具有2DOF铰接式躯干的仿猎豹四足奔跑机器人

为了探索脊柱运动对腿运动的增强机理,设计了具有2自由度铰接式躯干的仿猎豹四足奔跑机器人。对带腾空相的跳跃（bound）步态奔跑运动的力学过程进行描述,采用阻尼型弹性负载倒立摆（D-SLIP）模型建立了四足机器人动力学模型。依据猎豹的奔跑运动模式,对四足机器人脊柱关节与腿关节的耦合运动进行了轨迹规划。提出一种改进的粒子群优化（PSO）算法,解决了机器人脊柱关节驱动机构尺寸和运动轨迹控制参数之间目标互斥的嵌套优化问题。对四足机器人跳跃奔跑运动进行动力学仿真,结果表明：脊柱与腿的协调运动可以增大奔跑步幅,使机器人产生腾空相,从而提高机器人的奔跑速度。     

Fault-tolerant locomotion of the hexapod robot

In this paper, we propose a scheme for fault detection and tolerance of the hexapod robot locomotion on even terrain. The fault stability margin is defined to represent potential stability which a gait can have in case a sudden fault event occurs to one leg. Based on this, the fault-tolerant quadruped periodic gaits of the hexapod walking over perfectly even terrain are derived. It is demonstrated that the derived quadruped gait is the optimal one the hexapod can have maintaining fault stability margin nonnegative and a geometric condition should be satisfied for the optimal locomotion. By this scheme, when one leg is in failure, the hexapod robot has the modified tripod gait to continue the optimal locomotion.     

Planar legged walking of a passive-spine hexapod robot

This paper proposes a new legged walking method for a novel passive-spine hexapod robot. This robot consists of several body segments connected by passive body joints. Each of the body segments carries two 1-DoF (degree of freedom) actuated legs. The robot is capable of achieving planar legged walking by rapidly abducting and adducting its legs. To model the mobility of a robot based on this simple design, the candidate configurations from all possible configurations are first selected in a mobility analysis of the robot based on the screw theory. All the feasible sequences of these candidate configurations are then searched to form planar locomotion gaits. Next, locomotive performance of the gaits is analyzed. Finally, the proposed locomotion design and gait planning methods are verified through simulations and experiments.     

Self-organized adaptive legged locomotion in a compliant quadruped robot

In this contribution we present experiments of an adaptive locomotion controller on a compliant quadruped robot. The adaptive controller consists of adaptive frequency oscillators in different configurations and produces dynamic gaits such as bounding and jumping. We show two main results: (1) The adaptive controller is able to track the resonant frequency of the robot which is a function of different body parameters (2) controllers based on dynamical systems as we present are able to “recognize” mechanically intrinsic modes of locomotion, adapt to them and enforce them. More specifically the main results are supported by several experiments, showing first that the adaptive controller is constantly tracking body properties and readjusting to them. Second, that important gait parameters are dependent on the geometry and movement of the robot and the controller can account for that. Third, that local control is sufficient and the adaptive controller can adapt to the different mechanical modes. And finally, that key properties of the gaits are not only depending on properties of the body but also the actual mode of movement that the body is operating in. We show that even if we specify the gait pattern on the level of the CPG the chosen gait pattern does not necessarily correspond to the CPG’s pattern. Furthermore, we present the analytical treatment of adaptive frequency oscillators in closed feedback loops, and compare the results to the data from the robot experiments.

