【作者】 肖永茂；

【导师】 师帅兵；

【作者基本信息】 西北农林科技大学 ， 机械设计及理论， 2009， 硕士

【摘要】 随着汽车的大规模生产、使用和公路等级的不断提高,特别是高速公路的飞速发展,汽车的行驶速度逐渐加快,汽车安全问题越来越受到人们的重视。因此,研究汽车主动防撞系统对提高汽车行驶的安全性,避免事故的发生有重要的理论意义和实用价值。本论文在分析现有研究基础上,提出建立基于热红外成像传感器的汽车主动防撞系统,汽车正常行驶时,该系统不发生作用,当汽车与目标物之间距离小于相应行驶条件下设定的安全距离时,系统发出报警信息,提示驾驶员采取相应措施,及时避免事故的发生。在对汽车制动过程分析的基础上,得出汽车制动距离的一般模型。依据汽车制动距离模型的要求和目标物的运动类型得出汽车安全距离模型,给出安全距离的计算方法。道路附着系数是计算安全距离的重要参数之一,但道路附着系数由于路面条件不同很难实际测量。利用回忆对比的方法修改附着系数,将经典理论数据输入神经网络,并对模型进行训练巩固,在不同路面类型进行制动实验获得实验数据,将测得实验值与理论数值进行对比修正,使系统内的附着系数值和实际附着系数越来越接近,为准确计算汽车安全距离提供依据。将热红外成像传感器采集的图像用灰度增强法进行增强,并对图像进行分割以突出目标物。由于图像分割后,还可能有次目物体影响判断,因此对分割后的图像进行形态学处理,有效地将目标图像区域从场景区域中分割出来。通过投影成像关系,重新构建立体视图从而测出目标物的距离,并给出目标物的运动类型判断公式。建立了汽车防撞专家系统,设计了专家系统推理机和知识库。将汽车安全行驶的知识规则以及安全行驶的推理策略导入专家系统,以汽车驾驶知识作为辅助指导操作在危险情况下进行报警避免汽车碰撞。并依据汽车在实际行驶的需要设计了信号采集、信号处理、修改参数和报警提示等模块的系统程序。更多还原

【Abstract】 with the mass production of vehicles and the construction of the highway, the vehicle runs increasing fast. So, more people pay more attention to vehicle safety today. Therefore, the study of active collision avoidance system to improve the safety vehicle to avoid accidents has important theoretical and practical value.In this paper, the analysis of existing studies based on thermal infrared imaging based on vehicle active anti-collision sensor system. The system does not work when the vehicle runs normally. When the distance between object and the vehicle is less than a safe distance, this system works to give the driver alarming to avoid the accident.The model for calculating the distance of braking was set on the foundation of analyzing the brake procedure. The algorithm for calculate the safe distance was obtained according to the general models of the braking distance of car.The road adhere coefficient is a vital parameter for calculating the safety distance. However, the road adhere coefficient is difficult to be measured when the road condition changed. The artificial neural network model was set after studying the road adhere coefficient and utilize the existing theory. The signals gathered in the process of braking were transmitted to the network and the parameters were revised using these signals to make the parameter has the least difference with actual value.The image enhanced algorithm was applied in the enhancement of the infrared. Then, the image was segmented to detect the object. But there are other objects in the image after segmentation. The mathematical morphology was used to process the image after segment at the same time in order to improve the effect of object detection. The distance between the vehicle and the object was calculated after using the projection imaging model. And the moving model determination function was constructed at the same time.In addition, the vehicle security expert system was constructed after the studying of the expert system. The vehicle safe running rules and the deduce algorithm were input to the expert system after analyzing the knowledge of the expert system is the key method for anti-collision. This system provides an effect assistant drive method, which can escape the accident. At last, the signal capture and process, parameters modify and alarming models were designed according to the need of the vehicle running condition.更多还原