Rail transit vehicle inspection robot collaborative operation obstacle avoidance method and system

Abstract

The invention provides a rail transit vehicle inspection robot collaborative operation obstacle avoidance method and system. The method comprises the following steps: acquiring obstacle position angle information; extracting contour features of the obstacle; obstacle position information and contour features are used as control input parameters, fuzzy logic is adopted to divide a control interval, and a control instruction of obstacle avoidance motion of the inspection robot is generated through rule matching; establishing a multi-machine cooperative operation rule; the motion control instruction is analyzed according to the multi-stage cooperative operation rule, the target parameter is converted into the rotating speed of the driving wheel through the motion model, the motor control instruction is adjusted in real time to complete the obstacle avoidance action, and meanwhile closed-loop correction is achieved in combination with sensor feedback. According to the invention, technologies of multi-module integrated sensing, wireless remote communication, image processing, cooperative control and the like are adopted, and a control rule is designed based on an artificial intelligence algorithm, so that the inspection robot can automatically recognize and accurately control when facing an obstacle, and thus the obstacle avoidance performance of the inspection robot is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of vehicle inspection, in particular to a rail transit vehicle inspection robot cooperative operation obstacle avoidance method and system. BACKGROUND

[0002] With the rapid growth of EMUs and rail transit facilities, the maintenance task pressure of the maintenance depot applied in the railway field is also increasing. The traditional manual inspection has problems such as high labor cost, low inspection efficiency, large inspection workload, and complicated operation process. With the advantages of high efficiency, energy saving, and strong adaptability, the inspection robot is introduced into the daily inspection task of the maintenance depot, which can intelligently judge the abnormal state of key components such as bogies, couplers, underfloor electrical equipment, air compressors, and pneumatic pipelines. And it can be inspected and monitored all day long, thereby effectively reducing the working intensity of the staff.

[0003] In order to ensure the smooth completion of the inspection task, effective obstacle avoidance needs to be carried out according to the inspection environment, especially in the case of cooperative operation of the inspection robot system, the design of the obstacle avoidance method becomes the current research focus.

[0004] Some scholars complete hierarchical obstacle avoidance control through GVO (generalized velocity obstacle) technology, perform safety control on the collision risk faced in the identification process, and select the optimal obstacle avoidance scheme by using a target fitness evaluation function, but the calculation process of this method is too complex and performs poorly in the face of specific tasks. Another part of the scholars establishes a traffic rule reservation table in combination with the actual situation to realize the cluster obstacle avoidance of multiple robots, but the cooperative operation obstacle avoidance accuracy of this method is low, and the obstacle avoidance effect depends on the setting of conditions. Some scholars further analyze the completeness of the current obstacle avoidance method and clearly define the typical obstacle avoidance operation characteristics of intelligent robots. Through comprehensive analysis of various obstacle avoidance methods, the intelligent robot can judge the information of the dangerous area and the surrounding environment and plan a good motion path in the autonomous working process. However, the information obtained by this method is not detailed, resulting in poor obstacle avoidance effect. SUMMARY

[0005] The present application provides a rail transit vehicle inspection robot cooperative operation obstacle avoidance method and system, which can solve the technical problem of poor obstacle avoidance effect of the maintenance depot inspection robot system in the cooperative operation process in the prior art.

[0006] In a first aspect, the present application provides a rail transit vehicle inspection robot cooperative operation obstacle avoidance method, comprising the following steps: According to the motion model of the rail transit vehicle inspection robot, the pose parameters of the inspection robot are calculated by using the kinematics principle to obtain the position angle information of the obstacle; The collected environment image is preprocessed to extract the contour features of the obstacle; The obstacle position information and the contour feature are taken as control input parameters, fuzzy logic is used to divide the control interval, and the control instruction of the obstacle avoidance movement of the inspection robot is generated through rule matching. A multi-machine cooperative operation rule of the inspection robot is established. According to the multi-level cooperative operation rule, the motion control instruction is analyzed, the target parameter is converted into the driving wheel speed through the motion model, and the motor control instruction is adjusted in real time to complete the obstacle avoidance action.

[0007] Further, according to the motion model of the rail transit vehicle inspection robot, the pose parameters of the inspection robot are calculated by using the kinematics principle, and the obstacle position angle information is obtained, which specifically includes the following steps: Based on the double-wheel differential driving structure of the inspection robot, according to the geometric relationship between the left and right driving wheel speeds and the wheelbase, the robot motion model is established through the kinematics differential equation, and the pose parameters of the robot in the global coordinate system are obtained. Based on the pose parameters, the laser scanner ranging data and the environmental image information of the vision sensor are fused, the global coordinates of the obstacle are converted into the polar coordinates of the robot body through the homogeneous coordinate transformation matrix, and the distance r and the azimuth angle φ of the obstacle relative to the robot are obtained.

[0008] Further, the motion model is:

[0009] In the formula: is a pose state vector, , is the horizontal coordinate and the vertical coordinate of the robot, , is the left and right driving wheel speed, is the diameter of the left and right wheels of the robot, is the centroid motion speed of the robot; is the angular velocity of the centroid, is the angle between the forward direction of the robot and the X-axis.

