Full-Body Compliant Human–Humanoid Interaction: Balancing in the Presence of Unknown External Forces

This paper proposes an effective framework of human-humanoid robot physical interaction. Its key component is a new control technique for full-body balancing in the presence of external forces, which is presented and then validated empirically. We have adopted an integrated system approach to develop humanoid robots. Herein, we describe the importance of replicating human-like capabilities and responses during human-robot interaction in this context. Our balancing controller provides gravity compensation, making the robot passive and thereby facilitating safe physical interactions. The method operates by setting an appropriate ground reaction force and transforming these forces into full-body joint torques. It handles an arbitrary number of force interaction points on the robot. It does not require force measurement at interested contact points. It requires neither inverse kinematics nor inverse dynamics. It can adapt to uneven ground surfaces. It operates as a force control process, and can therefore, accommodate simultaneous control processes using force-, velocity-, or position-based control. Forces are distributed over supporting contact points in an optimal manner. Joint redundancy is resolved by damping injection in the context of passivity. We present various force interaction experiments using our full-sized bipedal humanoid platform, including compliant balance, even when affected by unknown external forces, which demonstrates the effectiveness of the method.     

Tracking People on a Torus

We present a framework for monocular 3D kinematic pose tracking and viewpoint estimation of periodic and quasi-periodic human motions from an uncalibrated camera. The approach we introduce here is based on learning both the visual observation manifold and the kinematic manifold of the motion using a joint representation. We show that the visual manifold of the observed shape of a human performing a periodic motion, observed from different viewpoints, is topologically equivalent to a torus manifold. The approach we introduce here is based on the supervised learning of both the visual and kinematic manifolds. Instead of learning an embedding of the manifold, we learn the geometric deformation between an ideal manifold (conceptual equivalent topological structure) and a twisted version of the manifold (the data). Experimental results show accurate estimation of the 3D body posture and the viewpoint from a single uncalibrated camera.     

Optimal traction control for heavy-duty vehicles

Heavy-duty vehicles such as tractors, bulldozers, certain construction and municipal vehicles, soil millers, forestry machinery etc. have a high demand for propulsion force and consequently a high fuel consumption. The current work presents a traction control approach based on motion dynamics estimation for optimizing propulsion force and energy efficiency according to a user-defined strategy. Unscented Kalman filter augmented with a fuzzy-logic system for adaptive estimation is used as the state observer. Simulation case study with an electrically driven tractor is presented. The new method of traction control showed considerable improvement of balancing energy efficiency and propulsion force.     

Acquisition of joint attention through natural interaction utilizing motion cues

Joint attention is one of the most important cognitive functions for the emergence of communication not only between humans, but also between humans and robots. In previous work, we have demonstrated how a robot can acquire primary joint attention behavior (gaze following) without external evaluation. However, this method needs the human to tell the robot when to shift its gaze. This paper presents a method that does not need such a constraint by introducing an attention selector based on a measure consisting of saliencies of object features and motion cues. In order to realize natural interaction, a self-organizing map for real-time face pattern separation and contingency learning for gaze following without external evaluation are utilized. The attention selector controls the robot gaze to switch often from the human face to an object and vice versa, and pairs of a face pattern and a gaze motor command are input to the contingency learning. The motion cues are expected to reduce the number of incorrect training data pairs due to the asynchronous interaction that affects the convergence of the contingency learning. The experimental result shows that gaze shift utilizing motion cues enables a robot to synchronize its own motion with human motion and to learn joint attention efficiently in about 20 min.     

Development of minimalist bipedal walking robot with flexible ankle and split-mass balancing systems

This paper presents a novel design of minimalist bipedal walking robot with flexible ankle and split-mass balancing systems.The proposed approach implements a novel strategy to achieve stable bipedal walk by decoupling the walking motion control from the sideway balancing control.This strategy allows the walking controller to execute the walking task independently while the sideway balancing controller continuously maintains the balance of the robot.The hip-mass carry approach and selected stages of walk implemented in the control strategy can minimize the efect of major hip mass of the robot on the stability of its walk.In addition,the developed smooth joint trajectory planning eliminates the impacts of feet during the landing.In this paper,the new design of mechanism for locomotion systems and balancing systems are introduced.An additional degree of freedom introduced at the ankle joint increases the sensitivity of the system and response time to the sideway disturbances.The efectiveness of the proposed strategy is experimentally tested on a bipedal robot prototype.The experimental results provide evidence that the proposed strategy is feasible and advantageous.     

