Dynamic Modeling Of Hot Film Type MAF Sensor Used For The Engine With Various Components Coal Bed Methane And Air-Fuel Ratio Control Based On The Observer

In order to make gas-engine work stable, it must be assure that the intake mixture have appropriate air-fuel ratio. Based on this conception, this dissertation is mainly about:Firstly, we make static and dynamic calibration experiment to the MAF sensor on the hot-film air mass flow sensor calibration board. And we used ARX, Hammerstein, Wiener and linear neural network model to make linear and dynamic non-linear identification to the MAF sensor. For linear neural network model, we achieved the highest fit precision by changing the delay. For Hammerstein and Wiener model, we used different strategy to make order selection for the linear part, and used the OE or ARX model construction, and with the polynomial identification, achieved factor estimation of the non-linear part.Secondly, we design and realize the estimation method of the gas-engine intake manifold pressure based on the state observer. The first step is the constitution of the intake manifold dynamical equation based on the average model. Then, constitute linear and non-linear slide mode observer. The simulated result of the MATLAB shows that the slide mode observer has better estimation effect than the linear observer.Thirdly, we constituted intake mass flow dynamic model and oxygen dynamic model based on the forward & feedback control precept of A/F. And constituted slide mode observer on this basis to estimate the air mass flow, gas mass flow and A/F. The simulation result demonstrates that this observer has good robusticity and astringency.Lastly, based on the A/F ratio control strategy, forward part adopt the air mass flow and the gas mass flow MAP, added controller in feedback part, in order to make precise control to A/F ratio. In this part, the feedback controller adopts two groups of data to make simulation by used of PID, SMC and RBF control strategy separately. The simulated result demonstrates that the A/F ratio feedback control based on observer has good effects on controlling A/F ratio around the stoichiometric point.