Robot obstacle avoidance inspection system based on visual monitoring

Abstract

The invention relates to the technical field of robot path planning, in particular to a robot obstacle avoidance inspection system based on visual monitoring, and the system comprises a detector which is used for scanning the feature information of an obstacle in a robot moving area; the modeling device is used for constructing an obstacle model according to the feature information of the obstacles and marking static obstacles and dynamic obstacles in the obstacle model; the visual monitor is used for monitoring the moving track of the obstacle within preset time and predicting the moving track; the obstacle avoiding device is used for planning a moving route of the robot, adjusting the advancing state of the robot and sending out an avoiding signal; the air spray gun device is used for receiving the evasion signal and sending an air flow to an obstacle; according to the robot obstacle avoidance inspection system based on visual monitoring, the obstacle recognition accuracy is improved by arranging the devices, and the accuracy of the robot obstacle avoidance inspection system based on visual monitoring is effectively improved while the success rate of avoiding moving obstacles is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of robot path planning, and in particular to a robot obstacle avoidance inspection system based on visual monitoring. Background Art

[0002] Robots are designed to provide a variety of life services for humans, such as home cleaning, elderly care, hotel reception, etc. In a home environment, robots may encounter various obstacles such as furniture, walls, stairs, pets, family members, etc.; in public places such as hotels and shopping malls, they will encounter pedestrians, shelves, exhibits, etc. These complex and diverse environments require service robots to have accurate obstacle avoidance capabilities to provide a good service experience and avoid causing any harm or inconvenience. Visual sensors can obtain visual images of the environment, and computer vision technology can be used to analyze and process the images to identify information such as the type, location, size, and color of obstacles. In recent years, with the rapid development of digital image processing technology, artificial intelligence technology, and especially deep learning technology, the application of visual sensors in robot obstacle avoidance has been greatly improved. Deep learning algorithms can train visual sensors to accurately identify obstacles in various complex environments, enabling robots to make more accurate obstacle avoidance decisions based on visual information. Modern robots have also made great improvements in hardware integration. The ability to organically combine multiple sensors, processors, actuators, etc. to form a complete system enables robots to work more coordinated. For example, multiple sensors such as lidar, visual sensors, and ultrasonic sensors can be integrated into a robot, and data fusion can be performed through software algorithms to obtain more accurate environmental perception and more effective obstacle avoidance capabilities.

[0003] China Patent Authorization Announcement No.: CN118534912B discloses a real-time path planning method for a multi-sensory fusion medical robot. The invention first performs three-dimensional modeling on the corridor where the medical robot that delivers medicines is located, thereby obtaining a real-time three-dimensional spatial model of the corridor and the real-time position of the medical robot; at the same time, the real-time dynamic distance and safety distance are calculated in combination with the real-time depth image, so that static objects and dynamic objects can be accurately distinguished, and then an accurate active obstacle avoidance path is provided for the medical robot, so that the medical robot can reach the destination safely and smoothly even in the presence of obstacles.

[0004] Chinese Patent Authorization Announcement No.: CN118550304B discloses an intelligent obstacle avoidance method and system for an inspection robot. The invention provides an intelligent obstacle avoidance method and system for an inspection robot, which belongs to the field of obstacle avoidance path planning; solves the problems of low data accuracy and computational redundancy; specifically as follows: collect inspection environment data to obtain scanning data points; obtain and calculate scanning data records and scanning data volume to obtain scanning data standard scores; screen scanning data standard scores to obtain point cloud data and construct a three-dimensional model; read and calculate the three-dimensional model to obtain total potential energy; obtain and calculate the path coordinate set to obtain the total path cost; obtain the cost value; process the cost value, total potential energy and three-dimensional model to obtain an obstacle avoidance path; collect the second environment data and compare it with the point cloud data to obtain the data change range; correct the obstacle avoidance path to obtain a safe obstacle avoidance route; the invention calculates the changes in the acquired three-dimensional point cloud data, reduces the amount of data calculation, and realizes intelligent obstacle avoidance planning.

[0005] However, the above method has the following problems: low obstacle recognition accuracy, low success rate of avoiding moving obstacles, inability to effectively avoid obstacles that appear at a very close distance, and failure to optimize subsequent action routes in the event of sudden collisions. Summary of the invention

[0006] To this end, the present invention provides a robot obstacle avoidance inspection system based on visual monitoring, which is used to overcome the problems in the prior art of low obstacle recognition accuracy, low success rate in avoiding moving obstacles, inability to effectively avoid obstacles that appear at a very close distance, and failure to optimize subsequent action routes in the event of sudden collisions.

