Motion tracking of a piezo-driven cell puncture mechanism using enhanced sliding mode control with neural network

Abstract

This study deals with the design of a cell puncture mechanism (CPM) driven by a piezoelectric actuator to complete the cell puncture process. Then, a novel robust controller is proposed to realize precise CPM motion control. For the convenience of engineering application, the complex nonlinear hysteresis effect and unmolded term in CPM are summed up as an unknown term. Then, a simplified dynamic model is developed. Moreover, a robust controller is constructed by synthesizing the sliding mode control (SMC) with a proportional–integral–differential (PID)-type manifold and radial basis function neural network (RBFNN). The closed-loop stability and finite-time convergence of this controller are proven by the Lyapunov theory. With the self-learning ability of RBFNN, online estimation and real-time compensation of the unknown term can be realized. Moreover, the gain of the robust term in the controller can be reduced. Furthermore, the PID-type manifold improves the transient response speed of the controller and reduces the steady-state error of the system. The advantages of the PID-type manifold and RBFNN complement each another, thereby ensuring the excellent control quality of the proposed controller. Computer simulation results show the feasibility of the proposed simplified model and the robustness of the controller. These results can help improve the intelligence of CPM, particularly when this mechanism buffers from a sophisticated unstructured engineering environment.