position;

[0009] A route planning module is used to obtain N first inspection routes, and the N first inspection routes cover all inspection points;

[0010] A route optimization module is used to identify the points to be inspected with repeated inspection robot identifications in the N first inspection routes, and connect the points to be inspected end to end to obtain M second inspection routes;

[0011] A route adjustment module, used for marking duplicate inspection points existing in at least two second inspection routes, and deleting duplicate inspection points in any second inspection route, to obtain H third inspection routes;

[0012] The battery life verification module is used to perform battery life verification based on the remaining power of the inspection robot and environmental conditions. If the verification passes, the inspection task will be performed using H third inspection routes and H matching inspection robots.

[0013] Preferably, the method for obtaining the first inspection route is:

[0014] Get all the points to be inspected;

[0015] Generate N first inspection routes based on the inspection robots closest to the inspection points;

[0016] The N first inspection routes are each set with a corresponding inspection robot number, and N is an integer greater than 0.

[0017] Preferably, the method for acquiring the second inspection route is:

[0018] Make a table of N first inspection routes and the corresponding N inspection robots;

[0019] Count the number of inspection points to be inspected that are repeatedly identified by the inspection robot in the N first inspection routes;

[0020] The inspection points corresponding to the repeated inspection robot identifications are merged into M inspection sets;

[0021] Edit the routes of the M sets to be inspected according to the electronic map, and obtain M second inspection routes by connecting them end to end;

[0022] The M sets to be inspected cover all inspection points, and M is an integer greater than 0, and .

[0023] Preferably, the method of editing the routes of the M sets to be inspected according to the electronic map and obtaining the M second inspection routes by connecting them end to end is:

[0024] Taking the location of the inspection robot as the starting point, construct F routes to be determined with all the points to be inspected included in the second inspection route;

[0025] Analyzing the road condition information and the length of the route to be determined to obtain the fitness of the route to be determined, including:

[0026] Obtain historical data, including but not limited to GPS records and sensor readings from historical inspection tasks of the inspection robot;

[0027] Data preprocessing, including missing value processing, outlier monitoring and data standardization;

[0028] Build prediction models and fitness assessment models;

[0029] Model training and validation: divide historical data into training set, validation set and test set, and use cross-validation method to evaluate model performance;

[0030] The characteristics of the route to be determined are input into the evaluation model to obtain the fitness of the route to be determined.

[0031] The characteristics of the route to be determined include road condition information and route length, wherein the road condition information includes the length and slope of a section with a slope in the route to be determined;

[0032] Compare the fitness of the F routes to be determined, and select the route to be determined with the largest fitness as the second inspection route.

[0033] Preferably, the method for acquiring the third inspection route is:

[0034] Obtain M second inspection routes;

[0035] Mark the repeated inspection points existing in the M second inspection routes;

[0036] Calculate the fitness of the second inspection route with the same repeated inspection point;

[0037] The repeated inspection points are retained in the second inspection route corresponding to the maximum fitness value, and the repeated inspection points are deleted from other second inspection routes;

[0038] Get H third inspection routes, where H is an integer greater than 0, and .

[0039] Preferably, the method for performing endurance verification according to the remaining power of the inspection robot and environmental conditions is:

[0040] Obtain the third inspection route and the inspection robot matching it;

[0041] Obtain the remaining power of the inspection robot based on the status monitoring module;

[0042] A prediction model for the remaining battery power and mileage of the inspection robot is built based on historical data, including:

[0043] Data collection: collects the historical driving data of the inspection robot, including the remaining power at the beginning of the task, the power consumption and mileage after the task is completed, the route characteristics and environmental conditions;

[0044] Data preprocessing, including missing value processing, outlier monitoring and data standardization;

[0045] Build predictive models based on machine learning algorithms;

[0046] Model training and verification: historical data is divided into training set, verification set and test set for model training and verification;

[0047] The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven;

[0048] Compare the mileage prediction result with the length of the third inspection route:

[0049] When the predicted mileage is greater than the sum of the length of the third inspection route and the length of the return route, the endurance check passes; otherwise, it fails.

[0050] The length of the return route is a line connecting the end point of the third inspection route and the initial position of the inspection robot.

[0051] Preferably, when the endurance check of the inspection robot fails, the third inspection route of the inspection robot is adjusted based on the route adjustment module to obtain a fourth inspection route.

