汽车有级自动变速器多性能综合最优个性化智能换挡规律

Aiming at the ever-increasing demands on optimal comprehensive performance, individualization and intelligentization for stepped automatic transmissions, the project is to investigate the performance optimization methodologies and modification strategies of gearshift rule that could take into account the power performance, fuel economy and emission together, represent individual driving styles, and adapt various driving conditions. Firstly, the models of various performance indices of engine will be established using artificial neural networks. Secondly, the manipulating parameters that could reflect diver's expectation on gearshift performance will be induced, and the quantification model used to map these parameters to diver's expectation on gearshift performance will be built using fuzzy reasoning method combined with optimization of fuzzy rules and membership functions. Thirdly, the evaluation function of comprehensive performance reflecting driver's preference will be constructed, and the computation method of gear-shift rule based on multi-performance optimization with driver's preference will then be put forward with application of the stochastic intelligent optimization theory,which uses the evaluation function as objective, the dynamic 3 parameters including throttle opening, velocity and acceleration as design parameters. Fourthly, the adaptive modification strategies under various driving conditions for the individualized optimal gear-shift rule will be proposed based on the analysis of the influence of the changes of driving conditions on gearshift performance. Finally, the proposed methodologies and strategies will be verified, improved and refined through system modeling, simulation and vehicle test. The outcome of this investigation will lay a good foundation for constructing a theoretical architecture for individualized intelligent gearshift rule satisfying all requirements of power performance, fuel economy and emission, which will improve the design and development capability of the control system of stepped automatic transmissions in China and also provide a novel method to further excavate the potential of energy saving and emission reduction of automatic gearshift systems.

针对综合性能最优、个性化和智能化的要求，研究全面考虑整车动力性、燃油经济性和排放性能并体现个性化驾驶风格的有级自动变速器智能换挡规律。研究发动机各性能指标的神经网络建模方法，并建立其性能模型；归纳反映驾驶员换挡性能期望的操纵特征物理量，运用模糊推理方法、结合对模糊规则及隶属函数的优化，建立驾驶员换挡性能期望量化模型；构造兼顾三类性能要求并体现个性化驾驶风格的综合性能评价函数，以该函数为优化目标、动态三参数为设计变量，应用随机性智能优化理论，设计一套多性能综合最优个性化换挡规律计算方法；研究行驶条件变化对换挡性能的影响，提出多性能综合最优个性化换挡规律的行驶条件适应性修正策略；通过建模仿真与实车试验，对提出的理论和方法进行验证、改进和完善；建立满足多种性能要求的个性化智能换挡理论体系，促进我国汽车自动变速器控制系统设计开发能力的提高，并为挖掘汽车自动变速系统节能与减排的潜力提供一种新方法。

为解决有级自动变速汽车在满足驾驶员换挡性能期望和适应道路行驶条件变化的前提下，实现整车动力、经济和排放性能综合最优的问题，研究了多性能综合最优个性化智能换挡规律的优化方法。通过建立基于神经网络的发动机多性能指标模型，为换挡规律优化过程中的性能计算提供了数据基础。提出了基于驾驶员性能期望加权的换挡性能综合评价函数，建立了驾驶员期望与综合性能优化目标之间的联系。设计了基于遗传算法的有级自动变速器多性能综合最优个性化换挡规律优化方法，优化后的换挡规律能够在体现驾驶员个性化需求的同时使汽车的综合性能达到最优。在对不同类型驾驶员操纵数据、车辆运行状态和主观性能评价数据进行采集分析的基础上，建立了基于模糊推理的驾驶员换挡性能期望量化模型，为根据驾驶风格识别结果进行个性化换挡规律的选择提供了依据。分别设计了适应坡度变化和适应弯道行驶的多性能综合最优个性化换挡规律优化方法，通过与坡度和弯道识别技术的结合，优化后的换挡规律可使有级自动变速汽车在减少非必要换挡次数和提高行车安全的同时，能够根据驾驶员的期望达到综合性能最优。