Study On Methods Of Occupants Counting And Weight-Identification Based On In-Vehicle Motion Sensors

For the sake of improving the fleet schedule and management of taxi companies, making the death statistics of car crash more complete, sticking to the trend of lightweight, intelligence and unmanned of vehicle and reinforcing the design of auto personality and car theft deterrent, it is proposed relatively high efficiency and extremely low cost methods of occupants counting and weight identification.Firstly, studying on the mathematical modeling of object of research. In order to represent the study of object rationally and accurately, the paper used a variety of previous vehicle dynamic models for reference. Based on the dynamic models and content of research, 7 DOFs, 4 DOFs and 3 DOFs vertical vehicle dynamic models are applied.Afterward, by analyzing the research of content and object, proper vertical vehicle model is selected. Furthermore, scenarios of single occupant boarding and alighting are simulated and analyzed by utilizing MATLAB/Simulink. So as to verify the mathematical model, fundamentals of design of experiment are introduced, and by which vehicle experiments are conducted. By comparing and analyzing simulation results and vehicle test results, the correctness of selected dynamic model is verified. Then, the most sensitive response signal is determined.By virtue of what have done, analyzing simulation and experiment results again, the solution of, based on threshold, occupants counting method is proposed. Additionally, the simulation model of occupants counting system is developed by applying MATLAB/Simulink, and also, simulation is processed with the help of experiment data.Finally, in order to achieve identification of occupants' weight, this paper derived and analyzed the transfer function of selected dynamic model. According to the property of linearity of linear, time-invariant system, the solution of occupant's weight identification method is developed based on in-vehicle motion sensors.