Self-adaptive force-position hybrid control method for duct piece pasting nine-axis linkage robot

Abstract

The invention provides a duct piece pasting nine-axis linkage robot self-adaptive force and position hybrid control method which is applied to a duct piece pasting nine-axis linkage robot, and the robot comprises a plurality of movement joints and a plurality of movement shafts. In a duct piece pasting task used in a shield tunnel, a variable impedance coefficient existing in the moving process of the robot can be determined according to the expected to-be-pasted posture and by combining actual control parameters at the current moment, then the variable impedance coefficient is introduced, and when the final pasting posture of the duct piece pasting nine-axis linkage robot is located at the expected posture, the duct piece pasting is completed. According to the method and the device, the joint angles of the corresponding moving joints and the expected moving positions of the moving shafts are determined, force control signals are generated according to the results, the joint angles of the robot are controlled to move, the moving shafts of the robot are controlled to move, and finally the robot can adaptively and accurately perform segment pasting according to the expected pose. And finally, the effects of high precision, high efficiency and high reliability of the duct piece pasting task in the shield tunnel construction process are achieved.

Description

Technical Field

[0001] The present application relates to the technical field of automatic control and service robots, and particularly to an adaptive force-position hybrid control method for a segment-pasting nine-axis linkage robot. Background Art

[0002] With the continuous development of automation technology and the wide promotion of robot applications, the application of robots in industries such as industry, manufacturing, and logistics has become increasingly common. Especially in recent years, with the rapid development of technologies such as artificial intelligence, visual recognition, and control algorithms, robot technology has been more widely applied.

[0003] The shield tunneling method is widely used in subways, railways, and other underground projects. The application of sealant at the segment joints is a key link to ensure the sealing performance after assembly. However, the existing manual glue application method has low efficiency, unstable quality, and is prone to problems such as missed coating or uneven thickness. Moreover, it has a high labor intensity, many safety hazards, and inaccurate glue volume control leads to material waste. Although some automated glue application equipment has been put into use, traditional robot control systems often face the following challenges when dealing with complex and dynamically changing environments. There are a large number of uncertain factors in the construction environment, such as uneven ground, material deformation, the action of external forces, etc.

[0004] Traditional control methods are difficult to cope with these uncertainties, resulting in a decline in control performance. In tasks such as segment pasting, the robot needs to accurately control force and position simultaneously. For example, during the pasting process, the robot not only needs to maintain an appropriate contact force but also ensure the accuracy of the pasting position. When traditional force control or position control is used alone, it cannot meet this requirement. Summary of the Invention

[0005] In view of this, the embodiments of the present application provide an adaptive force-position hybrid control method, device, and electronic device for a segment-pasting nine-axis linkage robot to achieve high-precision, high-efficiency, and high-reliability effects in the segment-pasting task during the shield tunnel construction process.

[0006] In a first aspect, the embodiments of the present application provide an adaptive force-position hybrid control method for a segment-pasting nine-axis linkage robot, wherein the method includes: The method is applied to a segment-pasting nine-axis linkage robot, and the segment-pasting nine-axis linkage robot includes a plurality of moving joints and a plurality of moving axes. The method includes: Obtain the desired trajectory , the actual control parameters at the current moment, and the actual control parameters include: the actual pose at the current moment , the actual speed , and the actual environmental acting force , the expected trajectory is the set moving trajectory of the nine-axis linkage robot for segment pasting; According to the preset intelligent impedance calculation model, it is determined that the nine-axis linkage robot for segment pasting follows the desired trajectory. Variable impedance coefficient when moving f 0 , wherein the preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters; Based on the variable impedance coefficient f 0 , determine the variable impedance coefficient in the f 0 Under the influence of ; According to the expected trajectory Solve the inverse kinematics of the robot arm to determine the expected trajectory of the nine-axis linkage robot for segment pasting After moving, the expected pose is , and based on the expected pose Determine the expected joint angle corresponding to each of the motion joints at the next moment and the expected moving position of each of the moving axes at the next moment; According to the expected posture The actual posture , generating force control signal , and based on the force control signal , control each of the moving joints to move to the expected joint angle, and control each of the moving axes to move to the expected moving position.

[0007] In combination with the first aspect, in a second possible embodiment, the method further includes: For the actual posture and the actual speed Solve the forward kinematics of the robot arm to determine the actual trajectory of the nine-axis linkage robot for segment pasting at the current moment ; According to the actual trajectory and the desired trajectory , determining the mobile tracking error of the nine-axis linkage robot for segment pasting; The mobile tracking error is input into the preset intelligent impedance calculation model, and the output result of the preset intelligent impedance calculation model is determined to be the variable impedance coefficient f 0 .

[0008] Combined with the second possible embodiment of the first aspect, in the third possible embodiment, the movement tracking error includes: position tracking error , speed tracking error , acceleration tracking error , force error , and the preset intelligent impedance calculation model satisfies the variable impedance coefficient formula constraint; The variable impedance coefficient formula is:

[0009] Wherein, is the variable impedance coefficient, is the actual inertia matrix varying with time, is the nominal parameter of the actual inertia matrix, is the actual damping matrix varying with time, is the nominal parameter of the actual damping matrix, is the actual stiffness matrix varying with the actual situation, is the nominal parameter of the actual stiffness matrix, is the stability coefficient.

[0010] Combined with the third possible embodiment of the first aspect, in the fourth possible embodiment, the method further includes: Based on the variable impedance coefficient formula, construct an RFWNN neural network model to obtain the preset intelligent impedance calculation model; Input the movement tracking error in the form of an input vector into the preset intelligent impedance calculation model, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient f 0 .

[0011] Combined with the fourth possible embodiment of the first aspect, in the fifth possible embodiment, the preset intelligent impedance calculation model includes: an input layer, a fuzzification and wavelet layer, a rule layer, a waveform layer, and an output layer; the step of inputting the movement tracking error in the form of an input vector into the preset intelligent impedance calculation model, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient f 0 , includes: Input the input vector into the input layer; Through the fuzzification and wavelet layer, with the help of the Gaussian membership function, determine the output result of the wavelet nodes of the fuzzification and wavelet layer; The rule layer calculates the activation degree of each fuzzy rule of the rule layer according to the output result of the wavelet nodes; Using the waveform layer and the output layer, perform defuzzification based on the activation degree of each fuzzy rule, and output the variable impedance coefficient f 0 。

[0012] Combined with the first aspect, in the sixth possible embodiment, the generating the force control signal according to the expected pose and the actual pose includes: 。 Taking the difference between the expected pose and the actual pose to generate a first error surface ; Performing virtual control on the first error surface to obtain a virtual control vector , and based on the actual velocity and the virtual control vector , generating a second error surface ; Based on the virtual control vector , the second error surface , combined with the actual dynamic model of the segment pasting nine-axis linkage robot, generating the force control signal 。

[0013] Combined with the sixth possible embodiment of the first aspect, in the seventh possible embodiment, the method further includes: Substituting the virtual control vector , the second error surface into a preset force control signal generation formula, and outputting the force control signal ; The preset force control signal generation formula satisfies:

[0014] Wherein, is the actual Coriolis force and centrifugal force matrix of the segment pasting nine-axis linkage robot, is the actual gravity matrix of the segment pasting nine-axis linkage robot, is the transposed matrix of the actual Jacobian matrix of the segment pasting nine-axis linkage robot, is the actual environmental acting force of the segment pasting nine-axis linkage robot, is the actual inertia matrix of the segment pasting nine-axis linkage robot, is the second error surface 's gain parameter, is the virtual control law, and t is the filtering time constant; Among them, the virtual control law satisfies:

[0015] wherein, is the gain parameter of the first error surface, is the expected pose corresponding expected velocity.