[0010] Further, the collected environmental image is preprocessed, and the contour feature of the obstacle is extracted, which specifically includes the following steps: The collected original color image is processed by gray scale conversion to remove color information and retain brightness features, and a gray scale image is obtained. Based on the obtained gray scale image, a binary processing method is used to convert the image into a binary image by setting a threshold value. For the binary image, an edge detection algorithm is used to analyze the pixel gray scale change and identify the edge information in the image. The distance data obtained is used to map edge information in the image to a three-dimensional coordinate system, to calculate the actual contour size and position of the obstacle, and to obtain the contour features of the obstacle.

[0011] Further, the obstacle position information and the contour features are used as control input parameters, fuzzy logic is used to divide the control interval, and a control instruction for obstacle avoidance movement of the inspection robot is generated through rule matching, specifically including the following steps: According to the obtained obstacle position information and the contour features, the angle and distance parameters of the obstacle relative to the inspection robot are extracted as input variables of the control system; Based on fuzzy control theory, the obstacle angle parameter is divided into multiple fuzzy intervals, each fuzzy interval is provided with a corresponding membership function, and a fuzzy processing result of the input variable is established; According to the movement characteristics of the inspection robot, the output control quantity is divided into multiple levels, each level corresponds to different movement speed and steering angle, and a fuzzy processing result of the output variable is established; The fuzzy processing results of the input variable and the output variable are combined to construct a fuzzy control rule library, and the input parameters are mapped to the output control quantity through rule matching; The output control quantity obtained through fuzzy reasoning is de-fuzzified to convert into accurate movement speed and steering angle parameters of the inspection robot, and a final control instruction is generated.

[0012] Further, the multi-robot cooperative operation rules include one-way driving traffic rules, distributed waiting mechanism, special collision processing rules, cooperative control architecture and abnormal processing mechanism.

[0013] Further, the movement control instruction is parsed according to the multi-level cooperative operation rules, the target parameters are converted into driving wheel rotation speed through a movement model, and the motor control instruction is adjusted in real time to complete the obstacle avoidance action, and a closed loop correction is realized by combining sensor feedback, specifically including the following steps: Target movement parameters including target speed and steering angle are extracted from the movement control instruction; Based on the movement model of the inspection robot, the target speed and steering angle are converted into target rotation speed of the left and right driving wheels; Combined with the multi-robot cooperative operation rules, the driving wheel rotation speed is dynamically adjusted to obtain adjusted left and right driving wheel rotation speed instructions; The adjusted left and right driving wheel rotation speed instructions are transmitted to the actuator to control the motor to operate according to the parameters, and the pose information is fed back in real time through the sensor for closed loop correction.

[0014] In a second aspect, a track traffic vehicle inspection robot cooperative operation obstacle avoidance system includes: The position angle information acquisition module is configured to calculate the position and posture parameters of the inspection robot by using kinematics principles according to a motion model of the rail transit vehicle inspection robot, and acquire the position angle information of the obstacle. The contour feature acquisition module is configured to preprocess the collected environment image and extract the contour feature of the obstacle. The control instruction acquisition module is in communication connection with the position angle information acquisition module and the contour feature acquisition module, configured to take the position information and the contour feature of the obstacle as control input parameters, divide the control interval by using fuzzy logic, and generate the control instruction of the obstacle avoidance motion of the inspection robot through rule matching. The cooperative operation rule establishment module is configured to establish the multi-robot cooperative operation rule of the inspection robot. The obstacle avoidance execution module is in communication connection with the control instruction acquisition module and the cooperative operation rule establishment module, configured to parse the motion control instruction according to the multi-stage cooperative operation rule, convert the target parameter into the driving wheel rotation speed through the motion model, and adjust the motor control instruction in real time to complete the obstacle avoidance action.

[0015] Further, the position angle information acquisition module comprises: The posture parameter acquisition unit is configured to establish the motion model of the robot by kinematics differential equations according to the geometric relationship between the speed of the left and right driving wheels and the wheelbase based on the differential driving structure of the inspection robot, and acquire the posture parameters of the robot in the global coordinate system. The position angle information acquisition unit is in communication connection with the posture parameter acquisition unit, configured to fuse the ranging data of the laser scanner and the environment image information of the vision sensor based on the posture parameters, convert the global coordinates of the obstacle into the polar coordinates of the robot body through the homogeneous coordinate transformation matrix, and acquire the distance and azimuth angle of the obstacle relative to the robot.

[0016] In a third aspect, the present application provides a computer readable storage medium, and the computer readable storage medium stores a rail transit vehicle inspection robot cooperative operation obstacle avoidance program, wherein when the rail transit vehicle inspection robot cooperative operation obstacle avoidance program is executed by a processor, the steps of the rail transit vehicle inspection robot cooperative operation obstacle avoidance method according to any one of claims 1 to 7 are implemented.

