External Force Estimation And Application Of A Hexapedal Robot

As an important division in robotics research,walking robot can archive good performance in negotiating with the terrain due to its discrete foot-ground contact points.Meanwhile,hexapedal robot,with the high load-carrying capacity provided by the parallel kinematics mechanism,has great potential in tasks that require carrying high payload and traversing through uneven terrain.Mostly,a walking robot requires the knowledge of external forces that helps it to detect the contact with the environment,which usually results the installation of foot-tip force sensors.However,a force sensor is usually an expensive and fragile device.Thus,developing an alternative approach to estimate the foot-tip contact force is needed.This research focuses on the high payload hexapedal robot with legs based on the parallel kinematics mechanism,proposed the distributed force estimation method.Focued on this topic,the pose estimation model and inertial force estimation model were established,the base inertial parameters were calibrated,an external force estimation method was established,and an adaptive gait was designed and tested in several experiments.More detailed descriptions are listed as follows:(1)For the parallel kinematics mechanism legs on the hexapedal robot,using screw representation and exponential map,the pose estimation model and the inertial force estimation model were established.The position model,velocity model,acceleration model and static force model of the spatial parallel kinematics mechanism leg were established.In these models.the position,velocity and acceleration of every parts of the leg can thus be evaluated.The relation between the actuation forces and the leg motion was analyzed,as well as the relation between the actuation forces and external forces.These models have been verified on the virtual prototype.(2)Based on the mathematical models established in this paper,as an alternative to the force sensor approach,a distributed force sensing method for the spatial parallel kinematics mechanism legs was proposed,which takes the motor current and position feedback as the input.In this method,the effects of the inertia force and the external force has been separated.The effect of the friction force was discussed,trajectories were selected to generate the calibration input data.Predetermined base inertial parameters were calibrated and then integrated to the mathematical model.Finally,the external force exerted on the foot can be estimated without any force sensor device installed on the foot-tip.Tests revealed that the calibration enhanced the external forece estimation result.Then,the relationship of external force exerted on the body and foot-tip forces were analyzed.The body external force estimation model was established.(3)From bionic point of view,an adaptive gait for the distributed force sensing hexapedal robot was proposed.Based on the similarity of human walking and robot walking,human blind walking were observed,recorded and analyzed.Some principle rules and pattern were extracted from the analysis.Counterpart of these rules and behaviors of the robot were designed.A set of rules were established to combine these behaviors during the robot walking.The proposed adaptive gait improves the hexapedal robot's ability to traverse the uneven terrain made of different textures and shapes.