[0016] In a second aspect, the present application provides a self-adaptive force-position hybrid control device for a segment-pasting nine-axis linkage robot. The device is applied to a segment-pasting nine-axis linkage robot, and the segment-pasting nine-axis linkage robot includes a plurality of motion joints and a plurality of moving axes. The device includes: An input module for obtaining a desired trajectory and actual control parameters at the current moment. The actual control parameters include: the actual pose at the current moment, the actual velocity and the actual environmental acting force . The desired trajectory is the set moving trajectory of the segment-pasting nine-axis linkage robot; A variable impedance coefficient determination module for determining the variable impedance coefficient existing when the segment-pasting nine-axis linkage robot moves according to the desired trajectory f 0 , wherein the preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters; An expected trajectory determination module for determining, based on the variable impedance coefficient f 0 , the expected trajectory f 0 of the segment-pasting nine-axis linkage robot under the influence of the variable impedance coefficient ; An inverse kinematics solution module for performing inverse kinematics solution of the robotic arm according to the expected trajectory to determine the expected pose obtained after the segment-pasting nine-axis linkage robot moves according to the expected trajectory , and determining the expected joint angles at the next moment corresponding to each of the motion joints and the expected moving positions at the next moment of each of the moving axes based on the expected pose ; A position control module for generating a force control signal according to the expected pose and the actual pose , and based on the force control signal Control each of the motion joints to move to the expected joint angles, and control each of the moving axes to move to the expected moving positions.

[0017] Combined with the second aspect, in a second possible embodiment, the input module is further configured to: For the actual pose and the actual speed Perform forward kinematics solution of the robotic arm to determine the actual trajectory of the segment pasting nine-axis linkage robot at the current moment ; According to the actual trajectory and the desired trajectory , determine the moving tracking error of the segment pasting nine-axis linkage robot; Input the moving tracking error into the preset intelligent impedance calculation model to determine the output result of the preset intelligent impedance calculation model as the variable impedance coefficient f 0 .

[0018] Combined with the second possible embodiment of the second aspect, in a third possible embodiment, the moving tracking error includes: position tracking error , speed tracking error , acceleration tracking error , force error , and the preset intelligent impedance calculation model satisfies the variable impedance coefficient formula constraint; The variable impedance coefficient formula is:

[0019] Wherein, is the variable impedance coefficient, is the actual inertia matrix varying with time, is the nominal parameter of the actual inertia matrix, is the actual damping matrix varying with time, is the nominal parameter of the actual damping matrix, is the actual stiffness matrix varying with the actual situation, is the nominal parameter of the actual stiffness matrix, is the stability coefficient.

[0020] Combined with the third possible embodiment of the second aspect, in a fourth possible embodiment, the variable impedance coefficient determination module is further configured to: Based on the variable impedance coefficient formula, construct an RFWNN neural network model to obtain the preset intelligent impedance calculation model; Input the moving tracking error in the form of an input vector into the preset intelligent impedance calculation model, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient based on the input vector. f 0 。

[0021] Combined with the fourth possible embodiment of the second aspect, in the fifth possible embodiment, the preset intelligent impedance calculation model includes: an input layer, a fuzzification and wavelet layer, a rule layer, a waveform layer, and an output layer; the variable impedance coefficient determination module is further configured to: Input the input vector into the input layer; Through the fuzzification and wavelet layer, by means of the Gaussian membership function, determine the output result of the wavelet nodes of the fuzzification and wavelet layer; The rule layer calculates the activation degree of each fuzzy rule of the rule layer according to the output result of the wavelet nodes; The waveform layer and the output layer are used to perform defuzzification processing based on the activation degree of each fuzzy rule and output the variable impedance coefficient f 0 。

[0022] Combined with the second aspect, in the sixth possible embodiment, the position control module is further configured to: According to the expected pose And the actual pose , generate a force control signal , including: For the expected pose And the actual pose Find the difference to generate a first error surface ; Perform virtual control on the first error surface to obtain a virtual control vector , and based on the actual velocity And the virtual control vector , generate a second error surface ; Based on the virtual control vector , the second error surface , combined with the actual dynamic model of the segment pasting nine-axis linkage robot, generate the force control signal 。

[0023] Combined with the sixth possible embodiment of the second aspect, in the seventh possible embodiment, the position control module is further configured to: The virtual control vector , the second error surface Substitute it into the preset force control signal generation formula to output the force control signal ; The preset force control signal generation formula satisfies:

[0024] where, is the actual Coriolis force and centrifugal force matrix of the segment pasting nine-axis linkage robot, is the actual gravity matrix of the segment pasting nine-axis linkage robot, is the transposed matrix of the actual Jacobian matrix of the segment pasting nine-axis linkage robot, is the actual environmental acting force of the segment pasting nine-axis linkage robot, is the actual inertia matrix of the segment pasting nine-axis linkage robot, is the second error surface The gain parameter of, is the virtual control law, and t is the filtering time constant; where, the virtual control law satisfies:

[0025] where, is the gain parameter of the first error surface, is the expected pose The corresponding expected velocity.

[0026] In a third aspect, an embodiment of the present application provides an electronic device, where the electronic device includes: a processor; and a memory for storing a program; where the program includes instructions that, when executed by the processor, cause the processor to execute the segment pasting nine-axis linkage robot adaptive force-position hybrid control method described in the first aspect.

[0027] In a fourth aspect, an embodiment of the present application provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to cause a computer to execute the segment pasting nine-axis linkage robot adaptive force-position hybrid control method described in the first aspect.

[0028] Advantages of the present application: The present application provides a method for adaptive force-position hybrid control of a nine-axis linkage robot for segment pasting. This method obtains the desired trajectory, the actual pose, actual speed, and actual environmental acting force at the current moment. Then, based on a preset intelligent impedance calculation algorithm, it determines the variable impedance coefficient existing when the nine-axis linkage robot for segment pasting moves along the desired trajectory. And based on this variable impedance coefficient, it determines the expected trajectory that the nine-axis linkage robot for segment pasting can be in under the influence of this variable impedance coefficient. Then, it performs inverse kinematics solution for the robotic arm according to this expected trajectory to determine the expected pose obtained after the nine-axis linkage robot for segment pasting moves along this expected trajectory. Then, based on this expected pose, it reversely infers the expected joint angles of the nine-axis linkage robot for segment pasting at the next moment and the expected moving positions of the moving axes. Finally, according to this expected pose and the actual pose, it generates a force control signal, and based on this force control signal, it controls the motion joints of the robot to move to the expected joint angles and controls the moving axes of the robot to move to the expected moving positions.

[0029] In this way, by selecting the embodiment of the present application, in the segment pasting task used in shield tunnels, according to the expected pose to be pasted, combined with the current actual pose, the variable impedance coefficient existing during the movement of the nine-axis linkage robot for segment pasting can be determined. Then, this variable impedance coefficient is introduced to reversely infer the poses corresponding to each motion joint and moving axis when the final pasting pose of the nine-axis linkage robot for segment pasting is at this target pose, as well as the joint angles of each motion joint corresponding to this pose and the expected moving positions of the moving axes. And according to this pose, it controls the joint angle movement and moving axis movement of the nine-axis linkage robot for segment pasting, so that finally the nine-axis linkage robot for segment pasting can adaptively and accurately execute segment pasting according to the expected pose to be achieved, and finally achieve the high-precision, high-efficiency, and high-reliability effects of the segment pasting task during the construction of shield tunnels. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In the following description of exemplary embodiments with reference to the accompanying drawings, more details, features, and advantages of the present application are disclosed. In the drawings: Figure 1 FIG. 1 shows a schematic structural diagram of a nine-axis linkage robot for segment pasting provided by an embodiment of the present application; Figure 2 FIG. 2 shows a schematic flowchart of a method for adaptive force-position hybrid control of a nine-axis linkage robot for segment pasting provided by an embodiment of the present application; Figure 3 FIG. 3 shows another schematic flowchart of a method for adaptive force-position hybrid control of a nine-axis linkage robot for segment pasting provided by an embodiment of the present application; Figure 4 FIG. 4 shows another schematic flowchart of a method for adaptive force-position hybrid control of a nine-axis linkage robot for segment pasting provided by an embodiment of the present application; Figure 5 Another schematic flowchart of the adaptive force-position hybrid control method for the segment pasting nine-axis linkage robot provided by the embodiment of the present application is shown; Figure 6 Another schematic flowchart of the adaptive force-position hybrid control method for the segment pasting nine-axis linkage robot provided by the embodiment of the present application is shown; Figure 7 A schematic logical structure diagram of the adaptive force-position hybrid control device for the segment pasting nine-axis linkage robot provided by the embodiment of the present application is shown; Figure 8 A structural block diagram of an exemplary electronic device that can be used to implement the embodiments of the present application is shown. Detailed implementation manners

[0031] Embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not used to limit the protection scope of the present application.

