Proposed Force Control Research Articles

714 接触作業時のロボットマニピュレータと未知環境間の摩擦補償(制御・ロボティクス・メカトロニクスIII)

This paper presents an effective intelligent control policy for position/force control of a robot manipulator in an unknown environment based on hybrid control using fuzzy-neural techniques. In the proposed force control policy, suitable fuzzy-neuro force control controllers are selected immediately among the initially prepared multiple fuzzy-neuro force controllers for various kinds of environments, and harmonized with a proper ratio using fuzzy reasoning according to the dynamic properties of the unknown environment. In the proposed position control policy, adaptive fuzzy friction models are applied to compensate for friction between a robot manipulator end-effector and the environment effectively.

A force control approach of robotic manipulators with neural network compensation

In this paper, the application of an impedance control scheme to drive a robot with a desired force profile in the presence of non-rigid environments is considered. However, for the usual impedance based control schemes, fine control of the contact force especially for non-rigid environments is difficult to obtain. Moreover, unmodeled robot model effects and environment uncertainties, can seriously affect the force tracking performance. In this paper, the nonlinear uncertainties in a model-based impedance controller are compensated by an on-line neural network compensation scheme. Further, a predictive algorithm is designed to compute the optimized virtual trajectory for the impedance controller in order to precisely track the desired force profile. The performance of the proposed control scheme is illustrated by simulation for a three degree-of-freedom PUMA 560 robot, which end-effector is forced to move along a surface located on the vertical plane. Simulation results reveal an accurate force tracking performance for the proposed force control scheme, in the presence of non-rigid environments.

TRACTION FORCE CONTROL OF AN ELECTRIC VEHICLE IN 2WS-4WD MODE USING FUZZY REASONING

This paper is concerned with the design of traction force control of an electric vehicle in 2WS–4WD mode using fuzzy reasoning. The electric vehicle is in a motor–in–wheel system, and the traction force in each wheel can be determined independently. The yaw moment is regulated by the motor torque to improve the handling and stability of the vehicle as well as the skidding of lyres. The input variables in the fuzzy control rules are treated as the error of yaw ratio and its time difference, and the slip ratio. The simulation results show that the proposed traction force control improves the performance more than the traction force without control.

Object Handling Using Two Arms Without Grasping

Two cooperative manipulators equipped with open-palm end effectors are particularly suited to handle large objects. This can be achieved by having one palm push from one end and the other from the opposite end. This article studies modeling and coordinated control of two-arm pushing operations. The performance of such tasks requires simultaneous control of the motion trajectory of the object and the interaction force. Control of the interaction force is essential to avoid dropping or crushing objects. The motion and force control problem is complicated by the presence of unilateral constraints, as the manipulators can only push the object. The control problem is first formulated in the state space. A coordinated control method is then presented that utilizes a state feedback to de couple the motion control subsystem from the force control subsystem. The unilateral constraints are satisfied during the entire motion period by using a proposed force control plan ning algorithm. The effectiveness of the control method is verified by simulations.

A Force Control of Multi-Degrees-of-Freedom Manipulator Based on Parameter Estimation.

The paper describes an adaptive force controller. In general force controller, force response depends on unknown parameters of the environment which the manipulator contacts. In case the target environment is hard, the force response of common system is oscillatory. To obtain stable force response which is independent of environment, it is necessary to adjust control gain according to the environment.In the proposed method, a disturbance observer is applied to each joint and the total disturbance effects are canceled out in joint space. In the observer-based manipulator, full dynamics including the environment is cleared up and the workspace force controller is constructed without any disturbance effect. Then the least-squares estimator of unknown parameters of environment is also introduced to adjust the control gain. Several experimental results are shown to verify the proposed force controller.

A Structure of Workspace Force Controller in Multi-Degrees-of-Freedom Manipulator.

The paper describes a workspace force controller of the multi-degrees-of-freedom manipulator. First, the disturbance observer is applied to each joint to improve the robustness against the disturbance effect and the joint acceleration controller is realized in the joint space. Secondly, the workspace motion equation is derived and the workspace force controller is constructed based on its motion equation.In the proposed force controller, the equivalent mass matrix of the workspace is defined clearly and the impedance of the manipulator is selected arbitrary. This is a remarkable point of the proposed method. Several experimental results are shown to verify the proposed controller.