CN-121973247-A - Multi-mode force feedback control method and system for nuclear facility robot operation

CN121973247ACN 121973247 ACN121973247 ACN 121973247ACN-121973247-A

Abstract

The invention belongs to the technical field of teleoperation and force sense interaction of robots, and provides a multi-mode force feedback control method and a multi-mode force feedback control system for nuclear facility robot operation, wherein state information of a master arm and a slave arm is collected in real time, and a force sense calculation strategy is dynamically scheduled according to task states: A high-precision resolution strategy is enabled in the fine mode of operation and a fast estimation strategy is enabled in the fast transit mode. And combining dynamically selected multi-mode force feedback parameters to synthesize and output a touch instruction to the main arm. Meanwhile, the admittance correction and feedforward compensation are carried out on the slave arm movement instruction by utilizing the net interaction force and the dynamic internal force, and the slave arm movement is controlled by synchronous output after safety optimization. The invention solves the problems of poor task adaptability, force sense distortion and insufficient safety in the teleoperation of the nuclear facilities, and realizes intelligent, high-fidelity and safe teleoperation of the robot.

Inventors

ZHANG PEI
XIE HUAGEN

Assignees

四川厚朴物联科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (8)

1. A multi-mode force feedback control method for nuclear facility robot operation, applied to a master-slave isomorphic robot system comprising a master arm and a slave arm, characterized in that the control method comprises the steps of: acquiring position and posture information of a master arm, six-dimensional force information of a slave arm and motion state information of each joint of the slave arm in real time, and generating a target motion instruction of the slave arm according to the position and posture information mapping of the master arm; dynamically scheduling a force sense calculation strategy according to a real-time task state, wherein the dynamic scheduling is to enable a resolving strategy when in a fine operation mode and enable an estimating strategy when in a fast transit mode; based on the determined force sense calculation strategy, calculating the dynamic internal force of the slave arm by utilizing the motion state information of each joint of the slave arm, and subtracting the dynamic internal force from the six-dimensional force information to obtain a net interaction force; Inputting the net interaction force into a preset resolving model for calculation, and combining preset force feedback mode parameters to synthesize a force feedback instruction fed back to the main arm; And the force feedback device is used for carrying out optimization correction on the target motion instruction of the slave arm based on the determined force sense calculation strategy, the dynamic internal force and the net interaction force, generating and synchronously outputting an optimized slave arm control instruction to a slave arm servo driver and outputting the force feedback instruction to a master arm.
2. The method for multi-mode force feedback control of a nuclear facility robot according to claim 1, wherein the real-time acquisition of the position and posture information of the master arm, the six-dimensional force information of the slave arm, and the movement state information of each joint of the slave arm, and the generation of the target movement command of the slave arm according to the position and posture information map of the master arm, comprises: acquiring position coordinates and attitude angle data of the tail end of a main arm in real time through a main arm sensor array to form a main arm pose vector; Based on a master-slave isomorphic mapping model, converting the master arm pose vector into an expected pose vector at the tail end of a slave arm; mapping the expected pose vector into target angle instructions of all joints of the slave arm by using a slave arm inverse kinematics calculation module; And carrying out kinematic compensation correction on the target angle instruction by combining joint angle and angular velocity data fed back by each joint encoder of the slave arm in real time to generate a target motion instruction of the slave arm.
3. The method of claim 2, wherein dynamically scheduling force sense calculation strategies based on real-time task status comprises: acquiring real-time task state information, wherein the real-time task state information comprises relative pose information of the tail end of the slave arm and a working object and tail end contact state information calculated by six-dimensional force information of the slave arm; Based on the real-time task state information, performing mode judgment on the current working stage, namely judging to enter a fine operation mode when the distance between the tail end of the slave arm and a working object is smaller than a dynamic approach threshold value or the tail end contact state information indicates that contact is generated; The method comprises the steps of dynamically scheduling a corresponding force sense calculation model and calculation frequency based on a mode judgment result, wherein when a fine operation mode is judged, a high-precision force sense solver based on a slave arm complete dynamics model is scheduled and calculated by adopting a first control frequency; outputting the scheduled force sense calculation model, the parameters thereof and the corresponding calculation frequency as the force sense calculation strategy.
