Design Of Biped Robot Control System And Real-time Gait Detection

Compared with traditional robots,bipedal robots can adapt to the harsh unstructured environment,which has been a hotspot in the field of robot research.As the "brain" of biped robot,the control system is an important part of biped robot.Aiming at the real-time and stable requirements of data acquisition during the walking of biped robot,this paper designed a distributed control system of biped robot and developed a gait detection algorithm to detect the gait state.The main research works are as follows:First of all,a distributed control system is designed,and a three-layer architecture is adopted,namely,the underlying driving layer,the data processing layer and the task planning layer.Ether CAT bus is used for data acquisition of the drive unit.In the data processing layer,the serial port based on RS232 is used to obtain the inertial data.In order to achieve remote wireless control,the task planning layer and the data processing layer transmit data in the form of wireless WIFI through UDP protocol,and finally movement data visualization has been build in the upper computer,which can monitor the movement data of biped robot.Secondly,a geodetic coordinate calculation method based on quaternion method and Euler Angle method is proposed.In order to obtain the inertial data in the geodetic coordinate system,euler Angle method is used to transform the coordinate system.Quaternion method is used to represent attitude transformation.By solving the quaternion differential equation and updating the attitude Angle,the accurate 3D Angle information of the foot of biped robot is obtained.Then,a gait parameter calculation method based on multi-level data fusion is proposed.The main parameters of gait are step number,step frequency,step speed,step length,etc.In order to improve the accuracy of step detection,a regional peak detection method is proposed to calculate the number of steps.The gait cycle is divided into support phase and swing phase,and the decision layer fusion method with multiple thresholds is used to judge the support phase.Zero speed updating algorithm is used to calibrate the step speed and step length errors,and more accurate gait parameters are obtained.In order to observe the walking trails more intuitively,the walking trails is reconstructed on MATLAB.Finally,an experimental platform for bipedal robot control system is built to verify the above work.The experiments of motion data visualization,system delay detection,signal distortion detection and gait data acquisition are carried out.The experimental results show that the biped robot control system designed in this paper has the characteristics of low delay and good reliability,and the developed gait detection algorithm can obtain accurate gait parameters.