[0007] To achieve the above object, the present invention provides a robot obstacle avoidance inspection system based on visual monitoring, comprising:

[0008] A detector for scanning characteristic information of obstacles in the robot's moving area;

[0009] The characteristic information of the obstacle includes the shape, outline, color, texture and collision distance of the obstacle;

[0010] A modeler connected to the detector, configured to construct an obstacle model according to the characteristic information of the obstacle and mark static obstacles and dynamic obstacles in the obstacle model;

[0011] A visual monitor, which is connected to the modeler and the detector, respectively, to monitor the movement trajectory of the obstacle within a preset time, and predict the predicted movement trajectory of the obstacle within a subsequent time in combination with the obstacle model;

[0012] an obstacle avoider, which is provided with an airbag and a pressure sensor, which are respectively connected to the detector, the visual monitor and the modeler, and are used to plan the robot's moving route in combination with the obstacle model and the predicted action trajectory and adjust the robot's moving state during the moving process, and send an avoidance signal when the detector scans that there is an obstacle within a preset distance or the pressure sensor detects that the robot collides with the obstacle;

[0013] The air spray gun device includes a horizontal air spray gun and a rotary air spray gun, which are respectively connected to the visual monitor and the obstacle avoider, for receiving the avoidance signal and starting the rotary air spray gun when no collision occurs, and rotating the horizontal air spray gun to emit an air flow toward the obstacle.

[0014] Furthermore, the obstacle avoider plans the movement route of the robot in combination with the obstacle model and the predicted motion trajectory, wherein:

[0015] If the obstacle model includes the static obstacle but does not include the dynamic obstacle, the obstacle avoider plans the movement route of the robot through a global planning algorithm;

[0016] If the obstacle model includes the dynamic obstacle and the static obstacle, the obstacle avoider constructs a grid map according to the obstacle model, generates a passage cost of each grid on the grid map corresponding to the predicted action trajectory, and plans the movement route of the robot according to the passage cost;

[0017] If the obstacle model does not include the dynamic obstacle and does not include the static obstacle, the obstacle avoider drives the robot to move along a straight line to the end point;

[0018] The grid map divides the traversable routes in the obstacle model into a plurality of grids, and the passage cost is the collision probability of the robot traveling to the grid.

[0019] Furthermore, the obstacle avoider adjusts the robot's traveling state according to the situation that an obstacle is scanned on the robot's moving route, wherein:

[0020] If the collision distance of the obstacle scanned by the detector is greater than or equal to the preset distance, reducing the movement speed of the robot;

[0021] If the collision distance of the obstacle scanned by the detector is less than the preset distance, the robot stops moving and sends the avoidance signal;

[0022] Wherein, the preset distance is positively correlated with the moving speed of the robot.

[0023] Furthermore, the visual monitor monitors the movement trajectory of the obstacle within the preset time when the obstacle avoider reduces the movement speed of the robot, and predicts the predicted movement trajectory of the obstacle within the next time in combination with the obstacle model, wherein:

[0024] If the obstacle does not move within the preset time, the visual monitor predicts that the position of the obstacle will not move within the next period of time;

[0025] If the obstacle moves within the preset time, the visual monitor records the movement trajectory of the obstacle on the obstacle model, and predicts the predicted movement trajectory of the obstacle in the next period of time according to the direction and speed of the movement trajectory;

[0026] The preset time is negatively correlated with the moving speed of the robot; the subsequent time is a time period that connects to the preset time and has the same duration as the preset time.

[0027] Furthermore, the obstacle avoider replans the movement route of the robot when the visual monitor predicts that the position of the obstacle will not move within a short period of time;

[0028] The obstacle avoider calculates a safe speed when the visual monitor predicts the predicted movement trajectory of the obstacle within a short period of time;

[0029] The safe speed is the moving speed of the robot along the moving route of the robot without collision.

[0030] Furthermore, the obstacle avoider compares the safety speed with a preset speed and determines whether to re-plan the robot's moving route according to the comparison result, wherein:

[0031] If the safety speed is greater than or equal to the preset speed, the obstacle avoider determines to maintain the movement route of the robot;

[0032] If the safety speed is less than the preset speed, the obstacle avoider determines to re-plan the movement route of the robot;

[0033] The preset speed is positively correlated with the distance of a single movement of the robot.

