Operating methods of robots for inspecting or replacing cutterhead cutting tools on tunnel boring machines (TBMs) and TBMs

Abstract

This invention relates to a method for operating a robot for detecting or replacing cutterhead cutting tools in a tunnel boring machine (TBM) and a TBM. The method includes: constructing a database of collision-free paths for the robot (3); obtaining the current posture and target posture of the robot arm; querying the database for collision-free paths from the current posture to the target posture and optimizing the paths; and controlling the robot arm to move from the current posture to the target posture according to the optimized paths.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of shield technology, in particular to a method for operating a robot for detecting or replacing cutters of a cutter head of a shield machine and the shield machine. BACKGROUND

[0002] During the construction process of the heading machine, the cutters on the cutter head cut the stratum, and the wear and loss of the cutters are inevitable, so the detection and replacement of the cutters are needed. At present, the detection and replacement of the cutters are mainly performed by manual operation. In the extreme environment of closed, narrow, high temperature, high humidity, high pressure and oxygen deficiency, workers perform the operations such as disassembly, carrying and installation of the cutter with a weight of up to 250 Kg, which leads to low efficiency, high cost and great safety risk, and cannot meet the requirements of safe and rapid excavation of the heading machine. With the increasingly prominent problem of cutter replacement, domestic and foreign researchers have conducted researches on the cutter head cutter detection / replacement robot, and the main function of the robot is to detect the worn cutters, disassemble the worn cutters from the cutter head, transport the worn cutters back to the robot cabin, grab new cutters from the robot cabin, and install the new cutters on the cutter head.

[0003] During the operation of the robot, in order to perform the operations such as cutter detection, disassembly, transportation and installation, the end effector of the robot needs to be operated with high precision in a specific posture. Since the robot mechanical system (including the end effector and the robot arm body) needs to avoid obstacles such as cabin doors and twisted legs, in order to meet the requirements of small space, large operation range and flexibility of the cabin body, the degrees of freedom of the cutter replacement robot are generally more, and the remote control operation has the problems of large operation scale, limited field of view, difficult operation, low positioning accuracy, slow construction speed, and operation effect depending on the technical level and proficiency of the operator.

[0004] When a conventional control method is used, the single-joint movement of the robot needs to be manually controlled, and each joint of the robot needs to be adjusted multiple times to reach the target position.

[0005] Automatic and intelligent cutter replacement not only can reduce the participation of manual labor, but also can improve the construction efficiency. In the narrow space with many obstacles, how to intelligently plan a path suitable for the cutter replacement robot to perform the cutter replacement operation to guide the operation of the cutter replacement robot is a technical difficulty to be solved. SUMMARY

[0006] The present application aims to provide a method for operating a robot for detecting or replacing cutters of a cutter head of a shield machine and the shield machine, which is beneficial to improve the operation efficiency.

[0007] According to one aspect of an embodiment of the present application, the present application provides a method for operating a robot for detecting or replacing cutters of a cutter head of a shield machine, and the method comprises:

[0008] constructing a database of collision-free paths of the robot;

[0009] obtaining a current pose of a robot arm and obtaining a target pose of the robot arm;

[0010] querying a database for a collision-free path from the current pose to the target pose and optimizing the path; and

[0011] controlling the robot arm to move from the current pose to the target pose along the optimized path.

[0012] In some embodiments, constructing a collision-free path for the robot comprises:

[0013] constructing a collision-free space for the robot arm;

[0014] obtaining a current pose q cur of the robot arm and taking the current pose as a starting point of a path Path ini ;

[0015] randomly generating an intermediate pose q rand of the robot arm;

[0016] finding a nearest neighboring pose q near to the intermediate pose q rand in Path ini ;

[0017] generating a new pose q new between the neighboring pose q near and the intermediate pose q rand at a fixed step size L;

[0018] if the new pose q new is within the collision-free configuration space, adding the new pose q new to the path Path ini ;

[0019] repeating the generating of the intermediate pose q rand , the finding of the neighboring pose q near , and the generating of the new pose q new until the new pose q new is within a set threshold ξ of the target pose q goal of the robot arm;

[0020] taking the current pose q cur , the target pose q goal , and the plurality of new poses q new as the path.

