Multi-robot path planning method and device in transfer station, and electronic equipment

Abstract

The invention discloses a multi-robot path planning scheme in a transfer station, and belongs to the technical field of intelligent hardware, and the method comprises the steps: obtaining the state information of a robot and the environment information of the transfer station; according to the state information and the environment information, a three-dimensional grid map of the transfer station is generated, and the position of an elevator is marked; judging whether the robot needs to cross floors from the current position to the target position or not to obtain a judgment result; planning a first path for the robot by adopting a path planning algorithm matched with the judgment result, and determining the time when the robot reaches each grid position under the first path; checking whether path conflict grids exist with other robots in the transfer system or not; adjusting the first path according to an inspection result to obtain a target path; and smoothing the target path based on a preset smoothing algorithm, and controlling the robot to move according to the smoothed target path. According to the scheme provided by the embodiment of the invention, rationalization and accuracy of the transfer guide path can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent hardware, and in particular, to a multi-robot path planning method and device in a transfer station, and an electronic device. Background Art

[0002] With the development of robot technology, more and more robots enter the indoor field, such as inspection service robots, transportation service robots, guiding service robots, etc. These mobile robots can be used in multiple scenarios such as hospitals, stations, restaurants, hotels, etc. For indoor mobile robots, due to the relatively complex and changeable operating environment, more accurate indoor positioning and path planning are required. Existing indoor robot path planning algorithms, such as the A*(A-Star) algorithm and the RRT (Rapidly-exploring Random Tree) algorithm, although they can handle path planning tasks in static environments, for the scenario of using multiple mobile robots in a transfer station, facing the complex scenario of cross-floor services, it is necessary to plan a safe and smooth movement path where collisions do not occur between each mobile robot, and the existing methods cannot meet this application requirement. Summary of the Invention

[0003] The purpose of the embodiments of the present invention is to provide a multi-robot path planning method and device in a transfer station, and an electronic device, which can solve the problems in the prior art that the indoor robot path planning lacks rationality and cannot plan a safe and smooth movement path for complex cross-floor scenarios.

[0004] To solve the above technical problems, the present invention provides the following technical solutions:

[0005] The embodiments of the present invention provide a multi-robot path planning method in a transfer station, which is applied to a guiding robot. Among them, the method includes:

[0006] Obtain the state information of the robot and the environmental information of the transfer station;

[0007] Generate a three-dimensional grid map of the transfer station according to the state information and the environmental information, where the three-dimensional grid map includes: information of each passable floor;

[0008] Mark the elevator position in the three-dimensional grid map;

[0009] Judge whether the robot needs to cross floors from the current position to the target position according to the marked three-dimensional grid map, and obtain a judgment result;

[0010] Use a path planning algorithm matching the judgment result to plan a first path for the robot, and determine the time when the robot reaches each grid position under the first path;

[0011] Check whether there are grid conflicts in the path with other robots in the transfer system;

[0012] Adjust the first path according to the inspection result to obtain a target path, and report the information of the target path and the time when the robot reaches each grid position under the target path to the transfer system;

[0013] Smooth the target path based on a preset smoothing algorithm, and control the robot to move along the smoothed target path;

[0014] Among them, the steps of smoothing the target path based on a preset smoothing algorithm and controlling the robot to move along the smoothed target path include:

[0015] Smooth the target path based on a preset smoothing algorithm to obtain the smoothed target path;

[0016] During the process of controlling the robot to move along the smoothed target path, for the next grid to go to, sense whether the robot will collide with an obstacle at the grid through a sensor;

[0017] If the robot will collide with an obstacle at the grid, control the robot to stop moving for a preset duration, and judge whether the robot will collide with an obstacle at the grid after the preset duration;

[0018] If the robot will collide with an obstacle at the grid after the preset duration, plan a local obstacle avoidance path;

[0019] Control the robot to drive along the obstacle avoidance path and then switch to continue moving on the smoothed target path.

[0020] Optionally, the steps of planning the first path for the robot by using a path planning algorithm matching the judgment result include:

[0021] In the case that the judgment result is that there is no need to cross floors, use a global path planning algorithm to plan a global path on the three-dimensional grid map;

[0022] For each global path, calculate the evaluation value of the global path;

[0023] Select the global path with the smallest evaluation value as the first path planned for the robot.

[0024] Optionally, the step of calculating the evaluation value of the global path for each global path includes:

[0025] For each global path, calculate the path distance, safety evaluation value, and turning evaluation value of the global path;

[0026] Perform a weighted sum of the path distance, the safety evaluation value, and the turning evaluation value to obtain the evaluation value of the global path.

[0027] Optionally, the step of planning a first path for the robot using a path planning algorithm that matches the judgment result includes:

[0028] In the case where the judgment result is that cross - floor movement is required, determine the available elevators from the floor where the robot is currently located to the floor where the target position is located;

[0029] For each available elevator, determine the entrance position of the available elevator on the floor where the robot is currently located as the first intermediate target position of the available elevator, and determine the exit position of the elevator on the floor where the target position is located as the second intermediate target position of the elevator;

[0030] Use a global path planning algorithm to plan a global path on the three - dimensional grid map that passes through the first intermediate target position and the second intermediate target position of the available elevator;

[0031] For each global path, calculate the evaluation value of the global path;

[0032] Select the global path with the minimum evaluation value as the first path planned for the robot.

