Research On Force Control Of Helmet Servo System In Virtual Cockpit Interaction Environment

In order to solve the deficiencies of helmet servo system in existing virtual cockpit interaction environment,such problems as uncomfortable and foreign body sensation obviously.Through the use of sEMG signal that advance in head and neck movement as the input of the control system,puts forward an improved algorithm of impedance control based on sEMG signal in this article.Realizing the coordination control of virtual reality helmet servo system and improving the response speed of the system effectively.Reducing the interaction force between head and helmet as well.The main content is as follows:(1)The analysis of skeletal muscle in head and neck:Having a detailed analysis of the three factors that affect the producing ability of the muscle force in this paper.According to the theory of anatomy,a detailed exposition of bone structure and kinematic features in cervical part has been done.Then in accordance with the order of the anatomy from shallow to deep,we sum up the type,position and function in different parts of the head and neck muscle.The muscular structure of the major muscles in the head and neck and its change in head movement have a detailed analysis.Analyzing major muscle activation methods related to the head and neck movement,so as to decide how to select different muscle in the different movement.(2)The construction of the neural network model based on the sEMG signal:This article first expounds the relationship between the sEMG signal and the muscle force briefly.Then,designing experiments and using the method of time and frequency feature extraction for the collected sEMG signal after noise reduction processing.Through comparing the extracted results,selecting the typical characteristics.By analyzing in detailed and comparing between biological mechanics model and the neural network model,the BPNN model is determined.Then designing experiments and collecting relevant data for model training.Choosing BP neural network that has the best prediction result as the sEMG-interaction force prediction model by comparing the performance of BP neural network with different the number of neurons in hidden layer nodes.(3)Putting forward an improved algorithm of impedance control based on sEMG signal and validating through simulation:This paper put forward an improved impedance control algorithm based on sEMG signal and use a BP neural network as the feedforward controller of impedance control to predict the interactive force.Setting up a signal selector in the control loop and fusing the predicting interactive force and measuring force as a control input.Using simulation module in matlab as the main tools.Constituting the circuit if improved impedance control and control model and head-helmet interactive model.Through force and position tracking simulation validation to the head-helmet interaction proved that improved impedance control algorithm can not only effectively realize accurate position tracking,but also the response time on the interaction force control is less than traditional impedance control and PD control and performance better.