[0017] The technical scheme provided by the embodiments of the present application has at least the following beneficial effects: By using the multi-module integrated perception, wireless remote communication, image processing, cooperative control and other technologies, and relying on the artificial intelligence algorithm to design the control rule, the inspection robot can automatically identify and accurately control when facing the obstacle, so as to improve the obstacle avoidance performance of the inspection robot, and prolong the service life of the robot while improving the work efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0019] Figure 1 A flowchart of the track traffic vehicle inspection robot cooperative operation obstacle avoidance method provided by the embodiment of the present application is shown in the figure. Figure 2 A motion diagram of the inspection robot provided by the embodiment of the present application is shown in the figure. Figure 3 A process diagram of image processing provided by the embodiment of the present application is shown in the figure. Figure 4 An artificial intelligence control structure diagram provided by the embodiment of the present application is shown in the figure. Figure 5 A curve diagram of input variable membership function provided by the embodiment of the present application is shown in the figure. Figure 6 A curve diagram of output variable membership function provided by the embodiment of the present application is shown in the figure. Figure 7 A system composition structure diagram provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present application more clear, the technical solutions in the present application will be described clearly and completely in the following with reference to the drawings in the present application. Obviously, the described embodiments are some embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] The terms "include" and "have" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but can optionally include steps or units not listed or can optionally include other steps or units inherent to the process, method, product or device. The terms "first", "second" and "third" and the like descriptions are used to distinguish different objects, and do not represent the order or limit the types of "first", "second" and "third".

[0022] In the description of the embodiments in this application, terms such as "exemplary," "for example," or "for instance" are used as examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a specific manner.

[0023] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0024] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0026] Firstly, such as Figure 1 As shown, this application provides a method for collaborative obstacle avoidance by rail transit vehicle inspection robots, including the following steps: Step S1: Based on the motion model of the rail transit vehicle inspection robot, calculate the pose parameters of the inspection robot using kinematic principles to obtain obstacle position and angle information; Step S2: Preprocess the acquired environmental images to extract the contour features of obstacles; Step S3: Using obstacle location information and contour features as control input parameters, fuzzy logic is used to divide the control interval, and control commands for the inspection robot's obstacle avoidance movement are generated through rule matching; Step S4: Establish multi-machine collaborative operation rules for the inspection robot; Step S5: Parse motion control commands according to multi-level collaborative operation rules, convert target parameters into drive wheel speeds through motion models, and adjust motor control commands in real time to complete obstacle avoidance actions. At the same time, combine sensor feedback to achieve closed-loop correction.

[0027] The application adopts multi-module integrated perception, wireless remote communication, image processing, cooperative control and other technologies, relies on artificial intelligence algorithm to design control rules, so that the inspection robot can automatically identify and accurately control when facing obstacles, thereby improving the obstacle avoidance performance of the inspection robot, and prolonging the working life of the robot while improving the working efficiency.

[0028] In an embodiment, the step S1: according to the motion model of the rail transit vehicle inspection robot, the pose parameters of the inspection robot are calculated by using the kinematics principle, and the position angle information of the obstacle is obtained, which specifically includes the following steps: Step S11: based on the double-wheel differential driving structure of the inspection robot, according to the geometric relationship between the speed of the left and right drive wheels and the wheel base, the robot motion model is established by the kinematics differential equation, and the pose parameters of the robot in the global coordinate system are obtained; Step S12: based on the pose parameters, the laser range finder ranging data and the environment image information of the vision sensor are fused, the global coordinates of the obstacle are converted into the polar coordinates of the robot body through the homogeneous coordinate transformation matrix, and the distance r and the azimuth angle φ of the obstacle relative to the robot are accurately obtained.

[0029] This embodiment realizes the accurate cooperative perception of the pose of the inspection robot and the information of the obstacle by fusing the double-wheel differential kinematics model and the multi-sensor data. Specifically, the kinematics differential equation established based on the geometric constraint of the differential driving wheel speed and the wheel base can solve the pose parameters (x, y, θ) of the robot in the global coordinate system in real time; by introducing the homogeneous coordinate transformation of the laser ranging data and the vision information, the global obstacle coordinates are dynamically mapped into the polar coordinates (r, φ) of the robot body, which not only solves the problem of limited view angle of the traditional single sensor, but also significantly improves the accuracy of the relative position detection of the obstacle and the azimuth angle resolution, providing a high-reliability spatial situation awareness basis for subsequent path planning and obstacle avoidance decision.

[0030] In a specific embodiment, step S1: according to the motion model of the rail transit vehicle inspection robot, the pose parameters of the inspection robot are calculated by using the kinematics principle, and the position angle information of the obstacle is obtained, which specifically includes the following steps: For the motion state of the inspection robot, an inspection robot motion diagram is established, as shown in Figure 2 The motion equation of the inspection robot is:

[0031] In the formula, is the centroid motion speed of the robot; is the angular velocity of the centroid, is the angle between the forward direction of the robot and the X-axis.

[0032] The relationship between the driving speed and the linear velocity, angular velocity of the robot is:

[0033] In the formula, the driving speed includes and , , are the moving speeds of the left and right driving wheels of the robot respectively; d are the diameters of the left and right wheels; if is equal to , it indicates that the running path of the inspection robot is a straight line; if is not equal to , it indicates that the inspection robot runs in rotation.