[0033] Optionally, the step of checking for path conflict grids with other robots in the transfer system includes:

[0034] Obtain the path planning information of other robots issued by the transfer system, where the path planning information includes: the planned path and the time when the robot reaches each grid position under the planned path;

[0035] Compare the path planning information of the other robots with the first path and the time when the robot reaches each grid position under the first path to determine the path conflict grids with other robots;

[0036] Among them, the path conflict grids include: the same passable grid reached by the robot and other robots at the same time, and the grids where the robot and other robots have passable exchanges in the previous and next two unit times.

[0037] Optionally, the step of adjusting the first path according to the inspection result to obtain the target path includes:

[0038] When it is detected that there are path-conflicting grids, determine the path-conflicting grids as obstacle grids and update the three-dimensional grid map;

[0039] Plan a first path for the robot according to the updated three-dimensional grid map.

[0040] An embodiment of the present invention further provides a multi-robot path planning device in a transfer station, which is applied to a guiding robot. Wherein, the device includes:

[0041] An acquisition module, configured to acquire the status information of the robot and the environmental information of the transfer station;

[0042] A generation module, configured to generate a three-dimensional grid map of the transfer station according to the status information and the environmental information, where the three-dimensional grid map includes: information of each passable floor;

[0043] A marking module, configured to mark the elevator position in the three-dimensional grid map;

[0044] A judgment module, configured to judge whether the robot needs to cross floors from the current position to the target position according to the marked three-dimensional grid map, and obtain a judgment result;

[0045] A planning module, configured to plan a first path for the robot by using a path planning algorithm matching the judgment result, and determine the time when the robot reaches each grid position under the first path;

[0046] An inspection module, configured to check whether there are path-conflicting grids with other robots in the transfer system;

[0047] A target path generation module, configured to adjust the first path according to the inspection result to obtain a target path, and report the information of the target path and the time when the robot reaches each grid position under the target path to the transfer system;

[0048] A control module, configured to perform smoothing processing on the target path based on a preset smoothing algorithm, and control the robot to move according to the smoothed target path;

[0049] Wherein, the control module includes:

[0050] A smoothing processing sub-module, configured to perform smoothing processing on the target path based on a preset smoothing algorithm to obtain the smoothed target path;

[0051] A collision judgment sub-module, configured to, during the process of controlling the robot to move according to the smoothed target path, for the next grid to go, sense whether the robot will collide with an obstacle at the grid through a sensor;

[0052] A stop sub-module, configured to control the robot to stop moving for a preset duration if the robot will collide with an obstacle at the grid, and determine whether the robot will collide with an obstacle at the grid after the preset duration;

[0053] An obstacle avoidance path planning sub-module, configured to plan a local obstacle avoidance path if the robot will collide with an obstacle at the grid after the preset duration;

[0054] An obstacle avoidance switching sub-module, configured to control the robot to switch to continue moving on the target path after smooth processing after driving along the obstacle avoidance path.

[0055] Optionally, the planning module includes:

[0056] A first sub-module, configured to use a global path planning algorithm to plan a global path on the three-dimensional grid map when the judgment result is that there is no need to cross floors;

[0057] A second sub-module, configured to calculate an evaluation value of each global path;

[0058] A third sub-module, configured to select the global path with the smallest evaluation value as the first path planned for the robot, and determine the time when the robot reaches each grid position on the first path.

[0059] Optionally, the second sub-module is specifically configured to:

[0060] For each global path, calculate the path distance, safety evaluation value, and turning evaluation value of the global path;

[0061] Perform a weighted sum of the path distance, the safety evaluation value, and the turning evaluation value to obtain the evaluation value of the global path.

[0062] Optionally, the planning module includes:

[0063] A fourth sub-module, configured to determine available elevators that can reach the floor where the target position is located from the floor where the robot is currently located when the judgment result is that it is necessary to cross floors;

[0064] A fifth sub-module, configured to, for each available elevator, determine the entrance position of the available elevator on the floor where the robot is currently located as the first intermediate target position of the available elevator, and determine the exit position of the elevator reaching the floor where the target position is located as the second intermediate target position of the elevator;

[0065] A sixth module, configured to use a global path planning algorithm to plan a global path passing through the first intermediate target position and the second intermediate target position of the available elevator on the three-dimensional grid map;

[0066] A seventh sub-module, configured to calculate an evaluation value of each global path.

[0067] An eighth sub-module, configured to select the global path with the smallest evaluation value as the first path planned for the robot, and determine the time for the robot to reach each grid position under the first path.

[0068] Optionally, the inspection module includes:

[0069] An information acquisition sub-module, configured to acquire path planning information of other robots sent by the transfer system, where the path planning information includes: a planned path and the time for the robot to reach each grid position under the planned path.

[0070] A comparison sub-module, configured to compare the path planning information of other robots with the first path and the time for the robot to reach each grid position under the first path, and determine grid positions where path conflicts occur with other robots.

[0071] Among them, the grid positions where path conflicts occur include: the same passable grid reached by the robot and other robots at the same time, and the grid positions where the robot and other robots have passable exchanges in the previous and next unit times.

[0072] Optionally, when the target path generation module adjusts the first path to obtain a target path according to the detection result, it is specifically configured to:

[0073] When it is detected that there are grid positions where path conflicts occur, determine the grid positions where path conflicts occur as obstacle grid positions and update the three-dimensional grid map.

[0074] Plan a first path for the robot according to the updated three-dimensional grid map.

