Search

Most of maritime accidents have been proved to be caused by human errors. In order to quantitatively analyze human factor of seafarers, we carried out the simulation test of ship operation. During the test, a brain instrument was used to collect Electroencephalogram (EEG) signals of the subjects. In the current research stage, EEG signals have been denoised by wavelet transform, and EEG features are extracted by finite impulse response(FIR) digital filter. The feature can be used as the feature vector of Support Vector Machine(SVM) classifier for training, and finally a classifier for workload recognition is established.

Aiming at traffic jams and potential traffic danger and other traffic problems at present, research on trajectory following control based on vehicle to vehicle communication on automatic drive is conducted. Through a simple and accurate vehicle kinematical model, it analyzes the state space of system, establishes model-based system state equations, combined with constraint conditions of practice and theory, it optimizes objective functions with the quadratic programming as the principal method, and avoids infeasibility solutions during its execution by properly improving objective functions, and then according to the updated status information of vehicle to vehicle communication it predicts the next time-domain output and feedbacks it to state space of system, repeat it until the control process is completed. Then, it establishes simulation models of tracking system through Simulink components in MATLAB, gains the movement data of front vehicle from Carsim software during vehicle to vehicle communication, and obtains the movement data of following vehicle accordingly. Finally, movement data of following vehicle and front vehicle is analyzed comparatively; the vehicle model predictive control predicts the feasibility of reference trajectory tracking in the quadratic programming method to complete the tracking strategy when moving.

Aiming at the status of the domestic and international wireless sensor network real-time acquisition system in the practical application, the program building real-time acquisition system by advanced technology of acceleration sensor detection, wireless network data transmission, and ZigBee protocol network is proposed. In view of the monitoring system automation, wireless transmission system network and modular system design, this wireless sensor network real-time acquisition system resolves the shortcomings, such as high cost of cable transmission system installation and environmental interference susceptible signals effectively. It has the advantages of easy installation and being real-time.