[0034] The calculation formula of the motion vector of the inspection robot is:

[0035] According to the above calculation formula and combined with motion analysis, the motion model of the inspection robot can be constructed, the motion of the inspection robot is clarified, and the position angle information of the obstacle is obtained, as shown in the following formula, so as to facilitate the subsequent obstacle avoidance control design:

[0036] In the formula, is a pose state vector, , are the horizontal coordinate and vertical coordinate of the robot respectively, , are the left and right driving wheel speeds respectively, are the diameters of the left and right wheels of the robot, is the motion speed of the center of mass of the robot; is the angular velocity of the center of mass, is the included angle between the forward direction of the robot and the coordinate axis X.

[0037] In an embodiment, as shown in Figure 3 , the step S2 of preprocessing the collected environmental image and extracting the contour features of the obstacle includes the following steps: Through the sensors and cameras of the perception module, the surrounding environment information and pictures are obtained; the original color image collected is subjected to gray scale conversion processing, the color information is removed and the brightness features are retained, the gray scale image is obtained, and the main features of the obstacle are retained; Based on the obtained gray scale image, a binary processing method is adopted, the image is converted into a binary image by setting a threshold value, specifically, the pixels of the gray scale image are set to 0 or 255, and the contrast between the target and the background is enhanced;​​ The regions with different gray values in the binary image have boundaries, based on which edge information can be detected, therefore, edge detection algorithm is used to analyze the change of pixel gray, and the edge information in the image is recognized; The laser ranging technology is used to obtain the wave beam reflected by the obstacle, the distance data obtained by calculating the distance between the receiving end and the obstacle through the distance formula, the edge information in the image is mapped to the three-dimensional coordinate system, the actual contour size and position of the obstacle are calculated, and the contour feature of the obstacle is obtained.

[0038] The embodiment realizes accurate extraction of the contour feature of the obstacle through the cooperative application of multi-modal perception and image processing technology. Firstly, the original color image is grayed and binarized, which effectively enhances the brightness feature of the target object and the background contrast; then, the edge detection algorithm is combined to accurately identify the boundary information of the object in the image. Through the spatial mapping of laser ranging data and visual information, the two-dimensional image edge is converted into three-dimensional coordinates, which not only overcomes the limitations of single visual sensor in depth perception, but also restores the three-dimensional contour feature and spatial position relationship of the obstacle, providing high-precision environmental geometric feature support for subsequent obstacle avoidance decision.

[0039] In an embodiment, the step S3: taking the obstacle position information and the contour feature as control input parameters, using fuzzy logic to divide the control interval, and generating the control instruction of the obstacle avoidance movement of the inspection robot through rule matching, specifically includes the following steps: According to the obtained obstacle position information and contour feature, the angle and distance parameters of the obstacle relative to the inspection robot are extracted as input variables of the control system; Based on the fuzzy control theory, the obstacle angle parameter is divided into multiple fuzzy intervals, each fuzzy interval sets a corresponding membership function, and the fuzzy processing result of the input variable is established; According to the motion characteristics of the inspection robot, the output control quantity is divided into multiple levels, each level corresponds to different motion speed and steering angle, and the fuzzy processing result of the output variable is established; The fuzzy control rule base is constructed by combining the fuzzy processing results of the input variable and the output variable, and the input parameters are mapped to the output control quantity through rule matching; The output control quantity obtained by fuzzy reasoning is de-fuzzied to convert into accurate motion speed and steering angle parameters of the inspection robot, and the final control instruction is generated.

[0040] The embodiment realizes intelligent processing of obstacle avoidance decision of the inspection robot by fuzzy control theory. Based on the relative angle and distance parameters of the obstacle, the input information is divided into multiple semantic intervals by using the fuzzy method, and the corresponding membership functions are established, effectively solving the problem of insufficient adaptability of traditional control methods to uncertain environment. By constructing a fuzzy rule base, the input parameters and the motion state of the robot are intelligently matched, so that the system can automatically adjust the motion speed and turning angle according to the environmental characteristics. The final output of the precise control instruction not only ensures the stability of the obstacle avoidance action, but also takes into account the response speed in the dynamic environment, significantly improving the autonomous obstacle avoidance ability and motion coordination of the robot in complex scenes.

[0041] In an embodiment, the operation of the two wheels of the inspection robot depends on the position angle and distance of the obstacle. For the design of the obstacle avoidance method, an artificial intelligence controller is constructed, as shown in Figure 4 .

[0042] In the controller, the processed obstacle recognition information is taken as an input variable, and the controller performs inference processing. The output result of the controller is taken as the control amount of the operation speed of the robot. For the running direction of the inspection robot, the negative direction is set as left running, and vice versa. The obstacle avoidance process of the inspection robot is analyzed, the position angle of the obstacle is divided into 7 quantities, and the membership of the input variable is shown in Figure 5 . Figure 5 Fig. (a) in the middle is a membership function curve of the position angle of the obstacle provided by the embodiment of the application; Figure 5 Fig. (b) in the middle is a membership function curve of the effective distance of the obstacle; the operation speed of the two wheels of the inspection robot is represented by 5 quantities, and the membership of the output variable is shown in Figure 6 .