[0034] Further, the air spray gun device turns on the horizontal air spray gun and / or the rotary air spray gun when receiving the avoidance signal, wherein:

[0035] If the detector scans that the obstacle exists within the preset distance, the rotating air spray gun is turned on and the horizontal air spray gun is rotated to point to the obstacle and emit a single jet of air flow;

[0036] If the pressure sensor detects that the robot collides with the obstacle, the horizontal air spray gun is rotated to point to the obstacle and continuously sprays air flow toward the obstacle until the detector scans that the collision distance between the robot and the obstacle is greater than the preset distance;

[0037] The air flow includes the single-jet air flow and the continuous-jet air flow.

[0038] Furthermore, when the obstacle avoider sends out the avoidance signal, the visual monitor monitors the direction of the obstacle detected by the detector or the pressure sensor, and cooperates with the air spray gun device to send out an air flow.

[0039] Furthermore, the modeler updates the obstacle model when the robot repeatedly passes through the robot's moving route, compares the obstacle models before and after the update, and re-marks the static obstacles and the dynamic obstacles in the obstacle model according to the comparison results.

[0040] Furthermore, when the pressure sensor detects that the robot collides with the obstacle, the obstacle avoider drives the detector to scan the characteristic information of the obstacle, compares the characteristic information with the obstacle model, and analyzes the cause of the collision based on the comparison result.

[0041] Compared with the prior art, the beneficial effect of the present invention lies in that the system of the present invention models the robot's travel route by setting a modeler, and can quickly search and generate multiple feasible paths from the starting point to the end point in the virtual model space. The modeling can not only generate feasible paths, but also evaluate and compare these paths. Each path can be quantitatively analyzed according to different standards, such as path length, number of turns, number of dangerous areas passed, etc. The robot can select the optimal path from them, and select the route with the shortest path length and the least number of turns, thereby improving the robot's travel efficiency, while reducing energy consumption and walking time, effectively improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0042] Furthermore, by setting up a visual monitor to predict the robot's route, by predicting the robot's possible future route, the path planning algorithm can comprehensively consider subsequent environmental changes, task requirements and other factors at the current moment, so as to plan a more forward-looking optimal path. The prediction of the robot's route can allow the robot to understand the laws and trends of human activities in advance, so as to better adapt to human activities and avoid inconvenience or interference to humans. Predicting the robot's route can help to reasonably arrange the robot's energy consumption. According to factors such as the length of the predicted route, the complexity of the terrain, and the expected walking speed, the amount of energy required for the robot to complete the task can be estimated, so as to make an energy replenishment plan in advance or optimize the robot's motion parameters to reduce energy consumption, ensuring that the task will not be interrupted due to insufficient energy, improving the accuracy of obstacle recognition, and improving the success rate of avoiding moving obstacles. Obstacles that appear at a very close distance can be effectively avoided, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0043] Furthermore, the present invention provides an air spray gun device to emit an air flow toward the obstacle when it receives the avoidance signal. Different from the traditional method of sensing and avoiding obstacles based on mechanical structure collision and contact, emitting air for obstacle avoidance is a non-contact method. The robot does not need to directly collide physically with the obstacle, which greatly reduces the risk of damage to the robot's own components due to collision, avoids the occurrence of shell deformation, sensor failure, joint wear, etc., and can also prevent damage to obstacles. It can be used in some scenarios with high requirements on the integrity of the environment or obstacles, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0044] Furthermore, the present invention timely updates the obstacle model and can promptly respond to the emergence of new obstacles. In the robot's operating environment, new obstacles may appear at any time. For example, in an indoor scene, someone suddenly places a box, or in an outdoor scene, a vehicle temporarily stops, etc. Through obstacle avoidance update modeling, the robot can use the new obstacle information perceived in real time to quickly update the model and incorporate the newly appeared obstacles into the obstacle avoidance consideration range, so as to timely adjust the route to avoid them, ensuring that the robot can still travel safely and efficiently in a complex and changeable environment, and can effectively optimize the subsequent action route in the event of a sudden collision, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a structural schematic diagram of a robot obstacle avoidance inspection system based on visual monitoring of the present invention;

[0046] Figure 2 A decision diagram for planning a robot movement route according to an embodiment of the present invention;

[0047] Figure 3 It is a schematic diagram of the top view of the air spray gun device according to an embodiment of the present invention;

[0048] Figure 4 It is a side structural schematic diagram of an air spray gun device according to an embodiment of the present invention;

