Research On Modeling Of Head Movement Based On Surface EMG Signals In Virtual Cockpit Interaction Environment

Visual lag is the main problem which affects virtual reality helmet simulated effect in cockpitenvironment. Traditional methods of measuring head movement can effectively solve this problemonly for the slow head movement, but not for the sudden rapid head movement. In view of this, weuse the method based on head and neck sEMG signal to study the pattern recognition andclassification of head movement. This paper provide valuable research methods and ideas for furtherstudy of head multi-degree freedom movement and act as the preliminary work and good reference forthe follow-up achievement for real-time prediction of head rapid movement.The major achievements of this paper are as follows:1. Select right and left sternocleidomastoid, capitis and trapezius of the head and neck as thetarget muscles and determine the location of these muscles and electrode placement method.Combined the motion characteristics of human head movement in cockpit environment with theprinciple of the sEMG signal generation, choose head rotation as the research object, makecomprehensive analysis of a large number of documents and experiments to study the target muscles.2. Build hardware and software experiment platform, prove the accuracy of the target muscles’position and electrode placement method proposed in this paper. Synchronously obtain the sEMG andinertial signals through a series of typical head rotation signal acquisition experiments to carry out theresearch.3. Acquire the head movement data samples in constant rotation angular velocity but differentrotation angles, establish a BP neural network(BPNN) model which maps head and neck sEMG torotation angle based on signal segmentation and feature extraction to achieve the identification andclassification for four different angles of head rotation.