[0052] Preferably, the method for acquiring the fourth inspection route is:

[0053] Obtain the remaining power of the third inspection route that failed the endurance check and the matching inspection robot;

[0054] The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven;

[0055] Adjusting the third inspection route according to the mileage prediction result, including but not limited to deleting at least one to-be-inspected point, to obtain a fourth inspection route;

[0056] The fourth inspection route satisfies that the sum of the length of the fourth inspection route and the length of the return route is less than or equal to the predicted drivable mileage result.

[0057] Preferably, the route adjustment module is further used to adjust the deleted to-be-inspected point to any unadjusted third inspection route;

[0058] The method for adjusting the deleted inspection point to the unadjusted third inspection route is:

[0059] Obtain all inspection robots adjacent to the deleted inspection points and record them as the inspection robot group to be adjusted;

[0060] Obtain the predicted drivable mileage of all inspection robots in the inspection robot group to be adjusted, and the sum of the length of the third inspection route and the length of the return route, and calculate the difference between the two, which is recorded as the compensable mileage;

[0061] The third inspection route corresponding to the inspection robot with the largest compensable mileage is selected, and the deleted points to be inspected are added to obtain the fifth inspection route.

[0062] Preferably, the machine learning algorithm used in the prediction model includes one or more combinations of a random forest algorithm, a support vector machine algorithm or a neural network algorithm.

[0063] The beneficial effects of the present invention are as follows:

[0064] A multi-inspection robot route collaborative control system described in the present invention deletes duplicate inspection points to avoid duplicate inspections and collisions that may exist in the inspection process during the route construction stage, and then obtains a third inspection route. The inspection robot now performs an endurance check according to the third inspection route. If the check passes, the inspection robot performs the inspection task according to the third inspection route, and can inspect all points to be inspected to avoid missed inspections. At the same time, duplicate inspection points are deleted based on the route adjustment module, which can avoid collision problems in the inspection process, and is also beneficial to reducing the inspection route length of the inspection robot, thereby optimizing the inspection efficiency of the inspection robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The present invention will be further described below in conjunction with the accompanying drawings.

[0066] Figure 1 is a flow chart of the present invention;

[0067] Figure 2 It is a schematic diagram of the inspection route of the inspection robot in the present invention. DETAILED DESCRIPTION

[0068] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0069] like Figure 1 to Figure 2 As shown, a multi-inspection robot route collaborative control system described in an embodiment of the present invention includes a state monitoring module, a route planning module, a route optimization module, a route adjustment module and a battery life verification module; the state monitoring module is used to record the state information of each inspection robot in the multi-inspection robot group, and the state information includes the inspection robot number, the remaining power and the initial position; the route planning module is used to obtain N first inspection routes, and the N first inspection routes cover all inspection points; the route optimization module is used to identify the inspection points to be inspected with repeated inspection robot marks in the N first inspection routes, and connect the inspection points to be inspected end to end to obtain M second inspection routes; the route adjustment module is used to mark the repeated inspection points that exist in at least two second inspection routes, and delete the repeated inspection points in any second inspection route to obtain H third inspection routes; the battery life verification module is used to perform battery life verification according to the remaining power of the inspection robot and the environmental conditions. If it passes, the H third inspection routes and the matching H inspection robots are used to perform the inspection task.

[0070] In the above-mentioned prior art, the method of controlling multiple inspection robots can be summarized as follows: first, multiple first routes are constructed based on all the inspection points to be inspected, and then the same inspection robots assigned to different first routes are counted, and the first routes belonging to the same inspection robot are merged and connected to form a first optimized route. Finally, the endurance check is performed on the first optimized route. If it passes, the inspection task is performed according to the first optimized route. However, since the positions of the inspection robots in the factory are fixed, there may be a same inspection point at the same distance from two inspection robots. When constructing the first route, multiple inspection points are assigned to one inspection robot in a range manner. This may result in an inspection point existing on two first optimized routes at the same time, which will increase the collision risk of the inspection robots. At the same time, it will also cause the inspection routes of multiple inspection robots to become longer, resulting in problems such as inability to return and low efficiency.