[0075] An embodiment of the present invention provides an electronic device, which includes a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of any one of the above multi-robot path planning methods in a transfer station are implemented.

[0076] An embodiment of the present invention provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of any one of the above multi-robot path planning methods in a transfer station are implemented.

[0077] The multi-robot path planning solution in a transfer station provided by the embodiment of the present invention obtains the status information of the robot and the environmental information of the transfer station; generates a three-dimensional grid map of the transfer station based on the status information and the environmental information; marks the elevator positions in the three-dimensional grid map; determines whether the robot needs to cross floors to reach the target position from the current position based on the marked three-dimensional grid map, and obtains a judgment result; uses a path planning algorithm matching the judgment result to plan a first path for the robot, and determines the time when the robot reaches each grid position under the first path; checks whether there are path conflict grids with other robots in the transfer system; adjusts the first path according to the inspection result to obtain a target path, and reports the information of the target path and the time when the robot reaches each grid position under the target path to the transfer system; performs smoothing processing on the target path based on a preset smoothing algorithm, and controls the robot to move along the smoothed target path. The solution provided by the embodiment of the present application can comprehensively determine the target path to guide the robot by combining multiple dimensions such as the position distribution of elevators in the transfer station and the conflict situation with the planned paths of other robots in the same transfer station, so as to improve the rationality and accuracy of the transfer guidance path. Description of the Drawings

[0078] Figure 1 is a flowchart showing the steps of a multi-robot path planning method in a transfer station according to an embodiment of the present application;

[0079] Figure 2 is a block diagram showing the structure of a multi-robot path planning device in a transfer station according to an embodiment of the present application. Detailed Embodiments

[0080] To make the technical problems, technical solutions, and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0081] The multi-robot path planning solution in a transfer station provided by the embodiment of the present application will be described in detail below with reference to the accompanying drawings, specific embodiments, and their application scenarios.

[0082] As shown in the Figure 1 accompanying drawings, the multi-robot path planning method in a transfer station according to an embodiment of the present application includes the following steps:

[0083] Step 101: Obtain the status information of the robot and the environmental information of the transfer station.

[0084] The multi-robot path planning method in a transfer station provided by the embodiments of the present application is applied to a guiding robot in the transfer station. This guiding robot (also known as a guiding service robot, hereinafter referred to as the robot for short) is an omnidirectional mobile robot that can move freely on the x and y planes and has the ability to rotate around the z axis. A control module (also known as the robot processor) is provided on the guiding robot to execute a preset multi-robot path planning algorithm in the transfer station to achieve optimal transfer path planning and guidance.

[0085] An intelligent transfer system (referred to as the transfer system for short) is set up in the transfer station, and each robot in the transfer station can interact with the transfer system for information. When the transfer system in the transfer station is started, it uses the robot's perception sensors to sense the environmental information of the transfer station and the status information of the robot.

[0086] The robot's perception sensors can include, but are not limited to, lidar, cameras, and inertial measurement units, etc.

[0087] Step 102: Generate a three-dimensional grid map of the transfer station based on the status information and environmental information.

[0088] Among them, the three-dimensional grid map includes: information about each passable floor.

[0089] In the actual implementation process, a three-dimensional grid map including multi-floor information can be generated for the operating environment of each passable floor in the transfer station according to the status information and environmental information. The grids in the three-dimensional grid map are divided into passable grids and obstacle grids.

[0090] Step 103: Mark the elevator positions in the three-dimensional grid map.

[0091] In the embodiments of the present application, the elevator positions available for the robot are accurately marked in the three-dimensional grid map, and the topological connection relationship between floors is established to ensure that the guiding robot can pass smoothly between different floors.

[0092] In addition to marking the elevator positions in the three-dimensional grid map, key areas such as the entrances and exits of the transfer station, waiting rooms, corridors, platforms, and safety exits can also be marked in the three-dimensional grid map. According to the historical visual information data collected by the cameras in the transfer station and the historical information data collected by the robot's cameras, the passenger density per unit area in the key areas is statistically calculated, and the areas with a passenger density exceeding the set threshold are marked as high-traffic areas.

[0093] Step 104: Judge whether the robot needs to cross floors to reach the target position based on the marked three-dimensional grid map, and obtain a judgment result.

[0094] The robot receives the user's guidance requirement information and determines the target position to reach according to the guidance requirement information. It judges whether the current position of the robot and the target position are on the same floor; if they are on the same floor, it is determined that the state between the current position of the robot and the target position to reach is not a cross-floor state; on the contrary, if they are not on the same floor, it is determined that the state between the current position of the robot and the target position to reach is a cross-floor state.

[0095] Through Step 102, 103 and Step 104, the elevator position in the transfer station can be determined, and whether the current navigation needs to cross floors can be determined. Based on this judgment result, a targeted path can be provided for the robot to plan a path, which can avoid the trouble brought by the robot's cross-floor movement. For example: if the robot needs to cross floors during navigation, but the elevator position is not determined and the cross-floor path is not planned when planning the navigation path, it will cause the robot to be unable to move accurately to the target position, which can improve the rationality of the transfer guidance path.

[0096] Step 105: Use a path planning algorithm matching the judgment result to plan the first path for the robot and determine the time when the robot reaches each grid position under the first path.

[0097] In the operating system of the robot, different path planning algorithms matching different judgment results are preset. After determining the judgment result, the matching path optimization algorithm is found to plan the first path for the robot. After the first path is planned, the transfer system simulates the process of the robot moving according to the first path plan based on the preset moving speed of the robot, so as to obtain the time when the robot reaches each grid position within the first path.

