行星混联客车系统实时复合优化算法与动态协调控制

Under the domestic background of energy-saving and emission-reducing, our country is devoted to the research and development of planetary power-split hybrid system for buses because it combines the advantages of series and parallel hybrid systems. However, the energy-saving optimization and the accurate control of the dynamic properties of the planetary system remain to be the bottleneck in this field. Faced with the fact that it is difficult to realize real-time performance and optimality of the planetary system at the same time, the applicants come up with a real-time composite optimal algorithm. The combination of genetic algorithm and dynamic programming aims at researching into the influence of the weighting factor in the penalty function on the global optimal solution. And it is able to realize composite optimal control with real-time performance and adaptation for complex driving cycles based on the prediction of operating cycles and extraction of control rules. Then, a gain-scheduling state-feedback controller is proposed to improve the weakness in effectiveness and uniformity of the dynamic coordination control of planetary hybrid system. Moreover, a planetary hybrid system including the engine harmonic model and the torsional vibration model of the drivetrain is established. At the same time, a non-linear engine torque observer which is excellent in robustness is designed and a dynamic coordination control algorithm based on the gain-scheduling state-feedback is developed for the control system. This project is anticipated to improve the fuel economy and longitudinal smoothness of the planetary hybrid bus system and provide theoretical guidance for the independent development and competence enhancement of the planetary hybrid buses.

在国内节能减排大背景下，由于行星混联式混合动力系统集成了串、并联系统的优点，促使国内大力研发混联式混动系统客车系统。混联动力系统的节能优化及系统动态品质精细控制是这一领域需重点突破的技术瓶颈。针对混联动力系统存在的节能优化的实时性和最优性难于兼顾问题，申请者提出一种实时复合优化算法。复合遗传算法与动态规划，探究储能系统电量状态罚函数权重因子对全局最优解的影响，进而，结合工况预测及控制规则提取方法，实现实时性及工况适应一致性的复合优化控制；针对混联动力系统动态协调控制效果与一致性差的问题，提出一种增益序列状态反馈协调控制算法。建立包含发动机谐波模型与传动系扭振模型的混联动力耦合系统，设计良好鲁棒性的非线性发动机转矩观测器，建立基于增益序列状态反馈的动态协调控制算法。期望有效提升行星混联客车产品的燃油经济性和纵向平顺性两个关键性能，为行星混联客车自主研发及提高其综合竞争力提供理论指导。

行星混联客车系统性能优异，应用优势明显，当前已得到各大客车公司的大力支持和持续资金投入。本项目主要研究行星混联客车系统实时优化控制与动态协调控制问题，目标是实现节能优化效果和动态控制品质的综合提升。首先通过分层优化算法获得全局最优解，并集成工况识别算法与规则提取方法在保证实时性的同时尽可能逼近全局最优解；该方法能够解决优化控制策略的最优性与实时性矛盾。其次，集成发动机谐波转矩模型与传动系扭振模型，为控制策略的研究提供可靠平台，开发非线性转矩观测器准确估计真实的发动机转矩。然后，分别针对发动机启动阶段与稳态阶段，建立基于模型预测控制架构的动态协调控制策略，从而解决行星混联系统高度耦合复杂特性的协调控制问题。最后，开展复合优化算法与动态协调算法集成测试，综合评价行星混联混合动力系统的燃油经济性和平顺性。圆满完成了项目计划书中的全部内容，达到并超过了预期目标，发表文章25篇（均标注本基金项目资助），其中SCI文章6篇，EI文章19篇，授权专利14项，出版中英文专著4部，参与国际/国内会议7次，培养博士/硕士生8人。本项目的研究成果为完善自主开发行星混联动力系统的控制理论做出贡献，也将为科学的优化控制算法走向实际工程应用提供基础理论支持。通过本课题的研究，最终期望有效提升行星混联混合动力客车产品的燃油经济性和纵向平顺性，提高我国混合动力客车产品的综合竞争力。