[0049] Among them, 1, horizontal air spray gun; 11, air outlet; 12, air inlet; 2, rotating air spray gun; 3, camera; 4, robot; 5, obstacle; 6, wheel. DETAILED DESCRIPTION

[0050] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0051] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0052] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0053] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0054] See also Figure 1 As shown, it is a structural schematic diagram of a robot obstacle avoidance inspection system based on visual monitoring of the present invention, a robot obstacle avoidance inspection system based on visual monitoring, comprising:

[0055] A detector for scanning characteristic information of obstacles in the robot's moving area;

[0056] Among them, the characteristic information of the obstacle includes the shape outline, color texture and collision distance of the obstacle;

[0057] A modeler, which is connected to the detector and is used to build an obstacle model according to characteristic information of the obstacles and mark static obstacles and dynamic obstacles in the obstacle model;

[0058] A visual monitor, which is connected to the modeler and the detector respectively, is used to monitor the movement trajectory of the obstacle within a preset time, and predict the predicted movement trajectory of the obstacle in the next period of time in combination with the obstacle model;

[0059] The obstacle avoider is provided with an airbag and a pressure sensor, which are respectively connected to the detector, the visual monitor and the modeler, and are used to plan the robot's moving route in combination with the obstacle model and the predicted action trajectory and adjust the robot's moving state during the movement, and send an avoidance signal when the detector scans an obstacle within a preset distance or the pressure sensor detects that the robot collides with the obstacle;

[0060] The air spray gun device comprises a horizontal air spray gun and a rotating air spray gun, which are respectively connected to a visual monitor and an obstacle avoider to receive an avoidance signal and start the rotating air spray gun when no collision occurs, and to aim the rotating horizontal air spray gun at the obstacle to emit an air flow.

[0061] Specifically, the system of the present invention models the robot's travel route by setting a modeler, and can quickly search and generate multiple feasible paths from the starting point to the end point in the virtual model space. The modeling can not only generate feasible paths, but also evaluate and compare these paths. Each path can be quantitatively analyzed according to different standards, such as path length, number of turns, number of dangerous areas passed, etc. The robot can then select the optimal path from them, and select the route with the shortest path length and the least number of turns, thereby improving the robot's travel efficiency. While reducing energy consumption and walking time, it effectively improves the accuracy and practicality of the robot's obstacle avoidance inspection system based on visual monitoring.

[0062] Specifically, the obstacle avoider combines the obstacle model and the predicted action trajectory to plan the robot's movement route, where

[0063] If the obstacle model contains static obstacles but no dynamic obstacles, the obstacle avoider plans the robot's moving route through a global planning algorithm;

[0064] If the obstacle model contains dynamic obstacles and static obstacles, the obstacle avoider constructs a grid map according to the obstacle model, generates the travel cost of each grid on the grid map corresponding to the predicted action trajectory, and plans the robot's movement route according to the travel cost;

[0065] If the obstacle model contains no dynamic obstacles and no static obstacles, the obstacle avoider drives the robot to move in a straight line to the end point;

[0066] During implementation, the obstacle avoidance system continuously scans the surrounding environment through lidar, camera and ultrasound to obtain relevant information about the environment. The lidar can accurately measure the distance between the robot and surrounding obstacles and present it in the form of point cloud data. The camera provides visual images for identifying objects.

[0067] In implementation, if the obstacle model contains static obstacles but no dynamic obstacles, the obstacle avoider plans the robot's moving route through a global planning algorithm. First, a corresponding map is constructed according to the robot's environment, and the corresponding nodes of the robot's starting position and target position in the map are determined. Then, the Di jkstra algorithm is used for calculation. During the calculation process, as the node closest to the source node is continuously selected for access, the distance information from each node on the map to the source node will be continuously updated, and the shortest distance path information from the starting position to other positions will be gradually formed. Finally, when the target node is visited, starting from the target node, by tracing back to each node on the path to the source node, the shortest path from the starting position to the target position can be obtained, which is the robot's moving route.