[0032] It should be understood that the steps recorded in the method embodiments of the present application can be executed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this regard.

[0033] The term "including" and its variants used herein are open-ended, that is, "including but not limited to". The term "based on" is "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description. It should be noted that the concepts such as "first" and "second" mentioned in the present application are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependent relationships.

[0034] It should be noted that the modifications of "one" and "multiple" mentioned in the present application are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly specified in the context, it should be understood as "one or more".

[0035] In a first aspect, the present application provides a method for adaptive force-position hybrid control of a segment pasting nine-axis linkage robot, which is applied to any electronic device with the function of adaptive force-position hybrid control of a segment pasting nine-axis linkage robot, and the electronic device can be a segment pasting robot.

[0036] As an example, the method can be applied to a segment pasting nine-axis linkage robot, and the structure of the segment pasting nine-axis linkage robot can be as Figure 1 shown, which consists of an XYZ pasting platform and a six-axis robotic arm. Figure 1 The numbers (number 1 to number 6) shown in it identify the joints of the six-axis robotic arm, and X, Y, and Z represent the three moving axes of the pasting platform of the segment pasting robot in space. Therefore, the specific position and angle of segment pasting are affected by the positions and angles of the joints of the six-axis robotic arm and the moving position of the pasting platform. That is, the specific pose of segment pasting depends on the poses of each joint and each moving axis. In the embodiments of the present application, pose can be understood as the general term of position + angle, and the pose change depends on the change of position and / or angle.

[0037] Therefore, Figure 1 There are a total of 9 factors that affect the position and angle of segment pasting. In this article, the factors that affect the position and angle of segment pasting (or collectively referred to as affecting the pose of the segment pasting robot) are uniformly referred to as degrees of freedom. That is, in the Figure 1 segment pasting robot, there are a total of 6 rotational degrees of freedom and 3 translational degrees of freedom. In the embodiments of the present application, the specific structure of the segment pasting nine-axis linkage robot can refer to the patent publication document CN118669156B, and the present application will not elaborate too much.

[0038] In some possible embodiments, the method is applied to a segment pasting nine-axis linkage robot, and the segment nine-axis linkage robot includes multiple motion joints and multiple moving axes, which can be as Figure 2 shown, and the method includes the following steps: S21. Obtain the desired trajectory and the actual control parameters at the current moment.

[0039] Among them, the desired trajectory is represented by the symbol , and the actual control parameters include: the actual pose , the actual speed , and the actual environmental acting force . The desired trajectory is the set moving trajectory of the segment pasting nine-axis linkage robot.

[0040] S22. Determine the variable impedance coefficient when the segment pasting nine-axis linkage robot moves according to the desired trajectory according to the preset intelligent impedance calculation model.

[0041] Among them, the variable impedance coefficient can be represented by the symbol f 0 The preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters.

[0042] S23. Based on the variable impedance coefficient, determine the expected trajectory of the segment-pasting nine-axis linkage robot under the influence of the variable impedance coefficient; among them, the expected trajectory can be represented by the symbol represent.

[0043] S24. Perform inverse kinematics solution of the robotic arm according to the expected trajectory, determine the expected pose obtained after the segment-pasting nine-axis linkage robot moves according to the expected trajectory, and determine the expected joint angle at the next moment corresponding to each motion joint and the expected moving position at the next moment of each moving axis based on the expected pose.

[0044] Among them, the expected pose can be represented by the symbol The expected joint angle at the next moment corresponding to the motion joint refers to the joint angle of the motion joint when the segment-pasting nine-axis linkage robot is in the target expected pose where the target expected pose is the expected pose at the next moment corresponding to the expected pose at the current moment K. Similarly, the expected moving position at the next moment of the moving axis refers to: the position of the moving axis when the segment-pasting nine-axis linkage robot is in the target expected pose where.

[0045] S25. Generate a force control signal according to the expected pose and the actual pose, and based on the force control signal, control each motion joint to move to the expected joint angle, and control each moving axis to move to the expected moving position.

[0046] This method obtains the desired trajectory, the actual pose, actual speed, and actual environmental force at the current moment. Then, based on the preset intelligent impedance calculation algorithm, determine the variable impedance coefficient existing when the segment-pasting nine-axis linkage robot moves according to the desired trajectory, and based on this variable impedance coefficient, determine the expected trajectory that the segment-pasting nine-axis linkage robot can be in under the influence of this variable impedance coefficient. Then, perform inverse kinematics solution of the robotic arm according to this expected trajectory to determine the expected pose obtained after the segment-pasting nine-axis linkage robot moves according to this expected trajectory. Then, based on this expected pose, deduce the expected joint angle at the next moment and the expected moving position of the moving axis of the segment-pasting nine-axis linkage robot. Finally, generate a force control signal according to this expected pose and the actual pose, and control the motion joints of the robot to move to the expected joint angle and control the moving axes of the robot to move to the expected moving position based on this force control signal.

[0047] Thus, by selecting the embodiment of the present application, in the segment pasting task for shield tunnels, according to the expected pose to be pasted and combined with the current actual pose, the variable impedance coefficient during the movement of the nine-axis linkage robot for segment pasting can be determined. Then, by introducing this variable impedance coefficient, the poses corresponding to each motion joint and moving axis can be determined when the final pasting pose of the nine-axis linkage robot for segment pasting is in the target pose. And according to this pose, the angular motion of the joints of the nine-axis linkage robot for segment pasting and the movement of the moving axis are controlled to calibrate and compensate the movement amounts of the motion joints and moving axis according to this variable impedance coefficient, so that finally the nine-axis linkage robot for segment pasting can adaptively and accurately perform segment pasting according to the expected pose to be achieved, and finally achieve the high-precision, high-efficiency and high-reliability effects of the segment pasting task during the construction of shield tunnels.

[0048] The above steps S21 to S25 will be described in detail below in combination with specific implementation examples: In the embodiment of the present application, the nine-axis linkage robot for segment pasting has the capabilities of data acquisition, data analysis, and data storage. During the operation of the nine-axis linkage robot for segment pasting, log records can be generated according to the specific operation conditions, and various motion parameters, program instructions, segment pasting results and other information during the execution of segment pasting by the segment pasting robot are clearly stored in the log records. In addition, the nine-axis linkage robot for segment pasting also has the program running function, and can execute the segment pasting task according to the requirements set by the program by loading, compiling, debugging, and running the set segment pasting program. During the operation of the nine-axis linkage robot for segment pasting, the specific pose of the current nine-axis linkage robot for segment pasting can be obtained in real time through the set position sensor and angle sensor, and the specific angles and positions of the pose are recorded in the log records in the form of motion parameters.

[0049] Based on this, as an implementation method, during the execution of the above step S21, the real-time data of the position sensor and angle sensor at the current moment can be obtained, and then based on this real-time data, the real-time control parameters at the current moment can be calculated to obtain the actual pose at the current moment Calculate the actual running speed of the nine-axis linkage robot for segment pasting , and the actual environmental acting force existing during the operation . As another implementation method, during the execution of the above step S21, the real-time data at the current moment recorded in the log record can be read, and by using the same calculation method, the actual pose at the current moment , actual speed , and actual environmental acting force can be calculated.

[0050] In the embodiment of the present application, the desired trajectory is a movement trajectory preset by the segment pasting program, which is used to indicate the trajectory that the nine-axis linkage robot for segment pasting needs to move to achieve the expected segment pasting effect. During the execution of step S21, this desired trajectory can be obtained by reading the parameters set in the program. This desired trajectory is the expected trajectory. Since there are actual interference factors such as friction and external forces during the operation of the nine-axis linkage robot for segment pasting, the nine-axis linkage robot for segment pasting cannot accurately reach the target pose corresponding to this desired trajectory. Therefore, it is necessary to use the method provided in the embodiment of the present application to perform motion compensation on the motion joints and moving axes of the nine-axis linkage robot for segment pasting, so that the final motion joints and moving axes are finally at the joint angles and moving axis positions corresponding to the desired trajectory, so that the final pose of the nine-axis linkage robot for segment pasting is consistent with the target pose expected in the desired trajectory.

