Research on Haptic Feedback Control System of Bionic Pulse Diagnosis Equipment

Pulse diagnosis is one of the most characteristic diagnostic methods in Traditional Chinese Medicine(TCM), as practiced and providing medical service in disease prevention, diagnosis and treatmentterritory. Lack of standardization and subjective of pulse diagnosis, is the main concern in modern Chinese medicine. There are problems in present pulse diagnosis equipment, such as low precision of pre-pressure control for pulse wave acquisition, weak anti-interference ability in the acquisition process and poor ability to adapt to different groups of people. In this study, we proposed an interactive model between pulse acquisition dexterity finger and flexible skin was identified by using nonlinear autoregressive networkwith exogenous inputs (NARX)neural network model, which was further used to simulate and test the entire control strategy. Secondly, an adaptive control strategy based on Radical Basis Function (RBF) neural network was proposed, and the stability of the closed-loop system was proved by Lyapunov stability theory. Finally, the simulation and laboratory experiments were conducted. The simulation experimentsshow that the mean square error(MSE)of the system identification model, which is based on NARX neural network model accurately describes the mathematical relationship between “Guan” wrist tissue and dexterous fingers, is 1.1878, and the percentage of relative mean square error is 0.36953%,and it could means that the model achieves the modeling accuracy needed for control purposes; The adaptive RBF neural network controller is at least 1.67 times faster than PID controller, 3.76 times smaller of maximum tracking negative error than PID controller, at the same time the laboratory experiments get the similar results. So it could be concluded that the adaptive RBF neural network controller has the characteristics of fast adjustment speed, small overshot and strong interference suppression ability, which achieves the precision and robustness of pre-pressure control in the process of pulse wave acquisition.