移动窗口域的VDO爆震燃烧识别扩展算法

Siemens VDO爆震识别算法广泛应用于汽油机爆震燃烧诊断,但对于发生在缸压峰值前的爆震循环,其在缸压下降沿确定爆震计算窗口的思想在一定程度上影响了爆震识别及评价的准确性。在VDO算法基础上提出了一种基于移动窗口域爆震识别算法,引入爆震窗口和参考窗口高频累积能量最大差值,利用在指定K区域内移动的参考窗口和爆震窗口,准确识别出爆震计算始点,使爆震窗口包含更多爆震能量,并用此时的爆震窗口和参考窗口累积能量之比作为爆震因子。结果表明:移动窗口域算法的爆震计算始点比VDO算法爆震计算始点提前,呈现更多的爆震能量信息,能弥补对发生在缸压峰值前的爆震事件的漏判。

VDO knock diagnosis expansion algorithm based on moving window domain

Siemens VDO knock detection algorithm is widely used in knock diagnosis in gasoline engine.However,the evaluating accuracy of knock is partly affected by the thought that the knock window is determined on the falling edge of cylinder pressure when knock occurs earlier than the peak pressure.VDO knock diagnosis expansion algorithm based on MWD (moving window domain) was proposed and the identification parameter △Emax was introduced to characterize the maximal accumulative high-frequency energy difference between a knock window and a reference window.By moving the reference window and the knock window within the area K,the knock calculation starting point can be detected accurately and more knock energy can be contained in the knock window.The ratio of the integral energy in the knock window and the reference window at this condition was defined as the knock factor.The result shows that the knock calculation starting point of MWD algorithm is earlier than the one of VDO method.It can present more knock energy information and avoid the missed diagnosis of the knock occurring earlier than the peak pressure.