混合动力汽车自适应巡航跟车预测控制方法

In recent years, more efficient and cleaner energy utilization technology has become a research hotspot due to the problem of oil, environment, and traffic. The problems have led to the development of adaptive cruise control (ACC) system which combines intelligent transportation system and hybrid electric vehicle (HEV) technology. This kind of ACC system can utilize road and traffic information, regenerate the dissipation kinematic energy during deceleration and optimize engine and motor operation points to realize safe, efficient, and environment friendly driving. However, the engine and motor optimal operating point depends on the road slopes, the traffic junction conditions, the preceding vehicle information, and so on which change every moment. It is necessary to develop car-following models and design control strategies which can be implemented in real-time. In this project, 3 car-following models and predictive control methods based on traffic signal information are proposed. In China, the research on real-time implementable ACC system for HEVs based on car-following models is still at the starting stage. This project focuses on three key issues, (1) the three car-following models which consider driver properties; (2) the fuel saving property using different car-following models based on real road slopes; (3) the fuel saving property using different car-following models based on real traffic light information, and the proposed method which includes switch algorithm between model predictive control and rule-based control can optimize both the engine operating points and the driving profile simultaneously. The control strategy considers model errors. Through the analysis of the property of the real-time control and the study of the technology characteristics, this project can provide academic and technical solutions for the ACC development of HEVs in China.

近年随着资源，环境和交通形势的日益严峻，智能交通系统与混合动力汽车技术相结合的环境友好型自适应巡航控制系统备受推崇。其可以利用道路交通信息，回收再生制动能量，优化驱动装置工作点，实现安全节能环保。但是，最优工作点受到道路坡度、路口交通信息、前行车辆信息等时变因素影响，现有的跟车模型及控制策略很难满足实时最优控制要求。针对该类型自适应巡航控制系统的研究还处于初步阶段，本项目拟建立3种跟车模型；采用预测控制方法实现混合动力车基于交通信号灯信息的实时最优控制。主要研究内容：(1) 建立3种跟车模型，并考虑驾驶员特性;（2）实际道路坡度条件下的节能特性；（3）实际交通信号灯控制条件下的节能特性，通过建立预测控制和规则控制切换策略，实现驱动装置工作点和车辆速度模式的最优化匹配。控制策略考虑模型误差，分析算法的实时控制特性，掌握其技术特征，为混合动力汽车自适应巡航控制系统开发提供有力的理论和应用指导。

跟车反馈行为模型和道路交通信息前馈行为模型的模态变化问题，是智能交通系统与混合动力汽车技术相结合的环境友好型自适应巡航控制领域研究的重点和难点之一。本研究采用动力学解耦建模方法，以混合动力汽车为研究对象，构建了自适应巡航系统的稳态和暂态理论框架，得到了自适应巡航系统跟车行为的暂态变化矩阵。研究了保证系统最优性能和稳定性能的参数数值判定方法和道路交通信息预测模型，给出了系统在边界条件，不完全条件，不确定条件下的变化模型。通过实验对模型参数进行了标定，分析了自适应巡航系统的跟车行为力学和运动学特性，建立了外界道路交通车辆条件的预测模型基本框架，给出了混合动力汽车系统对于外界条件的响应特性模型。项目取得的重要结果和关键数据及其关键科学意义如下：（1）对发动机工作点和车辆速度模式进行同时优化被证明是有效可行的；（2）对巡航驾驶的车辆间距进行自适应控制被证明是提高系统性能的主要手段；（3）对蓄电池的充放电进行预测最优控制可以大幅度提高系统稳态节能特性；（4）对道路坡度和交通信号序列信息离散预测模型的有效利用，可以实现新能源汽车的全部操作模式，对控制方法的通用性进行了实验验证。本研究给出了自适应巡航系统的动力学机理并标定了系统对于外界条件变化的转移参数矩阵，可以为新能源汽车自适应巡航系统的研究和开发提供理论和技术依据。