【作者】 林歆悠；

【导师】 孙冬野；

【作者基本信息】 重庆大学 ， 车辆工程， 2011， 博士

【摘要】 随着能源紧缺的预警逐渐增强,以及越来越庞大的城市交通燃油消耗带来的巨大压力,成为发展新能源汽车的外在压力和内在驱动力。综合考虑各种新能源技术产业化应用的成熟度及可行性,油电混合动力汽车由于具备很好的继承性和良好的燃油经济性,不仅成为了在现阶段的研究热点,而且甚至在未来的一个时期内均具有有很大应用空间和发展潜力。对于串联式混合动力汽车可改善低速工况时发动机恶劣的燃油消耗,并联式混合动力则在高速工况或高负荷工况具有较强的动力性;然而混联式混合动力汽车兼具了串联和并联的优点,所以更能适应各种各样的行驶工况,尤其适合城市的公交工况。把混联式混合动力系统作为新能源城市公交客车的首选,主要目的是为了节省燃油,提高燃油的利用效率,而欲实现这一目标,进行相应的多能源的协调控制就显得十分重要。对于多能源的控制,亦所谓的能量管理控制策略是一个涉及时变非线性系统控制和复杂问题决策的多纬度模型,其具体因素来自混合动力系统本身及其各部件间的协调工作极其复杂很难使用准确的数学模型进行表示;另外,运行工况及驾驶员操作均具有相当强的随机性,不同的行驶道路和不同的驾驶员风格习惯都需要相应的控制策略与之匹配;换言之,控制策略具有一定的通用性同时又具有独特性,层出不究,是混合动力系统的关键技术之一。本文针对一款混联式混合动力客车以改善其燃油经济性为目的,展开了能量管理优化和控制策略设计,其具体工作概括如下:根据其结构特点,以整车功率匹配及关键部件数值模型、整车动力学模型的建立为基础,制定以期获得燃油经济性的提高并且可应用于实际工程的能量管理控制算法,围绕着实现满足发动机运行于最佳效率曲线并兼顾考虑电池效率的功率均衡分配为核心思想,设计了等效燃油消耗最小的通用优化目标函数,分别进行基于庞特里雅金最小值原理为理论基础的等效燃油最小控制策略进行了功率均衡瞬时优化、动态规划算法进行了控制策略的全局优化,通过了随机动态规划获得了考虑驾驶员特性的功率分配策略并发展了自适应控制策略以期实现对电池能量的有效管理;接着基于优化结果进行统计分析,提取了控制策略的设计规则结合工况识别算法,进一步完善了控制策略;最后,进行了硬件在环实验以及整车道路实验,对所设计的控制策略进行验证。以下对各项工作进行具体介绍:整车功率参数匹配是能量管理控制器开发的基础,基于城市客车行驶工况的特点对混合动力系统功率进行匹配设计并根据匹配结果选定各关键动力部件的进行选型;采用实验为主、理论为辅的方法建立了各个动力部件子系统的数值模型以及基于matlab/simulink建立整车仿真模型,为实现整车能量管理控制策略的开发提供了仿真平台。基于对混联式混合动力系统的分析,制定以实现发动机运行于最佳效率曲线并考虑兼顾电池效率的功率均衡分配基本控制策略模型。但是采用经验式的规则设计方法很难实现这一控制目标,为此引入了优化目标函数。基于最小值优化控制理论推导出等效燃油消耗最小控制策略的理论模型,根据理论模型进行简化以期获得可用于实际应用的实时优化算法,确定电池等效燃油计算模型,以各时刻下客车燃油消耗率最小为优化目标,对电池和发动机功率进行实时优化均衡控制。最后为获得控制策略实现效果的参考依据,以动态规划算法进行功率均衡全局优化,经仿真结果分析表明采用动态规划全局优化控制策略的燃油经济性最好,相比原型车提高了34%;而基于此参考值,所制定的功率均衡等效燃油最小控制策略对改善整车燃油经济性是有效的,对实际工程的应用具有一定的价值。为满足控制策略对不同驾驶员风格的适应性,建立了马尔科夫驾驶员需求功率模型;在此基础上针对混联式混合动力客车,采用随机动态规划算法(又称马尔科夫决策理论)以在每一个SOC和车速的状态下对任何时刻的驾驶员需求功率以及根据统计规律预测的下一时刻的需求功率,以使得当前时刻及未来时刻下的期望成本累加值最小,即燃油消耗最小(这里同样也是采用等效燃油),对混联式混合动力客车动力系统功率分配进行优化,其所获得的发动机和电池之间功率分配的优化结果可直接应用于能量管理控制策略,而且该能量管理策略反映了城市道路行驶工况的特性。随机动态规划所获得的控制策略虽然可直接应用于实时控制,但是其在计算过程中仍然会面临“维数灾难”的问题,亦采用的求解方法所需的计算量和数据存储空间会随着状态的数目呈指数级增长,这对于应用在存储空间不大的控制器是一种致命的限制。为此,将随机动态规划所获得的控制策略拟合成数表并以此为基础引入等效燃油系数,通过分析等效燃油系数对对发动机工作点的影响以及在不同工况下对燃油经济性和电池SOC的影响,利用基于最小值原理对功率均衡控制策略模型进行离线优化,获得在各个等效燃油系数,不同需求功率和车速下的最优电池分配功率,并在此基础上利用自适应模糊滑模控制对等效燃油系数进行更新控制,从而制定了功率均衡自适应时实优化控制策略。通过仿真实验验证,自适应时实优化控制策略可对电池SOC进行有效平稳的控制,表明通过利用等效燃油系数改变电池电量的价值进而实现控制策略对行驶工况的适应性,也说证明了在每个道路工况下均存在一个等效燃油系数能够实现电池与发动机功率之间合理均衡的分配。基于优化算法的结果,进行能量管理控制系统设计。根据动态规划的全局优化计算结果,通过统计分析和多元非线性回归的方法总结最优控制下车辆能量流分配的宏观规律,确定动力总成工作模式的切换规则及各种模式下能量流分配规则;以“人-车-路”思想为指导,设计行驶道路识别及驾驶员意图识别规则并结合能量管理控制策略,形成基于工况识别的功率均衡控制策略。在上述基础上,以“人-车-路”思想为指导,设计行驶道路识别及驾驶员意图识别规则并结合能量管理控制策略,形成基于工况识别的功率均衡控制策略。最后经设计相应的工况进行仿真实验,其结果表明:工况识别的控制策略适用于实际路况的多变性,该方法具有较好的发展空间及应用前景。最后,论文进行了混联式混合动力客车功率均衡控制策略的实验验证。为验证模式切换、串联和并联模式下的功率分配的控制策略的正确性进行硬件在环实验;以动力性能及经济性进行了整车道路实验,最后并提出了对系统进一步的改进方向。实验结果表明:在满足驾驶员的操作需求下,能够实现较好的电池荷电状态维持能力以及提高燃油经济性,其相对原型车提高了23.73%。更多还原

