Research On Air-fuel Ratio Control Method For Gasoline Engine Base On Matlab

Air-fuel ratio control and Sparkle advance angle control are the two important problems in automotive control field. It is necessary to maintain the AFR close to its stoichiometric value and control the Sparkle advance angle to ensure vehicle driveability and low pollutant levels This is because the stoichiometry represents the maximum point of catalytic converter efficiency. Efficient performance of engine speed control is also a important problems in automotive control field,which can show the engine's power performance flexibly. Also, it can reduce the fuel consumption to enhance the engine economy.First,establishing the engine model and engine's control model using Matlab/simulink. As the engine's control system is a several imports and several exports nonlinear system, Air-fuel ratio control model is uncertainty in parameters and nonlinear in operation trait. The traditional model establishment in nonlinear system such as automobile engine is still very hard. So, This thesis builds the model by relevant of control arithmetic in each of work conditions.Neural network control has the virtue of simple algorithm and stronger robust. So, it is suit for controlling the nonlinear system such as engine which has pure time delay characters. This thesis choice the BPNN control method which is more robust in adapting by itself and more quickly in learn and more highly in efficient; Fuzzy auto tune PID control algorithm has well traits of excellent dynamic tracking quality and accurate running in stable state and unsensitivity to the control parameter;this thesis adopts MAP control method which be used comprehensive in the stable state and have better precision by NN arithmetic for optimizedSecond,engine simulative calibration model was established, in which the simulated calibration experiment was taken and the Primal oil MAP data was obtained. And artificial intelligence neural network ways, which can increase the efficiency to get the original MAP data. From the viewpoint of the process seems a bit tedious, but the use of neural networks MATLAB toolbox, making neural network design and training process are relatively simpler, more important is to use neural networks to the use of data and information. To consider the strong non-linear engine, so from that point speaking, the traditional linear interpolation is certainly inaccurate. And neural networks in nonlinear mapping areas it is precisely its advantage, theoretically speaking, the use of neural networks to predict the initial injection MAP data processing is more reliable Last, a set of graphical user interfaces is made with GUIDE, which is a graphical user interface development environment, inputing the data to the engine for the closed and opened loop control of the air/fuel ratio model and Engine simulative calibration model which can accomplish simulative calibration and original MAP to certain type of engine conveniently. In the interface design and control by calling the control of the callback function can be achieved engine modeling and simulation system interactive interface design.