[0021] In some embodiments,

[0022] q cur = [q cur_i ,..., q cur_n ], where qcur_i The amount of movement of the \(i\)-th joint out of \(n\) joints of the robotic arm representing the current posture relative to its original position, \(q\) cur_n The amount of movement of the \(n\)-th joint representing the current posture relative to its original position, where \(0 < i < n\).

[0023] In some embodiments,

[0024] q rand = [q rand_i , …… q rand_n , where \(q\) rand_i represents the amount of movement of the \(i\)-th joint of the robotic arm in the intermediate posture relative to its original position, and \(q\) rand_n represents the amount of movement of the \(n\)-th joint in the intermediate posture relative to its original position, where \(0 < i < n\).

[0025] In some embodiments,

[0026] q rand_i = \(q\) imin + (\(q\) imax - \(q\) imin ) × \(t\) i , where \(0 \leq t\) i \(\leq 1\), and where

[0027] q imin represents the minimum value of the movement range of the \(i\)-th joint of the robotic arm; ]>

[0028] q imax represents the maximum value of the movement range of the \(i\)-th joint of the robotic arm.

[0029] In some embodiments,

[0030] q near = [q near_i , …… q near_n , where \(q\) near_i represents the amount of movement of the \(i\)-th joint of the robotic arm in the neighboring posture relative to its original position, and \(q\) near_n represents the amount of movement of the \(n\)-th joint in the neighboring posture relative to its original position, where \(0 < i < n\), and where \(q\) near is:

[0031] Path ini_h represents the \(h\)-th path data existing currently in Path ini , and \(f\) i represents the movement cost weight of the \(i\)-th joint of the robotic arm and satisfies

[0032] In some embodiments,

[0033] q new= [q new_i , …… q new_n , where q new_i represents the amount of movement of the i-th joint of the robotic arm in the new posture relative to its original position, and q new_n represents the amount of movement of the n-th joint in the adjacent posture relative to its original position, 0 < i < n, where,

[0034]

[0035]

[0036] max[q near - q rand represents the maximum value of the distance between the intermediate posture and the adjacent posture of multiple joints of the robotic arm.

[0037] In some embodiments, generating the intermediate posture q rand includes taking the target posture q goal as the intermediate posture q rand .

[0038] In some embodiments, optimizing the path includes:

[0039] The starting point of the path is NP 1+0 , the m-th node after the starting point of the path is NP 1+m , and the nodes on the optimized path are NP 1+k , where:

[0040]

[0041] In some embodiments, constructing the collision-free space of the robotic arm includes:

[0042] Establishing a mathematical model of the working environment of the robot and a mathematical model of the robotic arm;

[0043] Randomly generating the configuration of the robotic arm in a posture;

[0044] Determining whether the robotic arm in this posture collides with obstacles in the working environment;

[0045] If the robotic arm in this posture does not collide with the obstacle, add this posture to the collision-free space.

[0046] In some embodiments,

[0047] Establishing a mathematical model of the working environment of the robot includes establishing a mathematical model of a regular shape containing obstacles.

[0048] In some embodiments, the regular shape includes at least one of a sphere, a cylinder, a cuboid, a prism, and a hemisphere.

[0049] In some embodiments, determining whether the robotic arm in the specified posture collides with an obstacle in the working environment includes determining whether the mathematical model of the robot and the mathematical model of the obstacle intersect; if there is no intersection, then the robot and the obstacle do not collide.

[0050] According to another aspect of the present invention, a tunnel boring machine is also provided, which includes a processor for performing the above-described operation method.

[0051] By applying the technical solution of this application, a robot collision-free path database is pre-established offline in the non-operational state of the tunnel boring machine. When the robot needs to perform cutter replacement or inspection operations, the path in the database can be called. Therefore, the operation method of this embodiment can effectively improve the robot's operation efficiency.

[0052] Other features and advantages of the invention will become clear from the following detailed description of exemplary embodiments of the invention with reference to the accompanying drawings. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 A schematic diagram of the robot and control system for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention is shown.