[0071] In one embodiment of the present invention, firstly, based on all the inspection points to be inspected, a first inspection route covering all the inspection points is obtained, and the first inspection route is a route from the initial position of the inspection robot to any inspection point. Then, a route optimization module is used to identify the inspection points to be inspected in the first inspection route that belong to the same inspection robot, and these inspection points to be inspected are connected to obtain a second inspection route belonging to the same inspection robot. The second inspection route is a route from the initial position of the inspection robot to the first inspection point, the second inspection point...the Pth inspection point, wherein the Pth inspection point is the last inspection point in the second inspection route. Then, the inspection robot performs a return mission, and then, According to the constructed second inspection route, the route adjustment module is used to identify the possible repeated inspection points in the second inspection route. By deleting the repeated inspection points, the possible repeated inspections and collisions in the inspection process can be avoided during the route construction stage. Subsequently, the third inspection route is obtained. At this point, the inspection robot performs an endurance check according to the third inspection route. If the check passes, the inspection robot performs the inspection task according to the third inspection route, and can inspect all the points to be inspected to avoid missed inspections. At the same time, deleting the repeated inspection points based on the route adjustment module can avoid collision problems in the inspection process, and is also beneficial to reducing the inspection route length of the inspection robot, thereby optimizing the inspection efficiency of the inspection robot.

[0072] In one embodiment, the method for obtaining the first inspection route is:

[0073] Get all the points to be inspected;

[0074] Generate N first inspection routes based on the inspection robots closest to the inspection points;

[0075] The N first inspection routes are each set with a corresponding inspection robot number, and N is an integer greater than 0.

[0076] Based on the above, in one embodiment of the present invention, the method for obtaining the first inspection route is to obtain all points to be inspected, and then obtain the first inspection route with the inspection robot closest to the inspection point. For example, when there are four inspection points A1, A2, A3, and A4, assuming that A1 and A2 are closest to the inspection robot X1, and A3 and A4 are closest to the inspection robot X2, four first inspection routes X1-A1, X1-A2, X2-A3, and X2-A4 are generated, and the inspection robots are numbered X1 and X2. As mentioned above, by obtaining the first inspection route, full coverage of all points to be inspected can be achieved, avoiding omissions of points to be inspected.

[0077] In one embodiment, the method for acquiring the second inspection route is:

[0078] Make a table of N first inspection routes and the corresponding N inspection robots;

[0079] Count the number of inspection points to be inspected that are repeatedly identified by the inspection robot in the N first inspection routes;

[0080] The inspection points corresponding to the repeated inspection robot identifications are merged into M inspection sets;

[0081] Edit the routes of the M sets to be inspected according to the electronic map, and obtain M second inspection routes by connecting them end to end;

[0082] The M sets to be inspected cover all inspection points, and M is an integer greater than 0, and .

[0083] Based on the above, when the first inspection route is constructed, full coverage of all inspection points to be inspected can be achieved to avoid omission of inspection points. Secondly, according to the obtained first inspection route, based on the summary of multiple first inspection routes belonging to the same inspection robot, the second inspection route is obtained by identifying multiple first inspection routes belonging to the same inspection robot, that is, identifying the number of inspection points to be inspected belonging to the same inspection robot, and editing all inspection points to be inspected with the same inspection robot identification in an end-to-end manner to obtain the second inspection route. For example, when there are four inspection points A1, A2, A3, and A4, assuming that A1 and A2 are closest to the inspection robot X1, and A3 and A4 are closest to the inspection robot X2, then X1-A1 is generated. , X1-A2, X2-A3, and X2-A4, and after editing, the two second inspection routes X1-A1-A2 and X2-A3-A4 are obtained according to the four first inspection routes. Through the above operations, the first inspection routes can be optimized. It is worth noting that the construction logic of the first inspection route is from the initial position of the inspection robot to the inspection point, and then returns from the inspection point to the initial position, and then repeats the inspection task according to another first inspection route. The optimized second inspection route avoids the problems of extended inspection time and inspection route caused by repeated returns of the inspection robot, improves the inspection efficiency, and is conducive to making full use of the remaining power of the inspection robot to perform inspection tasks, avoiding the inefficiency caused by frequent power replenishment of the inspection robot.

