Joint design and torque feedback experiment of rehabilitation robot

Abstract

The performance of the real-time dynamic force and torque compensation, flexible force interactive control, and the ability to compensate for the defect of the passive rehabilitation training are the important functions within the rehabilitation robot design process. In this investigation, the upper limb rehabilitation robot is designed, and the force sensor is used to measure the joint feedback torque with high precision, high sensitivity, and low cost. In the rehabilitation robot design process, the human–machine adaptability and lightweight flexible driving design are considered, and the static and dynamic moment detection performances of the driving joint are analyzed. Furthermore, the impedance control algorithm is used to control the force output of the single drive joint, and then the sinusoidal force output performance and step force output performance are tested under different amplitudes and frequencies. Finally, the passive rehabilitation mode of the prototype is tested to evaluate the performance of the rehabilitation robot. The results show that the force output accuracy and stability of the driving joint has a good performance, which can satisfy the force-assisted application of exoskeleton.

The modified extended Kalman filter based recursive estimation for Wiener nonlinear systems with process noise and measurement noise

“…Equations (19)- (26) and 51- (56) form the MEKF-RLS algorithm for estimating the parameters of the Wiener system in (1). The basic idea in this article can be combined the mathematical tools 41,42 such as the data filtering technique, 43,44 the particle filter, 45,46 and the iterative methods [47][48][49] to study the identification of other linear stochastic systems, 50 bilinear stochastic systems, 51,52 and nonlinear stochastic systems [53][54][55][56][57] and can be applied to other engineering areas. [58][59][60][61] Theorem 3.…”

“…[58][59][60][61] Theorem 3. For the Wiener nonlinear system in (1 )- (2 ) and the MEKF-RLS algorithm in (19 )- (26 ) and (51 )- (56 ), suppose that ( A1), ( A2), and ( A3) hold and there exist an integer N and positive constants 1 and 2 irrelevant to t such that for t ⩾ N, the following persistent excitation condition holds,…”

“…Collect the measurement data u t and y t . Form̂t,̂, and u using (54), (55), and (56). Compute − t using (19).…”

Three‐stage least squares‐based iterative estimation algorithms for bilinear state‐space systems based on the bilinear state estimator

“…The steps of the state estimator-based 3S-LSI algorithm in (36) to (49) for bilinear systems are as follows.…”

“…5. Readâ i,k fromâ k in (42) and readm i,k fromm k in (43), and construct k andM k using (48) and (49). 6.…”