Optimal Trajectories for a Quadruped Robot with Trot,Amble and Curvet Gaits for Two Energetic Criteria

In this paper, optimal cyclic reference trajectories aredesigned for three gaits of a quadruped robot, the curvet, theamble, and the trot, taking into account the actuatorscharacteristics. The gaits are composed of stance phases andinstantaneous double supports. The principle of virtual leg isused to obtain simpler dynamic model describing the motion ofthe quadruped. The impact phases are modeled by passive impactequations. For the curvet the step is composed of twodifferent half steps. For the amble and trot gaits twofollowing half steps are symmetrical.The optimization problem is solved with an algebraicoptimization technique. The actuated joint evolution is chosenas a polynomial function of time. The coefficients of thepolynomial functions are optimization parameters. Thequadruped studied has non-actuated ankles. The kineticmomentum theorem permits to define the evolution of this non-actuated variable in function of the actuated variables. Twoenergetic criteria are defined: a torque cost and an energetic cost. The first is represented by the integral of the torquenorm and the second by the absolute value integral of theexternal forces work. The two criteria are calculated for adisplacement of one meter. During the optimization process,the constraints on the ground reactions, on the validity ofimpact, on the torques, on the joints velocities and on themotion velocity of the robot prototype are taken into account.Simulation results are presented for the three gaits. Allmotions are realistic. Curvet is the less efficient gait withrespect to the criteria studied. For slow motion, trot is themore efficient gait. But amble permits the fastest motion withthe same actuators.     

Bionic Quadruped Robot Dynamic Gait Control Strategy Based on Twenty Degrees of Freedom

Quadruped robot dynamic gaits have much more advantages than static gaits on speed and efficiency, however high speed and efficiency calls for more complex mechanical structure and complicated control algorithm. It becomes even more challenging when the robot has more degrees of freedom. As a result, most of the present researches focused on simple robot, while the researches on dynamic gaits for complex robot with more degrees of freedom are relatively limited. The paper is focusing on the dynamic gaits control for complex robot with twenty degrees of freedom for the first time. Firstly, we build a relatively complete 3D model for quadruped robot based on spring loaded inverted pendulum (SLIP) model, analyze the inverse kinematics of the model, plan the trajectory of the swing foot and analyze the hydraulic drive. Secondly, we promote the control algorithm of one-legged to the quadruped robot based on the virtual leg and plan the state variables of pace gait and bound gait. Lastly, we realize the above two kinds of dynamic gaits in ADAMS-MATLAB joint simulation platform which testify the validity of above method.      

Development of wheel-less snake robot with two distinct gaits and gait transition capability

Snake robots are mostly designed based on single mode locomotion. However, single mode gait most likely could not work effectively when the robot is subject to an unstructured working environment with different measures of terrain complexity. As a solution, mixed mode locomotion is proposed in this paper by synchronizing two types of gaits known as serpentine and wriggler gaits used for non-constricted and narrow space environments, respectively, but for straight line locomotion only. A gait transition algorithm is developed to efficiently change the gait from one to another. This study includes the investigation on kinematics analysis followed by dynamics analysis while considering related structural constraints for both gaits. The approach utilizes the speed of the serpentine gait for open area locomotion and exploits the narrow space access capability of the wriggler gait. Hence, it can increase motion flexibility in view of the fact that the robot is able to change its mode of locomotion according to the working environment.     

Gait transition and modulation in a quadruped robot: A brainstem-like modulation approach

In this article, we propose a bio-inspired architecture for a quadruped robot that is able to initiate/stop locomotion; generate different gaits, and to easily select and switch between the different gaits according to the speed and/or the behavioral context. This improves the robot stability and smoothness while locomoting.We apply nonlinear oscillators to model Central Pattern Generators (CPGs). These generate the rhythmic locomotor movements for a quadruped robot. The generated trajectories are modulated by a tonic signal, that encodes the required activity and/or modulation. This drive signal strength is mapped onto sets of CPG parameters. By increasing the drive signal, locomotion can be elicited and velocity increased while switching to the appropriate gaits. This drive signal can be specified according to sensory information or set a priori.The system is implemented in a simulated and real AIBO robot. Results demonstrate the adequacy of the architecture to generate and modulate the required coordinated trajectories according to a velocity increase; and to smoothly and easily switch among the different motor behaviors.     

