基于补偿模糊神经网络的汽车双离合器式自动变速器起步控制策略研究

摘要

An engine mathematic model is established with neural network training data. A start control strategy for dual clutch transmission is put forward with throttle opening and it change rate as input variable based on compensation fuzzy neural network according to the current status that no serf-learning function is available for most start control strategies. Simulation results are examined through some indexes such as start time, slip energy, impact, engine maximum speed and synchronous speed. Simulation results suggest that the control strategy presented is better than the original one,and it is of great serf-adaptive ability.%利用神经网络训练数据建立了发动机数学模型.针对目前起步控制策略大多没有自学习功能的现状,基于补偿模糊神经网络,以油门开度及其变化率为输入变量,提出了一种汽车双离合器式自动变速器起步控制策略.采用起步时间、滑摩功、冲击度、发动机最高转速和同步转速等指标检验仿真结果.结果表明,基于补偿模糊神经网络的起步控制策略在各性能指标方面均优于原控制策略,并具有较强的自适应能力.