[0055] Figure 2 A schematic diagram of the cutterhead of a tunnel boring machine according to an embodiment of the present invention is shown;

[0056] Figure 3 This invention illustrates the working principle of a robot for a tunnel boring machine according to an embodiment of the present invention.

[0057] Figure 4 A schematic diagram illustrating the operation process of a robot for a tunnel boring machine according to an embodiment of the present invention is shown;

[0058] Figure 5 A schematic diagram of the robotic arm of a tunnel boring machine according to an embodiment of the present invention is shown;

[0059] Figure 6 A flowchart illustrating a collision-free space construction method for a robot operating a tunnel boring machine to detect or replace cutterhead tools, according to an embodiment of the present invention, is shown.

[0060] Figure 7A simplified model diagram of obstacles is shown for the operation method of a robot for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention.

[0061] Figure 8 A simplified model diagram of obstacles and the robot illustrating the operation method of a robot for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention is shown.

[0062] Figure 9 This diagram illustrates the obstacle and robot collision principle of a method for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention.

[0063] Figure 10 A schematic diagram of the initial path of a robot for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention is shown.

[0064] Figure 11 A schematic diagram of the optimized path of a robot for detecting or replacing cutterhead cutting tools in a tunnel boring machine according to an embodiment of the present invention is shown. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] like Figures 1 to 3 As shown, the tunnel boring machine in this embodiment includes: a monitor 10, a host computer 20, a slave computer 40, a robotic arm 50, and sensors 30, etc. The slave computer 40, the robotic arm 50, and the sensors 30 are the main components of the robot 3. The slave computer 40 controls the robotic arm 50 to perform work tasks based on the commands from the host computer 20 and the information detected by the sensors 30. The monitor 10 is signal-connected to the host computer 20.

[0067] The host computer 20 is used for the visualization and monitoring, motion control, mode selection, intelligent decision-making, algorithm operation, and data storage of the robot 3 for detecting or replacing cutterhead tools on the tunnel boring machine. The monitor 10 visualizes the posture of the robot 3 for detecting or replacing cutterhead tools and can communicate with the host computer. The slave computer 40 performs motion execution based on the information from the host computer 20 and uses the sensor 30 to provide feedback, thereby achieving high-precision control of the robot 3 for detecting or replacing cutterhead tools. The sensor 30 is used to provide feedback on the motion of each joint (arm segment) of the robot 3 for detecting or replacing cutterhead tools, thereby improving the accuracy of motion control. The sensor 30 can also be used for posture recognition of the robot 3 for detecting or replacing cutterhead tools.

[0068] When it is necessary to inspect or replace the cutting tool 2 at the front end of the cutter head 1, the robot 3 extends from the robot compartment 4 to the cutting tool 2 at the front end of the cutter head 1. The end of the robotic arm of the robot 3 is equipped with an end effector for replacing or inspecting the cutting tool 2.

[0069] A robotic arm consists of multiple arm segments (also called links) connected end-to-end. Adjacent arm segments are connected by joints, which can be rotary joints or linear joints. The axis of rotation of a rotary joint can be aligned with or perpendicular to the extension direction of the arm segment.

[0070] To improve the operational efficiency of robots, this embodiment provides a method for a robot to inspect or replace cutterhead cutting tools in a tunnel boring machine, referring to... Figure 4 As shown, the work methods include:

[0071] Construct a database of collision-free paths for robot 3;

[0072] Obtain the current posture of the robotic arm of the robot 3 and obtain the target posture of the robotic arm;

[0073] The database is queried to find a collision-free path from the current pose to the target pose, and the path is optimized.

[0074] Control the robotic arm to move from the current posture to the target posture along the optimized path.

[0075] In this embodiment, a collision-free path database for the robot is pre-established offline when the tunnel boring machine is not in operation. When the robot needs to perform cutter replacement or inspection operations, the path in the database can be called. Therefore, the operation method of this embodiment can effectively improve the robot's operation efficiency.

