A neural network-based approach for an assembly cell control

In several robotics applications, the robot must interact with the workspace, and thus its motion is constrained by the task. In this case, pure position control will be ineffective since forces appearing during the contacts must also be controlled. However, simultaneous position and force control called hybrid control is then required. Moreover, the nonlinear plant dynamics, the complexity of the dynamic parameters determination and computation constraints makes more difficult the synthesis of control laws. In order to satisfy all these constraints, an effective hybrid force/position approach based on artificial neural networks for multi-inputs/multi-outputs systems is proposed. This approach realizes, simultaneously, an identification and control of systems, and it is implemented according to two phases: At first, a neural observer is trained off-line on the basis of the data acquired during contact motion, in order to realize a smooth transition from free to contact motion. Then, an online learning of the neural controller is implemented using neural observer parameters so that the closed-loop system maintains a good performance and compensates for uncertain/unknown dynamics of the robot and the environment. A typical example on which we shall focus is an assembly task. Experimental results on a C5 links parallel robot demonstrate that the robot's skill improves effectively and the force control performances are satisfactory, even if the dynamics of the robot and the environment change.     

Stability of hybrid position and force control for robotic manipulator with kinematics and dynamics uncertainties

Most research so far on hybrid position and force control laws of robotic manipulator has assumed that knowledge of kinematics is known exactly. In the presence of modeling errors, it is unknown whether the stability of the constrained robot system can still be ensured. In this paper, stability and setpoint control problems of constrained robot with kinematics and dynamics uncertainties are formulated and solved. We shall show that the manipulator end-effector's motion is asymptotically stable even in the presence of such uncertainties.     

Human Interaction Dynamics for Its Use in Mobile Robotics:Impedance Control for Leader-follower Formation

A complete characterization of the behavior in human-robot interactions (HRI) includes both:the behavioral dynamics and the control laws that characterize how the behavior is regulated with the perception data. In this way, this work proposes a leader-follower coordinate control based on an impedance control that allows to establish a dynamic relation between social forces and motion error. For this, a scheme is presented to identify the impedance based on fictitious social forces, which are described by distance-based potential fields. As part of the validation procedure, we present an experimental comparison to select the better of two different fictitious force structures. The criteria are determined by two qualities:least impedance errors during the validation procedure and least parameter variance during the recursive estimation procedure. Finally, with the best fictitious force and its identified impedance, an impedance control is designed for a mobile robot Pioneer 3AT, which is programmed to follow a human in a structured scenario. According to results, and under the hypothesis that moving like humans will be acceptable by humans, it is believed that the proposed control improves the social acceptance of the robot for this kind of interaction.      

Adaptive hybrid force/position control of robot manipulators using an adaptive force estimator in the presence of parametric uncertainty

This study is devoted to sensorless adaptive force/position control of robot manipulators using a position-based adaptive force estimator (AFE) and a force-based adaptive environment compliance estimator. Unlike the other sensorless method in force control that uses disturbance observer and needs an accurate model of the manipulator, in this method, the unknown parameters of the robot can be estimated along with the force control. Even more, the environment compliance can be estimated simultaneously to achieve tracking force control. In fact, this study deals with three challenging problems: No force sensor is used, environment stiffness is unknown, and some parametric uncertainties exist in the robot model. A theorem offers control laws and updating laws for two control loops. In the inner loop, AFE estimates the exerted force, and then, the force control law in the outer loop modifies the desired trajectory of the manipulator for the adaptive tracking loop. Besides, an updating law updates the estimated compliance to provide an accurate tracking force control. Some experimental results of a PHANToM Premium robot are provided to validate the proposed scheme. In addition, some simulations are presented that verify the performance of the controller for different situations in interaction.     

双机械手对称协调力/位混合控制——模型、控制算法与实现

该文研究双手协调运动和力控制方法.基于一组面向对象的广义运动和力向量的定义, 考虑对象动力学,建立了面向对象的双手对称协调运动方程,该运动方程显式地表示了对象的 运动、内力及环境接触力与双手关节力矩间的关系.据此设计出广义工作空间一级的双手对 称协调力/位混合控制算法,并解决了算法的分解与并行实现问题.在两台PUMA562机械手 上进行的实验表明,本文研究的方法,可以在双手协调运动过程中实现对被操作对象的运动、 内力和环境接触力的混合控制.     