[0098] An optionally way to use a path planning algorithm matching the judgment result to plan the first path for the robot may include the following sub-steps:

[0099] Sub-step 1: In the case where the judgment result is that there is no need to cross floors, use the global path planning algorithm to plan the global path on the three-dimensional grid map.

[0100] When the state between the current position of the robot and the target position to reach is on the same floor rather than a cross-floor state, the robot uses the global path planning algorithm to plan the global path on the three-dimensional grid map. The global path planning algorithms used may include but are not limited to the Dijikstra algorithm (i.e., Dijkstra's algorithm), the RRT algorithm (i.e., Rapidly-exploring Random Tree algorithm), the deep reinforcement learning algorithm, etc.

[0101] After path planning through multiple global path planning algorithms, multiple global paths can be obtained. Therefore, it is necessary to screen out a global path from multiple global paths based on the evaluation values corresponding to each global path through Sub-step 2 and Sub-step 3.

[0102] Sub-step 2: For each global path, calculate the evaluation value of the global path.

[0103] An optionally way to calculate the evaluation value of each global path can be as follows:

[0104] For each global path, calculate the path distance, safety evaluation value, and turning evaluation value of the global path; perform a weighted sum of the path distance, safety evaluation value, and turning evaluation value to obtain the evaluation value of the global path.

[0105] Multiple global path schemes can be obtained by using a variety of different global path planning algorithms. The evaluation value of each global path scheme can be calculated using the following formula:

[0106] P i =λ1*D i +λ2*S i +λ3*Z i

[0107] where i is the number of the robot's global path scheme; P i is the evaluation value of the i-th robot global path scheme; D i is the path distance of the i-th robot global path scheme; S i is the safety evaluation value of the i-th robot global path scheme, and the safety evaluation value of the robot global path scheme is calculated based on the number of passable grids in the high pedestrian flow area passed by the path scheme; Z i is the turning evaluation value of the i-th robot global path scheme. Among them, the turning evaluation value of the robot global path can be calculated based on the number of turns of the robot in the global path. λ1, λ2, and λ3 are the weight coefficients of the path distance, safety evaluation value, and turning evaluation value respectively.

[0108] Sub-step 3: Select the global path with the smallest evaluation value as the first path planned for the robot.

[0109] An optionally way to plan the first path for the robot by using a path planning algorithm matching the judgment result can include the following sub-steps:

[0110] S1: In the case where the judgment result is that cross-floor movement is required, determine the available elevators on the floor where the target position can be reached from the floor where the robot is currently located.

[0111] When it is in a cross-floor state between the current position of the robot and the target position to be reached, obtain the elevator information set on the current floor of the robot, and set the elevators that can reach the floor where the target position is located as available elevators. Set the corresponding numbers of the available elevators as e1, e2,.........

[0112] S2: For each available elevator, determine the entrance position of the available elevator on the floor where the robot is currently located as the first intermediate target position of the available elevator, and determine the exit position of the elevator when it reaches the floor of the target position as the second intermediate target position of the elevator.

[0113] That is, take the entrance position of the e1-th available elevator on the floor where the robot is currently located as the first intermediate target position of the e1-th available elevator, and take the exit position of the e1-th available elevator when it reaches the floor of the target position as the second intermediate target position of the e1-th available elevator.

[0114] S3: Use the global path planning algorithm to plan a global path on the three-dimensional grid map that passes through the first intermediate target position and the second intermediate target position of the available elevator.

[0115] The robot uses the global path planning algorithm to plan a global path on the three-dimensional grid map from the current position to the first intermediate target position of each available elevator, then to the corresponding second intermediate target position of the available elevator, and then to the target position to be reached. After path planning through multiple global path planning algorithms, multiple global paths can be obtained. Therefore, it is necessary to select a global path from multiple global paths based on the evaluation values corresponding to each global path through S4 and S5.

[0116] S4: For each global path, calculate the evaluation value of the global path.

[0117] The calculation method of the evaluation value of the global path is the same as that of the global path in sub-step 2, and will not be elaborated here.

[0118] S5: Select the global path with the smallest evaluation value as the first path planned for the robot.

[0119] Step 106: Check whether there are path conflict grids with other robots in the transfer system.

[0120] An optional way to check whether there are path conflict grids with other robots in the transfer system can be as follows:

[0121] Obtain the path planning information of other robots issued by the transfer system, compare the path planning information of other robots with the first path and the time when the robot reaches each grid position under the first path, and determine the path conflict grids with other robots.

[0122] After the robot obtains the first path plan and the time information for reaching each grid position within the first path, it acquires the final planned paths of other robots sent by the transfer system and the time information for other service robots to reach each grid position within the final planned paths. Therefore, by comparing the time when it reaches each grid with the time when other robots reach the corresponding grid, it can determine whether there are grid positions in the first path that conflict with the paths of other robots.

[0123] Among them, the path planning information includes: the planned path and the time for the robot to reach each grid position under the planned path; the grid positions with path conflicts include: the same passable grid that the robot and other robots reach at the same time, and the grid where the robot and other robots have a passable exchange within the previous and next two unit times.

[0124] Checking in step 106 whether there are grid positions with path conflicts with other robots in the transfer system can effectively avoid collisions with other robots during movement.

