摘要
标定系统的主要任务是对MAP数据进行优化,通过建立发动机动力模型,利用遗传算法对动力模型进行优化计算,寻找满足发动机最佳性能的点火提前角和最小喷气脉宽,应用BP神经网络的方法,对优化后的点火提前角和喷气脉宽进行检验,防止优化后的控制参数可能导致排放性能的恶化。结果表明:初始化标定工作的方法是可行的。
Data optimization is the major property of calibration system. This paper builds up an engine dynamic model and optimizes the model using genetic algorithm to determine the control parameter of spark-advance angle and minimum specific fuel consumption at certain test condition which keeps the dynamic performance of the engine. The authors also develop exhausting model on the basis of experiment exhausting data using the BP network and check up the MAP data to assure that the exhausting performance will not be worse. The results obtained show that the initial calibration method is feasible.
出处
《西华大学学报（自然科学版）》
CAS
2005年第6期6-9,共4页
Journal of Xihua University:Natural Science Edition