基于神经网络的机械臂的力控制方法

本文提出一种基于神经网络的力控制方法，由两个串联的神经网络构成机械臂的力控制器，其中一个网络用来学习机械臂本身的逆动力学系统，而另一网络用来学习未知的被接触环境的动力学特征，这种方法避免了困难的接触环境建模问题。一个 双连杆机械臂的力控制的仿真实验描述了种种方法的有效性。     

Control of Robots with Elastic Joints Interacting with Dynamic Environment

In this paper, the control of robots with elastic joints in contact with dynamic environment is considered. It is shown how control laws synthesized for the robots with rigid joints interacting with dynamic environment can also be used in the case of robots with elastic joints. The proposed control laws are based on a robot model interacting with dynamic environment, including the dynamics of actuators and the elasticity of joints. The proposed control laws possess two feedback loops: the outer, serving for on-line calculation of the motor shaft angle based on the position error or the contact force error, and the inner one, serving for performing stabilization around the calculated motor shaft angle. Simulation results which exhibit the application of the appropriate control laws are also presented.     

How to Control Robots Interacting with Dynamic Environment

The goal of this paper is to shed light on the control problem of constrained robot motion from the aspect of the dynamical nature of the environment with which the robot is in contact. Therefore, the criticism of traditional hybrid control which allows position/force feedback loops to split into independent control with respect to position and force, is not the main point we want to make. Reference to the papers written by the founders of hybrid control and their numerous followers served only to better understand the reason and motivation for suggesting a different approach to control of robots interacting with environment.The paper has a predominantly review character, based on recently published work. It also contains some new, unpublished results in the framework of the unified approach to the position/force control of robots, proposed by the present author and his co-workers. By pointing to the possibility of introducing an environment dynamics in the contact tasks of the machining type, the author emphasizes that the proposed dynamically interactive control can be applied to a completely different class of tasks, in which a contact is made between the system (constructions or structure) and very specific kinds of dynamic environments.     

Control system design of a 3-DOF upper limbs rehabilitation robot

This paper presents the control system design of a rehabilitation and training robot for the upper limbs. Based on a hierarchical structure, this control system allows the execution of sequence of switching control laws (position, force, impedance and force/impedance) corresponding to the required training configuration. A model-based nonlinear controller is used to impose the desired environment to the patient's arm. The knowledge of robot kinematics and dynamics is thus necessary to ensure haptic transparency and patient safety. The identification process of robot dynamics is emphasised and experimental identification results are given for the designed robot. The paper also presents a particular rehabilitation mode named Active-Assisted. Simulation results of this rehabilitation mode illustrate the potentialities of the overall control scheme, which can also be applied to other rehabilitation robots.     

Cooperative Control of a 3D Dual-Flexible-Arm Robot

This paper discusses cooperative control of a dual-flexible-arm robot to handle a rigid object in three-dimensional space. The proposed control scheme integrates hybrid position/force control and vibration suppression control. To derive the control scheme, kinematics and dynamics of the robot when it forms a closed kinematic chain is discussed. Kinematics is described using workspace force, velocity and position vectors, and hybrid position/force control is extended from that on dual-rigid-arm robots. Dynamics is derived from constraint conditions and the lumped-mass-spring model of the flexible robots and an object. The vibration suppression control is calculated from the deflections of the flexible links and the dynamics. Experiments on cooperative control are performed. The absolute positions/orientations and internal forces/moments are controlled using the robot, each arm of which has two flexible links, seven joints and a force/torque sensor. The results illustrate that the robot handled the rigid object damping links' vibration successfully in three-dimensional space.     

Decoupling force and motion control in industrial robots

A theoretical approach to force control design for industrial robots involved in surface-following tasks is described in this paper, assuming an infinitely stiff environment. Independent Joint Control techniques, based on standard (PID) algorithms, are adopted for position control. Force control acts as an outer loop, by adding a bias to the position set points in the joint space. A simple model and compensation of the joint flexibility effects, that play an important role in determining the dynamic behavior of the system, are also presented, leading to a force control decoupled from motion control. Some experimental results are discussed, with reference to the industrial robot SMART.     

A motion control for a spherical robot with pendulum drive

Problems of controlling a spherical robot with pendulum drive are considered. A mathematical model of the dynamics of this robot is constructed and control laws in the form of state feedback that provide robot motion along a given trajectory are synthesized. Results of computer simulation that demonstrate efficiency of the proposed control laws are presented.     