[0084] In one embodiment, the method of editing the routes of the M sets to be inspected according to the electronic map and obtaining the M second inspection routes by connecting them end to end is as follows:

[0085] Taking the location of the inspection robot as the starting point, construct F routes to be determined with all the points to be inspected included in the second inspection route;

[0086] Analyzing the road condition information and the length of the route to be determined to obtain the fitness of the route to be determined, including:

[0087] Obtain historical data, including but not limited to GPS records and sensor readings from historical inspection tasks of the inspection robot;

[0088] Data preprocessing, including missing value processing, outlier monitoring and data standardization;

[0089] Build prediction models and fitness assessment models;

[0090] Model training and validation: divide historical data into training set, validation set and test set, and use cross-validation method to evaluate model performance;

[0091] The characteristics of the route to be determined are input into the evaluation model to obtain the fitness of the route to be determined.

[0092] The characteristics of the route to be determined include road condition information and route length, wherein the road condition information includes the length and slope of a section with a slope in the route to be determined;

[0093] Compare the fitness of the F routes to be determined, and select the route to be determined with the largest fitness as the second inspection route.

[0094] When constructing the second inspection route, since the routes that the inspection robot can travel in the factory are diverse and multi-selective, the inspection robot may not have only one route from the initial position to the point to be inspected. Therefore, when constructing the second inspection route of the inspection robot, it is necessary to consider whether the selection of different inspection routes has different impacts on the inspection robot. Specifically, in one embodiment of the present invention, fitness is used to evaluate the impact of selecting different inspection routes on the inspection robot. The inspection route is the route to be determined mentioned above. The fitness of the route to be determined can be obtained by analyzing the driving road condition information and the route length of the route. The fitness of the route to be determined is calculated by an evaluation model. By inputting the characteristics of the route to be determined into the evaluation model, the evaluation model can calculate and output the fitness of the route to be determined based on the characteristics of the route to be determined. That is, when the fitness of the route to be determined is higher, it means that when the inspection robot performs the inspection task according to the route to be determined with higher fitness, the inspection efficiency is higher and the endurance is more guaranteed, which reduces the situation where the inspection robot is difficult to complete or fails in the middle of the inspection task diagram. It is worth noting that in one embodiment of the present invention, the characteristics of the route to be determined include road condition information and route length, and the road condition information specifically refers to the length of the section with a slope in the route to be determined and the corresponding slope. In simple terms, the greater the total length of the section with a slope in the route to be determined and the greater the corresponding slope, the more difficult it is for the inspection robot to pass smoothly, the greater the difficulty of the inspection, and the lower the inspection efficiency. Based on this, by obtaining the route to be determined with the highest fitness as the second inspection route, the inspection efficiency of the inspection robot is further improved, and the problem of low inspection efficiency of the inspection robot is avoided.

[0095] In one embodiment, the method for acquiring the third inspection route is:

[0096] Obtain M second inspection routes;

[0097] Mark the repeated inspection points existing in the M second inspection routes;

[0098] Calculate the fitness of the second inspection route with the same repeated inspection point;

[0099] The repeated inspection points are retained in the second inspection route corresponding to the maximum fitness value, and the repeated inspection points are deleted from other second inspection routes;

[0100] Get H third inspection routes, where H is an integer greater than 0, and .

[0101] Based on the above, after constructing the second inspection route, it is also necessary to identify the repeated inspection points that may exist in the second inspection route. By deleting the repeated inspection points, the risk of collision between multiple inspection robots during the inspection process can be reduced. In one embodiment of the present invention, the method for obtaining the third inspection route is to obtain M second inspection routes, and then mark the repeated inspection points in the M second inspection routes. The fitness of multiple second inspection routes with the same repeated inspection points is obtained through the above-mentioned evaluation model, which can be used to evaluate the inspection difficulty of multiple second inspection routes with the same repeated inspection points, retain the repeated inspection points in the second inspection route corresponding to the maximum fitness value, and delete the repeated inspection points from other second inspection routes, which is conducive to balancing the inspection difficulty of multiple second inspection routes and promoting the smooth progress of the inspection task.

[0102] In one embodiment, the method for performing endurance verification based on the remaining power of the inspection robot and environmental conditions is:

[0103] Obtain the third inspection route and the inspection robot matching it;

[0104] Obtain the remaining power of the inspection robot based on the status monitoring module;

[0105] A prediction model for the remaining battery power and mileage of the inspection robot is built based on historical data, including:

[0106] Data collection: collects the historical driving data of the inspection robot, including the remaining power at the beginning of the task, the power consumption and mileage after the task is completed, the route characteristics and environmental conditions;

[0107] Data preprocessing, including missing value processing, outlier monitoring and data standardization;

[0108] Build predictive models based on machine learning algorithms;

[0109] Model training and verification: historical data is divided into training set, verification set and test set for model training and verification;

[0110] The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven;

[0111] Compare the mileage prediction result with the length of the third inspection route:

[0112] When the predicted mileage is greater than the sum of the length of the third inspection route and the length of the return route, the endurance check passes; otherwise, it fails.

