ABSTRACT For the ‐link upper limb flexible exoskeleton robots, an adaptive constraint control strategy is introduced in this work, which does away with reliance on the feasibility requirements of virtual controllers and input saturation. Firstly, a nonlinear transformation that relies solely on the constrained variables is constructed to directly handle the full‐state constraints with asymmetric property during the operation of exoskeleton robots, ensuring that all states are within a reasonable range. Secondly, considering the complexity of the exoskeleton robot system and the requirement for high‐precision control, a new coordinate transformation is put forward and incorporated into the framework of backstepping algorithm which is based on dynamic surface control, effectively avoiding the stringent feasibility condition limitations of the virtual controller and optimizing the control process. The above method eliminates the cumbersome offline feasibility verification calculations, granting designers greater freedom in parameter selection. Ultimately, the validity of the suggested method is confirmed via simulation study.
Chen et al. (Thu,) studied this question.