[0043] Based on the input and output variables, artificial intelligence control rules are constructed, and are presented in the form of a table, as shown in Table 1. The position angle and effective distance of the obstacle are used as reference conditions and basis for inquiry, and through the inquiry result of Table 1, the operation speed of the two wheels of the inspection robot is obtained and intelligently processed into an actual speed control amount. By controlling the running speed of the inspection robot, the function of avoiding obstacles is achieved.

[0044] Table 1 Artificial intelligence control rules

[0045] In an embodiment, the multi-machine cooperative operation rule in step S4 includes: One-way driving path planning rule: each inspection road is strictly specified as a one-way driving mode, and the robot must travel according to the preset direction. For example, Figure 7Based on the motion model shown, the system forces the flow direction of the inspection path through the central scheduling center to avoid path conflicts caused by bidirectional travel. The inspection robot needs to form a closed-loop travel route around the detection point to ensure coverage of all areas to be inspected.

[0046] Step-by-step waiting verification mechanism: After completing the inspection of each path node (such as a detection point), the robot must enter a stationary waiting state. At this time, the central scheduling system will receive real-time position data of all robots through the communication module, such as Figure 1 The system architecture shown is based on artificial intelligence algorithms for collision risk assessment. Only when the system confirms that the next move will not cause a collision will it issue a continue moving instruction through the result feedback module; Crossing intersection weight priority rule: When multiple robots approach the intersection path at the same time, the system will calculate the work weight value of each robot based on the path planning algorithm. The weight calculation is based on the remaining inspection path length, and the robot with a larger path value will have priority. If the weights are the same, a random allocation mechanism is used to determine the order of passage. This rule is supported by the Figure 3 The artificial intelligence control structure shown achieves dynamic decision-making; Tail collision prevention mechanism: For robots traveling in the same direction, the system strictly implements the "first come, first served" principle. The robot that arrives later at the path node must remain stationary until the robot in front completes the inspection of that section and moves away to a safe distance before being authorized to move. This mechanism is triggered by real-time ranging data from the laser scanner, such as Figure 2 The distance calculation process shown provides technical support.

[0047] Emergency handling of abnormal states: When the system detects communication interruption or sensor failure, it will immediately start the pre-set emergency path. All robots in the affected area will follow the Figure 4 The membership function curve shown automatically switches to the lowest speed mode and moves to the emergency stop point along the nearest safe boundary.

[0048] This embodiment realizes efficient and safe operation of the inspection robot cluster by establishing a multi-level collaborative control mechanism. Based on one-way path planning and closed-loop detection route design, the conflict of opposite travel is avoided in space; combined with step-by-step waiting verification and real-time risk assessment, the timing safety of the moving instruction is ensured; for the intersection path scenario, a dynamic weight decision algorithm is introduced to optimize the efficiency of multi-machine passage; through tail distance monitoring and emergency handling plan, the risk of collision in the same direction and sudden failure is effectively prevented. The whole set of rule system integrates the advantages of central scheduling and autonomous decision-making, while ensuring full coverage inspection, significantly reducing the conflict probability of multi-machine collaborative work, and improving the system robustness in complex environments.

[0049] In an embodiment, the step S5: analyzing the motion control instruction according to the multi-stage cooperative operation rule, converting the target parameter into the driving wheel rotating speed through the motion model, and adjusting the motor control instruction in real time to complete the obstacle avoidance action, and combining the sensor feedback to realize closed-loop correction, specifically includes the following steps: extracting the target motion parameters from the motion control instruction, including the target speed and the steering angle; converting the target speed and the steering angle into the target rotating speed of the left and right driving wheels based on the motion model of the inspection robot; dynamically adjusting the driving wheel rotating speed to obtain the adjusted left and right driving wheel rotating speed instructions in combination with the multi-robot cooperative operation rule; transmitting the adjusted left and right driving wheel rotating speed instructions to the actuator to control the motor to operate according to the parameters, and feeding back the pose information in real time through the sensor to realize closed-loop correction.

[0050] In this embodiment, the intelligent motion control and real-time feedback adjustment are organically combined to realize the precise and reliable obstacle avoidance navigation of the inspection robot. The system converts the high-level decision instruction into the driving wheel rotating speed parameter, dynamically adjusts based on the kinematic model, and ensures that the robot moves according to the expected trajectory; at the same time, through the closed-loop feedback mechanism of multi-sensor fusion, the deviation between the actual pose and the target state is continuously corrected. This control method not only ensures the rapid response of the obstacle avoidance action, but also overcomes the error caused by external interference through real-time adjustment, so that the robot can stably perform the cooperative inspection task in a complex and variable environment, significantly improving the control accuracy and operation reliability of the system.