[0068] In implementation, if the obstacle model contains dynamic obstacles and static obstacles, the obstacle avoider constructs a grid map according to the obstacle model, generates the pass cost of each grid on the grid map corresponding to the predicted action trajectory, plans the robot's moving route according to the pass cost, and divides the robot's environmental space into several square grids of equal size according to certain rules. Each grid has a unique two-dimensional coordinate identifier in the map, which is used to indicate its position in the map. The environmental area corresponding to each grid is marked as different states. The states include: passable, which means that there are no obstacles in the grid area and the robot can theoretically pass smoothly. Obstacle, if there is an obstacle in the grid area, then the grid is marked as an obstacle state, unknown, when the robot has not fully scanned or detected a certain area, the grid corresponding to the area will be marked as an unknown state. As the robot moves and the sensor continues to scan the environment, the grid in the unknown state will gradually be updated to a passable or obstacle state. The probability of the robot encountering an obstacle in different grids is recorded as the pass cost of the corresponding grid, and the path with the smallest pass cost is selected as the robot's moving route.

[0069] If the obstacle model does not contain dynamic obstacles and does not contain static obstacles, the obstacle avoider drives the robot to move in a straight line to the end point. At this time, the obstacle avoider determines that there are no obstacles from the starting point to the end point.

[0070] The grid map divides the traversable routes in the obstacle model into several grids, and the travel cost is the collision probability of the robot traveling to the grid.

[0071] Specifically, the obstacle avoider adjusts the robot's travel state according to the obstacles scanned on the robot's moving route, where:

[0072] If the collision distance of the obstacle scanned by the detector is greater than or equal to the preset distance, the robot's moving speed is reduced;

[0073] If the collision distance of the obstacle scanned by the detector is less than the preset distance, the robot stops moving and sends an avoidance signal;

[0074] In the implementation, the preset distance is set to 1 meter. If the collision distance of the obstacle scanned by the detector is 1.5 meters, which is greater than the preset distance of 1 meter, the movement speed of the robot is reduced;

[0075] If the collision distance of the obstacle scanned by the detector is 0.8 meters, which is less than the preset distance of 1 meter, the robot stops moving and sends an avoidance signal.

[0076] Among them, the preset distance is positively correlated with the robot's moving speed.

[0077] It is understandable that the faster the robot moves, the greater its inertia, the longer it takes to stop moving when the detector detects an obstacle, the greater the braking distance at this time, and the greater the preset distance.

[0078] Optionally, the robot moves at a speed of 0.1 m / s and the preset distance is 0.5 m;

[0079] The robot moves at a speed of 0.5 meters per second and the preset distance is 1 meter;

[0080] The robot moves at a speed of 1 meter per second and the preset distance is 2 meters.

[0081] Specifically, the visual monitor monitors the movement trajectory of the obstacle within a preset time when the obstacle avoider reduces the robot's movement speed, and predicts the predicted movement trajectory of the obstacle in the next period of time in combination with the obstacle model, where:

[0082] If the obstacle does not move within the preset time, the visual monitor predicts that the obstacle will not move within the next period of time;

[0083] If the obstacle moves within the preset time, the visual monitor will record the obstacle's moving trajectory on the obstacle model, and predict the obstacle's predicted movement trajectory in the next period of time based on the direction and speed of the moving trajectory;

[0084] During implementation, when the obstacle avoider reduces the robot's moving speed, the visual monitor monitors the moving trajectory of the obstacle within a preset time, and predicts the predicted movement trajectory of the obstacle in the next period of time in combination with the obstacle model. The visual monitor continuously collects the position coordinates of the obstacle at regular time intervals within the preset time. Under the premise of ensuring the accuracy of the monitoring results, this time interval can be selected as 0.1 seconds and 0.5 seconds. According to the direction and speed of the obstacle's moving trajectory, the predicted movement trajectory of the obstacle in the next period of time is predicted.

[0085] Among them, the preset time is negatively correlated with the robot's moving speed; the subsequent time is a time period that connects to the preset time and has the same duration as the preset time.

[0086] It is understandable that the faster the robot moves, the shorter the time it takes to approach an obstacle, the shorter the robot's prediction time, and the shorter the preset time.

[0087] Optionally, the robot moves at a speed of 0.1 m / s and the preset time is 10 seconds;

[0088] The robot moves at a speed of 0.5 meters per second, and the preset time is 3 seconds;

[0089] The robot moves at a speed of 1 meter per second and the preset time is 1 second.