[0076] The collision-free path for constructing robot 3 includes:

[0077] Construct a collision-free space for the robotic arm;

[0078] Obtain the current pose q of the robotic arm. cur and use the current pose as the path. ini The starting point;

[0079] Randomly generate the intermediate pose q of the robotic arm rand ;

[0080] Find Path ini Mid-range q rand Recent approaching stance q near ;

[0081] With a fixed step size L, approaching the attitude q near and intermediate posture q rand Generate a new pose q between them new ;

[0082] If the new posture q new If the new posture q is located within the collision-free configuration space, then... new Add to the path ini ;

[0083] Repeat the generation of the intermediate pose q rand Find the nearest pose q near And generate the new pose q new Until the new posture q new With respect to the target pose q of the robotic arm goal The distance is less than the set threshold ξ;

[0084] The current posture q cur The target posture q goal And several new stances q new Make the path described above.

[0085] In some embodiments, q cur =[q cur_i , ...q cur_n ], where q cur_i q represents the motion of the i-th joint among n consecutive joints of the robotic arm in its current posture, relative to its original position. cur_n This represents the motion of the nth joint in the current pose relative to its original position, 0 <i<n。

[0086] In some embodiments, q rand =[q rand_i , ...q rand_n ], where q rand_i q represents the amount of motion of the i-th joint of the robotic arm in its intermediate posture relative to its original position. rand_nThe amount of movement of the nth joint representing the intermediate posture relative to its original position, where 0 < i < n.

[0087] In some embodiments, q near = [q near_i , …… q near_n , where q near_i represents the amount of movement of the ith joint of the robotic arm in the adjacent posture relative to its original position, and q near_n represents the amount of movement of the nth joint in the adjacent posture relative to its original position, where 0 < i < n, and where q near is:

[0088] Path ini_h represents the hth path data currently existing in the list, and f i represents the motion cost weight of the ith joint of the robotic arm, and satisfies

[0089] In some embodiments, q new = [q new_i , …… q new_n , where q new_i represents the amount of movement of the ith joint of the robotic arm in the new posture relative to its original position, and q new_n represents the amount of movement of the nth joint in the adjacent posture relative to its original position, where 0 < i < n, and where,

[0090]

[0091]

[0092] max[q near - q rand represents the maximum value of the distances between the intermediate posture and the adjacent posture of multiple joints of the robotic arm.

[0093] In some embodiments, generating the intermediate posture q rand includes using the target posture q goal as the intermediate posture q rand .

[0094] In some embodiments, optimizing the path includes:

[0095] The starting point of the path is NP 1+0 , the mth node after the starting point of the path is NP 1+m , and the nodes on the optimized path are NP 1+k , where:

[0096]

[0097] like Figure 6 As shown, in some embodiments, the collision-free space for constructing the robotic arm includes:

[0098] Establish a mathematical model of the robot's working environment and a mathematical model of the robotic arm;

[0099] Randomly generate the configuration of the robotic arm in one of its postures;

[0100] Determine whether the robotic arm in this posture collides with an obstacle in the working environment;

[0101] If the robotic arm in this posture does not collide with the obstacle, then this posture is added to the collision-free space.

[0102] Combination Figure 5 As shown, the mathematical model of the robot can optionally (but is not limited to) use the DH parameter method. Since the joints of the tool detection / tool ​​changing robot are mainly divided into two types: translational joints and rotational joints, the standard DH parameter method is used as an example to establish the robot's mathematical model. The robot's link coordinate system is established based on the standard DH parameter method. (n is the number of robot degrees of freedom, and the meaning of T can be found in robotics theory books or related journal articles) Describe the homogeneous transformation relationship between adjacent links.

[0103] Then the i-th joint of the robot can be represented in the base coordinate system as:

[0104]

[0105] According to robotics theory and the concept of the T matrix, its fourth column represents position coordinates. Therefore, the coordinates of the i-th joint end in the base coordinate system can be extracted using the following formula:

[0106]

[0107] Then the coordinates of the two ends of the i-th joint in Cartesian space are:

[0108]

[0109] The central axis of the i-th joint can then be described by the expression of a straight line passing through two points in space.

[0110]

[0111] To simplify the joint model, the i-th joint is simplified to a model with p... i-1 and p i With r as the center, i With radius, A cylinder with the axis as its center. This completes the mathematical representation of the tool turret inspection / tool ​​changing robot in space.

