Research On Control Method Of Walking Assistant Robot Base On Sensing Information Of Flexible Biped

Wearable assistant robot is a intelligent interactive system of man-machine combination, which can determine the intent of human motion with human motion interactive information acquisition by the perception system of robot. Then the control system sends commands to power-driven by trajectory planning. Wearable assistant robot can enhance the capacity of human motion, extend the scope of human movement and has great potentials in health care, expedition travel, military and other areas.This dissertation is to study walking motion control of the wearable assistant robot. As the human gait shows significant seasonal changes during walking, and human plantar pressure information can be well mapped the features of cycle, so this dissertation selected plantar pressure distribution information as wearable assistant robot control signal input.The main content of this dissertation is as follows:1. According to characteristics of human lower limb movement, this dissertation analyzes the human walking rules and gait phase, provides theoretical support for the control system of wearable assistant robot.2. This dissertation designs a robot flexible biped structure. The structure fits the natural structure of human biped and it can reduce human-machine interference, improve the measurement of human-machine interaction information.3. This dissertation designs a plantar pressure distribution information retrieval system. The measurement accuracy, timeliness and reliability of the system can meet the test requirements by experiments. Then this dissertation quantitative analyzes the experimental results with multi-sensor information fusion theory, divides human gait phase cycle and provides data support for control system of wearable assistant robot.4. This dissertation designs the control system for lower limb assistant robot from motion planning level, gait synthesis layer and execution layer. These three levels are from the control concepts of traditional autonomous biped robot. For the nonlinear relationship between control system input-plantar pressure distribution information and control system output- drive information, this dissertation associates the nonlinear relationship with fuzzy controller in gait synthesis level. Through the study based on a theoretical level, we develop an assistant robot experimental platform.5. This dissertation verifies the security and reliability of platform by experiments. Finally, This dissertation using control system to complete two-phase walking reproduction by plantar pressure distribution information.