【作者】 张西雅；

【导师】 孔金生；

【作者基本信息】 郑州大学 ， 检测技术与自动化装置， 2010， 硕士

【摘要】 信息融合技术是上世纪80年代兴起的新技术,随着计算机技术、信息技术的发展,它逐渐的被应用到各行各业中。多传感器信息融合的特点就是将来自不同方面的信息进行融合,避免了单一传感器测量带来的信息丢失和片面等问题。目前,车辆的防撞技术主要是被动的比较多,主动的防撞装置在西方一些国家相对较多,而我国这方面还需要进一步的研究。汽车防追尾避撞系统主要是采集目标车辆的运动状态,帮助驾驶人员更好的判断车辆的运行状态。本文结合多传感器信息融合及汽车防追尾避撞系统特点,对基于信息融合的汽车防追尾避撞系统的目标识别方法进行了深入研究,其主要工作和内容如下：论文在对汽车防追尾避撞系统组成和工作原理进行了分析和探讨的基础上,提出了基于信息融合的汽车防追尾避撞系统目标识别问题,并对其目标识别和测距原理进行了深入的分析和讨论。在对多传感器信息融合原理及其融合模型分析的基础上,对多传感器信息融合与目标识别工作原理和算法进行了深入的分析和探讨,为基于信息融合的汽车防追尾避撞目标识别研究提供了坚实的基础。在对联合卡尔曼滤波器的原理、结构和算法介绍和分析的基础上,针对汽车防追尾避撞系统,给出了基于联合卡尔曼滤波的信息融合方法。仿真结果表明基于联合卡尔曼滤波的信息融合方法可以使雷达和红外传感器得到较好的融合效果,提高了数据的准确性,从而增强汽车防碰撞系统的安全性。给出了基于信息融合的汽车防追尾避撞目标识别方法,该方法将联合卡尔曼滤波器中局部滤波和全局滤波的输出进行融合得到目标的状态,用于汽车防追尾避撞系统的目标识别中,可有效提高目标识别效果。仿真结果表明了所提出方法的有效性和可靠性。更多还原

【Abstract】 Information fusion technology is a new technology appeared in 1980s in the 20th century, it is used in every walk of life along with the development of computer technology and information technology. The characteristic of multi-sensor information technology is that it cans fusion information coming from different sensors, so the information losing and unilateralism problem caused by single sensor can be avoided. Now, automotive anti-collision is mainly about passive equipments, and active anti-collision equipments are more in some western countries, but they need another more research in our country. The mainly function of automation anti-collision is collecting athletic information of target automation to help the driver to estimate the state of auto car effectively. In this paper, combined the characteristics of multi-sensor information fusion and automation anti-collision system, target recognition algorithms based on information fusion in automation anti-collision system is deeply researched. Main work and contents are as follows:In this paper, the structure and work elements of the automation anti-collision system are analyzed, then the problem of target recognition based on information fusion in automation anti-collision system is put forward, and target recognition and measurement theory are deeply analyzed and discussed.Based on the analysis of elements of multi-sensor information fusion and its fusion model, multi-sensor information fusion and the elements, algorithms of target recognition are deeply analyzed, providing strong foundation for automation anti-collision system’s information fusion target recognition.The principle, structure and algorithm of federal Kalman filter are introduced and analyzed, and then the algorithm of information fusion based on federal Kalman filter is given. Simulation results show that this algorithm is very useful in fusion the radar and IRI data, data veracity is improved and the safety of automotive anti-collision system is enhanced.The target recognition method based on information fusion in automotive anti-collision is given. In this method, the outputs of partial filter and overall filter in federal Kalman filter are fused to get the state of target auto car, which is used in the automotive anti-collision system to improve the effect of target recognition. Simulation results show the method discussed in this paper is effective and reliable.更多还原