[0125] Step 107: Adjust the first path according to the inspection result to obtain the target path, and report the information of the target path and the time for the robot to reach each grid position under the target path to the transfer system.

[0126] In an optional embodiment, the method of adjusting the first path according to the inspection result to obtain the target path can be as follows:

[0127] When the detection result is that the first path does not contain grid positions with path conflicts, the first path is determined as the target path, that is, there is no need to adjust the first path. That is to say, when the robot does not detect grid positions with path conflicts, the first path plan of the robot is used as the final planned path, and the final planned path information and the time information for guiding the service robot to reach each grid position within the final planned path are uploaded to the intelligent transfer system.

[0128] When it is checked that there are grid positions with path conflicts, the grid positions with path conflicts are determined as obstacle grid positions and the three-dimensional grid map is updated; a first path is planned for the robot according to the updated three-dimensional grid map. The first path planned according to the updated three-dimensional grid map also needs to return to step 106 to check whether there are grid positions with path conflicts with other robots in the transfer system and step 107. If there are still grid positions with path conflicts after this adjustment, the path adjustment is repeated until a target path that does not contain grid positions with path conflicts is planned.

[0129] That is to say, when the robot detects a path conflict grid, it regards the path conflict grid as an obstacle grid to update the three-dimensional grid map of the robot, then re-uses the global path planning algorithm to plan the global path on the updated three-dimensional grid map, and calculates the first path; then detects the path conflict grid with other robots again until no path conflict grid is detected. The first path plan of the robot without path conflict grid is used as the final planned path (i.e., the target path), and the final planned path information and the time information of the robot reaching each grid position within the final planned path are uploaded to the intelligent transfer system.

[0130] Step 108: Smooth the target path based on a preset smoothing algorithm, and control the robot to move along the smoothed target path.

[0131] The robot smooths the target path using a path smoothing algorithm, and the path smoothing algorithm used may include, but is not limited to, a path smoothing algorithm based on a cubic spline curve, a path smoothing algorithm based on a Bezier curve, etc. The robot uses a path tracking algorithm and a real-time obstacle stopping and avoidance combination strategy to make the robot move along the smoothed target path and finally reach the target position.

[0132] Among them, the path tracking algorithm adopted in the embodiments of the present application may include, but is not limited to, a PID (proportional-integral-differential) path tracking algorithm, a Pure Pursuit path tracking algorithm, i.e., a pure tracking algorithm.

[0133] Among them, the real-time obstacle stopping and avoidance combination strategy is that when the robot moves to the next passable grid, it uses the data collected by the lidar and camera sensors to sense whether it will collide with obstacles such as people, robots or other objects in the transfer station. If no collision will occur, the robot continues to move to the next passable grid. If a collision will occur, the robot needs to plan a local obstacle avoidance path to avoid the grid at the possible collision location, and then return to the target path to continue moving after avoidance. The above judgment is executed before moving to each grid in the target path until it moves to the target position along the target path.

[0134] In an optional embodiment, the method of smoothing the target path based on a preset smoothing algorithm and controlling the robot to move along the smoothed target path may include the following sub-steps:

[0135] Sub-step 1: Smooth the target path based on a preset smoothing algorithm to obtain the smoothed target path.

[0136] Sub-step 2: During the process of controlling the robot to move along the smoothed target path, for the next grid to be traveled to, use sensors to sense whether the robot will collide with an obstacle at the grid.

[0137] Sub-step 3: If the robot will collide with an obstacle at the grid, control the robot to stop moving for a preset duration, and determine whether the robot will collide with an obstacle at the grid after the preset duration.

[0138] In the actual implementation process, the preset duration can be flexibly set by those skilled in the art according to actual needs, and no specific limitation is made in this embodiment of the present application.

[0139] Sub-step 4: If the robot will collide with an obstacle at the grid after the preset duration, plan a local obstacle avoidance path.

[0140] If the robot will not collide with an obstacle at the grid after the preset duration, there is no need to plan a local obstacle avoidance path, and it can directly move to the grid.

[0141] If it is sensed by the sensor that the robot will collide with an obstacle at the grid, stop moving in time first. If the obstacle moves within the preset implementation time of stopping and will no longer collide with the robot, the robot continues to move to the next passable grid. If the person, robot, or other object to be collided with within a certain preset time of stopping does not move or moves but still will collide with the guiding robot, the robot uses the global path planning algorithm to plan a local obstacle avoidance path that can bypass the obstacle, and then returns to the grid of the target path and continues to move.

[0142] Further, in sub-step 4, a feasible local obstacle avoidance path planning process is as follows:

[0143] (1) According to the current position of the robot, the next position to be reached, and the grid map, use the method of randomly sampling in the passable grids to generate N candidate obstacle avoidance paths, where N is a system preset parameter.

[0144] (2) The robot sets a local grid map with a size of a×a around the current position of the robot by real-time sensing of the environment near the current position, where a is a system preset parameter. Calculate the occupancy ratio OD of the obstacle grids in the local grid map:

[0145]

[0146] (3) Calculate the distance DO from the robot's current position to the nearest obstacle.

[0147] (4) Perceive the grid map in the fan-shaped area along the current movement direction of the robot, and calculate the lateral width PH of the largest continuous obstacle-free area.

[0148] (5) Calculate the risk parameters of the robot at the current position:

[0149]

[0150] Among them, α1, α2, and α3 are system preset coefficients. OD max is the maximum occupancy ratio of the obstacle grid preset by the system. DO min is the minimum distance to the nearest obstacle preset by the system. PH min is the minimum lateral width of the obstacle-free area preset by the system.