[0113] The length of the return route is a line connecting the end point of the third inspection route and the initial position of the inspection robot.

[0114] When the third inspection route is constructed, it means that the inspection robot can perform the inspection task, and the collision risk of multiple inspection robots has been reduced in advance, and repeated inspections of the same inspection point have been avoided. However, the inspection robot's endurance needs to be verified, specifically by obtaining the remaining power of the inspection robot and environmental conditions, and comprehensively considering whether the inspection robot can perform the inspection task according to the corresponding third inspection route. It is worth noting that it is also necessary to supplement the return route of the inspection robot from the last inspection point to the initial position. In summary, in one embodiment of the present invention, it is first necessary to obtain the remaining power and environmental conditions of the inspection robot, and construct a prediction model based on the inspection history data and historical environmental conditions of the inspection robot. The prediction model can output the drivable mileage prediction result by inputting the remaining power and environmental conditions of the inspection robot, that is, output the drivable mileage of the inspection robot, and compare the drivable mileage of the inspection robot with the task mileage to determine whether the inspection robot's endurance check has passed. The task mileage represents the sum of the length of the third inspection route and the length of the return route.

[0115] For example, when the predicted drivable mileage of the inspection robot X1 is 5KM, and the mileage of the inspection task that the inspection robot X1 is about to perform is 4.5KM, it means that the cruise verification of the inspection robot X1 has passed. On the contrary, when the mileage of the inspection task that the inspection robot is about to perform is 6KM, it means that the endurance verification of the inspection robot X1 has failed. If the third inspection route of the inspection robot X1 is not modified, the inspection robot X1 will inevitably terminate the task or return due to lack of power during the inspection process or the return process when performing the inspection task.

[0116] In one embodiment, when the endurance check of the inspection robot fails, the third inspection route of the inspection robot is adjusted based on the route adjustment module to obtain a fourth inspection route.

[0117] Based on the above, if the inspection robot's endurance check fails, considering that the current inspection robots are unmanned when performing inspection tasks, if the inspection robot terminates due to power outage during the inspection process or the return process, it will inevitably affect the inspection efficiency. Therefore, in one embodiment of the present invention, when the inspection robot's endurance check fails, it is necessary to adjust the inspection robot's third inspection route to obtain the fourth inspection route, and the fourth inspection route must pass the endurance check.

[0118] In one embodiment, the method for acquiring the fourth inspection route is:

[0119] Obtain the remaining power of the third inspection route that failed the endurance check and the matching inspection robot;

[0120] The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven;

[0121] Adjusting the third inspection route according to the mileage prediction result, including but not limited to deleting at least one to-be-inspected point, to obtain a fourth inspection route;

[0122] The fourth inspection route satisfies that the sum of the length of the fourth inspection route and the length of the return route is less than or equal to the predicted drivable mileage result.

[0123] Based on the above, how to adjust the third inspection route of the inspection robot that fails the endurance check. In one embodiment of the present invention, by obtaining the remaining power of the inspection robot, the prediction model can be used to obtain the prediction result of the drivable mileage corresponding to the inspection robot. According to the prediction result of the drivable mileage of the inspection robot, the third inspection route corresponding to the inspection robot is adjusted. The logic of the adjustment includes deleting at least one point to be inspected, thereby shortening the mission mileage of the inspection robot. When the mission mileage of the inspection robot is shortened, it can pass the endurance check.

[0124] In one embodiment, the route adjustment module is further used to adjust the deleted to-be-inspected point to any unadjusted third inspection route;

[0125] The method for adjusting the deleted inspection point to the unadjusted third inspection route is:

[0126] Obtain all inspection robots adjacent to the deleted inspection points and record them as the inspection robot group to be adjusted;

[0127] Obtain the predicted drivable mileage of all inspection robots in the inspection robot group to be adjusted, and the sum of the length of the third inspection route and the length of the return route, and calculate the difference between the two, which is recorded as the compensable mileage;

[0128] The third inspection route corresponding to the inspection robot with the largest compensable mileage is selected, and the deleted points to be inspected are added to obtain the fifth inspection route.

