Mechanical arm decentralized optimal fault-tolerant control method considering actuator faults

Abstract

The invention provides a mechanical arm decentralized optimal fault-tolerant control method considering actuator faults The method comprises the following steps that a modular mechanical arm system sub-joint model based on the joint torque sensor feedback technology under the actuator faults is constructed, a Roburg state observer is designed to estimate unknown actuator faults, based on information measured by a joint torque sensor, a comprehensive approximate improved cost function of the unknown actuator faults, tracking performance and controller output is established, an optimal fault-tolerant controller is approximated by utilizing an evaluation neural network structure, output of control torque is reduced while system tracking performance is optimized, and finally a decentralized approximate optimal controller is obtained. Finally, under the different actuator faults, the effectiveness of the method is verified through experimental exploration of modular mechanical arms with different configurations.

Description

technical field [0001] The invention relates to a decentralized optimal fault-tolerant control method of a modular manipulator considering unknown actuator faults, and belongs to the field of robot control systems and control algorithms. Background technique [0002] Modular manipulator is a kind of manipulator with standard modules and interfaces, which can recombine and configure its own configuration according to different task requirements. According to the concept of modularization, the joint module of the modular manipulator includes communication, drive, control, sensing and other units, which can make the manipulator change its own configuration according to the task under different external environments and constraints, so that after reconstruction The robotic arm can have better adaptability to the new working environment. It is precisely because of such structural characteristics that the modular manipulator has the ability to integrate the latest machinery, sens...

AI Technical Summary

Problems solved by technology

However, this solution highlights the shortcomings in actual operation. Since the neural network recognizer needs a large amount of data for training, the output of the actuator fluctuates greatly when the task is first executed, and even exceeds the maximum output limit of the actuator. , causing irreparable damage to the

Images