Numerical and experimental study of the virtual quadrupedal walking robot-semiquad

SemiQuad is a prototyped walking robot with a platform and two double-link legs. Thus, it is a five-link mechanism. The front leg models identical motions of two quadruped's front legs, the back leg models identical motions of two quadruped's back legs. The legs have passive (uncontrolled) feet that extend in the frontal plane. Due to this the robot is stable in the frontal plane. This robot can be viewed as a "virtual" quadruped. Four DC motors drive the mechanism. Its control system comprises a computer, hardware servo-systems and power amplifiers. The locomotion of the prototype is planar curvet gait. In the double support our prototype is statically stable and over actuated. In the single support it is unstable and under actuated system. There is no flight phase. We describe here the scheme of the mechanism, the characteristics of the drives and the control strategy. The dynamic model of the planar walking is recalled for the double, single support phases and for the impact instant. An intuitive control strategy is detailed. The designed control strategy overcomes the difficulties appeared due to unstable and under actuated motion in the single support.Due to the control algorithm the walking regime consists of the alternating different phases. The sequence of these phases is the following. A double support phase begins. A fast bend and unbend of the front leg allows a lift-off of the front leg. During the single support on the back leg the distance between the two leg tips increases. Then an impact occurs and a new double support phase begins. A fast bend and unbend of the back leg allows the lift-off of the back leg. During the single support on the front leg the distance between the two leg tips decreases to form a cyclic walking gait.The experiments give results that are close to those of the simulation.Funding for SemiQuad was primarily provided by the CNRS and the region Pays de La Loire.     

Implementation of omnidirectional crawl for a quadruped robot

As a reptile animal crawls in a cluttered environment, so a quadruped robot should be able to crawl on an irregular ground profile with its static stability by adopting the straightgoing and standstill-turning free gaits. The generalized and explicit formulations for the automatic generation of straight-going gaits and various standstill-turning gaits are presented in this paper. The maximized stride for the straight-going gait and the maximum turning angle for the turning gait of a quadruped robot named TITAN-VIII in a gait cycle are discussed by considering the robot's mechanism constraints and the irregularities of the ground profile. The control algorithm, including control of the joint positions of the robot, is described to implement the desired walking path of the quadruped robot. The effectiveness of the proposed method is demonstrated through experimental result.     

Piecewise linear spine for speed–energy efficiency trade-off in quadruped robots

We compare the effects of linear and piecewise linear compliant spines on locomotion performance of quadruped robots in terms of energy efficiency and locomotion speed through a set of simulations and experiments. We first present a simple locomotion system that behaviorally resembles a bounding quadruped with flexible spine. Then, we show that robots with linear compliant spines have higher locomotion speed and lower cost of transportation in comparison with those with rigid spine. However, in linear case, optimal speed and minimum cost of transportation are attained at very different spine compliance values. Moreover, it is verified that fast and energy efficient locomotion can be achieved together when the spine flexibility is piecewise linear. Furthermore, it is shown that the robot with piecewise linear spine is more robust against changes in the load it carries. Superiority of piecewise linear spines over linear and rigid ones is additionally confirmed by simulating a quadruped robot in Webots and experiments on a crawling two-parts robot with flexible connection.     

Two walking gaits for a planar bipedal robot equipped with a four-bar mechanism for the knee joint

The design of a knee joint is a key issue in robotics and biomechanics to improve the compatibility between prosthesis and human movements, and to improve the bipedal robot performances. We propose a novel design for the knee joint of a planar bipedal robot, based on a four-bar linkage. The dynamic model of the planar bipedal robot is calculated. Two kinds of cyclic walking gaits are considered. The first gait is composed of successive single support phases with stance flat-foot on the ground separated by impacts. The second gait is a succession of finite time double support phases, single support phases, and impacts. During the double support phase, both feet rotate. This phase is ended by an impact of the toe of the forward foot, while the rear foot is taking off. The single support phase is ended by an impact of the swing foot heel, the other foot keeping contact with the ground through its toe. For both gaits, the reference trajectories of the rotational joints are prescribed by cubic spline functions in time. A parametric optimization problem is presented for the determination of the parameters corresponding to the optimal cyclic walking gaits. The main contribution of this paper is the design of a dynamical stable walking gait with double support phases with feet rotation, impacts, and single support phases for this bipedal robot.     

