Font Size: a A A

The Design Of Vehicle Anti-theft System Based On Multi-sensor Information Fusion Theory

Posted on:2017-09-17Degree:MasterType:Thesis
Country:ChinaCandidate:G ChenFull Text:PDF
GTID:2322330482486495Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Car is the product of industrial civilization in modern transport. With the increase of the number of cars, high-grade car theft rates rise year by year. Reliable and effective guard against theft system become a necessary means of guarantee property owners. Network gradually replace the traditional mechanical anti-theft anti-theft system into the mainstream. It is a use of modern electronic information technology and network technology to realize the car and the owner of the real-time information feedback of guard against theft system. This design based on this background.This paper design of car alarm system USES STM32 microcontroller CPU, through multi-sensor real-time detect the state of the car, car real-time status by multi-sensor detection, and carries on the fusion processing by the CPU, the fusion results through the GSM network to send to the owner the mobile end. System hardware design consists of CPU module, sensor module and mobile communication module and CAN bus module. Using the AT commands to realize communication central processing unit and communication module. This system software design mainly includes : mobile communication module program design; Sensor detection program design; Information fusion algorithm design; Alarm and control program design. Finally, the system hardware and software debugging.This paper, by using multi-sensor information fusion technology effectively reduce the risk of car theft and false alarm rate. First put forward the criterion of the improved algorithm based on D-S evidence, to improve the information fusion of conflict evidence. By this algorithm can realize auto security early warning, and effectively improve the system accuracy. Then based on SOM neural network information fusion algorithm is put forward, this is a no mentor self-organizing feature mapping neural network algorithm. This is verified by the experiment, this method can achieve input clustering characteristic of evidence, so as to realize the information fusion. Through the establishment of SOM network model of car alarm, simulation results in Matlab platform. The results show that the algorithm meet the design requirements. After debugging operation observation found that the system reliable operation is normal. So as to realize the real-time security monitoring for cars.
Keywords/Search Tags:Vehicle anti-theft system, Information Fusion Theory, SOM neural network, D-S theory
PDF Full Text Request
Related items