旋翼飞行机器人研究进展

旋翼飞行机器人是面向空中自主作业需求,将旋翼飞行器与多自由度机械臂相结合所提出的新型机器人.该机器人作业过程中旋翼飞行器、机械臂与作业目标之间的动态相对运动以及与作业目标接触过程中未建模外力、力矩扰动使自主控制受到极大挑战.本文将针对旋翼飞行机器人的结构演变及关键技术、作业机构集成技术进行综述.从动力学建模及动力学特性分析、动态运动约束/力约束下的协调规划、非结构环境下的运动和作业控制、面向任务动态操作的环境感知、面向任务的实验系统构建与实验验证五个方面初步构建了旋翼飞行机器人自主作业理论体系.     

对未知环境的机器人力控自律跟踪及建模

在考虑摩擦力的情况下，利用传感器所感知的机器人和未知轮廓间的相互作用力确接触处轮廓的切矢和法矢，据此建立端点约束坐标系，在该坐标系中沿切矢进行位置控制并沿法矢进行力控制。实现机器人对未知轮廓的自律跟踪运动。由跟踪运动所确定每一点处切矢信息及该点位置信息构造未知轮廓几何模型。在机器人学开放研究实验室的ＰＵＭＡ５６２机器人上实现了上述自律运动并建立了环境模型。     

具有参数不确定性的轮式移动机器人自适应backstepping控制

针对参数不确定的轮式移动机器人的轨迹跟踪问题,设计自适应跟踪控制器.基于移动机器人的动力学模型,采用backstepping积分方法,通过逐步递推选择适当的Lyapunov函数,设计基于状态反馈的自适应控制器,并进行了相应的稳定性分析.与传统PID控制进行仿真对比,结果表明提出的自适应控制策略能较好地补偿系统参数摄动的影响,提高了移动机器人的轨迹跟踪性能和鲁棒性.     

Neural Network Force Control for Industrial Robots

In this paper, we present a hierarchical force control framework consisting of a high level control system based on neural network and the existing motion control system of a manipulator in the low level. Inputs of the neural network are the contact force error and estimated stiffness of the contacted environment. The output of the neural network is the position command for the position controller of industrial robots. A MITSUBISHI MELFA RV-M1 industrial robot equipped with a BL Force/Torque sensor is utilized for implementing the hierarchical neural network force control system. Successful experiments for various contact motions are carried out. Additionally, the proposed neural network force controller together with the master/slave control method are used in dual-industrial robot systems. Successful experiments are carried out for the dual-robot system handling an object.     

Estimating the Domain of Admissible Parameters of a Control System of a Vibratory Robot

We consider the rectilinear motion of a vibratory robot on a plane; the robot is presented by a rigid body and a pendulum inside it. The motion is carried out in the gravity field; the force of dry friction acts between the body and the plane. The robot is controlled by choosing the angular acceleration of the pendulum. Two modes of the robot’s control that correspond to various constraints on the choice of the control are investigated. Each of the studied control laws ensures a periodic displacement of the robot; here, the robot moves only in one direction (the motion is irreversible). We discuss the problem of finding the boundaries of the dry friction parameter and the control parameter; we find the boundaries with which the proposed control modes are feasible.     

Motion Control of Caster-Type Passive Mobile Robot with Servo Brakes

In this paper, we introduce a passive mobile robot called prototype Caster-Type Passive Robot Porter (C-PRP), which is developed on the basis of a concept of passive robotics. This mobile robot consists of two casters with servo brakes and one passive rigid wheel. Prototype C-PRP has passive dynamics with respect to the force applied by a human and controls its appropriate motion with the servo brakes. We derive the feasible braking force/moment applied to the robot on the basis of the characteristics of the servo brakes. This paper especially focuses on a fundamental motion control algorithm based on the feasible braking force/moment. We realize the path tracking function and the collision avoidance function as examples by applying the proposed algorithm to prototype C-PRP. These functions are implemented to prototype C-PRP actually, and experimental results confirm its validity.     

液压驱动型四足机器人电液伺服控制系统仿真建模与实验分析

复杂的作业环境和艰巨的作业任务使液压驱动型四足机器人对其伺服系统的精度、速度和力量均比一般机器人在普通情况下有更高的要求。为掌握液压驱动型四足机器人在多种路况下行走时各液压缸的受力情况以及液压系统内流量、压力的变化情况,需要对其虚拟样机进行机械动力系统和液压伺服系统的联合仿真,定性分析电液伺服系统位置、速度等被控对象的特性,并分析PID控制器在四足机器人伺服控制方面的特性与不足。针对传统控制算法在四足机器人控制存在的短板问题,设计了一种非对称前馈补偿模糊自适应PID算法,并利用物理样机进行了实际验证。实验结果为四足机器人电液伺服控制系统硬件、软件和控制算法的设计与优化指明了方向,还为研究四足机器人平稳步态控制策略提供了决策依据和数据支持。