[0112] The mathematical model for establishing the robot's working environment includes establishing a mathematical model containing regular shapes of obstacles.

[0113] refer to Figure 7 As shown, information about the surrounding environment is obtained, and the environment is mathematically described (irregular obstacles can be expanded into spheres, cylinders, etc. to simplify the expression and improve the efficiency of the algorithm, but this measure will reduce the robot's range of motion because it expands the obstacles).

[0114] The sphere can be represented using the following methods, but is not limited to them (Note: A single obstacle can also be represented by multiple regular shapes, such as a sphere, a cylinder, etc., to improve the accuracy of the model).

[0115] according to Figure 7 As shown, the coordinates of the obstacle center in the base coordinate system are P. obs_k (x k ,y k ,z k Its spherical radius is R. obs_k .

[0116] Combination Figure 8 and 9 As shown, collision detection is performed based on the robot model and simplified obstacle model established above.

[0117] By simplifying obstacles into spheres, the collision detection problem is transformed into: calculating the center P of each obstacle k sphere. obs_k (x k ,y k ,z k ) to the tool inspection / tool ​​changing robot links The distance d,

[0118] 1) If the distance d is greater than the radius R of the simplified obstacle obs_k The radius r of link i i If they are aligned, no collision will occur;

[0119] 2) When the distance d is less than the simplified obstacle radius R obs_k The radius r of link i i And, we need to simplify the distances dd1 and dd2 between the center points of the two ends of the link by judging the distance of the obstacle sphere, and discuss the cases according to the shortest distance dd (dd = min(dd1, dd2)):

[0120] a) When the calculated distance dd is greater than the obstacle radius R obs_k +ri a) When the conditions are met, no collision occurs; b) otherwise, a collision occurs.

[0121] Combination Figure 6 As shown, due to the large number of degrees of freedom in the tool detection / changing robot, it is difficult to obtain a collision-free configuration space. Specifically, the method for constructing the collision-free space of the robotic arm in this embodiment includes:

[0122] Step 01: Use random sampling to generate a large number of joint variables within the joint variables;

[0123] First, the range of motion q of each joint of the robot is determined based on the robot's structure, electrical wiring, oil pipes, etc. i ∈[q imin ,q imax ], i = 1, 2, 3, ..., n (n is the number of links) Establish the configuration space according to the range of joint variables.

[0124] Q space =[q1,q2,q3,...q i ...q n (6)

[0125] Where, q i ∈[q imin ,q imax ],

[0126] The robot's configuration space can then be constructed using the following formula:

[0127] q 1_rand =q 1min +(q 1max -q 1min )×t rand_1

[0128] q 2_rand =q 2min +(q 2max -q 2min )×t rand_2

[0129] q 3_rand =q 3min +(q 3max -q 3min )×t rand_3

[0130]

[0131] q i_rand =q imin +(q imax -q imin )×t rand_i

[0132]

[0133] q n_rand =q nmin +(q nmax -q nmin )×t rand_n (7)

[0134] In the formula, t rand_i The value of t is a random number ranging from 0 to 1; the more values ​​of t, the more configurations are obtained, and the more accurate the configuration space is constructed, but the amount of computation also increases, and the computation time increases; since this step is an offline construction and does not occupy the time of path search, the number of random numbers constructed can be increased in the offline process to make the model more accurate.

[0135] Step 02: Based on the established linkage model and the tool head detection / tool ​​changing robot configuration obtained in Step 1, calculate the position coordinates of each joint and linkage of the robot under this configuration;

[0136] Step 03: Based on the collision detection method described in step 3 above, determine whether the tool detection / tool ​​changing robot under this joint variable has collided with an obstacle. According to the determination result, if a collision or interference occurs, add it to the collision configuration C-obs database; if no collision or interference occurs, add it to the C-free no-collision database. This completes the construction of the offline configuration space.

[0137] Step 04: Determine whether a collision or interference will occur in the next configuration;

[0138] Step 05: Repeat the above steps until all joint configurations have completed collision detection;

[0139] Step 06: Output C-free and C-obs.

