Design And Walking Control Algorithm Analysis Of A Foot-Pad Locomotion Interface

Virtual reality provides a new effective means to solve the complex problems which are very difficult to be done in the experiment, in military, medical, industrial, and cultural areas. Foot-pad locomotion interface provides an ideal technical scheme, achieving the users walking free in the virtual environment fully immersively and interactively. While there are some problems needed to be solved within foot-pad locomotion interface,such as covering too much space, low speed, unnatural walking control algorithm and so on, that restrict seriously the development of virtual reality hardware technology. This thesis covers an area of study around the above problems.This thesis sets up a 20-DOF 8-rigid-body model of human lower limbs, with simplifying the foot movement to the rotation of rigid body around the unfixed point of ankle, draws an law of ankle kinematics during interactively walking under ideal conditions by operating motion capture data of normal walking. Comparing Law of ankle kinematics during interactively walking with that during normal walking, proposes a more natural walking control, based on tracking and canceling of ankle’s position.This thesis proposes a more compact, more commonality total design scheme of foot-pad locomotion interface and conducts parameter optimization design of lifting mechanism parts, making the mechanism covering minimum space. Design variables includes length, length ratio and angle. Target function considers covering space and the mechanical kinematics properties. And technics of pedal flexural equation to establish multiple constraints.This thesis proposes different control algorithm based on walking phase, adapting to different walking condition, such as walking straight line, turning, climbing up or down and uphill or downhill slopes. During double stance, both foot-pads keep still relative to ankle. During single stance (swing phase), swing pad follow the ankle of swing limbs while standing pad carry body back inversely, that is the key of walking control algorithm. For planar walking straight line, use current displacement, velocity and acceleration of swing ankle to predict the next ankle’s position which is used to control the motion of pads, that is the key of the whole control algorithm of foot-pad locomotion interface. For planar turning, proposes minor angle phased continuous turning control algorithm, which work well both in spin turn and in step turn. We draw a criteria to distinguish turning from walking straight line by coordinate conversion and set different control algorithm for different turning phase. For spatial walking control algorithm, propose the rule to full use the virtual terrain information to pre-judging and to control the pad motion. This algorithm divided the pad motion into horizontal movement, which can be controlled by the walking straight control algorithm, vertical movement and pitching motion. We control pad move steady in vertical movement and pitching motion, with information of last swing phase and virtual terrain.At last, sets ADAMS-MATLAB co-simulation of straight walking control algorithm prediction of ankle’s position based, to analyze the effectiveness of the walking control algorithm under different sampling frequency. The best sampling frequency of the foot-pad locomotion interface is between 30Hz and 50Hz.