[0051] On this basis, how to accurately evaluate the interference of factors such as external environmental forces and friction on the movement of the nine-axis linkage robot for segment pasting has become the key to realizing accurate segment pasting. In the embodiment of the present application, a variable impedance coefficient is introduced to determine the interference situation existing during the operation of the nine-axis linkage robot for segment pasting through actual control parameters.

[0052] Among them, the variable impedance coefficient specifically refers to the parameter of the change in the impedance characteristics when the nine-axis linkage robot for segment pasting interacts with the external environment, and parameters such as inertia, damping, and stiffness existing during the movement of the robot can be dynamically adjusted through this variable impedance coefficient. As an implementation method, by continuously dynamically adjusting this variable impedance coefficient, the nine-axis linkage robot for segment pasting can flexibly adjust its own mechanical characteristics according to different segment pasting task requirements and external environmental conditions, so as to achieve accurate, efficient, and flexible segment pasting operations.

[0053] In some possible embodiments, as shown in Figure 3 the schematic flow diagram, the desired trajectory and the current actual control parameters can be input into a preset intelligent impedance calculation model. Based on the input desired trajectory and the current actual control parameters (including the actual pose , the actual speed , and the actual environmental force ), the variable impedance coefficient f 0 is calculated, that is, the variable impedance coefficient is determined by executing the above step S22 f 0 . As an implementation method, this step S22 can be implemented through the following steps S22-1 to S22-3: S22-1. For the actual pose and the actual speed perform the forward kinematics solution of the robotic arm to determine the actual trajectory of the segment pasting nine-axis linkage robot at the current moment .

[0054] Among them, the forward kinematics solution refers to solving the pose (position and attitude) of the end of the segment pasting nine-axis linkage robot in space according to the motion parameters of the joints and moving axes of the segment pasting nine-axis linkage robot (such as joint angles, joint speeds, moving axis displacements, moving axis moving speeds, etc.). The end of the segment pasting nine-axis linkage robot specifically refers to the operating part of the segment pasting robot that performs segment pasting. Among them, the pose points at multiple consecutive moments are connected in chronological order to obtain the actual trajectory at the current moment .

[0055] S22-2. According to the actual trajectory and the desired trajectory , determine the moving tracking error of the segment pasting nine-axis linkage robot

[0056] In some possible embodiments, the moving tracking error includes: position tracking error , speed tracking error , acceleration tracking error , force error .

[0057] In the embodiments of the present application, in order to enable the joint angle at which the moving joint of the segment pasting nine-axis linkage robot is located at the current moment, and the position of the moving axis at which the moving axis is located at the current moment to adjust its stiffness, damping, and inertia according to the requirements of the segment pasting task to adapt to different environmental interactions, the present application specifically designs variable impedance control for the segment pasting nine-axis linkage robot, so that when the segment pasting nine-axis linkage robot interacts with an unknown changing environment, it can adapt to the impedance parameters corresponding to the unknown environment

[0058] When the external environment change is known, the desired impedance of the segment pasting nine-axis linkage robot provided by the present application (that is, the impedance in the ideal case should satisfy the mathematical model of the following formula (1)) can be described by the following formula (1): Formula (1) Among them respectively represent: represents the desired inertia matrix that changes with time represents the desired damping matrix that changes with time represents the desired stiffness matrix that changes with time, and all three are matrix. To ensure linear response and decoupling in each degree of freedom, these impedance parameters are selected as a diagonal matrix.

[0059] Among them, in formula (1), respectively represent the position tracking error, velocity tracking error, and acceleration tracking error existing in the segment-pasting nine-axis linkage robot. represents the force error. This position tracking error , velocity tracking error , acceleration tracking error , force error can be calculated by the following formula (2), that is, when performing the above step S22-2, the movement tracking error of the segment-pasting nine-axis linkage robot can be determined by the following formula (2): Formula (2) Among them, in the implementation of this application , respectively represent a(t) the first derivative and second derivative of is the expected environmental force, is the actual environmental force.

[0060] Based on the above formula (1) and formula (2), the expected impedance mathematical model of the segment-pasting nine-axis linkage robot can be obtained. Further, considering the actual unknown environment, it is necessary to ensure that the robot can adapt to its impedance parameters. By improving the above formula (1) and formula (2), the actual impedance mathematical model of the segment-pasting nine-axis linkage robot is obtained, and this actual impedance mathematical model satisfies the following formula (3): Formula (3) Among them, is the actual inertia matrix that changes with time, is the nominal parameter of the actual inertia matrix, is the actual damping matrix that changes with time, is the nominal parameter of the actual damping matrix, is the actual stiffness matrix that changes with the actual situation, is the nominal parameter of the actual stiffness matrix.

[0061] Among them , , are known. , , are dynamic parameters that change with time and are unknown.

[0062] By transforming formula (3), formula (3) can be rewritten as formula (4) as follows: Formula (4) Among them, formula (4) can be simplified to formula (5) as follows: Formula (5) By transforming formula (5), the following variable impedance coefficient formula can be obtained, that is, the corresponding variable impedance coefficient can be calculated through formula (6) f 0 : Formula (6) Among them, is the stability coefficient, which is used to represent the stability of the variable impedance characteristic and is affected by the variable impedance coefficient f 0 Specifically, it can be calculated through the following formula (7): Formula (7) Among them, is a constant or a designed parameter, usually greater than 0, Specifically, a preset intelligent impedance calculation model can be constructed based on formula (6), and the following step S22-3 is further executed to determine the variable impedance coefficient of the segment-pasting nine-axis linkage robot.

[0063] S22-3. Input the movement tracking error into the preset intelligent impedance calculation model, and determine that the output result of the preset intelligent impedance calculation model is the variable impedance coefficient f 0 .

[0064] As an implementation manner, in the process of executing step S22-3, it can be realized through the following steps S22-31 and S22-32: S22-31. Based on the variable impedance coefficient formula, construct an RFWNN neural network model to obtain the preset intelligent impedance calculation model; S22-32. Input the movement tracking error into the preset intelligent impedance calculation model in the form of an input vector, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient based on the input vector f 0 .

[0065] Selecting the embodiment of the present application to adaptively adjust the variable impedance coefficient through RFWNN f 0 , it can enable the segment-pasting nine-axis linkage robot to interact with a changing and unknown environment.

[0066] Specifically, based on the variable impedance coefficient formula corresponding to the above formula (6), an RFWNN (Recurrent Fuzzy wavelet Neural Network) neural network model can be constructed to obtain a preset intelligent impedance calculation model. Then, the above-mentioned mobile tracking error is input into the preset intelligent impedance calculation model in the form of an input vector. The preset intelligent impedance calculation model constructed by the RFWNN neural network model adaptively outputs the variable impedance coefficient based on the input vector. f 0 : Among them, the model architecture of the preset intelligent impedance calculation model constructed based on RFWNN includes: an input layer, a fuzzification and wavelet layer, a rule layer, a waveform layer, and an output layer. Based on this, the above steps S22-31 and S22-32 can be specifically implemented through the following steps 1) to 4): 1) Input the input vector into the input layer.

[0067] Among them, each input node of the input layer corresponds to an input variable, receives each mobile tracking error generated by the nine-axis linkage robot of the segment pasting robot, and provides the original data for the data processing of subsequent layers. Specifically, the input signal of this input layer is usually represented in the form of an input vector. Exemplarily, the input signal is , where each input variable in the vector corresponds to a mobile tracking error. Exemplarily, x 1 is the position tracking error , x 2 is the speed tracking error , x 3 is the acceleration tracking error , x 4 is the acting force error .

[0068] 2) Through the fuzzification and wavelet layer, with the help of the Gaussian membership function, determine the output result of the wavelet nodes of the fuzzification and wavelet layer.