[0051] In an embodiment, for general cases, the traffic rules are used for reference to stipulate that each inspection road is one-way and travels around the detection point. When multiple inspection robots cooperate, each step according to the traffic rules needs to be waited in place. The central scheduling system makes decisions and issues instructions to determine whether a collision will occur, and then continues to the next step until the inspection task is completed.

[0052] For special cases such as crossroad collision and rear-end collision, the avoidance control of the inspection robot is based on weight. Relying on the path planning algorithm, the weight of the robot is calculated through the path length of the robot inspection task. The greater the calculated inspection path value, the greater the weight of the robot, which can pass through the crossroad collision first. If the weights are the same, a random pass is made. For rear-end collision, the robot that appears later must wait in place, and only after the robot that appears first passes through can the next operation be performed, thereby completing the obstacle avoidance of the cooperative operation of multiple inspection robots.

[0053] In an embodiment, the track inspection robot cooperative operation obstacle avoidance method provided by the present application is specifically implemented as follows: ​​​​Firstly, a motion model of the inspection robot is constructed. The motion model of the inspection robot is established according to the motion state of the inspection robot, so that the position and angle information of the obstacle is obtained, thereby facilitating the subsequent obstacle avoidance control design.

[0054] Secondly, the image information of the obstacle is obtained and the distance is calculated. The surrounding environment information and pictures are obtained through the sensors and cameras of the perception module. The image processing module processes the sensing image, and the obstacle contour features are extracted by applying grayscale processing, binary processing and obstacle edge extraction technology. The laser ranging technology is used to obtain the reflected beam of the obstacle through the laser scanner of the perception module, and the distance between the receiving end and the obstacle is calculated through the distance formula.

[0055] Thirdly, the artificial intelligence control rule is designed. The artificial intelligence controller is constructed with the artificial intelligence control algorithm as the core. The processed obstacle recognition information is taken as the input variable in the controller, and the controller is used for inference processing. The output result of the controller is taken as the robot running speed control quantity, and the artificial intelligence control rule is constructed based on the input and output variables.

[0056] Fourthly, the multiple inspection robots are cooperatively operated to avoid obstacles. Considering the cooperative operation of the multiple inspection robots, in order to avoid the collision of two robots or multiple robots, the traffic rules are referred to, and it is stipulated that each inspection road is one-way and travels around the detection point. When the multiple inspection robots are cooperatively operated, each step according to the traffic rules needs to be waited in place. The central dispatching system makes a decision and issues an instruction to determine whether a collision will occur, and then the next step is continued until the inspection task is completed.

[0057] Secondly, the application provides a track inspection robot cooperative operation obstacle avoidance system, comprising a position and angle information acquisition module, a contour feature acquisition module, a control instruction acquisition module, a cooperative operation rule establishment module and an obstacle avoidance execution module. The position and angle information acquisition module is used to calculate the pose parameters of the inspection robot by using the kinematics principle according to the motion model of the track inspection robot, and to obtain the position and angle information of the obstacle. The contour feature acquisition module is used to preprocess the collected environment image and extract the contour features of the obstacle. The control instruction acquisition module is in communication connection with the position and angle information acquisition module and the contour feature acquisition module, and is used to take the position information and contour features of the obstacle as control input parameters, divide the control interval by using fuzzy logic, and generate the control instruction of the obstacle avoidance motion of the inspection robot through rule matching. The cooperative operation rule establishment module is used to establish the multi-robot cooperative operation rule of the inspection robot. The obstacle avoidance execution module is in communication connection with the control instruction acquisition module and the cooperative operation rule establishment module, and is configured to parse the motion control instruction according to the multi-level cooperative operation rule, convert the target parameter into the driving wheel rotating speed through the motion model, and adjust the motor control instruction in real time to complete the obstacle avoidance action, and realize closed-loop correction in combination with the sensor feedback.

[0058] In an embodiment, the position angle information acquisition module comprises: The running state acquisition unit is configured to establish a motion model according to the mechanical structure and the motion mode of the inspection robot, and calculate the motion state of the robot by analyzing the motion relationship between the left and right driving wheels. The pose parameter acquisition unit is in communication connection with the running state acquisition unit, and is configured to calculate the linear speed and the angular speed of the robot's center of mass based on the established robot motion model, and derive the pose parameters of the robot in a two-dimensional plane. The relative position relationship acquisition unit is in communication connection with the pose parameter acquisition unit, and is configured to calculate the relative position relationship between the robot and the obstacle according to the calculated pose parameters of the robot and in combination with the environmental data of the set. The position information acquisition unit is in communication connection with the relative position relationship acquisition unit, and is configured to calculate the motion trajectory and the attitude change of the robot in real time through the established mathematical model and the kinematic equation, and acquire the position angle information of the obstacle.

[0059] The functions of the various modules in the rail transit vehicle inspection robot cooperative operation obstacle avoidance system correspond to the steps in the rail transit vehicle inspection robot cooperative operation obstacle avoidance method, and the functions and implementation processes will not be repeated here.

[0060] In an embodiment, the present application provides a system for implementing the rail transit vehicle inspection robot cooperative operation obstacle avoidance method, as shown in Figure 7 The system mainly comprises an inspection robot body, a perception module, an image processing module, a communication module, a central regulation system, a result feedback module, and the like.