[0151] (6) Calculate the local obstacle avoidance path evaluation parameters c 1 and c 2 :

[0152]

[0153] c 1，b 、c 2，b 、c 1，z 、c 2，z 、c 1，j 、c 2，j 、F b 、F J are system preset parameters.

[0154] (7) Calculate the evaluation values of N candidate local obstacle avoidance paths, and select the path with the minimum evaluation value as the local obstacle avoidance path. The calculation formula for the evaluation value of the nth candidate local obstacle avoidance path is as follows:

[0155] p n = c 1 × DIS n + c 2 × ODS n

[0156] Among them, DIS n is the length of the nth candidate local obstacle avoidance path. ODS n is the sum of the distances to the nearest obstacles on each passable grid included in the nth candidate local obstacle avoidance path. Through the above feasible local obstacle avoidance path planning method, the local obstacle avoidance path can be determined efficiently and reliably. The determined local obstacle avoidance path facilitates the robot to reasonably avoid collisions with other robots when driving on the target path.

[0157] Sub-step 5: Control the robot to drive along the obstacle avoidance path and then switch to continue moving on the smoothed target path.

[0158] In the actual implementation process, after the robot travels along the obstacle avoidance path, it can switch to the grid that has been avoided, or move to the next grid of the avoided grid in the smoothed target path, so as to switch to the smoothed target path and continue to move.

[0159] The solution for controlling the robot to move along the target path shown in Sub-step 1 - Sub-step 5 can effectively avoid collisions between robots. Moreover, when it is determined that there may be a collision, a local obstacle avoidance path is planned, which can control the robot to avoid in a short distance and then return to the target path, ensuring the safety of the robot's movement and minimizing the extra movement distance of the robot as much as possible, and improving the rationality and accuracy of the transfer guidance path.

[0160] The multi-robot path planning method in the transfer station provided by the embodiment of the present application obtains the state information of the robot and the environmental information of the transfer station; generates a three-dimensional grid map of the transfer station based on the state information and the environmental information; marks the elevator position in the three-dimensional grid map; determines whether the robot needs to cross floors from the current position to the target position based on the marked three-dimensional grid map, and obtains a judgment result; uses a path planning algorithm matching the judgment result to plan a first path for the robot, and determines the time when the robot reaches each grid position under the first path; checks whether there are path conflict grids with other robots in the transfer system; adjusts the first path according to the check result to obtain the target path, and reports the information of the target path and the time when the robot reaches each grid position under the target path to the transfer system; performs smoothing processing on the target path based on a preset smoothing algorithm, and controls the robot to move according to the smoothed target path. The method provided by the embodiment of the present application can comprehensively determine the target path to guide the robot by combining multiple dimensions such as the position distribution of elevators in the transfer station and the conflict situation with the planned paths of other robots in the same transfer station, thereby improving the rationality and accuracy of the transfer guidance path.

[0161] Figure 2 It is a structural block diagram of a multi-robot path planning device in a transfer station for implementing the embodiment of the present application.

[0162] The multi-robot path planning device in the transfer station provided by the embodiment of the present application is applied to a guiding robot and includes the following functional modules:

[0163] An acquisition module 201, configured to acquire the state information of the robot and the environmental information of the transfer station;

[0164] A generation module 202, configured to generate a three-dimensional grid map of the transfer station based on the state information and the environmental information, where the three-dimensional grid map includes: information of each passable floor;

[0165] A marking module 203 for marking the elevator position in the three-dimensional grid map;

[0166] A judgment module 204 for judging whether the robot needs to cross floors from the current position to the target position based on the marked three-dimensional grid map, and obtaining a judgment result;

[0167] A path planning module 205 for planning a first path for the robot using a path planning algorithm matching the judgment result, and determining the time when the robot reaches each grid position on the first path;

[0168] An inspection module 206 for checking whether there are path conflict grids with other robots in the transfer system;

[0169] A target path generation module 207 for adjusting the first path according to the inspection result to obtain a target path, and reporting the information of the target path and the time when the robot reaches each grid position on the target path to the transfer system;

[0170] A control module 208 for smoothing the target path based on a preset smoothing algorithm, and controlling the robot to move along the smoothed target path;

[0171] Wherein, the control module includes:

[0172] A smoothing processing sub-module for smoothing the target path based on a preset smoothing algorithm to obtain the smoothed target path;

[0173] A collision judgment sub-module for, during the process of controlling the robot to move along the smoothed target path, for the next grid to be traveled to, sensing through a sensor whether the robot will collide with an obstacle at the grid;

[0174] A stop sub-module for, if the robot will collide with an obstacle at the grid, controlling the robot to stop moving for a preset duration, and judging whether the robot will collide with an obstacle at the grid after the preset duration;

[0175] An obstacle avoidance path planning sub-module for, if the robot will collide with an obstacle at the grid after the preset duration, planning a local obstacle avoidance path;

[0176] An obstacle avoidance switching sub-module for controlling the robot to switch to continue moving on the smoothed target path after traveling along the obstacle avoidance path.

[0177] Optionally, the path planning module includes:

[0178] The first sub-module is used to plan a global path on the three-dimensional grid map using a global path planning algorithm when the judgment result is that there is no need to cross floors;

[0179] The second sub-module is used to calculate the evaluation value of each global path for each global path;

[0180] The third sub-module is used to select the global path with the minimum evaluation value as the first path planned for the robot and determine the time when the robot reaches each grid position under the first path.

