机电复合传动多功率流耦合作用机制与最优协同调控研究

Electro-Mechanical Transmission (EMT) system is a feasible solution of hybrid driving for the heavy and off-road vehicles. EMT is splitting and coupling the driving power by the planet gear machine. The multi power has the intense coupling characteristics and its distribution and control is affected markedly by the stochastic driving requirement and the dynamic characteristics of the power component such as the engine and the motors. The coupling mechanism and dynamic response is complex. So the coupling characteristics and stochastic response-based optimal distribution and coordinated control of the multi power are the key for realizing the stability and optimization of system state, and enhancing the performance such as dynamic and economy. Aiming to the core problem of the coordinated control and online optimization of the multi power such as the coupling mechanism and the optimal coordinated control, the electro-mechanical coupling model is built and the coupling mechanism is analyzed and described. The probability prediction method for the stochastic request power of the driving condition is researched based on the statistical learning theory and the prediction model of the multi power states for the local driving condition and power states information are researched. Based on the prediction information and the finite horizon-based optimization ideology, feedforward-feedback compound structure-based optimal coordinated strategy is proposed for the power optimal distribution and dynamic coordination of EMT. The research will provide the theoretic and technical foundation for developing the high-quality EMT system.

机电复合传动是重型或非道路车辆实现混合动力的可行方案，以行星齿轮机构实现功率分流汇流的机电复合传动多功率流具有强耦合性，且受到随机行驶工况需求和动力部件非线性时变输出特性的显著影响，功率流耦合作用机制和动态响应复杂。基于多功率流耦合作用特性和不确定性响应特点，进行功率最优分配与动态协调控制是实现机电复合传动工作状态稳定和优化、提升动力性和经济性等性能的关键。本项目针对机电复合传动多功率流耦合作用机制与最优协同调控等最优能量管理和部件协调控制核心问题，建立机电耦合作用模型分析多功率流耦合作用机制，得到耦合作用表达模型；研究基于统计学习理论的行驶工况随机需求预测方法和部件动态特性预测模型预测局部工况和状态信息；基于预测信息和有限时域优化思想建立复合结构多功率流最优协同调控策略，实现机电复合传动多功率流的最优分配与多动力部件的协调控制。研究将为自主研发高性能的机电复合传动系统提供理论与技术基础。

机电复合传动能够满足重型与非道路车辆调速范围宽、驱动功率大、辅助系统和特定功能系统用电功率大等特殊需求，降低对车用电机的功率要求，具有重要研究意义。但机电功率的分流与汇流通过行星齿轮组组成的功率耦合机构实现，功率流耦合问题给能量优化和功率元件状态调控带来了很大的难度，如果精确、稳定的调控机电复合传动多功率流状态严重影响机电复合传动系统优良的动力、经济等性能实现，制约机电复合传动系统的推广应用。.本项目针对机电复合传动多功率流耦合作用机制与最优协同调控等最优能量管理和部件协调控制核心问题开展研究工作，综合机械转矩与电磁转矩耦合、发电与用电功率耦合、转速转矩等机械参数与电压电流等电气参数耦合等多重耦合作用，建立机电耦合动力学模型，分析了多功率流耦合作用关系与功率分配规律。建立基于高阶马尔科夫与基于工况特征参数学习与识别的工况预测与识别方法，结合动态预测模型，预测机电复合传动的多功率流复杂动态行为。基于工况预测和系统内部动态特性预测结果，建立了基于非线性模型预测控制能量管理作为前馈输入、以鲁棒协调控制策略作为反馈调节的复合结构多功率流最优协同调控策略，实现机电复合传动多能量源的优化管理与多动力部件的协调控制。搭建了机电复合传动分体式台架实验系统和全尺寸实车实验系统，对研究的功率流调控策略进行了台架与试验验证。将研究成果初步应用于国际重工（洛阳）有限公司的机电复合传动高速推土机，为公司相关产品立项、研发提供了技术支撑。.项目执行期内，在国内外学术期刊发表论文30篇，其中SCI检索论文14篇、EI检索论文16篇；项目负责人作为第一或通讯作者的SCI检索论文12篇、EI检索论文12篇，1篇论文获中国精品科技期刊顶尖学术论文奖，1篇论文获中国汽车工程学会优秀青年论文奖；申请并获批中国发明专利2项，申请中国发明专利12项；培养博士生5名，硕士生4名，1人获中国汽车工程学会优秀博士学位论文奖。