[0061] The perception module comprises a temperature / humidity sensor, a laser scanner, and an image acquisition camera. The integration of the perception module enables the robot to monitor the temperature, humidity, position, and other parameters of the equipment to be inspected in the vehicle running part of the maintenance library in real time through the mounted sensors, so as to transmit the environmental monitoring data to the inspection robot in real time.

[0062] The image processing module includes image input, image storage, image processing and analysis, etc. First, the camera of the perception module collects images, and converts the continuous images into digital images that can be processed by the computer. According to the actual task requirements, the key points, edges, textures and other features in the image are detected, and the corresponding feature descriptors are extracted, which are used for image recognition, control scheduling and other tasks in the later stage.

[0063] The communication module includes interfaces, power supplies, connectors, circuits, etc. It is connected with the central scheduling center through WIFI or Bluetooth. It is mainly responsible for the normal communication of the inspection robot, and transmits the images collected by the perception module to the central scheduling center, and accepts the alarm, abnormal display and other information from the result feedback module.

[0064] The central scheduling system processes and decides the results according to the environmental data and image processing results monitored by the perception module in real time through the designed artificial intelligence algorithm, traffic rules and reservation table.

[0065] The result feedback module mainly accepts the instructions from the central scheduling system, and transmits the instructions to the inspection robot, so as to realize the cooperative work and precise obstacle avoidance of the robot.

[0066] In a third aspect, the embodiments of the present application provide a rail transit vehicle inspection robot cooperative work obstacle avoidance device. The rail transit vehicle inspection robot cooperative work obstacle avoidance device can be a personal computer (PC), a notebook computer, a server or other device with data processing function.

[0067] The communication interface includes input / output (input / output, I / O) interface, physical interface and logical interface, etc. for realizing the interconnection of devices inside the rail transit vehicle inspection robot cooperative work obstacle avoidance device, and for realizing the interconnection of the rail transit vehicle inspection robot cooperative work obstacle avoidance device and other devices (such as other computing devices or user devices). The physical interface can be an Ethernet interface, a fiber interface, an ATM interface, etc. The user device can be a display (Display), a keyboard (Keyboard), etc.

[0068] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), and the like.

[0069] The processor can be a general-purpose processor, which can invoke the rail transit vehicle inspection robot cooperative operation obstacle avoidance program stored in the memory and execute the rail transit vehicle inspection robot cooperative operation obstacle avoidance method provided by the embodiments of the present application. For example, the general-purpose processor can be a central processing unit (CPU). The method executed by the rail transit vehicle inspection robot cooperative operation obstacle avoidance program when invoked can refer to various embodiments of the rail transit vehicle inspection robot cooperative operation obstacle avoidance method of the present application, which will not be described here.

[0070] In a fourth aspect, the embodiments of the present application further provide a readable storage medium.

[0071] The readable storage medium of the present application stores the rail transit vehicle inspection robot cooperative operation obstacle avoidance program, wherein the rail transit vehicle inspection robot cooperative operation obstacle avoidance program is executed by the processor to realize the steps of the rail transit vehicle inspection robot cooperative operation obstacle avoidance method as described above.

[0072] The method realized by the rail transit vehicle inspection robot cooperative operation obstacle avoidance program when executed can refer to various embodiments of the rail transit vehicle inspection robot cooperative operation obstacle avoidance method of the present application, which will not be described here.

[0073] It should be noted that the above sequence numbers of the embodiments of the present application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0074] Those skilled in the art can clearly understand the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a plurality of instructions for making a terminal device execute the method described in each embodiment of the present application.

[0075] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for collaborative obstacle avoidance by rail transit vehicle inspection robots, characterized in that, Includes the following steps: Based on the motion model of the rail transit vehicle inspection robot, the pose parameters of the inspection robot are calculated using kinematic principles to obtain obstacle position and angle information; The acquired environmental images are preprocessed to extract the contour features of obstacles; Using obstacle location information and contour features as control input parameters, fuzzy logic is used to divide the control interval, and control commands for the obstacle avoidance movement of the inspection robot are generated through rule matching. Establish rules for multi-robot collaborative operation of inspection robots; The motion control commands are analyzed according to the multi-level collaborative operation rules. The target parameters are converted into the drive wheel speed through the motion model, and the motor control commands are adjusted in real time to complete the obstacle avoidance action.

2. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The process of calculating the robot's pose parameters and obtaining obstacle position and angle information based on the motion model of the rail transit vehicle inspection robot using kinematic principles includes the following steps: Based on the dual-wheel differential drive structure of the inspection robot, the robot motion model is established by kinematic differential equations according to the geometric relationship between the speed of the left and right drive wheels and the wheel track, and the robot pose parameters in the global coordinate system are obtained. Based on these pose parameters, the system integrates laser scanner ranging data and visual sensor environmental image information, and uses a homogeneous coordinate transformation matrix to convert the global coordinates of the obstacle into the polar coordinates of the robot body, thereby obtaining the distance and azimuth of the obstacle relative to the robot.

3. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The motion model is as follows: In the formula: This is the pose state vector. , These are the robot's horizontal and vertical coordinates, respectively. , These are the speeds of the left and right drive wheels, respectively. The diameter of the robot's left and right wheels. The velocity of the robot's center of mass. Let be the angular velocity of the center of mass. The angle between the robot's forward direction and the X-axis.

4. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The preprocessing of the acquired environmental images to extract the contour features of obstacles specifically includes the following steps: The original color image is converted to grayscale to remove color information and retain brightness features, thus obtaining a grayscale image. Based on the obtained grayscale image, a binarization processing method is adopted to convert the image into a binary image by setting a threshold; For binarized images, edge detection algorithms are used to analyze pixel grayscale changes and identify edge information in the image; The acquired distance data is mapped to the edge information in the image into a three-dimensional spatial coordinate system to calculate the actual contour size and position of the obstacle, thereby obtaining the contour features of the obstacle.

5. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The process of using obstacle location information and contour features as control input parameters, employing fuzzy logic to divide the control interval, and generating control commands for the obstacle avoidance movement of the inspection robot through rule matching specifically includes the following steps: Based on the acquired obstacle location information and contour features, the angle and distance parameters of the obstacle relative to the inspection robot are extracted and used as input variables for the control system. Based on fuzzy control theory, the obstacle angle parameter is divided into multiple fuzzy intervals, and a corresponding membership function is set for each fuzzy interval to establish the fuzzification processing result of the input variable. Based on the motion characteristics of the inspection robot, the output control quantity is divided into multiple levels, each level corresponding to different motion speed and turning angle, and the fuzzy processing result of the output variable is established. By combining the fuzzification results of input and output variables, a fuzzy control rule base is constructed, and input parameters are mapped to output control quantities through rule matching; The output control quantity obtained from fuzzy inference is defuzzified and transformed into precise parameters of the inspection robot's movement speed and turning angle, generating the final control command.

6. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The multi-machine collaborative operation rules include one-way traffic rules, distributed waiting mechanisms, special collision handling rules, collaborative control architecture, and anomaly handling mechanisms.

7. The obstacle avoidance method for collaborative operation of rail transit vehicle inspection robots as described in claim 1, characterized in that, The process of parsing motion control commands according to multi-level cooperative operation rules, converting target parameters into drive wheel speeds through a motion model, and adjusting motor control commands in real time to complete obstacle avoidance actions specifically includes the following steps: Extract target motion parameters from motion control commands, including target speed and steering angle; Based on the motion model of the inspection robot, the target speed and steering angle are converted into the target rotation speed of the left and right drive wheels; Based on the multi-machine collaborative operation rules, the speed of the drive wheels is dynamically adjusted to obtain the adjusted speed commands for the left and right drive wheels; The adjusted left and right drive wheel speed commands are transmitted to the actuator to control the motor to operate according to the parameters, and the position and posture information is fed back in real time by the sensor for closed-loop correction.

8. A collaborative obstacle avoidance system for rail transit vehicle inspection robots, characterized in that, include: The position and angle information acquisition module is used to calculate the pose parameters of the inspection robot based on the motion model of the rail transit vehicle inspection robot and the kinematic principle, and to obtain the position and angle information of the obstacle. The contour feature acquisition module is used to preprocess the acquired environmental images and extract the contour features of obstacles. The control command acquisition module is communicatively connected to the position angle information acquisition module and the contour feature acquisition module. It is used to take the obstacle position information and contour features as control input parameters, divide the control interval using fuzzy logic, and generate control commands for the obstacle avoidance movement of the inspection robot through rule matching. The collaborative operation rule establishment module is used to establish multi-machine collaborative operation rules for inspection robots; The obstacle avoidance execution module is communicatively connected to the control command acquisition module and the cooperative operation rule establishment module. It is used to parse motion control commands according to multi-level cooperative operation rules, convert target parameters into drive wheel speeds through motion models, and adjust motor control commands in real time to complete obstacle avoidance actions.

9. The collaborative obstacle avoidance system for rail transit vehicle inspection robots as described in claim 8, characterized in that, The position angle information acquisition module includes: The pose parameter acquisition unit is used to obtain the pose parameters of the robot in the global coordinate system based on the dual-wheel differential drive structure of the inspection robot. According to the geometric relationship between the speed of the left and right drive wheels and the wheel track, the robot motion model is established through kinematic differential equations. The position angle information acquisition unit is communicatively connected to the pose parameter acquisition unit. Based on the pose parameter, it fuses the laser scanner ranging data and the environmental image information of the vision sensor, and converts the global coordinates of the obstacle into the polar coordinate representation of the robot body through a homogeneous coordinate transformation matrix to obtain the distance and azimuth angle of the obstacle relative to the robot.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a collaborative obstacle avoidance program for rail transit vehicle inspection robots, wherein when the collaborative obstacle avoidance program for rail transit vehicle inspection robots is executed by a processor, it implements the steps of the collaborative obstacle avoidance method for rail transit vehicle inspection robots as described in any one of claims 1 to 7.

Images