Walking algorithm for a robotic transfemoral prosthesis capable of walking pattern recognition and posture stabilization

This paper describes a walking pattern generation algorithm for a robotic transfemoral prosthesis that is synchronized with the walking motion of a transfemoral amputee, and posture stabilization for ground adaptation and maintaining balance on inclined grounds. The developed robotic transfemoral prosthesis in this study has a knee joint and ankle roll/pitch joints for walking on complex slopes. The walking motion data obtained from the motion capture system are used as the standard walking pattern data to accurately imitate the inherent gait of the wearer. Walking intention, percent of gait cycle (PGC), and walking stride are predicted through two inertial sensors attached at both thighs, and the joint angles of the robotic transfemoral prosthesis are then generated in real-time from the PGC and the standard walking pattern data. Additionally, variable impedance control and zero moment point (ZMP) control are carried out with a force/torque sensor for posture stabilization against variable ground slopes, and ground slope compensation and disturbance rejection are also done with the use of inertial sensors at the foot and shank. Consequently, the performance of the walking pattern generation algorithm and posture stabilization control was verified through walking experiments of an author on an inclined treadmill.     

Periodic motion synthesis and Fourier compression

Periodic motion is an important class of motion to synthesize, but it is not easy to compute it robustly and efficiently. In this paper we propose a simple, robust and efficient method to compute periodic motion from linear equation systems. The method first calculates the response of the system when an external periodic force is applied during one period, and then sums up the periodically shifted versions of the system response to provide the periodic solution. It is also shown that Fourier decomposition is very effective to compress the motion data without a drop in visual fidelity. © 1998 John Wiley & Sons, Ltd.     

Intelligent Adaptive Motion Control Using Fuzzy Basis Function Networks for Electric Unicycle

This paper presents two intelligent adaptive controllers, called self‐balancing and speed controllers, for self‐balancing and motion control, respectively, of an electric unicycle using fuzzy basis function networks (FBFN), which are employed to approximate model uncertainties and unknown friction between the wheel and the terrain surface. Both controllers are established based on the linearized model of the vehicle whose model uncertainties and parameter variations are caused by different riders and terrain. An adaptive backstepping controller together with online learning FBFN and sensing information of the rider's body inclination then is presented to achieve self‐balancing motion control. By adding an electronic throttle as the input device of speed commands, a decoupling sliding‐mode controller with online learning FBFN is proposed to accomplish self‐balancing and speed control. The performance and merit of the two proposed control methods are exemplified by conducting four simulations and three experiments on a laboratory‐built electric unicycle.     

Three different lifting strategies for controlling the motion patterns of the external load

Co-ordination of various components of the human body during the course of lifting are very complex and difficult to control. This study hypothesized that strategies used to control the motion patterns of the external load may be applied to control co-ordination and also to control the level of compressive force on the lumbosacral joint. A simulation of lifting based on the optimization approach was introduced to generate three classes of unique dynamic motion patterns of the external load directed by three different objective functions. The first objective function was to maximize the smoothness of the motion pattern of the external load. The second objective function was to minimize the sudden change of the centre of gravity of the body-load system. The third objective was to minimize the integration over time of the sum of the square of the ratio of the predicted joint moments to the corresponding joint strength during the course of lifting. Eight subjects were recruited to perform 40 lifts using each of the three optimal motion patterns of the load. Compressive forces on the lumbosacral joint were computed and compared. The data showed with statistical significance that subjects using the motion patterns of the external load suggested by the first objective function had the lowest compressive force peaks. Thus, this study satisfied two goals: (1) it indexed and synthesized three motion patterns of the external load by three biomechanically unique objective functions, and (2) it established the association between the spinal loading and the control of the motion patterns of the external load during lifting.     

Data‐guided Model Predictive Control Based on Smoothed Contact Dynamics

In this paper, we propose an efficient data‐guided method based on Model Predictive Control (MPC) to synthesize a full‐body motion. Guided by a reference motion, our method repeatedly plans the full‐body motion to produce an optimal control policy for predictive control while sliding the fixed‐span window along the time axis. Based on this policy, the method computes the joint torques of a character at every time step. Together with contact forces and external perturbations if there are any, the joint torques are used to update the state of the character. Without including the contact forces in the control vector, our formulation of the trajectory optimization problem enables automatic adjustment of contact timings and positions for balancing in response to environmental changes and external perturbations. For efficiency, we adopt derivative‐based trajectory optimization on top of state‐of‐the‐art smoothed contact dynamics. Use of derivatives enables our method to run much faster than the existing sampling‐based methods. In order to further accelerate the performance of MPC, we propose efficient numerical differentiation of the system dynamics of a full‐body character based on two schemes: data reuse and data interpolation. The former scheme exploits data dependency to reuse physical quantities of the system dynamics at near‐by time points. The latter scheme allows the use of derivatives at sparse sample points to interpolate those at other time points in the window. We further accelerate evaluation of the system dynamics by exploiting the sparsity of physical quantities such as Jacobian matrix resulting from the tree‐like structure of the articulated body. Through experiments, we show that the proposed method efficiently can synthesize realistic motions such as locomotion, dancing, gymnastic motions, and martial arts at interactive rates using moderate computing resources.     