[0181] Optionally, the second sub-module is specifically used for:

[0182] For each global path, calculate the path distance, safety evaluation value, and turning evaluation value of the global path;

[0183] Weighted sum the path distance, the safety evaluation value, and the turning evaluation value to obtain the evaluation value of the global path.

[0184] Optionally, the planning module includes:

[0185] The fourth sub-module is used to determine the available elevators that can reach the floor where the target position is located from the floor where the robot is currently located when the judgment result is that it is necessary to cross floors;

[0186] The fifth sub-module is used to, for each available elevator, determine the entrance position of the available elevator on the floor where the robot is currently located as the first intermediate target position of the available elevator, and determine the exit position of the elevator reaching the floor where the target position is located as the second intermediate target position of the elevator;

[0187] The sixth module is used to plan a global path passing through the first intermediate target position and the second intermediate target position of the available elevator on the three-dimensional grid map using a global path planning algorithm;

[0188] The seventh sub-module is used to calculate the evaluation value of each global path for each global path;

[0189] The eighth sub-module is used to select the global path with the minimum evaluation value as the first path planned for the robot and determine the time when the robot reaches each grid position under the first path.

[0190] Optionally, the inspection module includes:

[0191] The information acquisition sub-module is used to acquire the path planning information of other robots sent by the transfer system, where the path planning information includes: the planned path and the time when the robot reaches each grid position under the planned path;

[0192] A comparison sub-module, configured to compare the path planning information of the other robots with the first path and the time when the robot reaches each grid position on the first path, and determine grid positions where there are path conflicts with the other robots;

[0193] Wherein, the grid positions with path conflicts include: the same passable grid that the robot and other robots reach at the same time, and the grid where the robot and other robots have passable exchanges in two adjacent unit times.

[0194] Optionally, when the target path generation module adjusts the first path according to the detection result to obtain a target path, it is specifically configured to:

[0195] When it is detected that there are grid positions with path conflicts, determine the grid positions with path conflicts as obstacle grid positions and update the three-dimensional grid map;

[0196] Plan a first path for the robot according to the updated three-dimensional grid map.

[0197] The multi-robot path planning device in a transfer station provided by an embodiment of the present invention acquires the state information of the robot and the environmental information of the transfer station; generates a three-dimensional grid map of the transfer station according to the state information and the environmental information; marks the elevator positions in the three-dimensional grid map; determines whether the robot needs to cross floors when reaching the target position from the current position according to the marked three-dimensional grid map, and obtains a judgment result; uses a path planning algorithm matching the judgment result to plan a first path for the robot, and determines the time when the robot reaches each grid position on the first path; checks whether there are grid positions with path conflicts with other robots in the transfer system; adjusts the first path according to the inspection result to obtain a target path, and reports the information of the target path and the time when the robot reaches each grid position on the target path to the transfer system; performs smoothing processing on the target path based on a preset smoothing algorithm, and controls the robot to move according to the smoothed target path. The device provided by the embodiment of the present application can comprehensively determine the target path from multiple dimensions such as the position distribution of elevators in the transfer station and the conflict situation with the planned paths of other robots in the same transfer station to guide the robot, thereby improving the rationality and accuracy of the transfer guidance path.

[0198] In the embodiment of the present application Figure 2 The multi-robot path planning device in the transfer station shown is arranged in the control system of an indoor robot. The control system provided with this device can be a device with an operating system. The operating system can be an Android operating system, can be an iOS operating system, or can be other possible operating systems, which are not specifically limited in the embodiment of the present application.

[0199] The multi-robot path planning device in the transfer station provided by the embodiment of the present application Figure 2 can implement each process implemented by the method embodiment shown above. To avoid repetition, it will not be elaborated here. Figure 1 To avoid repetition, it will not be elaborated here.

[0200] Optionally, the embodiment of the present application further provides an electronic device, including a processor, a memory, and a program or instruction stored on the memory and executable on the processor. When the program or instruction is executed by the processor, it implements each process executed by the above-mentioned guiding robot path planning device in the transfer station and can achieve the same technical effect. To avoid repetition, it will not be elaborated here.

[0201] It should be noted that the electronic device in the embodiment of the present application includes the above-mentioned server.

[0202] Wherein, the processor is the processor in the electronic device in the above-mentioned embodiment. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disc, etc.

[0203] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover a non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of another identical element in the process, method, article or device including that element.

[0204] The above is the preferred embodiment of the present invention. It should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A multi-robot path planning method in a transfer station, applied to a guiding robot, characterized in that The method includes: Obtaining the state information of the robot and the environmental information of the transfer station; Generating a three-dimensional grid map of the transfer station based on the state information and the environmental information, where the three-dimensional grid map includes: information of each passable floor; Marking the elevator positions in the three-dimensional grid map; Judging whether the robot needs to cross floors to reach the target position from the current position based on the marked three-dimensional grid map, and obtaining a judgment result; Using a path planning algorithm matching the judgment result to plan a first path for the robot, and determining the time when the robot reaches each grid position under the first path; Checking whether there are grid positions with path conflicts with other robots in the transfer system; Adjusting the first path according to the check result to obtain a target path, and reporting the information of the target path and the time when the robot reaches each grid position under the target path to the transfer system; Performing smoothing processing on the target path based on a preset smoothing algorithm, and controlling the robot to move according to the smoothed target path; Among them, the step of performing smoothing processing on the target path based on a preset smoothing algorithm and controlling the robot to move according to the smoothed target path includes: Performing smoothing processing on the target path based on a preset smoothing algorithm to obtain the smoothed target path; During the process of controlling the robot to move according to the smoothed target path, for the next grid to go to, sensing through a sensor whether the robot will collide with an obstacle at the grid; If the robot will collide with an obstacle at the grid, controlling the robot to stop moving for a preset duration, and judging whether the robot will collide with an obstacle at the grid after the preset duration; If the robot will collide with an obstacle at the grid after the preset duration, planning a local obstacle avoidance path; After controlling the robot to drive along the obstacle avoidance path to bypass the grid, switching to continue moving on the smoothed target path.