[0129] Based on the above, when any third inspection route that has not passed the endurance test is adjusted, a deleted inspection point will inevitably be generated. If the deleted inspection point is not planned, the inspection point will be missed. In one embodiment of the present invention, the above-mentioned multiple inspection robots that have passed the endurance test and the corresponding third inspection routes, based on the comparison of the drivable mileage and the mission mileage, there must be an inspection robot whose drivable mileage is far greater than the mission mileage. Therefore, by obtaining all the inspection robots adjacent to the deleted inspection point, that is, the inspection robots that are relatively close to the deleted inspection point, By marking the inspection robot, it is possible to avoid classifying the inspection robot that is far away from the deleted inspection point into the inspection robot group to be adjusted, thereby causing the problem of too low inspection efficiency. Through the above operation, according to the difference between the predicted mileage of the inspection robot group to be adjusted and the task mileage, the mileage can be compensated, and the third inspection route corresponding to the appropriate inspection robot is selected for adjustment, that is, the deleted inspection point to be inspected is added to the third inspection route, and the fifth inspection route is obtained, so that the endurance check of all inspection robots is passed, and the purpose of balancing the inspection efficiency of all robots to be inspected is achieved;

[0130] It is worth noting that the fifth inspection route also needs to undergo an endurance check. If it fails, the inspection robot corresponding to the fourth inspection route will be controlled to complete the inspection task and then perform a secondary inspection task. The secondary inspection task is that the inspection robot will perform the inspection task again after recharging according to the initial position to the deleted inspection point and the return route.

[0131] In one embodiment, the machine learning algorithm used by the prediction model includes one or more combinations of a random forest algorithm, a support vector machine algorithm, or a neural network algorithm.

[0132] like Figure 2 As shown in the figure, the inspection points marked with circles are at the same distance from the two inspection robots. Therefore, allocating inspection points and corresponding inspection robots according to distance may lead to repeated inspection points in the inspection route, which invisibly increases the collision risk of the inspection robots. Based on this, by pre-modifying the inspection route to ensure that there are no repeated inspection points in different inspection routes, the risk of collision of the inspection robots during the inspection process can be reduced, while shortening the inspection route length of the inspection robots and improving the inspection efficiency.

[0133] Working principle: By deleting duplicate inspection points, possible duplicate inspections and collisions during the inspection process can be avoided during the route construction phase, and then the third inspection route is obtained. At this point, the inspection robot performs an endurance check according to the third inspection route. If the check passes, the inspection robot performs the inspection task according to the third inspection route, and can inspect all points to be inspected to avoid missed inspections. At the same time, deleting duplicate inspection points based on the route adjustment module can avoid collision problems during the inspection process, and is also beneficial to reducing the inspection route length of the inspection robot, thereby optimizing the inspection efficiency of the inspection robot.

[0134] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A multi-inspection robot route collaborative control system, characterized in that: include: A status monitoring module is used to record the status information of each inspection robot in the multi-inspection robot group, wherein the status information includes the inspection robot number, remaining power and initial position; A route planning module is used to obtain N first inspection routes, and the N first inspection routes cover all inspection points; A route optimization module is used to identify the points to be inspected with repeated inspection robot identifications in the N first inspection routes, and connect the points to be inspected end to end to obtain M second inspection routes; A route adjustment module, used for marking duplicate inspection points existing in at least two second inspection routes, and deleting duplicate inspection points in any second inspection route, to obtain H third inspection routes; The battery life verification module is used to perform battery life verification according to the remaining power of the inspection robot and the environmental conditions. If the verification passes, the inspection task is performed using H third inspection routes and H matching inspection robots; When the endurance check of the inspection robot fails, adjusting the third inspection route of the inspection robot based on the route adjustment module to obtain a fourth inspection route; The method for obtaining the fourth inspection route is: Obtain the remaining power of the third inspection route that failed the endurance check and the matching inspection robot; The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven; Adjusting the third inspection route according to the mileage prediction result, including but not limited to deleting at least one to-be-inspected point, to obtain a fourth inspection route; The fourth inspection route satisfies that the sum of the length of the fourth inspection route and the length of the return route is less than or equal to the predicted drivable mileage result; The route adjustment module is also used to adjust the deleted inspection point to any unadjusted third inspection route, specifically: Obtain all inspection robots adjacent to the deleted inspection points and record them as the inspection robot group to be adjusted; Obtain the predicted drivable mileage of all inspection robots in the inspection robot group to be adjusted, and the sum of the length of the third inspection route and the length of the return route, and calculate the difference between the two, which is recorded as the compensable mileage; The third inspection route corresponding to the inspection robot with the largest compensable mileage is selected, and the deleted points to be inspected are added to obtain the fifth inspection route.

