【作者】 王俊；

【导师】 王庆年；

【作者基本信息】 吉林大学 ， 车辆工程， 2015， 博士

【摘要】 混合动力车辆控制是车辆性能的保障与核心，如何让车辆满足动力性需求同时降低油耗一直是新能源汽车研究人员的研究重点，据统计来看大量混合动力相关论文均是集中于混合动力的各种控制算法进行研究和探讨。笔者将稳态下混合动力能量管理策略分为控制策略设计、控制规则制定以及控制参数优化三个部分，进行系统性的介绍AMT混合动力客车的设计方法。同时将瞬态下的协调控制策略分为模式切换协调控制与换挡协调控制中这两个部分介绍，最终解决混合动力车辆稳态与瞬态的控制问题。针对混合动力系统参数匹配问题，本文提出以动力性需求作为约束条件，燃油经济性作为优化目标，采用优化算法进行动力系统的Map选型，解决了仅依据动力系统功率匹配的不足。与此同时，本文搭建了整车仿真模型，将其仿真结果与商业Cruise仿真模型结果对比验证，来确保后续仿真优化的精确度与可信度。混合动力系统的运行是一个随机的复杂过程，传统的基于规则设计方法很难满足车辆较优的性能需求。针对该系统的优化控制问题，提出将优化算法应用于混合动力车辆控制策略开发中，通过将车辆燃油经济性优化问题简化为基于时间历程的数学函数优化问题，制定函数目标、约束条件与计算方法进行优化计算获取最优解。本文分别根据瞬时最优与全局最优两种求解方法来探讨车辆的优化控制问题，其控制与优化结果为后续的控制规则提取提供了有效的依据。针对混合动力控制策略规则设计问题，本文提出一种基于优化算法的控制规则设计方法，通过结合上述应用的全局优化算法与瞬时最优算法各自优点，分别提取模式切换规则、换挡规则及转矩分配规则，根据AMT混合动力系统的结构特点，确定换挡控制与转矩分配的优先级，实现AMT混合动力系统控制规则的集成，为逻辑门限控制算法的规则制定提供有效的指导意义，同时提出的控制规则在实车上应用时具有较好的实时性与稳定性。该方法解决了传统控制规则设计依靠设计者经验和试凑法导致花费大量的时间重复调试的问题。针对混合动力控制参数优化问题，本文提出基于迭代算法的控制参数优化方法，通过将待优化参数编码、设定参数区域、优化目标等，启动优化算法自动计算获取最优的控制参数组，为控制参数优化提供了新的思路。该算法解决了传统控制参数的标定方法耗时，没有理论依据，往往给标定工程师带来很大的工作量，当控制参数数量增加时，其调试的复杂度也呈指数增加等问题。同时本文提出了基于滚动时间窗的工况识别算法可根据工况识别结果进行实时更新控制参数，实现能量管理控制参数自适应调整，解决了单一控制参数针对复杂实际工况的局限性。对于混合动力车辆因模式切换导致的冲击度及换挡动力中断等问题，本文提出采用模式切换协调控制解决驱动模式切换时产生的车辆冲击度，采用无离合器操作的换挡控制技术解决AMT变速器换挡过程中的动力中断等问题。针对离线开发控制算法在实时系统上应用的控制性能问题，本文提出硬件在环仿真系统，通过将dSPACE控制器作为车辆环境模拟、TTC200控制器作为实际控制器，并搭建CAN网络等。经过实验结果知，整车控制算法实时性好、可行性高，为后续的实车验证提供前提与保障。最后采用本课题组的试验台架进行功能性验证。更多还原

【Abstract】 Hybrid electric vehicle control is key technology to optimize vehicle performances.How to reduce vehicle fuel consumption and meet the power demand, these sections havebeen the researchers’ focus. According to the statistic results known a large number ofhybrid electric vehicle related papers are more focused on hybrid control algorithmstudying. In this paper, the steady state hybrid energy management strategies weredivided into three parts of control strategy designing, control rules making, and controlparameters optimization to introduce hybrid powertrain design methods. Meanwhile, thecoordinated control strategies in transient condition were divided into transient modeswitching control and gear-shifting coordinated control, so as to solve the vehicle controlproblems under the steady-state and transient conditions.For the parameters matching problems of hybrid powertrain system, in this paper,power requirement as constraints condition, fuel economy as the optimization objective toselect the best powertrain Map by optimization algorithm, which avoid the shortage ofpower matching for hybrid powertrain. Meanwhile, vehicle simulation model was built,its simulation results were compared with commercial Cruise vehicle model results toensure the subsequent simulation accuracy and reliability.Hybrid powertrain system of the running process is a complex stochastic process, itis difficult to meet the vehicle performance demand with the traditional rule-based designmethodology. For the optimal control problems of hybrid powertrain system, optimizationalgorithms were used in the vehicle control strategy development. Vehicle fuel economyoptimization problem was simplified as the mathematical function optimization problembased on time-history, formulate function goals, constraints, optimization and calculationmethods to obtain the optimal solution. In this paper, instantaneous and globaloptimization methods were provided to solve optimal vehicle control problem, its controland optimization results provides effective evidence for subsequent control rulesextraction.For the control rules problems of hybrid electric vehicle, this paper proposes acontrol rules designing method that combines the advantages of instantaneous and globaloptimization algorithms to extract mode switching rules, gear-shifting rules and torquedistribution rules. According with the hybrid powertrain characteristics to determine thepriorities of gear-shifting control and torque distribution and achieve system integration control, which provide effective guidance for logic threshold control rules designing andprovide the good real-time and stability of control rules for application in real vehicle..Meanwhile, these methods avoid the shortage of traditional design methods that dependon designer experiences and trial and error method takes a lot of time to repeatdebugging.For the control parameters optimization problems of hybrid electric vehicle, in thispaper, iterative optimization method was proposed to encode the optimization parameters,set the parameters areas, optimization goals to obtain the optimal control parameters,which provides a new ideal for the control parameter optimization. These methods solvethe traditional control parameters calibration method time consuming and no theoreticalsupports, the complexity increases exponentially when the control parameter increases.Meanwhile, this paper proposes driving cycle recognition algorithm based on the rollingtime window to update the control parameters and achieve adaptive controlling. Thesemethods avoid the limitations of single control parameter for actual driving cycle.For the problems of hybrid electric vehicle jerking of mode transition and powerinterruption of gear-shifting, this paper provides the coordinated control algorithm ofmode switching to decrease vehicle jerk, and no-clutch shift operation control technologyto avoid the power interruption of transmission gear-shifting.For verifying the performance of offline control algorithm was used in the real-timesystem. In this paper, hardware-in-the-loop simulation system was proposed, dSPACEcontroller as a vehicle environment simulation, TTC200controller as the actual controllerand CAN networks and other structures were building. According with experimentalresults known, vehicle control algorithms have good real-time performance and highfeasibility, providing the prerequisite and guarantee for subsequent verification of theactual vehicle. Finally, test bench experiment was provided for functional verification.更多还原