| With the progress of the society, the amount of the automobile runs up, while consuming a large amount of fuel, the automobile exhaust has already become the main atmosphere pollution sources of our times. Engine electronic control technology is most important means to reduce the exhaust pollution of the engine, improve power and the economic. Its research has been paid highly attention to by developer all the time. The development of engine electronic control system can be divided into two basic stages of design and application. Hardware and software design of engine electronic control system is finished in the first stage. While in the application stage control parameters are calibration. Whether the parameters are calibrated well determines the engine's performance to a great extend. It is essential in the course of developing of the engine.Air fuel ratio and spark advance angle are the most important control parameters of electronic control gasoline engine. In order to satisfy requirement of emission regulations, Theory air fuel ratio is generally a goal of modern engine control, and carried out by close-loop control. The spark advance become the very important factor which affect the engine's power, economic and exhaust emission.,while the engine work on this region. So abstaining the best spark advance when the engine is in stable state operating mode is the important component during the calibration course .While the modern gasoline engine pursues good power and economy, will meet the request for certain emission regulation, So will consider power, economy and emission performance at the same time when calibration the spark advance angle, namely multi-objective optimization.The thesis is on the basis of further investigating calibration technology of the gasoline engine. According to the on-line calibration requirement, developed a calibration system of gasoline engine .Designed the hardware and software of the calibration computer and ECU .On-line calibration is achieved. A method of multi-objective optimization of spark advance angel on basis of statistic model is proposed. The method is composed of establishment of experimental statistic model and finding a solution of the multi-objective optimization problem. Good prediction performance shows when establishing orthogonal polynomial regress model of engine performance. The linear weighted sum method and group sorting method are adopted in solving of the multi-objective optimization problem, and Genetic Algorithm is applied in finding the minimum value of estimation function. The method is effective. |