Kinematic Constraints on Fault-Tolerant Gaits for a Locked Joint Failure

Fault-tolerant gaits in legged locomotion are defined as gaits with which legged robots can continue their walking after a failure event has occurred to a leg of the robot. For planning an efficient fault-tolerant gait, kinematic constraints and remaining mobility of the failed leg should be closely examined with each other. This paper addresses the problem of kinematic constraints on fault-tolerant gaits. The considered failure is a locked joint failure which prevents a joint of a leg from moving and makes it locked in a known place. It is shown that for the existence of the conventional fault-tolerant gait for forward walking on even terrain, the configuration of the failed leg must be within a range of kinematic constraints. Then, for coping with failure situations where the existence condition is not satisfied, the conventional fault-tolerant gait is adopted by including the adjustment of the foot trajectory of the failed leg. The foot trajectory adjustment procedure is analytically derived to show that it can help the fault-tolerant gait avoid dead-lock resulting from the kinematic constraint. To demonstrate its effectiveness, the proposed method is applied to the fault-tolerant gait generation for a quadruped robot walking with the wave-crab gait before a locked joint failure.     

一种多模型融合的仿猎豹四足机器人复杂运动控制方法

针对具有2自由度主动脊柱关节的仿猎豹四足机器人,基于任务分解思想和生物神经系统机理,提出多模型融合的控制方法。该方法以弹簧负载倒立摆模型实现单腿跳跃控制,通过中枢模式发生器（CPG）实现4条腿之间以及脊柱―腿之间的协调控制,利用虚拟模型控制实现机器人与环境交互,采用基于CPG输出的有限状态机来融合3个控制模型,构建仿猎豹四足机器人的多模型分层运动控制器。参考猎豹脊柱运动特征,设计了机器人脊柱关节运动模式,给出脊柱与腿的协调控制策略。最后,在Webots仿真环境中搭建了仿猎豹四足机器人虚拟样机,实现了不同步态下的脊柱―腿的协调控制、在崎岖地形上稳定奔跑,以及平滑的对角―疾驰―对角步态转换,仿真结果验证了所提出的多模型融合的四足机器人运动控制方法的有效性。     

Mechanical Design of a Trunk with Redundant and Viscoelastic Joints for Rhythmic Quadruped Locomotion

Passive mechanisms, such as free joints and viscoelastic components, enable natural oscillation of the robot body, which allows rhythmic locomotion with low energy and computational costs. In particular, joint viscoelasticity can be a powerful candidate for changing natural oscillation and so influence the operation performance of locomotion. The present study considers the passive mechanism of a trunk, and investigates the contributions of a trunk mechanism with redundant joints and tunable viscoelasticity to quadruped locomotion. A physical quadruped robot with a trunk mechanism is developed, and the walking performance of this robot for various gait patterns and joint viscoelasticities is investigated. A simulation model is also constructed based on the physical robot, and the contribution of the viscoelasticity to trunk oscillation and the appropriate joint viscoelasticity and number of trunk joints are discussed. Experimental results obtained using the physical robot indicate that the proposed trunk mechanism contributes to successful locomotion as compared to a robot with a rigid trunk and that the velocity is influenced by not only the gait pattern, but also the joint viscoelasticity (i.e., there are appropriate couplings of the joint viscoelasticity and gait pattern). The simulation results indicate that the trunk mechanism requires joint viscoelasticity in order to achieve oscillation and that a greater number of joints having a smaller joint viscoelasticity enables higher velocity. These results suggest that, in addition to the leg mechanism and the controller design, the design of the trunk mechanism is also important.     