[0140] In some embodiments, the regular shape includes at least one of a sphere, cylinder, cuboid, prism, and hemisphere.

[0141] In some embodiments, detecting whether the robot and the obstacle collide includes determining whether the mathematical model of the robot and the mathematical model of the obstacle intersect; if there is no intersection, then the robot and the obstacle do not collide.

[0142] In addition to constructing a collision-free space for the robotic arm, constructing a collision-free path for the robot also requires obtaining the current pose q of the robotic arm. cur and target attitude q goal and the current attitude q cur and target attitude q goal As the starting and ending points of a collision-free path.

[0143] Robot tool change target posture q goal The calculations include:

[0144] Step 1: Obtain the position and orientation of the tool that needs to be replaced through detection and other means, and convert it to the robot's base coordinate system;

[0145] Step 2: Based on the tool posture and robotics theory, inverse kinematics calculations are used to deduce the motion q of each joint during the tool removal and installation process of the tool head tool inspection / tool ​​changing robot. goal (Tool detection / target posture of tool changer robot);

[0146] q goal =[q goal_i , ...q goal_n ], where q goal_i q represents the amount of motion of the i-th joint of the robotic arm in its intermediate posture relative to its original position. goal_n This represents the motion of the nth joint in the intermediate pose relative to its original position, 0 <i<n。

[0147] In this embodiment, the robotic arm has 8 joints, therefore:

[0148] q goal =[q goal_1 ,q goal_2 ,q goal_3 ,q goal_4 ,q goal_5 ,q goal_6 ,q goal_7 ,q goal_8 ].

[0149] The current pose q of the robotic arm cur Identification is achieved through sensors, specifically...

[0150] q cur =[q cur_1 ,q cur_2 ,q cur_3 ,q cur_4 ,q cur_5 ,q cur_6 ,q cur_7 ,q cur_8 ].

[0151] Construct an empty path ini In the constructed offline collision-free path, the following algorithm is used for path lookup and output as the robot's initial motion path; the specific constructed path includes:

[0152] Step 1: Construct an empty list Path ini, will the current configuration type q cur =[q cur_i ,…,q cur_n As the initial node of the path, add it to Path. ini ;

[0153] Step 2: Obtain q through random sampling as described in step 01 above. rand , i.e. q rand Satisfying q rand_i =q imin +(q imax -q imin )×t i Where i represents the i-th joint, t i The result is a random number, and 0 ≤ t i ≤1, because the motion precision of each joint is not uniform, therefore the random number t for each joint is ≤1. i Different levels of precision can be used to represent this. (e.g., joint q) i The range of motion is 0 to 10°, and its motion accuracy is 0.01°. Then, the t of this joint... i (A value of 0.001 can be used, meaning three decimal places are retained.)

[0154] Find the Path using the following formula. ini Mid-range q rand The nearest node as q near , where q near for:

[0155]

[0156] Path ini_h Represents Path ini The h-th path data currently existing in the list. i Let represent the motion cost weight of joint i, and satisfy .

[0157] With a fixed step size L in q near With q rand Generate a new node q in the direction of the connection. new , i.e. q new The i-th element q new_i It can be calculated using the following formula:

[0158]

[0159]

[0160] In the formula, L is the joint movement step length, which is set as follows: q rand Represents the randomly generated joint vector, q near The distance q on the random treerand The joint vector q that is closest to the nearest point rand-i -q near-i Represents a random tree starting from q near to q rand The distance between the connecting lines and joint i, max[q] near -q rand The symbol ] represents the maximum value among all joint distances, therefore After normalizing the joint increment, the result of the formula is the newly generated node q. new The maximum value of the joint increment is 1, while the joint movement step size L is a constant. By changing the value of L, the maximum joint increment can be changed to L. This method allows the robot to keep the maximum joint increment at L when performing path search, which can improve the convergence speed of the algorithm.

[0161] Step 3: If q new If not in the collision-free space C-free, then q is regenerated. rand Conversely, if the node does not meet the requirements, add it to the Path list as a candidate node for the next motion pose. ini ;

[0162] Step 4: Repeat steps 2 and 3 until q new With target point q goal The algorithm terminates when the distance is less than the set threshold ξ.

