针对搭载湿式DCT的混合动力汽车换挡过程存在的随机多时滞现象，及以往建模和控制策略多侧重于非时滞系统的情况，本项目拟深入开展随机多时滞的混合动力汽车换挡过程的建模研究工作，揭示多种时滞的特性，在此基础上建立换挡过程的控制策略。项目首先提出多模型的建模思路，将四种换挡方式建模为相互独立和离散状态的时滞马尔可夫随机过程。其次提出“策略层+控制层”的控制策略。策略层通过模糊推理得到的调节因子向量,去调节换挡品质评价指标最优目标泛函的权矩阵，从而得到动态平衡各评价指标矛盾的目标控制信号。控制层通过构建时滞动态输出观测器反馈系统，使离合器油压等实际值快速跟随目标控制信号。为此项目从理论上提出时滞动态输出观测器的设计方法和稳定性分析方法。项目建立的模型和提出的控制策略将在硬件在环平台和样车上进行验证和优化。上述研究将为具有相似结构的混合动力汽车换挡过程建模和控制提供新的理论和方法及切实有效的实现途径。

Since random multi-time-delay phenomenon is existed in shift process of hybrid vehicle with wet dual-clutch transmission (DCT), and pervious references focus on non-delay system, this project will deeply research the modeling work with random multi-time-delay in shift process of DCT hybrid vehicle, and post the character of time-delay. Based on above work, the project will establish the control strategy of shift process. In order to achieve these purposes, the project first presents a multi-model idea. Four shift modes will be modeled as independent and discrete state Markov random process with time-delay. Then the project presents the control strategy ,which is composed of strategy layer and control layer.The strategy layer uses fuzzy inference to obtain adjustment factor vector. The adjustment factor vector will be used to regulate the weight matrix of optimal objective function, which is composed of shift quality evaluation indexes. This method can get the target singals to eliminate the contradiction between evaluation indexes dynamically. In order to get better trackability, control layer will build a feedback system with time-delay dynamic output observer. The project will present a new design method and stability analysis method of dynamic output observer system with time-delay. The models and control strategy, which this project presents, will be validated and optimized on hardware-in-loop plantform and hybrid vehicle. The above researches present a novel theory and method of modeling and control for other similar hybrid vehcles. The proposed methods are effective for the practical application.