Gait recognition and trajectory prediction of Lower limb Load Exoskeleton

In this paper system design of the CAEP-SWJTU Lower Load Exoskeleton (CASWELL) is presented. CASWELL has five DOF per leg, two of which are powered by linear hydraulic actuators. Details of design of the electronic system and sensor network are represented, too. The electronic system consists of a group of MCUs, and the sensor network is composed of four main perceptive modules. Diverse sensors including attitude, pressure and length transducers are used to collect motion information. Recognition of the human gait and prediction algorithm of motion pattern are major concerns of this paper. Two methods are utilized in this regard. These are realized using local information of joint motion collected by the sensors. The trajectory prediction solves the time-delay problem caused by the electronic system as well as the mechanical system. Performance of the algorithms is represented by experimental results.