【作者】 张志辉；

【导师】 郭成安；

【作者基本信息】 大连理工大学 ， 信号与信息处理， 2009， 硕士

【摘要】 随着科学技术的飞速发展,世界各国政府对汽车领域节能与环保的要求不断提高,高压共轨技术正是为了适应这一潮流和趋势而发展起来的新技术。在我国采用高压共轨柴油发动机的机动车数量大幅增长,高压共轨系统的故障诊断问题就摆在了人们面前。但是由于我国还没有掌握高压共轨系统的故障诊断技术,故障车辆必须依靠国外指定的少数维修站来高价检测,因此这极大的影响了我国汽车产业对节能环保技术的推广与应用。成为影响我国汽车产业发展的瓶颈。研发具有自主知识产权的高压共轨故障诊断系统迫在眉睫。在高压共轨故障诊断系统的研发中,需要解决高压共轨系统高压压力的控制与故障诊断的问题。本文在研究现有控制算法的基础上,提出了三种压力控制算法,分别为基于PID的压力控制算法、基于卡尔曼滤波与PID的压力控制算法、基于BP神经网络的压力控制算法。通过试验分析表明,以上三种压力控制算法都可以较好的完成高压共轨系统的高压压力控制,得到较高的压力控制精度,完全可以满足在高压共轨智能诊断系统中压力控制的需要。在高压共轨系统的故障诊断方面,本文分别提出了基于阈值判别的故障诊断算法与基于BP神经网络的高压共轨系统故障诊断算法。由于BP神经网络具有良好的非线性映射特性,非常适合用于复杂系统的故障诊断。通过试验分析表明,基于BP神经网络的高压共轨系统的故障诊断算法可以很好的完成对高压共轨系统中的高压压力、系统自带传感器以及低压泵的智能故障诊断,其诊断的正确率可以达到99.89%以上,完全可以满足汽车产业对燃油发动机高压共轨系统故障诊断的需要。更多还原

【Abstract】 As the rapid development of science and technology, all countries have raised the standard of energy saving and environment protection, which makes high pressure common rail system the perfect technology for satisfying all these. In China, number of vehicles using high pressure common rail engines is increasing remarkably, while it comes up with problem of diagnosis of the system. However, China has no command of diagnosis technology of high pressure common rail engines, therefore, designated reparation stations are needed and the price of these stations is very high, which affects the development and implementation of this new technology. It has became a threshold in Chinese automobile industry. It is urgent to develop a self-designed diagnosis of high pressure common rail engine system.In the design of diagnosis of high pressure common rail engine system, the control and intelligent diagnosis of high pressure needs to be tackled.Based on the existing control algorithms, this paper presents three pressure control algorithms, namely, pressure control algorithm based on PID, pressure control algorithm based on Kalman filter and PID, pressure control algorithm based on BP neural networks. Experimental results show that these three algorithms can control the pressure of the high pressure common rail system, to obtain higher precise of the pressure. It is proved that they can all satisfy the application of diagnosis of high pressure common rail engine system.In the diagnosis, this paper presents two algorithms for intelligent diagnosis of high pressure common rail engine system based on thresholds and based on BP neural networks. Because of the great nonlinear projection ability, it is fit for complicated diagnosis. Experimental results show that this algorithm can better diagnosis the high pressure in common rail system, sensors and low-pressure. The preciseness can be higher than 99.89%, which is satisfied in diagnosis of high pressure common rail engine system.更多还原