[0163] Step 5: Output the initial path. ini .

[0164] Obtain the initial path ini See Figure 10 (In reality, path planning is performed in joint space, so the dimension of its path information is consistent with the robot's degrees of freedom; since high-dimensional space is difficult to represent graphically, a simplified illustration is provided here for ease of understanding.)

[0165] It was found that the initial path points generated using a random strategy have the following two drawbacks: a. Unnecessary redundant nodes exist; b. Node connections have corners and are not smooth. This prolongs the job cycle, affects operational accuracy, and may even affect the lifespan of the tool inspection / changing robot. Therefore, the following steps are used to remove redundant nodes from the path and optimize it.

[0166] See Figure 11 Output the initial path. ini After that, the path also needs to be... ini Optimization of the path includes:

[0167] Let the starting point of the path be Path ini_1 NP 1+0 Let Path be the starting point. ini_1 The last m nodes are NP 1+m Let NP be the newly formed point between two points. 1+1 NP 1+2 …NP 1+k ...NP 1+(m-1) but:

[0168]

[0169] Where k = 1, 2, 3, m.

[0170] Where, Path ini_i This represents the i-th point in the initial path.

[0171] Then, check the nodes NP one by one. 1+(m-1) ... NP 1+2 NP 1+1 Is it within the collision-free space (C-free)? If so, then set NP. 1+0 With NP 1+m Delete the original path points between them, and NP 1+0 NP 1+1 NP 1+2 ..., NP 1+(m-1) NP 1+m Use it as a new waypoint; otherwise, retain the original waypoint and use NP. 1+m Starting from the point, based on the above method, determine NP. 1+m To NP 1+2m Can the original path between them be deleted? Repeat the above steps until it is determined that the path is... ini_n Stop optimization. And use the optimized path points as the final path.

[0172] Through multiple iterations, a better path suitable for robot operations was finally obtained. final .

[0173] The optimized path can be further improved by using B-spline curves and other methods to enhance its smoothness and flexibility.

[0174] Finally, based on the Path generated above final As the motion path for tool replacement in the tool head tool inspection / tool ​​changing robot, it can be directly used for robot joint control without the need for inverse kinematics calculation, thereby improving the real-time performance of robot control and enabling intelligent operation of the robot.

[0175] In this embodiment, a model of the robot and its surrounding environment is built offline, and a collision-free C-free path for the robot is constructed based on the random sampling method proposed in this patent. When a tool change operation is required, the current joint variables of the robot are obtained through sensors, and its initial configuration is constructed. After determining the position of the tool to be replaced in the robot's base coordinate system, it is converted into the target posture of the robot operation. The target configuration of the robot is obtained through robot inverse kinematics calculation. The improved fast expanding random tree proposed in this paper is used to query the path information from the initial configuration to the target configuration in the offline C-free collision-free path, and an initial collision-free operation path for the tool detection / tool ​​change robot is output. The path optimization method proposed in this paper is used to optimize the robot's motion path, and finally a better path suitable for the robot's intelligent operation is obtained. Finally, the host computer uses the planned path as the input for the robot's joint motion and sends the information to the slave computer. The slave computer controls the robot to complete the tool detection / change operation based on the feedback from the host computer and sensors.

[0176] According to another aspect of the present invention, a tunnel boring machine is also provided, which includes a processor for performing the above-described operation method.