[0069] The fuzzification and wavelet layer calculates the membership degree of each signal of the input layer through the Gaussian membership function, and uses the wavelet transform combined feedback mechanism for signal processing. Among them, the Gaussian membership function satisfies the constraint of the following formula (8): Formula (8) Among them, represents the The membership function of the input variable, where m is the number of input variables in the input signal and n is the number of fuzzy rules. Among them, and are the center and width of the Gaussian membership function respectively. Among them, the fuzzy rule refers to the relationship between the input and the output described in the form of "IF-THEN". Exemplarily, in the fuzzy rule: if the contact force is very large and the speed is very fast, then increase the damping coefficient. At this time, "the contact force is very large" and "the speed is very fast" are fuzzy sets of an input variable, defined by the Gaussian membership function, and "increase the damping coefficient" is the fuzzy set of the output variable. Among them, the fuzzy rule can be expressed as: indicating the nth fuzzy rule, where n represents the nth fuzzy rule.

[0070] Based on the above formula (8), the output results of each wavelet node of the fuzzification and wavelet layer can be determined satisfying the following formula (9): Formula (9) Among them, and are the translation and scaling factors of the mother wavelet respectively, is the input of the wavelet node. The input of this wavelet node can be calculated by the following formula (10): Formula (10) Among them, is the information storage parameter, is the previous value of the mother wavelet function in the fuzzification and wavelet layer.

[0071] 3) The rule layer calculates the activation degree of each fuzzy rule of the rule layer according to the output result of the wavelet node.

[0072] Among them, each neural node in the rule layer calculates the activation intensity of the fuzzy rule associated with the neural node, and combines the calculated activation intensity with the output result of the above wavelet node for weighted processing, providing a feature representation after rule matching for the subsequent waveform layer and output layer.

[0073] Specifically, the number of nodes in the rule layer is equal to the number of fuzzy rules and wavelet nodes. Each node in the rule layer corresponds to a fuzzy rule, and each node in the rule layer corresponds to a wavelet node. The activation degree of each rule can be calculated by using the AND (product) operation , and the activation degree of each fuzzy rule can be calculated by the following formula (11): Formula (11) Correspondingly, the jThe output of the wavelet layer can be calculated by the following formula (12): Formula (12) where is the output result of the j th wavelet layer, is the weight of the output of the j th rule part, is the output result of the defuzzification and each wavelet node of the wavelet layer: 4) Using the waveform layer and the output layer, perform defuzzification based on the activation degree of each fuzzy rule, and output the variable impedance coefficient f 0 .

[0074] where the waveform layer can calculate the corresponding output result by executing the following formula (13) y : Formula (13) On this basis, the output layer can defuzzify the output y of the waveform layer according to the following formula (14), and finally adaptively output the variable impedance coefficient: Formula (14) In the process of executing the above steps 1) to 4), the present application defines a learning cost function for measuring the learning cost of the RFWNN model, and measures the difference between the predicted output of the RFWNN model and the variable impedance coefficient of the actual target output, which is equivalent to the loss function in the model training process. The model parameters of the RFWNN model are optimized by minimizing the difference between the predicted output and the variable impedance coefficient of the actual target output.

[0075] Specifically, the learning cost function E(t) can satisfy the constraint of the following formula (15): Formula (15) where is the gain function of the speed error , is the gain function of the force error , which can be set according to requirements and can be variable or constant. Among them, the speed error can be calculated by the following formula (16): Formula (16) The force error can be calculated by the following formula (17): Formula (17) Furthermore, the gradient descent algorithm can be used to adjust the network weights of the neural network model of the RFWNN to optimize the cost function. E(t) Specifically, assume the following parameter vector p (see the following formula (18)) contains the parameters to be adjusted in the RFWNN network model:

[0076] Among them, represents the weight of the nth fuzzy rule, represents the mean of the Gaussian membership function of the mth input variable in the nth fuzzy rule, represents the width (also called the standard deviation) of the Gaussian membership function of the mth input variable in the nth fuzzy rule, represents the size parameter of the wavelet of the mth input variable in the nth fuzzy rule, which is used to control the width of the waveform, represents the translation parameter of the wavelet of the mth input variable in the nth fuzzy rule, which is used to control the position of the waveform, represents the recursive parameter of the mth input variable in the nth fuzzy rule, which is used to store historical information.

[0077] Specifically, the parameters to be adjusted can be updated through the following formula (19): Formula (19) Among them, , , respectively represent t+ the parameter vectors at time 1, t time and t-1 time p . represents the learning rate of the fuzzy processing function part of the RFWNN, which is used to control the update step size of the fuzzy part parameters (membership function weight, center, width), represents the learning rate of the wavelet output part of the RFWNN, which is used to control the update step size of the wavelet part parameters (scale, translation, recursive parameter). Among them, the wavelet parameters need to be adjusted more finely, is much smaller than . is the momentum term coefficient of the RFWNN model, which is used to accelerate the model convergence speed and prevent oscillation. is the gradient of the cost function with respect to the parameters calculated by the chain rule.

[0078] Furthermore, the specific weight update rule can be expressed as the following formula (20) and formula (21): For the parameters of the fuzzy part (such as: ), can be updated through the following formula (20): Formula (20) For the parameters of the wavelet part (such as ), can be updated through the following formula (21): Formula (21) Selecting the embodiment of the present application, the RFWNN neural network model can be trained based on the above formula (6), so that the trained RFWNN model can accurately output the variable impedance coefficient of the segment pasting nine-axis linkage robot in response to the current environment based on the input desired trajectory and actual control parameters, and then adjust the current pose of the segment pasting nine-axis linkage robot with the help of the variable impedance coefficient, so that the nine-axis linkage robot moves to the target pose expected to be reached in the desired trajectory for segment pasting, thereby achieving high-precision, high-reliability and high-efficiency completion of the segment pasting task.

[0079] In some possible embodiments, refer to the flowchart as shown in Figure 3 When the preset intelligent impedance calculation model outputs the corresponding variable impedance coefficient based on the input desired trajectory x d (t) and the current actual control parameters f 0 After that, by executing the above step S23, variable impedance processing can be performed on the desired trajectory x d (t) to obtain the desired trajectory x d (t) At the expected trajectory corresponding to the variable impedance coefficient f 0 and based on the expected trajectory x a , the corresponding expected pose x a can be determined x I .

[0080] As an implementation manner, as shown in Figure 4 , by using the variable impedance coefficient output by the preset intelligent impedance calculation model f 0 and the acting force error e f to perform variable impedance solution on the desired trajectory x d (t) the expected trajectory corresponding to the desired trajectory can be calculated x aSpecifically, based on the above formula (5), the following formula (22) can be obtained: Formula (22) The expected trajectory x d (t) , force error e f , variable impedance coefficient f 0 After substituting into formula (22), the variable impedance solution formula can be obtained, that is, the following formula (23): Formula (23) in, is the expected acceleration corresponding to the expected trajectory, for x a The corresponding acceleration is the actual acceleration generated by the segment pasting robot after passing through the variable impedance characteristic. The expected trajectory x a The second-order derivative of x a .

[0081] Further, step S24 is executed, according to the expected trajectory x a Solve the inverse kinematics of the robot arm to determine the expected trajectory of the nine-axis linkage robot for segment pasting x a After moving, the expected pose is , and based on the expected pose Determine the expected joint angle corresponding to each of the motion joints at the next moment and the expected moving position of each of the moving axes at the next moment.

[0082] In the embodiments of the present application, posture is a general term for position and posture. From a macroscopic perspective, the posture can be the position and posture of the entire segment pasting robot. From a microscopic perspective, the posture specifically refers to the position and posture of each motion joint, and the position and posture of each moving axis. When the posture is specifically a general term for the position and posture of each motion joint and each moving axis, the trajectory and position are expressed in the form of a vector, and each element in the vector corresponds to the value of the nine degrees of freedom of the nine-axis linkage robot, that is, the degrees of freedom of six motion joints and three moving axes. That is, illustratively, the posture can be expressed in the form of a vector as follows: =[M1,M2,M3,M4,M5,M6,X,Y,Z], where M1~M6 correspond to the degrees of freedom of the six motion joints, and X, Y, and Z are the degrees of freedom of the three moving axes.

[0083] Based on this, in some possible embodiments, as shown in Figure 4 by performing inverse kinematics solution on the expected trajectory the expected pose corresponding to the movement of the segment pasting nine-axis linkage robot along the expected trajectory can be determined After moving. x I In the embodiments of the present application, forward kinematics solution can be understood as determining the actual movement trajectory of the robot based on the actual pose of the robot. Inverse kinematics solution is opposite to forward kinematics solution, which is to reversely determine the pose of the robot according to the movement trajectory of the robot. Specifically, it can be known the target poses of the moving joints and moving axes of the robot, and solve the joint angles of the moving joints corresponding to the target pose and the moving positions of the moving axes.

