The nonlinearity and uncertainty of dynamics pose significant challenges to ensuring the tracking performance of joint trajectories, especially time-varying effects on the load and temperature. In this paper, we present an adaptive neural computed torque control scheme to improve the tracking accuracy of the robot joint towards various tasks, which is a novel semiparametric model including a parametric friction model and a nonparametric compensator trained with multiple radial basis function neural networks $(\rm{MRBFNNs})$ . Specifically, the asymmetric model considers velocity-, load-, and temperature-dependent friction phenomena. The computed torque controller integrates the sliding mode method and the proposed friction model to reduce the boundary layer of fluctuated disturbances and achieve globally asymptotic convergence. MRBFNNs are trained separately to further compensate for the unmodeled nonlinearity and parameter uncertainty in real time during the trajectory tracking process. The comparative experiments were carried out on a robot joint, validating that our asymmetric model significantly improves correspondence to reality in terms of friction; the proposed control strategy exhibits the superior tracking performance of joints with variable payloads.

The accurate estimation of the end-effector load force is essential in dynamic robotic scenarios, especially when the end-effector payload varies, to ensure safe and stable physical interaction among humans, robots, and environments. Currently, most applications still rely on payload calibration schemes, but existing calibration techniques often struggle to balance efficiency and accuracy. Moreover, current-based payload estimation methods, which are a commonly used and low-cost technique, face practical challenges such as non-negligible noise. To handle these issues, we propose a sensorless scheme based on a modified disturbance Kalman filter for accurately estimating the load force exerted on robots. Specifically, we introduce the dynamic model of robots that incorporates the nonlinear friction related to velocity and load. Subsequently, a generalized disturbance observer for the robot dynamics is adopted to avoid the measurement noise of joint acceleration. Considering the influence of friction and velocity on the noise parameters in the Kalman filter, a variable-parameter noise model is established. Finally, experimental results demonstrate that the proposed method achieves better performance in terms of accuracy, response, and overshoot suppression compared to the existing methods.