基于虚拟传感系统的内燃机智能多变量自适应控制的研究

As more and more control variables of internal combustion engines including compression ignition engines and spark ignition engines for increasingly stringent emissions and fuel costs requirements, conventional MAP-based control is facing more and more challenges. The reason using MAP-based control is that many performance parameters of engine is very difficult to measure directly. This study first developed a neural network-based virtual sensing system that is able to predict real-time engine performance parameters such as engine power output, fuel consumption, especially, emissions. Then, a virtual sensing system-based multi-variables intelligent adaptive controller and the way of self-produced control rule was be researched. The study will realize real optimal control and can efficiently solve the problem of future electronic controlled engine calibration.

常规基于MAP的开环控制方法已忧为内燃机综合应用多种技术、实现更高效更清洁燃烧的瓶薄１狙芯磕饨⒁恢帜苁盗吭げ饽谌蓟餍阅懿问男槟獯衅飨低常⒀芯烤哂凶允视ψ匝澳芰Φ亩啾淞恐悄芸刂品椒ǎ允迪帜谌蓟恼嬲呕Ｕ舛酝诰蛉忌湛刂萍际醯那绷褪视δ谌蓟际醴⒄沟男枰哂兄匾南质狄庖搴蜕钤兜睦砺垡庖濉