4. A multi-mode force feedback control method of nuclear facility robot operation as claimed in claim 3, wherein the generation of the dynamic approach threshold comprises: based on the real-time task state information, extracting the required level of pose precision of the current operation task, the surface characteristics and the size parameters of the operation object and the expected safe initial contact force range; And calculating and generating the dynamic approach threshold adapting to the current task scene through a preset threshold mapping rule according to the requirement level, the surface characteristics and the size parameters of the operation object and the expected safe initial contact force range.
5. The method of claim 4, wherein the calculating the dynamic internal force of the slave arm using the motion state information of each joint of the slave arm based on the determined force sense calculation strategy and subtracting the dynamic internal force from the six-dimensional force information to obtain the net interaction force comprises: calculating to obtain theoretical dynamic internal force by using real-time motion state information of each joint of the slave arm and the dynamic model determined according to the force sense calculation strategy; Six-dimensional force information of the slave arm is introduced as feedback, and model parameters of the dynamic model are identified and compensated on line to generate parameter compensation quantity; Correcting the theoretical dynamic internal force by using the parameter compensation quantity to obtain a compensated dynamic internal force; Subtracting the compensated dynamic internal force from the six-dimensional force information of the slave arm acquired in real time to obtain a preliminary net interaction force, and performing low-pass filtering on the preliminary net interaction force to obtain a final net interaction force.
6. The method of claim 5, wherein inputting the net interactive force into a preset solution model for calculation, and combining preset force feedback mode parameters to synthesize a force feedback command fed back to the main arm, comprises: inputting the net interaction force into a statics resolving model constructed based on a virtual work principle, calculating to obtain a basic force feedback vector, and dynamically selecting a force feedback mode and parameters according to the real-time task state information and the force sense calculating strategy; Based on the selected force feedback mode and parameters, carrying out modeling modulation and dynamic response adjustment on the basic force feedback vector to synthesize a modeling force feedback component; And carrying out synchronization and time domain smoothing on the modeling force feedback component to generate a final force feedback instruction matched with the servo period of the main arm.
7. The method of claim 6, wherein the optimizing calibration of the target motion command of the slave arm based on the determined force sense calculation strategy, the dynamic internal force, and the net interaction force, generating and synchronizing output of the optimized slave arm control command to the slave arm servo driver, and the force feedback device outputting the force feedback command to the master arm, comprises: Based on the net interaction force and the dynamic internal force, respectively carrying out admittance correction and dynamic feedforward compensation on a target motion instruction of a slave arm to generate a preliminary corrected motion expected instruction; carrying out safety limiting and smoothing treatment on the joint moment instruction based on the physical limit of each joint of the slave arm to generate an optimized slave arm control instruction; And performing time synchronization alignment on the optimized slave arm control instruction and the final force feedback instruction according to the calculation frequency in the force sense calculation strategy, and respectively and synchronously outputting the time synchronization alignment to the slave arm servo driver and the force feedback device of the master arm.
8. A nuclear facility robot-operated multi-mode force feedback control system for performing a nuclear facility robot-operated multi-mode force feedback control method as claimed in any one of claims 1 to 7, comprising a master-slave isomorphic robot system of master and slave arms, characterized in that the control system comprises: The motion mapping module is used for acquiring the position and posture information of the master arm, the six-dimensional force information of the slave arm and the motion state information of each joint of the slave arm in real time, and generating a target motion instruction of the slave arm according to the position and posture information mapping of the master arm; the strategy dynamic scheduling module is used for dynamically scheduling the force sense calculation strategy according to the real-time task state, wherein the dynamic scheduling is to enable the calculation strategy when in a fine operation mode and enable the estimation strategy when in a fast transit mode; The net force extraction module is used for calculating the dynamic internal force of the slave arm by utilizing the motion state information of each joint of the slave arm based on the determined force sense calculation strategy, and subtracting the dynamic internal force from the six-dimensional force information to obtain net interaction force; the instruction synthesis module is used for inputting the net interaction force into a preset resolving model for calculation, and synthesizing a force feedback instruction fed back to the main arm by combining preset force feedback mode parameters; The command output module is used for carrying out optimization correction on the target motion command of the slave arm based on the determined force sense calculation strategy, the dynamic internal force and the net interaction force, generating and synchronously outputting an optimized slave arm control command to a slave arm servo driver, and outputting the force feedback command to a force feedback device of the master arm.