Motion/force control scheme for electrically driven cooperative multiple mobile manipulators

This paper presents the motion and force control problem of rigid-link electrically driven cooperative mobile manipulators handling a rigid object. Although, the motion/force control problem of cooperative mobile manipulators has been enthusiastically studied. But there is little research on the motion/force control of electrically driven cooperative mobile manipulators. Due to the inclusion of the actuator dynamics with the manipulator’s dynamics, the controller exhibits some important characteristics. For the electromechanical system, we have designed a novel controller at the dynamic level as well as at the actuator level. In the proposed control scheme, at the dynamic level, uncertain non-linear mechanical dynamics is approximated with a hybrid controller containing model-based control scheme combined with model-free neural network based control scheme together with an adaptive bound. The adaptive bound is used to suppress the effects of external disturbances, friction terms, and reconstruction error of the neural network. At the actuator level, for the approximation of the unknown electrical dynamics, the model-free neural network is utilized. The developed control scheme provides that the position tracking errors, as well as the internal force, converge to the desired levels. Additionally, direct current motors are also controlled in such a way that the desired currents and torques can be attained. In order to make the overall system to be asymptotically stable, online learning of the weights and the parameter adaptation of the parameters is utilized in the Lyapunov function. The superiority of the developed control method is carried out with the numerical simulation results and its superior robustness is shown in a comparative manner.     

Adaptive motion/force control of uncertain nonholonomic mobile manipulator with estimation of unknown external force

This paper describes a stable adaptive motion/force control of uncertain nonholonomic mobile manipulator with the consideration of external force. As it is well known, unexpected external force makes the motion of the system unstable since there are no fixed points in the stationary coordinate. Here, a novel adaptive control scheme is utilized to estimate and compensate the unknown external force exerted to the end-effector even if the parameters of the system are uncertain. The important advantages of this approach are to achieve estimation without the requirement of force-sensing feedback and the knowledge of the system dynamic model. The update laws for the force and the parameters are derived from a Lyapunov function to guarantee the control system stability. Furthermore, a unified operational space dynamic formulation is presented to solve the problem of redundancy. As a result, the desired end-effector and platform trajectories are simultaneously tracked with a perfect coordination between the two subsystems. Therefore, the proposed controller proves that it can not only guarantee the stability, but also the tracking performance of the system in the task space. The effectiveness of the proposed algorithm is evaluated through extensive simulations and they demonstrate the stability, tracking trajectories and feasibility in estimating the external force and the dynamic uncertainties.     

A Learning Behavior Based Controller for Maintaining Balance in Robotic Locomotion

The Behavior Based Locomotion Controller (BBLC) extends the applicability of the behavior based control (BBC) architecture to redundant systems with multiple task-space motions. A set of control behaviors are attributed to each task-space motion individually and a reinforcement learning algorithm is used to select the combination of behaviors which can achieve the control objective. The resulting behavior combination is an emergent control behavior robust to unknown environments due to the added learning capability. Hence, the BBLC is applicable to complex redundant systems operating in unknown environments, where the emergent control behaviors can satisfy higher level control objectives such as balance in locomotion. The balance control problem of two robotic systems, a bipedal robot walker and a mobile manipulator, are used to study the performance of this controller. Results show that the BBLC strategy can generate emergent balancing strategies capable of adapting to new unknown disturbances from the environment, using only a small fixed library of balancing behaviors.     

VSC coordinated control of two manipulator arms in the presence ofenvironmental constraints

A variable structure control (VSC) method is developed for motion, internal force, and constrained force control of two manipulators grasping a common constrained object. Based on a transformed dynamic equation of the entire system in the joint space, a motion and force control are designed together via a VSC method with robustness to parametric uncertainties and external disturbances. The proposed controller guarantees the system with prescribed qualities in the sliding mode and during the reaching transient. Simulation results illustrate the method     

应用规划生成系统的柔顺性控制

通常的柔顺性控制采取连续的力伺服控制。我们提出用开关力监控来进行柔顺性控制、它只反映机械手各关节受力是否达到允许阈值,因而检测手段大大简化;这样的柔顺性控制过程成了离散的动作序列,只要通过分析与示教,提出状态分析与运动规划规则,并运用规划生成系统和归纳学习方法来自动编制动作序列并转为程序,就能在同一类型操作中适应各种变化。这种机械手不是简单地再现人的示救动作,它能总结示教中的规律,以适应多变的操作过程因而其示教具有智能。     