2. The method according to claim 1, wherein The step of using a path planning algorithm matching the judgment result to plan a first path for the robot includes: In the case where the judgment result is that no floor crossing is required, using a global path planning algorithm to plan a global path on the three-dimensional grid map; Calculating the evaluation value of each global path; Selecting the global path with the minimum evaluation value as the first path planned for the robot.

3. The method according to claim 2, wherein The step of calculating the evaluation value of each global path includes: Calculating the path distance, safety evaluation value, and turning evaluation value of each global path; Performing weighted summation of the path distance, the safety evaluation value, and the turning evaluation value to obtain the evaluation value of the global path.

4. The method according to claim 1, wherein The step of using a path planning algorithm matching the judgment result to plan a first path for the robot includes: In the case where the judgment result is that floor crossing is required, determining the available elevators from the floor where the robot is currently located to the floor where the target position is located; For each available elevator, determine the entrance position of the available elevator on the floor where the robot is currently located as the first intermediate target position of the available elevator, and determine the exit position of the elevator on the floor where the target position is located as the second intermediate target position of the elevator; Use a global path planning algorithm to plan a global path on the three-dimensional grid map that passes through the first intermediate target position and the second intermediate target position of the available elevator; For each global path, calculate the evaluation value of the global path; Select the global path with the smallest evaluation value as the first path planned for the robot.

5. The method according to claim 1, characterized in that, The steps of checking for grid cells with path conflicts with other robots in the transfer system include: Obtain the path planning information of other robots issued by the transfer system, where the path planning information includes: the planned path and the time when the robot reaches each grid position on the planned path; Compare the path planning information of the other robots with the first path and the time when the robot reaches each grid position on the first path to determine the grid cells with path conflicts with other robots; Among them, the grid cells with path conflicts include: the passable grid cells that the robot and other robots reach at the same time, and the grid cells where the robot and other robots have passable exchanges in the previous and next two unit times.

6. The method according to claim 1, characterized in that, The steps of adjusting the first path according to the inspection result to obtain the target path include: When it is checked that there are grid cells with path conflicts, determine the grid cells with path conflicts as obstacle grid cells and update the three-dimensional grid map; Plan the first path for the robot according to the updated three-dimensional grid map.

7. A multi-robot path planning device in a transfer station, which is applied to a guiding robot, and is characterized in that The device includes: An acquisition module for acquiring the state information of the robot and the environmental information of the transfer station; A generation module for generating a three-dimensional grid map of the transfer station according to the state information and the environmental information, where the three-dimensional grid map includes: information on each passable floor; A marking module for marking the elevator position in the three-dimensional grid map; A judgment module for judging whether the robot needs to cross floors from the current position to the target position according to the marked three-dimensional grid map, and obtaining a judgment result; A planning module for planning the first path for the robot using a path planning algorithm matching the judgment result, and determining the time when the robot reaches each grid position on the first path; An inspection module for checking whether there are grid cells with path conflicts with other robots in the transfer system; A target path generation module for adjusting the first path according to the inspection result to obtain the target path, and reporting the information of the target path and the time when the robot reaches each grid position on the target path to the transfer system; A control module for smoothing the target path based on a preset smoothing algorithm, and controlling the robot to move according to the smoothed target path; Among them, the control module includes: A smoothing sub-module for smoothing the target path based on a preset smoothing algorithm to obtain the smoothed target path; A collision judgment sub-module, which is used to sense whether the robot will collide with an obstacle at the grid when the robot moves along the smoothed target path, for the next grid to be traveled; A stop sub-module, which is used to control the robot to stop moving for a preset duration if the robot will collide with an obstacle at the grid, and judge whether the robot will collide with an obstacle at the grid after the preset duration; An obstacle avoidance path planning sub-module, which is used to plan a local obstacle avoidance path if the robot will collide with an obstacle at the grid after the preset duration; An obstacle avoidance switching sub-module, which is used to control the robot to switch to the smoothed target path to continue moving after driving along the obstacle avoidance path.

8. The device according to claim 7, wherein The planning module includes: A first sub-module, which is used to plan a global path on the three-dimensional grid map using a global path planning algorithm when the judgment result is that there is no need to cross floors; A second sub-module, which is used to calculate the evaluation value of each global path; A third sub-module, which is used to select the global path with the smallest evaluation value as the first path planned for the robot, and determine the time when the robot reaches each grid position on the first path.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a program or instruction stored on the memory and executable on the processor, and the program or instruction is executed by the processor to perform the steps of any one of the multi-robot path planning methods in the transfer station according to claims 1-6.

Images