Energy-optimal gait analysis of quadruped robots

It is important for walking robots such as quadruped robots to have an efficient gait. Since animals and insects are the basic models for most walking robots, their walking patterns are good examples. In this study, the walking energy consumption of a quadruped robot is analyzed and compared with natural animal gaits. Genetic algorithms have been applied to obtain the energy-optimal gait when the quadruped robot is walking with a set velocity. In this method, an individual in a population represents the walking pattern of the quadruped robot. The gait (individual) which consumes the least energy is considered to be the best gait (individual) in this study. The energy-optimal gait is analyzed at several walking velocities, since the amount of walking energy consumption changes if the walking velocity of the robot is changed. The results of this study can be used to decide what type of gait should be generated for a quadruped robot as its walking velocity changes. This work was presented, in part, at the Sixth International Symposium on Artificial Life and Robotics, Tokyo, Japan, January 15–17, 2001.     

基于改进蚁群算法的四足机器人步态规划

四足机器人关节众多、运动方式复杂,步态规划是四足机器人运动控制的基础。传统的算法多基于仿生原理,缺乏广泛适应性。 在建立运动学方程的基础上,提出了一种基于改进蚁群算法的步态规划算法。该算法利用了四足机器人4条腿运动的线性无关性,将步态规划问题转换为在四维空间里求取最长路径问题。仿真结果表明,该算法得出了满足约束条件的所有步态,最后通过机器人样机检验,验证了该算法求取结果的有效性和合理性。     

Autonomous evolution of dynamic gaits with two quadruped robots

A challenging task that must be accomplished for every legged robot is creating the walking and running behaviors needed for it to move. In this paper we describe our system for autonomously evolving dynamic gaits on two of Sony's quadruped robots. Our evolutionary algorithm runs on board the robot and uses the robot's sensors to compute the quality of a gait without assistance from the experimenter. First, we show the evolution of a pace and trot gait on the OPEN-R prototype robot. With the fastest gait, the robot moves at over 10 m/min, which is more than forty body-lengths/min. While these first gaits are somewhat sensitive to the robot and environment in which they are evolved, we then show the evolution of robust dynamic gaits, one of which is used on the ERS-110, the first consumer version of AIBO.     

The dynamic optimization approach to locomotion dynamics: human-like gaits from a minimally-constrained biped model

This work presents a unifying framework to study energy-efficient optimal gaits for a bipedal model without elastic elements. The model includes a torso, flat feet, and telescoping legs, equipped with rotational hip and ankle joints. Two general types of gaits are studied: with and without a flight phase. The support surface can be level ground, sloped, or staircase. The algorithm achieves the optimum within the admissible space by using a minimal set of realistic physical constraints, and avoiding a priori assumptions on kinetic and kinematic parameters such as extended or instantaneous double-support, collisional or collisionless foot-ground contact, step length, step period, etc. The gait optimization for this simple model predicts many features of human locomotion including the optimality of pendular walking and impulsive running at slow and fast progression speeds, ankle push-off prior to touch-down, swing leg retraction, landing on a near vertical leg in gaits with flight phase, and burst hip torques at both ends of the swing phase.     

The effect of spine on the bounding dynamic performance of legged system

The effect of the spinal joint on the dynamic performance of a legged system in bounding gait is presented in this paper. We employ the simplified quadruped passive model to discuss the underlying mechanism of the bounding gait pattern. A comparison between the dynamic model with and without the spinal joint suggests that the model with a spinal joint is superior to the model with a rigid body in dynamic indicators including the vertical fluctuation of the centre of mass (CoM), leg forces and energy consumption. The effects of different initial included angles of the spinal joint and different initial vertical positions of the CoM on the dynamics of bounding are also analysed. We found that a low level of the energy consumption and leg forces can be obtained with the proper initial included angle and initial vertical position of the CoM, and there exists an optimal situation for each horizontal running velocity that the lowest energy consumption and the lowest leg forces could be obtained at the same time.