[0084] In the embodiments of the present application, the expected pose can be specifically represented as the expected displacement vectors of each joint in the joint space, and can represent the expected velocity vectors of each joint in the joint space. Specifically, it can be obtained by inverse kinematics solution of and where, is the end pose of the robotic arm of the segment pasting nine-axis linkage robot, is the pose of the segment pasting nine-axis linkage robot at the current joint angle. As an implementation manner, based on the Newton numerical iteration method, the inverse kinematics solution process can be embodied as the following formula (24): Formula (24) where, is the corresponding generalized inverse at the current joint angle, which is specifically calculated by the following formula (25): Formula (25) where, J is the nine-axis Jacobian matrix, is its transpose matrix, and the vector product method is used for solution.

[0085] Based on this, during the execution of step S24 and step S25, the expected joint angles of each moving joint and the expected moving positions of each moving axis at the next moment can be determined through inverse kinematics solution of steps S51 to S57 as shown in Figure 5 : S51. Obtain the set expected pose X I . The expected pose X I here is the expected trajectory determined by using the above step S23 x aThe expected pose obtained by inverse kinematics solution x I ; S52. Set the starting joint angle and the starting position of the moving axis. The starting joint angle and the starting position of the moving axis can be initial values and are represented by vectors = [0 0 0 0 0 0 0 0 0].

[0086] S53. Perform forward kinematics solution to determine the actual joint angles at the current moment and the current pose corresponding to the actual position of the moving axis at the current moment. Exemplarily, the actual pose can be directly obtained x R as the current pose.

[0087] S54. Calculate the pose difference △X between the current pose and the expected pose I . Specifically, the pose difference can be calculated in the form of vector subtraction.

[0088] S55. Perform inverse kinematics solution to update the current actual joint angles and the current actual position of the moving axis, and obtain the target pose X corresponding to the next moment I(K+1) = X I(K) + △X I . Specifically, the target pose corresponding to the next moment can be determined using the above formula (24).

[0089] S56. Determine whether the pose difference < ɛ. If it is greater than ɛ, return to step S53 and perform iterative calculation and judgment , where ɛ represents the allowed error, which can be flexibly set according to actual needs and is not strictly limited in this application. If it is less than ɛ, it means that the difference between the actual pose and the expected pose of the segment pasting nine-axis linkage robot is small enough, and the segment pasting accuracy is high enough. At this time, step S57 can be executed, and a force control signal is generated based on the target pose X I(K+1) corresponding to the target joint angle and the position of the target moving axis.

[0090] As a possible implementation manner, during the execution of the above step S57 or step S25, as shown in Figure 4 , by inputting the expected pose x I , the actual pose , and the actual speed into the controller constructed based on the dynamic surface fuzzy mechanism, the controller constructed based on the dynamic surface fuzzy mechanism generates a force control signal based on the input information . Among them, the operating principle of the controller is to construct a force control signal generator using the actual dynamic model of the segment pasting nine-axis linkage robot.

[0091] Among them, in the ideal state, the nominal dynamic model of the segment pasting nine-axis linkage robot can be expressed by the following formula (27): Formula (27) Wherein, is the joint displacement vector of the corresponding moving joint, or the displacement vector of the moving axis, or the displacement vector composed of the combination of the moving joint and the moving axis. is q 's velocity vector, is q 's acceleration vector. is the joint torque or force control signal applied by the segment pasting nine-axis linkage robot. is the inertia matrix, is the Coriolis force and centrifugal force matrix, is the gravity matrix, is the friction force vector, which can also be the actual environmental acting force.

[0092] Due to the existence of modeling errors and external disturbances in the actual environment. Therefore, the above inertia matrix, Coriolis force and centrifugal force matrix, gravity matrix, and friction force vector will all have uncertainties. Separating this uncertainty, the following formula (28) can be obtained: Formula (28) Wherein, is the actual inertia matrix of the segment pasting nine-axis linkage robot, is the uncertainty matrix of the actual inertia matrix, is the actual Coriolis force and centrifugal force matrix of the segment pasting nine-axis linkage robot, is the uncertainty matrix of the actual Coriolis force and centrifugal force matrix, is the actual gravity matrix of the segment pasting nine-axis linkage robot, is the uncertainty matrix of the actual gravity matrix, is the actual friction force matrix, is the uncertainty matrix of the actual friction force matrix, and each uncertainty matrix is unknown.

[0093] Furthermore, the actual dynamic equation of the segment pasting nine-axis linkage robot can be determined based on the formula (28), and the actual dynamic equation is shown as the following formula (29): Formula (29) Combining the above formula (27), the formula (29) can be rewritten as the following formula (30): Formula (30) Wherein, ​​​​​​​​​​It represents the dynamic uncertainty and interference effect of the segment pasting nine-axis linkage robot, satisfying the constraints of the following formula (31):

[0094] Formula (31) Furthermore, substitute the q in the formula with the actual pose of the segment pasting nine-axis linkage robot , and substitute the in the formula with the actual velocity of the segment pasting nine-axis linkage robot . Thus, the actual dynamic model of the nine-axis linkage robot of the segment pasting machine can be obtained as the following formula (32): Formula (32) Wherein, , specifically representing the unknown dynamic uncertainty and interference effect.

[0095] On this basis, the actual dynamic model constrained by the above formula (32) can be used to construct a force control signal generator to generate a force control signal x I based on the input expected pose , actual pose , and actual velocity .

[0096] In the embodiment of the present application, in order to eliminate the uncertainty and interference terms, the present application proposes a controller constructed based on a dynamic surface fuzzy mechanism, as shown in Figure 6 . Through the internal fuzzy mechanism and virtual control mechanism, the gain of the dynamic surface is adjusted to generate an accurate force control signal.

[0097] As an implementation manner, the above step S25 can be specifically implemented through the following steps: S25-1. Calculate the difference between the expected pose and the actual pose to generate a first error surface ; S25-2. Perform virtual control on the first error surface to obtain a virtual control vector , and generate a second error surface based on the actual velocity and the virtual control vector ; S25-3. Generate the force control signal based on the virtual control vector , the second error surface , and in combination with the actual dynamic model of the segment pasting nine-axis linkage robot.

[0098] When performing step S25-1, the expected pose can be calculated according to the following formula (33) x I , the actual pose to find the difference and determine the pose difference : Formula (33) This pose difference can be determined as the first error surface, which specifically represents the position error of each degree of freedom in the joint space, and the corresponding derivative obtained by differentiation corresponds to the velocity error and satisfies the constraint of formula (34): Formula (34) Furthermore, in order to make the pose difference approach 0, the virtual control law corresponding to the following formula (35) can be used to perform virtual control on this pose difference : , where >0 Formula (35) To avoid the disadvantages of multi-surface sliding mode control, can be passed through a first-order filter with a time constant of t ( t＞0 ) (this first-order filter can satisfy the constraint of the following formula (36)) for transmission: Formula (36) In this way, the filtered virtual control vector can be output through this virtual control law, and this is smoother than and is more convenient for the actual controller to generate accurate force control signals based on .

[0099] Furthermore, when performing step S25-2, it can be achieved through the following process: First, construct the second error surface through the following formula (37) : Formula (37) Furthermore, differentiate this second error surface through the following formula (38): Formula (38) Then, substitute the actual dynamic model of the above segment lining pasting nine-axis linkage robot into the formula (36) and formula (38), that is, substitute the actual dynamic model constrained by the above formula (32) into the formula (36) and formula (38), and the following formula (39) can be obtained: Formula (39) Wherein, t is the filtering time constant.

[0100] Furthermore, as an implementation manner, the virtual control vector , the second error surface can be substituted into a preset force control signal generation formula to output the force control signal . Specifically, in the above formula (39), by controlling the second error surface to approach 0, the pose difference between the actual pose and the expected pose can be made to approach 0, so as to achieve precise control of the pose of the segment lining pasting nine-axis linkage robot.