基于ai深度学习的机器人碰撞预估计控制器设计

机器人运动过程中与外部障碍物之间容易发生碰撞,当碰撞作用力过大时会造成机器零件损坏的问题,为解决这一问题,设计基于ai深度学习的机器人碰撞预估计控制器。建立人机交互电路与串口通信电路,将伺服电机设备、运动控制器、PC感应装置分别接入既定作用区域内,完成预估计控制器的整体应用结构设计。以PyTorch深度学习框架为基础,定义激活函数,再根据预估计参数的实际取值范围,实现对目标机器人对象的精准检测。按照力矩控制条件表达式,确定碰撞行为的表现强度,完成对机器人运动路径的规划,联合相关应用设备,实现基于ai深度学习的机器人碰撞预估计控制器设计。实验结果表明,ai深度学习算法作用下,机器人与障碍物碰撞部位的接触面积不会超过0.25m2,由碰撞行为导致的外部作用力相对较小,不会造成严重的机器零件损坏问题。     

Two-stage adaptive robot position/force control using fuzzy reasoning and neural networks

Many studies have been performed on the position/force control of robot manipulators. Since the desired position and force required to realize certain tasks are usually designated in the operational space, the controller should adapt itself to an environment and generate the control force vector in the operational space. On the other hand, the friction of each joint of a robot manipulator is a serious problem since it impedes control accuracy. Therefore, the friction should be effectively compensated for in order to realize precise control of robot manipulators. Recently, soft computing techniques (fuzzy reasoning, neural networks and genetic algorithms) have been playing an important role in the control of robots. Applying the fuzzy-neuro approach (a combination of fuzzy reasoning and neural networks), learning/adaptation ability and human knowledge can be incorporated into a robot controller. In this paper, we propose a two-stage adaptive robot manipulator position/force control method in which the uncertain/unknown dynamic of the environment is compensated for in the task space and the joint friction is effectively compensated for in the joint space using soft computing techniques. The effectiveness of the proposed control method was evaluated by experiments.     

COM motion estimation of a biped robot based on kinodynamics and torque equilibrium

A novel technique to estimate motion of the center of mass (COM) for a biped robot is proposed. A Kalman filter is synthesized where the time evolution of COM is predicted from the external force and corrected based on kinematic estimation and torque equilibrium. They complementarily work to compensate the initial estimation offset, the error accumulation, and errors in modeled mass properties. It makes use of the authors’ previous method to estimate the translational and rotational motion of the base body from inertial information and joint angle measurements. The information about torque equilibrium helps to reduce an uncertainty of the height of COM and to improve the estimation accuracy of it by utilizing an interference of the horizontal and vertical motion of COM. The parameters are tuned based on error analyses in mass properties and sensor signals. A comparative study showed a better performance of the proposed method than other methods through dynamics simulations.     

Adaptive fuzzy-neural-network control for a DSP-based permanent magnet linear synchronous motor servo drive

An adaptive fuzzy neural network (AFNN) control system is proposed to control the position of the mover of a field-oriented control permanent magnet linear synchronous motor (PMLSM) servo-drive system to track periodic reference trajectories in this paper. In the proposed AFNN control system, an FNN with accurate approximation capability is employed to approximate the unknown dynamics of the PMLSM, and a robust compensator is proposed to confront the inevitable approximation errors due to finite number of membership functions and disturbances including the friction force. The adaptive learning algorithm that can learn the parameters of the FNN on line is derived using Lyapunov stability theorem. Moreover, to relax the requirement for the value of lumped uncertainty in the robust compensator, which comprises a minimum approximation error, optimal parameter vectors, higher order terms in Taylor series and friction force, an adaptive lumped uncertainty estimation law is investigated. Furthermore, all the control algorithms are implemented in a TMS320C32 DSP-based control computer. The simulated and experimental results due to periodic reference trajectories show that the dynamic behaviors of the proposed control systems are robust with regard to uncertainties.     

采用扰动估计的气动人工肌肉系统非线性控制

气动人工肌肉系统凭借其材质轻便、输出力大及柔顺性好等优势, 其运动控制研究近年来逐渐成为热点问题. 然而, 气动人工肌肉(pneumatic artificial muscle, PAM)系统所固有的特性(如迟滞、蠕变、非线性时变等), 为其控制方法设计与实现带来了挑战. 考虑到实际工作过程中, 系统往往遭受未知干扰的影响, 本文针对气动人工肌肉系统, 提出了一种基于干扰估计的非线性控制策略, 可在系统存在持续不确定干扰的情况下, 在线进行扰动抑制, 实现精确的跟踪控制. 具体而言, 本文先通过模型变换, 将系统不确定性、未建模动态、外部扰动等处理成集总扰动的形式. 随后, 结合自适应更新律及正则化最小二乘算法, 在线估计未知系统参数及扰动; 在精确扰动代数估计的基础上, 通过所提基于干扰估计的非线性控制器, 消除未知扰动对系统造成的影响, 并确保跟踪误差收敛至零. 此外, 经稳定性分析证明了跟踪误差的渐近收敛性. 最后, 通过硬件实验验证了本文方法的有效性及鲁棒性.