[0090] Specifically, by setting up a visual monitor to predict the robot's route, and by predicting the robot's possible future route, the path planning algorithm can comprehensively consider subsequent environmental changes, task requirements and other factors at the current moment, thereby planning a more forward-looking optimal path. The prediction of the robot's route allows the robot to understand the laws and trends of human activities in advance, so as to better adapt to human activities and avoid inconvenience or interference to humans. Predicting the robot's route can help to reasonably arrange the robot's energy consumption. According to factors such as the length of the predicted route, the complexity of the terrain, and the expected walking speed, the amount of energy required for the robot to complete the task can be estimated, so that energy replenishment plans can be made in advance or the robot's motion parameters can be optimized to reduce energy consumption, ensuring that the task will not be interrupted due to insufficient energy, improving the accuracy of obstacle recognition, and improving the success rate of avoiding moving obstacles. Obstacles that appear at a very close distance can be effectively avoided, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0091] Specifically, the obstacle avoider replans the robot's moving route when the visual monitor predicts that the obstacle will not move in the next period of time;

[0092] The obstacle avoider calculates the safe speed when the visual monitor predicts the predicted movement trajectory of the obstacle in the next period of time;

[0093] In implementation, the safe speed is the ratio of the distance of the predicted action trajectory predicted by the obstacle avoider to the immediate following time.

[0094] The safe speed is the speed at which the robot can move along its path without causing a collision.

[0095] See also Figure 2 As shown, it is a decision diagram for planning the robot's moving route according to an embodiment of the present invention. The obstacle avoider compares the safe speed with the preset speed and determines whether to re-plan the robot's moving route according to the comparison result.

[0096] If the safe speed is greater than or equal to the preset speed, the obstacle avoider determines to maintain the robot's moving route;

[0097] If the safe speed is less than the preset speed, the obstacle avoider will decide to re-plan the robot's moving route;

[0098] In implementation, the preset rate is 1 meter per second. If the safe rate is 1.5 meters per second, which is greater than the preset rate of 1 meter per second, the obstacle avoider determines to maintain the robot's moving route;

[0099] If the safety speed is 0.7 meters per second, which is less than the preset speed of 1 meter per second, the obstacle avoider will decide to replan the robot's moving route.

[0100] The preset speed is positively correlated with the distance the robot moves in a single time.

[0101] It can be understood that, when the robot's moving time is constant, the longer the robot's single moving distance is, the greater its moving speed is, and the greater the preset speed is.

[0102] Optionally, the robot can move 10 meters in one go, with a preset rate of 0.1 seconds per second;

[0103] The robot can move 20 meters at a time, with a preset rate of 0.3 seconds per second;

[0104] The robot can move 30 meters at a time, with a preset speed of 0.5 seconds per second.

[0105] Please cooperate Figure 3 See also Figure 4As shown, they are respectively a top view structural schematic diagram of the air spray gun device according to an embodiment of the present invention and a side view structural schematic diagram of the air spray gun device according to an embodiment of the present invention, the horizontal air spray gun 1 includes an air outlet 11 and an air inlet 12; the horizontal air spray gun 1 is arranged at the lower part of the robot 4; the rotating air spray gun 2 is arranged below the horizontal air spray gun 1; the camera 3 is arranged above the horizontal air spray gun 1; and the wheel 6 is arranged at the bottom of the robot 4.

[0106] Specifically, the air spray gun device opens the horizontal air spray gun and / or the rotating air spray gun when receiving the avoidance signal, wherein:

[0107] If the detector scans an obstacle within the preset distance, the rotating air spray gun is turned on and the rotating horizontal air spray gun is pointed at the obstacle and emits a single jet of air flow;

[0108] If the pressure sensor detects that the robot collides with an obstacle, the horizontal air spray gun is rotated to point to the obstacle and continuously sprays air flow toward the obstacle until the detector scans that the collision distance between the robot and the obstacle is greater than the preset distance;

[0109] The air flow includes a single jet air flow and a continuous jet air flow.

[0110] Specifically, when the obstacle avoider sends out an avoidance signal, the visual monitor detects the direction of the obstacle through the monitoring detector or pressure sensor, and cooperates with the air spray gun device to send out an air flow.

[0111] In implementation, the obstacle avoider is provided with an airbag and a pressure sensor. The airbag is provided at a key position on the robot that may collide with external objects, and the pressure sensor is installed inside the airbag or at a position closely connected to the airbag. When the airbag is squeezed due to a collision, the pressure inside the airbag will change. The pressure sensor can accurately sense this pressure change and convert the pressure value into a corresponding electrical signal based on its internal conversion mechanism. This electrical signal can be received and processed by the robot's obstacle avoider.

