摘要
传统的汽车尾气监控将尾气排放的数据采集后,简单表示出排放是否超标,无法将尾气排放信息智能融合后形成诊断模型;设计并实现了一款使用与内部的汽车尾气超标智能监控系统,通过嵌入式的微控制器I-8431与Modle6800传感器群上下位机形成汽车排气管处的硬件互联,对尾气成分分别检测,防止混合气体采集时带来的数据噪声,将采集的排气数据加载到经过优化权重的神经网络信息融合模型中,通过神经网络强大的非线性推理汽车尾气排出前的污染情况,实际的系统测试中,通过不同配比与浓度的汽车尾气分析,这种方法尾气检测准确率高达99.7%,具有很高的实用价值。
Traditional automobile exhaust monitoring will exhaust after data collection, simple show excessive emissions, to exhaust e mission information intelligent diagnosis model is formed after fusion. Was designed and implemented an internal use and car exhaust excee ding intelligent monitoring system, through the embedded micro controller I 8431 and Modle6800 sensor group of upper and lower place machine hardware connected auto exhaust pipe, the exhaust gas composition was detected, prevents the mixed gas gathering of data noise, the exhaust of the data is loaded into the optimized weights of neural network information fusion model, through the neural network's power ful nonlinear reasoning in front of the car exhaust pollution situation, the actual system testing, by different ratio and concentration of auto motive exhaust analvsis, this method emissions testing accuracy is as high as 99. 7%, has the very high oractical value.
出处
《计算机测量与控制》
北大核心
2013年第11期3001-3003,共3页
Computer Measurement &Control
关键词
尾气智能监控
优化神经网络
非线性推理
intelligent monitoring of tail gas
optimization of neural network nonlinear reasoning