[0101] Specifically, the preset force control signal generation formula can be obtained from the above formula (39) by controlling →0, and the following formula (40) is obtained, that is, the preset force control signal generation formula satisfies the constraint of the following formula (40): Formula (40) Wherein, is the actual Coriolis force and centrifugal force matrix of the segment lining pasting nine-axis linkage robot, is the actual gravity matrix of the segment lining pasting nine-axis linkage robot, is the transposed matrix of the actual Jacobian matrix of the segment lining pasting nine-axis linkage robot, is the actual environmental acting force of the segment lining pasting nine-axis linkage robot, is the actual inertia matrix of the segment lining pasting nine-axis linkage robot, is the second error surface 's gain parameter, is the virtual control law, t is the filtering time constant.

[0102] In the embodiment of the present application, in order to further optimize the parameters of the controller, a closed-loop system equation is proposed to optimize the parameters of the controller. First, the above formula (40) can be substituted into the formula (39), and the following formula (41) can be obtained: Formula (41) Furthermore, the above formula (38) can be substituted into the formula (34) to obtain the following formula (42): Formula (42) That is, formula (42) is the analytical expression of the closed-loop system.

[0103] Furthermore, if the boundary layer error is defined for: Formula (43) Then, by differentiating the boundary layer error, we can obtain the following formula (44): Formula (44) Thus, after substituting the above formulas (42) and (43) into formula (44), the following formula (45) can be obtained: Formula (45) in, .

[0104] Furthermore, the above formula (41), formula (42) and formula (43) can generate the dynamic surface equation shown in the following formula (46): Formula (46) Therefore, the following closed-loop system equation can be obtained, that is, the following formula (47): Formula (47) In the embodiment of the present application, the gain parameter of the first error surface can be adjusted The second error surface Gain parameter , in order to reduce the impact of unknown dynamic uncertainty and interference effect U on actual control. To achieve intelligent control and optimization of the nine-axis linkage robot through redundant manipulator inverse kinematics algorithm combined with fuzzy gain dynamic surface control and cyclic fuzzy wavelet neural network, it can adapt to the complex and changeable construction environment and ensure the safety and quality of the pipe segment pasting process.

[0105] Based on the method described in the first aspect, in the second aspect, the present application provides an adaptive force-position hybrid control device for a nine-axis linkage robot for pipe segment pasting, wherein the device is applied to a nine-axis linkage robot for pipe segment pasting, wherein the nine-axis linkage robot for pipe segment pasting includes multiple motion joints and multiple moving axes, which can be Figure 7 As shown, the device 70 includes: Input module 701, used to obtain the desired trajectory , the actual control parameters at the current moment, the actual control parameters include: the actual posture at the current moment , Actual speed , Actual environmental forces , the expected trajectory is the set moving trajectory of the segment pasting nine-axis linkage robot; The variable impedance coefficient determination module 702 is configured to determine the variable impedance coefficient existing when the segment pasting nine-axis linkage robot moves along the expected trajectory according to a preset intelligent impedance calculation model ; f 0 , where the preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters; The expected trajectory determination module 703 is configured to determine the expected trajectory of the segment pasting nine-axis linkage robot under the influence of the variable impedance coefficient based on the variable impedance coefficient f 0 ; f 0 The inverse kinematics solution module 704 is configured to perform inverse kinematics solution of the robotic arm according to the expected trajectory to determine the expected pose obtained after the segment pasting nine-axis linkage robot moves along the expected trajectory, and determine the expected joint angle at the next moment corresponding to each of the motion joints and the expected moving position at the next moment of each of the moving axes based on the expected pose; ; The position control module 705 is configured to generate a force control signal according to the expected pose and the actual pose, and control each of the motion joints to move to the expected joint angle and control each of the moving axes to move to the expected moving position based on the force control signal. ; ; ; ; ; ; ; ; ;

[0106] Combined with the second aspect, in the second possible embodiment, the input module is further configured to: Perform forward kinematics solution of the robotic arm for the actual pose and the actual speed to determine the actual trajectory of the segment pasting nine-axis linkage robot at the current moment; ; ; ; Determine the moving tracking error of the segment pasting nine-axis linkage robot according to the actual trajectory and the expected trajectory; ; ; Input the moving tracking error into the preset intelligent impedance calculation model to determine that the output result of the preset intelligent impedance calculation model is the variable impedance coefficient f 0 .

[0107] Combined with the second possible embodiment of the second aspect, in the third possible embodiment, the movement tracking error includes: position tracking error , speed tracking error , acceleration tracking error , force error , and the preset intelligent impedance calculation model satisfies the variable impedance coefficient formula constraint; The variable impedance coefficient formula is:

[0108] wherein, is the variable impedance coefficient, is the actual inertia matrix varying with time, is the nominal parameter of the actual inertia matrix, is the actual damping matrix varying with time, is the nominal parameter of the actual damping matrix, is the actual stiffness matrix varying with actual changes, is the nominal parameter of the actual stiffness matrix, is the stability coefficient.

[0109] Combined with the third possible embodiment of the second aspect, in the fourth possible embodiment, the variable impedance coefficient determination module is further configured to: Based on the variable impedance coefficient formula, construct an RFWNN neural network model to obtain the preset intelligent impedance calculation model; Input the movement tracking error in the form of an input vector into the preset intelligent impedance calculation model, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient based on the input vector f 0 .

[0110] Combined with the fourth possible embodiment of the second aspect, in the fifth possible embodiment, the preset intelligent impedance calculation model includes: an input layer, a fuzzification and wavelet layer, a rule layer, a waveform layer, and an output layer; the variable impedance coefficient determination module is further configured to: Input the input vector into the input layer; Through the fuzzification and wavelet layer, with the help of the Gaussian membership function, determine the output result of the wavelet nodes of the fuzzification and wavelet layer; The rule layer calculates the activation degree of each fuzzy rule of the rule layer according to the output result of the wavelet nodes; The waveform layer and the output layer are used to perform defuzzification processing based on the activation degree of each fuzzy rule and output the variable impedance coefficient f0 。

[0111] Combined with the second aspect, in the sixth possible embodiment, the position control module is further configured to: Based on the expected pose and the actual pose , generate a force control signal , including: For the expected pose and the actual pose calculate the difference to generate a first error surface ; Perform virtual control on the first error surface to obtain a virtual control vector , and based on the actual speed and the virtual control vector , generate a second error surface ; Based on the virtual control vector , the second error surface , combined with the actual dynamic model of the segment pasting nine-axis linkage robot, generate the force control signal .

[0112] Combined with the sixth possible embodiment of the second aspect, in the seventh possible embodiment, the position control module is further configured to: Substitute the virtual control vector , the second error surface into a preset force control signal generation formula to output the force control signal ; The preset force control signal generation formula satisfies:

[0113] where is the actual Coriolis force and centrifugal force matrix of the segment pasting nine-axis linkage robot, is the actual gravity matrix of the segment pasting nine-axis linkage robot, is the transposed matrix of the actual Jacobian matrix of the segment pasting nine-axis linkage robot, is the actual environmental acting force of the segment pasting nine-axis linkage robot, is the actual inertia matrix of the segment pasting nine-axis linkage robot, is the second error surface , is the virtual control law, t is the filtering time constant; where the virtual control law Satisfy:

[0114] Wherein, is the gain parameter of the first error surface.

[0115] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0116] In a third aspect, an exemplary embodiment of the present application further provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, it is used to cause the electronic device to execute the method according to the embodiments of the present application. Specifically, as a preferred embodiment, the electronic device may be a segment pasting robot. As an example, the electronic device may be a nine-axis linkage segment pasting robot.

[0117] An exemplary embodiment of the present application further provides a non-transitory computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor of a computer, it is used to cause the computer to execute the method according to the embodiments of the present application.

[0118] An exemplary embodiment of the present application further provides a computer program product, including a computer program, wherein when the computer program is executed by a processor of a computer, it is used to cause the computer to execute the method according to the embodiments of the present application.

[0119] Referring to Figure 8 , the following will describe the structural block diagram of the electronic device 800 that can be used as the server or client of the present application, which is an example of the hardware device applicable to various aspects of the present application. The electronic device is intended to represent various forms of digital electronic computer devices, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processors, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present application described herein and / or claimed.