[0177] The above description is merely an exemplary embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of operating a robot for inspecting or replacing a cutter of a cutterhead of a tunnel boring machine, characterized in that comprises: constructing a database of collision-free paths for the robot (3); acquiring a current pose of a robot arm of the robot (3) and acquiring a target pose of the robot arm; querying the database for a collision-free path from the current pose to the target pose and optimizing the path; and controlling the robot arm to move from the current pose to the target pose along the optimized path, constructing a collision-free space for the robot arm comprises: constructing a collision-free space for the robot arm comprises: obtaining a current pose q of the robot arm cur， and using the current pose as a starting point for a path Path ini of the robot arm. randomly generating intermediate poses q of the robot arm rand ; Finding the Path ini Intermediate pose q rand Nearest neighbor pose q near ; A new pose q is generated between the nearby pose q near and the intermediate pose q rand with a fixed step size L new ; If a new pose q new is located within the collision-free configuration space, then the new pose q new is added to the path Path ini ; repeatedly generating said intermediate poses q rand , finding said neighboring poses q near and generating said new poses q new until a new pose q new is within a distance of a set threshold from a target pose q goal of the robot arm The current pose q cur , the target pose q goal , and a plurality of new poses q new are made as the path, establishing a mathematical model of a working environment of the robot and a mathematical model of the robot arm; Let Path be the path start point ini_1 For NP 1+0 , let Path be the start point ini_1 The last m nodes are NP 1+m , record the new point between two points as NP 1+1 , NP 1+2 … NP 1+k … NP 1+(m-1) Then: randomly generating a configuration of the robot arm in a pose; wherein Path ini_i represents the i-th point in the initial path, Then, it is judged in turn whether the nodes NP 1+(m-1) ,..., NP 1+2 , NP 1+1 are in the collision-free space C-free, if all are in, the original path points between NP and NP are deleted, and NP 1+0 , NP 1+1 , NP 1+2 ,..., NP 1+(m-1) , NP 1+m are taken as new path points; otherwise, the original path points are kept, and it is judged whether the original path between NP 1+m and NP 1+2m can be deleted based on the above method with NP as the initial point; the above steps are repeated until the path between NP ini_n is judged, the optimization is stopped, and the optimized path points are taken as the final path, q rand =[q rand_i，…… q rand_n ], where q rand_i denotes the amount of motion of the i-th joint of the robot arm for the intermediate pose relative to its original position, q rand_n denotes the amount of motion of the n-th joint for the intermediate pose relative to its original position, 0 < i < n, q near =[q near_i，…… q near_n ], where q near_i denotes the amount of motion of the i-th joint of the robot arm in proximity of the pose with respect to its original position, q near_n denotes the amount of motion of the n-th joint in proximity of the pose with respect to its original position, 0 < i < n. determining whether the robot arm in the pose collides with an obstacle in the working environment; q cur = [q cur_i，…… cur_n ], wherein q cur_i denotes the amount of motion of the i-th joint of the n joints of the robot arm in the current pose relative to its original position, q cur_n denotes the amount of motion of the n-th joint relative to its original position, 0 < i < n.​ if the robot arm in the pose does not collide with the obstacle, adding the pose to the collision-free space. , wherein, q imin qmin,i represents the minimum value of the range of motion of the i-th joint of the robot arm; q imax denotes the maximum value of the range of motion of the i-th joint of the robot arm.

4. The work method according to claim 1, characterized in that, 8. The method according to claim 7, wherein q near is: ,Path ini_h represents Path ini the hth path data currently existing in Path, represents the motion cost weight of the ith joint of the robot arm, and satisfies . the establishing a mathematical model of a working environment of the robot comprises establishing a mathematical model of a regular shape including an obstacle. q new =[q new_i，…… q new_n ], where q new_i denotes the amount of motion of the i-th joint of the robot arm in the new pose relative to its original position, q new_n denotes the amount of motion of the n-th joint in the next pose relative to its original position, 0 < i < n, wherein, ， represents the maximum of the intermediate pose and the proximate pose distance of the plurality of joints of the robot arm.

6. The work method according to claim 1, characterized in that, generating intermediate poses q rand comprises determining a target pose q goal as the intermediate pose q rand .

7. The work method according to claim 1, characterized in that, The regular shape includes at least one of a sphere, a cylinder, a cuboid, a prism, and a hemisphere. Determining whether the robot arm in the pose collides with the obstacle in the working environment comprises determining whether the mathematical model of the robot and the mathematical model of the obstacle have an intersection, and if there is no intersection, the robot and the obstacle do not collide. comprises a processor that executes the method of any one of claims 1 to 10. ​ ​ ​ ​ 9. The method of claim 8, wherein, ​ 10. The method of claim 8, wherein ​ 11. A tunneling machine characterized by, ​

Images