摘要
为揭示发动机排气噪声的影响因素及程度,采用正交试验设计法对X2105C型发动机排气噪声及其参数进行实验研究,确定了影响发动机排气噪声的主次因素及其交互影响。采用遗传算法的神经网络对试验样本进行学习,建立噪声预测方程。结果表明:正交试验可以有效地对发动机排气噪声影响因素进行测试。对于X2105型柴油机及指定消声器水平,排气噪声强度可用转速、扭矩、排气管长近似描述,建立线性方程。经对比检验,遗传神经网络预测法有较高的精确度。
Orthogonal experimental study on engine exhaust noise and its influence factors is carried out on type X2105C diesel engine. This is to study the influence factors and degree on internal combustion engine exhaust noise, and determine the main and secondary influence factors on diesel exhaust noise and their interaction. A noise prediction equation is established by studying the test samples based on genetic algorithm neural network. The results of the study show that orthogonal test can be used to test engine exhaust noise influence factors effectively. For X2105 type diesel engine and designated muffler level, exhaust noise intensity can be approximate described by using engine speed, torque, pipe length to build a linear equation. The contrast test, genetic neural network prediction method is more accurate.
出处
《实验室研究与探索》
CAS
北大核心
2014年第4期29-32,共4页
Research and Exploration In Laboratory
基金
江苏省自然科学基金项目(BK2012278)
关键词
发动机
排气噪声
正交试验设计
遗传神经网络
噪声预测
engine
exhaust noise
orthogonal experimental design
genetic algorithm neural network
noise prediction
