Research Of Humanoid Robot’s Walking Control Based On NAO

Robots have made a great development since it entered the human society in the 20th century, and the robot’s technology represents the most advanced automatic control technology. The humanoid robots have received high attention from all walks of life with the humanoid appearance. They can complete various tasks instead of people. Now the study about the humanoid robots is only limited in the ideal environment, while robots cannot deal with harsh environment intelligently. Because of humanoid robots’ nonlinear, high dimension, strong coupling, biped walking has been an important and difficult technology in the field of robotics. When dealing with the surrounding environment, the robot will be distracted inevitably. How to cope with these disturbances will be very important.This thesis is aimed to study biped walking of humanoid robot’s control, and some relevant technology has been applied in the robot. First, in order to get the NAO robot posture information more accurately, some work has been done to overcome the gyroscope and accelerometer’s own limitations. With the method of Kalman filter, the information of gyroscope and accelerometer will be fused together to obtain the torso angle rotating around axis.Then, this thesis puts forward a gait planning method based on model of biped robot, namely, the linear inverted pendulum mode which is applied to the humanoid robot NAO platforms. Based on the introduction of the principle of linear inverted pendulum model, the data for humanoid robot NAO gait planning was finished. And the planning gait was tested on experiment platform of NAO robot.In this thesis, the walking planning method break the traditional constant ZMP model using cubic function smooth trajectory in the feet support moment. At the same time, the ZMP trajectory and the corresponding COG trajectory is planned out synchronously in each control cycle. Under such condition, the problem of ZMP trajectory’s connection is avoided. The rate of movement of the center of mass can be a smooth transition. In addition, the traditional inverse kinematics model was changed, and the foot’s lift height is considered in the process of planning. We have obtained the good experimental results.In this paper, a series of feedback strategies in the work of NAO’s standing and walking is described. We put forward three kinds of feedback strategies, respectively for the ankle, hip strategy, strategy of upper limb and lower limb joints. The combination of the feedback strategy can make robot resist the interference of the external environment when standing or walking. At last, the robot can complete the task smoothly. In addition, in view of the complicated structure and various states about humanoid robot NAO, we designed a real-time monitoring system in order to use robot safely and efficiently.