基于惯性姿态参量量化融合的机器腿仿生步态控制

Bionic gait control of robot leg based on quantization fusion of inertial attitude parameters

为了提高机器腿仿生步态控制的稳定性,提出一种基于惯性姿态参量量化融合的机器腿仿生步态控制方法。构造机器腿仿生步态的运动学模型,进行机器腿仿生步态控制约束参量建模,采用陀螺仪和加速度计等位姿传感器进行姿态参量采集。采用扩展卡尔曼滤波方法进行机器腿的惯性姿态参量融合并输入到时控制执行器中。针对未知扰动对机器腿步态参量控制的误差,采用自回归更新方法进行姿态参量误差反馈修正,实现机器腿仿生步态稳定控制。仿真结果表明:提出的方法对机器腿仿生控制的稳定性较好,姿态参量的定位跟踪能力较强,提高机器腿步行的稳健性。

In order to improve stability of bionic robotic leg gait control,put forward a kind of bionic robotic leg gait control method based on quantization fusion of parameters of inertial attitude.The kinematics model for bionic robotic leg gait is constructed,bionic leg gait control constraint parameter modeling is carried out,attitude parameters are acquired using position sensors such as gyro and accelerometer.Inertial parameter fusion of robot leg using extended Kalman filtering method and input fusion control actuator.Aiming at unknown error of disturbance to robot leg gait parameter control,using autoregressive updating method for correction of the attitude parameter error feedback,realize the stable control of robot gait.The simulation results show that the method has good stability,and the positioning and tracking ability of the attitude parameters are stronger,and the robustness of robot leg walking is improved.