Description

Multi-mode force feedback control method and system for nuclear facility robot operation Technical Field The invention relates to the technical field of teleoperation and force sense interaction of robots, in particular to a multi-mode force feedback control method and system for nuclear facility robot operation. Background In the field of nuclear industry, in high-radioactivity and high-risk working environments such as nuclear power station reactor internal maintenance, radioactive waste treatment, facility retirement and demolition, direct personnel operation is not realistic, and robots are required to be relied on for remote teleoperation. The master-slave isomorphic teleoperation robot system can remotely reproduce the actions of operators and feed back the force sense information of the working environment to the operators, so that the master-slave isomorphic teleoperation robot system becomes key equipment for completing the fine and complex tasks. However, the operating environment inside the nuclear facility is extremely specific and task-versatile, providing a serious challenge to the existing robot-fed-back teleoperation technology: 1. Task scenarios and requirements vary, and the task may include rapid transit in a wide space, precision assembly in a narrow space, compliant manipulation of fragile components, forced breaking of hard structures, and the like. It is difficult for a single force feedback and control mode to meet contradictory requirements of "fast", "accurate", "compliant", "powerful", etc. simultaneously. The operator needs to frequently carry out cognitive switching and manually adjust control parameters among different tasks, and the method has the advantages of heavy burden, low efficiency and easy error. 2. Environmental perception and model uncertainty, namely complex internal structure of nuclear facilities, equipment state change possibly caused by long-term irradiation, and poor adaptability of the traditional control method based on a preset accurate model. Factors such as robot body dynamics and joint friction can pollute the reading of the end force sensor, so that the force sense fed back to an operator comprises interference of the motion of the robot, and judgment of fine contact and force state is not real and pure enough and is seriously influenced. 3. The operation safety and the presence sense are balanced, on one hand, the robot must be ensured not to damage fragile equipment or the robot when in accidental contact or operation failure, and the force feedback is required to provide enough flexibility or warning, and on the other hand, the robot must provide real and delay-free force feedback when in high-rigidity contact in order to ensure the operation efficiency and the operation precision. Existing systems often have difficulty achieving a dynamic, intelligent balance between security and operational authenticity. Therefore, there is a need to provide a multi-mode force feedback control method and system for nuclear facility robot operation that solves the above-mentioned problems. Disclosure of Invention In order to solve the technical problems, the invention provides the multi-mode force feedback control method and the system for the nuclear facility robot operation, which effectively solve the problems of poor task adaptability, force feedback distortion, insufficient man-machine cooperative safety and the like in the teleoperation of the nuclear facility robot and improve the overall operation performance of the system in a complex and high-risk nuclear environment through the collaborative design of dynamic dispatching of force sense calculation strategies, multi-mode force feedback synthesis and closed loop optimization of motion instructions. The invention provides a multi-mode force feedback control method for nuclear facility robot operation, which is applied to a master-slave isomorphic robot system comprising a master arm and a slave arm, and comprises the following steps: acquiring position and posture information of a master arm, six-dimensional force information of a slave arm and motion state information of each joint of the slave arm in real time, and generating a target motion instruction of the slave arm according to the position and posture information mapping of the master arm; dynamically scheduling a force sense calculation strategy according to a real-time task state, wherein the dynamic scheduling is to enable a resolving strategy when in a fine operation mode and enable an estimating strategy when in a fast transit mode; based on the determined force sense calculation strategy, calculating the dynamic internal force of the slave arm by utilizing the motion state information of each joint of the slave arm, and subtracting the dynamic internal force from the six-dimensional force information to obtain a net interaction force; Inputting the net interaction force into a preset resolving model for calculation, and combining preset force f