Research On The Ignition Control System Of HCNG Engine Based On Fuzzy Neural Network

With the increasing amount of automobiles in China, the problems of the pollutioncaused by automobiles are becoming more and more serious, so are the situations ofresources exhaustion. Therefore, the exploration of new clean energy sources forautomobile is increasingly a cause for concern. Hydrogen enriched with compressednatural gas, or HCNG for short, is a new type of gaseous fuel, which is mixed with CNGand hydrogen in a certain proportion. HCNG can make the most of the advantages ofCNG and hydrogen, and it is a kind of alternative fuel for automobiles, which caneffectively relieve the energy crisis and the pollution caused by automobiles.In this paper, the research on the control of HCNG engine ignition system wascarried out to obtain the optimal ignition advance angle of HCNG engine by using thecontrol method of fuzzy neural network. All of the research was based on theoreticalstudies, simulation analysis and experimental verification. A portable device wasdesigned to generate the ignition advance angle of HCNG engine (hereinafter referred toas portable device), which can be used to guide bench test of HCNG engine.The control method of neural network and fuzzy inference system were investigated,and the research on neural network control was improved, which was based on back-propagation neural network. In order to realize the simulation of improved neural network,some relevant programs were written in the software of MATLAB. At the same time, theimproved neural network was used to control the ignition advance angle of HCNG engine.It proves that the improved neural network is correct and superior comparing with thetraining result of back-propagation neural network.Adaptive-network-based fuzzy inference system, or ANFIS for short, was used torealize the fuzzy neural network control of the HCNG engine ignition advance anglebased on the investigation of the neural network control and fuzzy control. In the softwareof MATLAB, a control model of HCNG engine ignition advance angle was establishedby programming the control method of adaptive-network-based fuzzy inference system.The calibrating experiment data of HCNG engine ignition advance angle was taken as thetraining sample for the simulation. It proves that the ANFIS meets the accuracyrequirement of the control system.MC9S12XS128SCM was chosen as the main chip, and the corresponding hardwarecircuits were designed to manufacture the portable device. The software program was designed in the software of CodeWarrior IDE by using the simulation parameters ofadaptive-network-based fuzzy inference system. The BDM circuit was used to debug&download the program from the software of CodeWarrior IDE to the portable device. Theresults show that the portable device can output the corresponding ignition advance angleof HCNG engine with different engine speed and intake manifold pressure. It proves thatthe portable device is correct and steady.The portable device was used to guide bench test of HCNG engine. The referencevalue of ignition advance angle was obtained by the portable device before the test. Thetest value of ignition advance angle was set every other crank angle around the referencevalue. Then, the bench test of HCNG engine was proceeded, and the experiment data wasrecorded. The test results show that the optimum ignition advance angle of HCNG engineis almost the same with the reference value, and it proves that the portable devicepossesses some value of practicality.