Research on Recognition and Estimation Algorithm for Real time Changes in the Center of Gravity of Human Sitting Quality

Abstract

This article explores the interference estimation problem of human body mass changes and shaking in the sitting posture of medical transport nursing robots in medical scenarios. Due to the significant time-varying and uncertain load characteristics caused by patients' vital signs and behaviors, including strong randomness and non-linearity, high requirements are placed on the speed tracking accuracy and anti-interference ability of the robots. When developing high-performance decision-making and control algorithms for transport robots, the estimation of the mass of the seated human body and the shaking angle are basic parameters that must be considered. Due to the fact that the shaking of the seated human body can cause changes in the mass of the seated human body, there is a tight coupling problem, and the existing recognition algorithms for the shaking angle of the seated human body have insufficient adaptability to working conditions, which affects the accuracy and convergence speed of the estimation algorithm. However, existing research on incomplete wheeled robots mainly focuses on the analysis methods represented by Newton vector mechanics systems and Lagrange equations. However, there are still shortcomings in analyzing the uncertainty of the robot's internal dynamics model and external disturbances, especially in dealing with parameter mismatch and nonlinear oscillation problems. To address this issue, this paper proposes a dynamic modeling method based on the Lagrangian equation of the dissipation function. By incorporating threshold switching to improve the adaptive recursive least squares method and combining it with the adaptive unscented Kalman filtering algorithm, real-time adaptive estimation of human sitting posture mass and difficult to measure human body sway pitch angle is performed using drive wheel torque data. Through comparison between the pre - and post improvement algorithms, the improved estimation algorithm improves speed and accuracy.