[0120] As Figure 8As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to computer programs stored in the read-only memory (ROM 802) or computer programs loaded from the storage unit 808 into the random access memory (RAM 803). In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other through a bus 804. The input / output interface (I / O interface 805) is also connected to the bus 804.

[0121] Multiple components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, an output unit 807, a storage unit 808, and a communication unit 809. The input unit 806 can be any type of device capable of inputting information into the electronic device 800. The input unit 806 can receive input digital or character information, and generate key signal inputs related to the user settings and / or function controls of the electronic device. The output unit 807 can be any type of device capable of presenting information, and can include but is not limited to a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 808 can include but is not limited to a magnetic disk, an optical disk. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks, and can include but is not limited to a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a BluetoothTM device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0122] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 801 executes the various methods and processes described above. For example, in some embodiments, the aforementioned segment pasting nine-axis linkage robot adaptive force-position hybrid control method can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 800 via the ROM 802 and / or the communication unit 809. In some embodiments, the computing unit 801 can be configured to execute the aforementioned segment pasting nine-axis linkage robot adaptive force-position hybrid control method in any other appropriate manner (e.g., by means of firmware).

[0123] The program code for implementing the methods of the present application can be written in any combination of one or more programming languages. These program codes can be provided to the processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, such that when executed by the processor or controller, the program codes cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, executed partially on the machine as an independent software package and partially on a remote machine, or executed entirely on a remote machine or server.

[0124] In the context of the present application, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0125] As used in the present application, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., a disk, an optical disk, a memory, a programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal for providing machine instructions and / or data to a programmable processor.

[0126] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, speech input, or tactile input).

[0127] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0128] A computer system can include clients and servers. Clients and servers are generally remote from each other and typically interact through a communication network. The client-server relationship is generated by computer programs running on the respective computers and having a client-server relationship with each other.

Claims

1. An adaptive force-position hybrid control method for a nine-axis linkage robot for segment pasting, characterized in that: The method is applied to a nine-axis linkage robot for pipe segment pasting, wherein the nine-axis linkage robot for pipe segment pasting includes a plurality of motion joints and a plurality of moving axes, and the method includes: Get the expected trajectory , the actual control parameters at the current moment, the actual control parameters include: the actual posture at the current moment , Actual speed , Actual environmental forces , the expected trajectory is the set moving trajectory of the nine-axis linkage robot for segment pasting; According to the preset intelligent impedance calculation model, it is determined that the nine-axis linkage robot for segment pasting follows the desired trajectory. Variable impedance coefficient when moving f 0, wherein the preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters; Based on the variable impedance coefficient f 0, determine the variable impedance coefficient in the f 0, the expected trajectory of the nine-axis linkage robot for segment pasting ; According to the expected trajectory Solve the inverse kinematics of the robot arm to determine the expected trajectory of the nine-axis linkage robot for segment pasting After moving, the expected pose is , and based on the expected pose Determine the expected joint angle corresponding to each of the motion joints at the next moment and the expected moving position of each of the moving axes at the next moment; According to the expected posture The actual posture , generating force control signal , and based on the force control signal , control each of the moving joints to move to the expected joint angle, and control each of the moving axes to move to the expected moving position.

2. The method according to claim 1, characterized in that The method further comprises: For the actual posture and the actual speed Solve the forward kinematics of the robot arm to determine the actual trajectory of the nine-axis linkage robot for segment pasting at the current moment ; According to the actual trajectory and the desired trajectory , determining the mobile tracking error of the nine-axis linkage robot for segment pasting; The mobile tracking error is input into the preset intelligent impedance calculation model, and the output result of the preset intelligent impedance calculation model is determined to be the variable impedance coefficient f 0.

3. The method according to claim 2, characterized in that The mobile tracking error includes: position tracking error , speed tracking error , acceleration tracking error , force error , the preset intelligent impedance calculation model satisfies the variable impedance coefficient formula constraint; The variable impedance coefficient formula is: in, is the variable impedance coefficient, is the actual inertia matrix that varies with time, are the nominal parameters of the actual inertia matrix, is the actual damping matrix that varies with time, are the nominal parameters of the actual damping matrix, is the actual stiffness matrix that varies with the actual conditions, are the nominal parameters of the actual stiffness matrix, is the stability coefficient.

4. The method according to claim 3, characterized in that The method further comprises: Based on the variable impedance coefficient formula, a RFWNN neural network model is constructed to obtain the preset intelligent impedance calculation model; The mobile tracking error is input into the preset intelligent impedance calculation model in the form of an input vector, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient based on the input vector. f 0.

5. The method according to claim 4, characterized in that The preset intelligent impedance calculation model includes: an input layer, a fuzzification and wavelet layer, a rule layer, a waveform layer, and an output layer; the mobile tracking error is input into the preset intelligent impedance calculation model in the form of an input vector, and the preset intelligent impedance calculation model adaptively outputs the variable impedance coefficient based on the input vector. f 0, including: Inputting an input vector into the input layer; Determining the output results of the wavelet nodes of the fuzzy and wavelet layers by using the Gaussian membership function; The rule layer calculates the activation degree of each fuzzy rule of the rule layer according to the output result of the wavelet node; The waveform layer and the output layer are used to perform defuzzification processing based on the activation degree of each fuzzy rule, and the variable impedance coefficient is output. f 0.

6. The method according to claim 1, characterized in that According to the expected posture The actual posture , generating force control signal ,include: For the expected posture The actual posture Take the difference and generate the first error surface ; The first error surface is virtually controlled to obtain a virtual control vector , and based on the actual speed With the virtual control vector , generating the second error surface ; Based on the virtual control vector The second error surface , combined with the actual dynamics model of the nine-axis linkage robot for segment pasting, the force control signal is generated .

7. The method according to claim 6, characterized in that The method further comprises: The virtual control vector The second error surface Substitute into the preset force control signal generation formula and output the force control signal ; The preset force control signal generation formula satisfies: in, The actual Coriolis force and centrifugal force matrix of the nine-axis linkage robot for pasting the pipe segment, The actual gravity matrix of the nine-axis linkage robot for pasting the pipe segment, The transposed matrix of the actual Jacobian matrix of the nine-axis linkage robot for pasting the pipe segment, The actual environmental force of the nine-axis linkage robot for pasting the pipe segment, The actual inertia matrix of the nine-axis linkage robot for pasting the pipe segment, The second error surface The gain parameter, is the virtual control law, t is the filtering time constant; Among them, the virtual control law satisfy: in, is the gain parameter of the first error surface, is the expected pose The expected speed.

8. An adaptive force-position hybrid control device for a nine-axis linkage robot for segment pasting, characterized in that: The device is applied to a nine-axis linkage robot for pipe segment pasting, wherein the nine-axis linkage robot for pipe segment pasting includes a plurality of motion joints and a plurality of moving axes, and the device includes: Input module, used to obtain the desired trajectory , the actual control parameters at the current moment, the actual control parameters include: the actual posture at the current moment , Actual speed , Actual environmental forces , the expected trajectory is the set moving trajectory of the nine-axis linkage robot for segment pasting; The variable impedance coefficient determination module is used to determine the desired trajectory of the nine-axis linkage robot for pasting pipe segments according to the preset intelligent impedance calculation model. Variable impedance coefficient when moving f 0, wherein the preset intelligent impedance calculation model is a neural network model constructed based on the actual control parameters; An expected trajectory determination module is used to determine the expected trajectory based on the variable impedance coefficient f 0, determine the variable impedance coefficient in the f 0, the expected trajectory of the nine-axis linkage robot for segment pasting ; An inverse kinematics solving module is used to solve the expected trajectory according to the Solve the inverse kinematics of the robot arm to determine the expected trajectory of the nine-axis linkage robot for segment pasting After moving, the expected pose is , and based on the expected pose Determine the expected joint angle corresponding to each of the motion joints at the next moment and the expected moving position of each of the moving axes at the next moment; A position control module is used to control the position of the The actual posture , generating force control signal , and based on the force control signal , control each of the moving joints to move to the expected joint angle, and control each of the moving axes to move to the expected moving position.

9. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing a program; wherein the program comprises instructions, and when the instructions are executed by the processor, the processor executes the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to make a computer execute the method according to any one of claims 1-7.

Images