2. A multi-inspection robot route collaborative control system according to claim 1, characterized in that: The method for obtaining the first inspection route is: Get all the points to be inspected; Generate N first inspection routes based on the inspection robots closest to the inspection points; The N first inspection routes are each set with a corresponding inspection robot number, and N is an integer greater than 0.

3. A multi-inspection robot route collaborative control system according to claim 2, characterized in that: The method for obtaining the second inspection route is: Make a table of N first inspection routes and the corresponding N inspection robots; Count the number of inspection points to be inspected that are repeatedly identified by the inspection robot in the N first inspection routes; The inspection points corresponding to the repeated inspection robot identifications are merged into M inspection sets; Edit the routes of the M sets to be inspected according to the electronic map, and obtain M second inspection routes by connecting them end to end; The M sets to be inspected cover all inspection points, and M is an integer greater than 0, and M≤N.

4. A multi-inspection robot route collaborative control system according to claim 3, characterized in that: The method of editing the routes of the M sets to be inspected according to the electronic map and obtaining the M second inspection routes by connecting them end to end is as follows: Taking the location of the inspection robot as the starting point, construct F routes to be determined with all the points to be inspected included in the second inspection route; Analyzing the road condition information and the length of the route to be determined to obtain the fitness of the route to be determined, including: Obtain historical data, including but not limited to GPS records and sensor readings from historical inspection tasks of the inspection robot; Data preprocessing, including missing value processing, outlier monitoring and data standardization; Build prediction models and fitness assessment models; Model training and validation: divide historical data into training set, validation set and test set, and use cross-validation method to evaluate model performance; Input the characteristics of the route to be determined into the evaluation model to obtain the fitness of the route to be determined; The characteristics of the route to be determined include road condition information and route length, wherein the road condition information includes the length and slope of a section with a slope in the route to be determined; Compare the fitness of the F routes to be determined, and select the route to be determined with the largest fitness as the second inspection route.

5. A multi-inspection robot route collaborative control system according to claim 4, characterized in that: The method for obtaining the third inspection route is: Obtain M second inspection routes; Mark the repeated inspection points existing in the M second inspection routes; Calculate the fitness of the second inspection route with the same repeated inspection point; The repeated inspection points are retained in the second inspection route corresponding to the maximum fitness value, and the repeated inspection points are deleted from other second inspection routes; H third inspection routes are obtained, where H is an integer greater than 0, and H≤M.

6. A multi-inspection robot route collaborative control system according to claim 5, characterized in that: The method for performing endurance verification according to the remaining power of the inspection robot and environmental conditions is as follows: Obtain the third inspection route and the inspection robot matching it; Obtain the remaining power of the inspection robot based on the status monitoring module; A prediction model for the remaining battery power and mileage of the inspection robot is built based on historical data, including: Data collection: collects the historical driving data of the inspection robot, including the remaining power at the beginning of the task, the power consumption and mileage after the task is completed, the route characteristics and environmental conditions; Data preprocessing, including missing value processing, outlier monitoring and data standardization; Build predictive models based on machine learning algorithms; Model training and verification: historical data is divided into training set, verification set and test set for model training and verification; The remaining power and environmental conditions of the inspection robot are input into the prediction model to obtain the prediction result of the mileage that can be driven; Compare the mileage prediction result with the length of the third inspection route: When the predicted mileage is greater than the sum of the length of the third inspection route and the length of the return route, the endurance check passes; otherwise, it fails. The length of the return route is a line connecting the end point of the third inspection route and the initial position of the inspection robot.

7. The multi-inspection robot route collaborative control system according to claim 1, characterized in that: The machine learning algorithm used in the prediction model includes one or more combinations of a random forest algorithm, a support vector machine algorithm, or a neural network algorithm.

Images