Vehicle Classification Research Based On Piezoelectricity Sensor

Sustained and rapid growth of China's socialist economy, and the continue improvement of people's living standards, led to the development of the automobile industry and increasing purchases of cars. Car ownership has become a very common phenomenon. This has indeed brought to people's daily lives greatly facilitated, while promoting the further development of the social economy. But it also produced a series of social problems, such as worsening traffic environment, continue to rise of traffic accident rates, and increasing urban road traffic congestion. Meanwhile, driven by the interests and the impact of transport competition, transport vehicles of overloaded and overrun increased year by year, and the destruction of their road and other economic and social losses causing by that is shocking.Comprehensive and systemic research is done in the existing method to classify vehicles automatically. On the basis of analyzing the merits and demerits of diverse methods, the thesis puts up its own scheme. My main work contains:(1)This article designed a vehicle identification system based on the piezoelectric film. We measured data as accurate as possible, through reasonable lying of the high sensitivity of piezoelectric thin shaft. After signal enhancement and filtering, we can sort out the wheelbase, axle load, axle number, round number, vehicle weight and other parameters of the vehicle.(2)We collated the effect of vehicle parameters for vehicle classification, and selected useful features for the classification. This article focuses on exploring axle load distribution of the vehicle of different models and different axes. We effectively improved the recognition rate by introduction of axle load characteristics.(3)In order to enhance the use of the vehicle features as possible, we design a two-tier classification. The first layer classifier makes completely accurate classification or limit utilizing the obvious features of models for the first time. The second layer is a fine classifier based on optimization features and kernel function and parameter selection of support vector machine. This classifier further subdivided models based on the discrimination results of first layer.Experiments show, the two-intelligent classifier can effectively utilize data measured by piezoelectric film. At the same time, axle load data which article focuses can play a good role in the classification classifier.