[0112] In implementation, if the detector scans an obstacle within the preset distance, the rotating air spray gun is turned on. At this time, the rotating air spray gun changes the robot's moving direction, and the rotating horizontal air spray gun points to the obstacle and emits a single jet of air flow. The robot obtains a directional force to slow down, and at the same time, it issues a warning to the moving obstacle;

[0113] If the pressure sensor detects that the robot collides with an obstacle, the horizontal air spray gun is rotated to point at the obstacle and continuously sprays air towards the obstacle until the detector scans that the collision distance between the robot and the obstacle is greater than the preset distance. While issuing a warning to the obstacle, the robot is driven away from the obstacle.

[0114] Specifically, the present invention provides an air spray gun device to emit an air flow to the obstacle when it receives an avoidance signal. Different from the traditional method of sensing and avoiding obstacles based on mechanical structure collision and contact, emitting air for obstacle avoidance is a non-contact method. The robot does not need to directly collide physically with the obstacle, which greatly reduces the risk of damage to the robot's own components due to collision, avoids the occurrence of shell deformation, sensor failure, joint wear, etc., and can also prevent damage to obstacles. It can be used in some scenarios with high requirements on the integrity of the environment or obstacles, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0115] Specifically, the modeler updates the obstacle model when the robot repeatedly passes through the robot's moving route, compares the obstacle models before and after the update, and re-marks the static obstacles and dynamic obstacles in the obstacle model based on the comparison results.

[0116] Specifically, when the pressure sensor detects that the robot collides with an obstacle, the obstacle avoider drives the detector to scan the characteristic information of the obstacle, compares the characteristic information with the obstacle model, and analyzes the cause of the collision based on the comparison results.

[0117] In implementation, the detector includes a camera for collecting the texture and color of the obstacle, a lidar for determining the distance between the robot and the obstacle and the shape of the obstacle, and an ultrasonic sensor for measuring the distance between the obstacle and the robot within a short distance. After a collision occurs, the detector scans the obstacle to obtain characteristic information of the obstacle. The characteristic information includes the color, texture, shape contour and mobility of the obstacle. By comparing the obstacle with similar obstacles in the obstacle model, it is determined whether the obstacle is a static obstacle or a dynamic obstacle, thereby obtaining the cause of the collision.

[0118] Specifically, the present invention updates the obstacle model in a timely manner and can respond to the emergence of new obstacles in a timely manner. In the robot's operating environment, new obstacles may appear at any time. For example, in an indoor scene, someone suddenly places a box, or in an outdoor scene, a vehicle temporarily stops, etc. Through obstacle avoidance update modeling, the robot can use the new obstacle information perceived in real time to quickly update the model, and include the newly appeared obstacles in the obstacle avoidance consideration range, so as to adjust the route in time to avoid them, ensuring that the robot can still travel safely and efficiently in a complex and changeable environment, and can effectively optimize the subsequent action route in the event of a sudden collision, further improving the accuracy and practicality of the robot obstacle avoidance inspection system based on visual monitoring.

[0119] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A robot obstacle avoidance inspection system based on visual monitoring, characterized in that: include: A detector for scanning characteristic information of obstacles in the robot's moving area; The characteristic information of the obstacle includes the shape, outline, color, texture and collision distance of the obstacle; A modeler connected to the detector, configured to construct an obstacle model according to the characteristic information of the obstacle and mark static obstacles and dynamic obstacles in the obstacle model; A visual monitor, which is connected to the modeler and the detector, respectively, to monitor the movement trajectory of the obstacle within a preset time, and predict the predicted movement trajectory of the obstacle within a subsequent time in combination with the obstacle model; an obstacle avoider, which is provided with an airbag and a pressure sensor, which are respectively connected to the detector, the visual monitor and the modeler, and are used to plan the robot's moving route in combination with the obstacle model and the predicted action trajectory and adjust the robot's moving state during the moving process, and send an avoidance signal when the detector scans that there is an obstacle within a preset distance or the pressure sensor detects that the robot collides with the obstacle; The air spray gun device includes a horizontal air spray gun and a rotary air spray gun, which are respectively connected to the visual monitor and the obstacle avoider, for receiving the avoidance signal and starting the rotary air spray gun when no collision occurs, and rotating the horizontal air spray gun to emit an air flow toward the obstacle.

2. The robot obstacle avoidance inspection system based on visual monitoring according to claim 1 is characterized in that: The obstacle avoider plans the robot's moving route in combination with the obstacle model and the predicted motion trajectory, wherein: If the obstacle model includes the static obstacle but does not include the dynamic obstacle, the obstacle avoider plans the movement route of the robot through a global planning algorithm; If the obstacle model includes the dynamic obstacle and the static obstacle, the obstacle avoider constructs a grid map according to the obstacle model, generates a passage cost of each grid on the grid map corresponding to the predicted action trajectory, and plans the movement route of the robot according to the passage cost; If the obstacle model does not include the dynamic obstacle and does not include the static obstacle, the obstacle avoider drives the robot to move along a straight line to the end point; The grid map divides the traversable routes in the obstacle model into a plurality of grids, and the passage cost is the collision probability of the robot traveling to the grid.

3. The robot obstacle avoidance inspection system based on visual monitoring according to claim 2 is characterized in that: The obstacle avoider adjusts the robot's traveling state according to the situation that an obstacle is scanned on the robot's moving route, wherein: If the collision distance of the obstacle scanned by the detector is greater than or equal to the preset distance, reducing the movement speed of the robot; If the collision distance of the obstacle scanned by the detector is less than the preset distance, the robot stops moving and sends the avoidance signal; Wherein, the preset distance is positively correlated with the moving speed of the robot.

4. The robot obstacle avoidance inspection system based on visual monitoring according to claim 3 is characterized in that: The visual monitor monitors the movement trajectory of the obstacle within the preset time when the obstacle avoider reduces the movement speed of the robot, and predicts the predicted movement trajectory of the obstacle within the next time in combination with the obstacle model, wherein: If the obstacle does not move within the preset time, the visual monitor predicts that the position of the obstacle will not move within the next period of time; If the obstacle moves within the preset time, the visual monitor records the movement trajectory of the obstacle on the obstacle model, and predicts the predicted movement trajectory of the obstacle in the next period of time according to the direction and speed of the movement trajectory; The preset time is negatively correlated with the moving speed of the robot; the subsequent time is a time period that connects to the preset time and has the same duration as the preset time.

5. The robot obstacle avoidance inspection system based on visual monitoring according to claim 4 is characterized in that: The obstacle avoider replans the movement route of the robot when the visual monitor predicts that the position of the obstacle will not move within a short period of time; The obstacle avoider calculates a safe speed when the visual monitor predicts the predicted action trajectory of the obstacle within a short period of time; The safe speed is the moving speed of the robot along the moving route of the robot without collision.

6. The robot obstacle avoidance inspection system based on visual monitoring according to claim 5 is characterized in that: The obstacle avoider compares the safety speed with the preset speed and determines whether to re-plan the robot's moving route according to the comparison result, wherein: If the safety speed is greater than or equal to the preset speed, the obstacle avoider determines to maintain the movement route of the robot; If the safety speed is less than the preset speed, the obstacle avoider determines to re-plan the movement route of the robot; The preset speed is positively correlated with the distance of a single movement of the robot.

7. The robot obstacle avoidance inspection system based on visual monitoring according to claim 6 is characterized in that: The air spray gun device turns on the horizontal air spray gun and / or the rotary air spray gun when receiving the avoidance signal, wherein: If the detector scans that the obstacle exists within the preset distance, the rotating air spray gun is turned on and the horizontal air spray gun is rotated to point to the obstacle and emit a single jet of air flow; If the pressure sensor detects that the robot collides with the obstacle, the horizontal air spray gun is rotated to point to the obstacle and continuously sprays air flow toward the obstacle until the detector scans that the collision distance between the robot and the obstacle is greater than the preset distance; The air flow includes the single-jet air flow and the continuous-jet air flow.

8. The robot obstacle avoidance inspection system based on visual monitoring according to claim 7 is characterized in that: When the obstacle avoider sends out the avoidance signal, the visual monitor monitors the direction of the obstacle detected by the detector or the pressure sensor, and cooperates with the air spray gun device to send out an air flow.

9. The robot obstacle avoidance inspection system based on visual monitoring according to claim 8 is characterized in that: The modeler updates the obstacle model when the robot repeatedly passes through the robot's moving route, compares the obstacle models before and after the update, and re-marks the static obstacles and the dynamic obstacles in the obstacle model according to the comparison results.

10. The robot obstacle avoidance inspection system based on visual monitoring according to claim 9, characterized in that: When the pressure sensor detects that the robot collides with the obstacle, the obstacle avoider drives the detector to scan the characteristic information of the obstacle, compares the characteristic information with the obstacle model, and analyzes the cause of the collision based on the comparison result.

Images