【Abstract】 With the energy becoming less and less, the fuel comsuption of urban transportation is required more and more, therefore this contradiction is turn into the internal drives and exterior pressures to investigate and develop the new energy sources vehicles. The various elements should be taken into the comprehensive consideration, such as the industrialization maturity and feasibility of all kind muti-energy vehicle technology, so far the hybrid electric vehicle provide enough efficiency improvement and keep the character inherit from the traditional vichle, and not only has gradually been the research focus of science and industry but also with great potential and promising prospect. The series hybrid electric vehicles can be prodived with certain adantages to improve the fuel efficiency of engine when operating on low speed condition, and while the parallel hybrid electric vehicle with the excellence to drive powerfully on high speed or laod condition. Logically, the series-parallel electric powertrain with both merit of the series and parallel, so it is adaptive to various driving cycle for a series-parallel HEV, especially, it is suitable no more than as an urban transportation vehicle. The reason to select the series-parallel for the optimal choice is its best fuel economy and high fuel efficiency, and it is very important to to acheave this purpose by coordinated controlling between the muti-energy components. Muti-energy management control strategy is a time varying、nonlinear and multidimensional model involving decision making of complex problem, thery are derived from the complexity architecture of the hybrid powertrain itself and the synergetic operation of different components, it is difficult to construct an accurate mathematical model of the hybrid powertrain; Another reason is the unpredictability of driving conditions and driver’s operation and the difficulty of driving intention judgment resulted from the diversity of driving style enhance the difficulty of the conresponding control strategy for the engineer. In a word, energy management control strategy, as one of the key techiques of HEV, is the algorithm to realize vehicle energy management and power distribution control for the powertrain; it is unique for the corresponding configuration.For the sake of improving fuel economy of a series-parallel hybrid electric bus (SPHEB), this dissertation addresses the vehicle optimal energy managenment and design of the control strategy, and the main contents may be briefly summed up as follows: according to the characteristic of the novel series-parallel architecture, based on the power matching、the theoretical model and data model of key components and kinetic model, an energy management is proposed to expect applied to the engineering practice, which is focus on power balancing distribution control strategy between engine and battery not only advance the engine operating efficiency but also take the battery efficiency into consideration. Aim to the purpose, the equivalent fuel model of battery is implemented and combined with the fuel of engine to constitute the objective function that is to minimize the fuel consumption at each sampled time, an optimal control law is proposed based on Principle of Pontryagin, and then proved to be the theoretic model of equivalent consumption minimization strategy, then a global optimization control strategy based on dynamic programming is proposed. Since there are many factors are great influence on the fuel economy of vehicle in real world, such as unpredictability of driving conditions and diversity of driving style. A Markov model for driver’s power requirements based on statistical data from several driving cycle, an optimal power management control strategy based on stochastic dynamic programming (SPD) is proposed, and then an adaptive supervisory control strategy for the battery management is developed. At last, an integrated control strategy that combined the design rules are derived from the results of DP with the driving pattern recognition approach. To validate the proposed strategy effective and reasonable, simulated test based on a forward model, hardware-in-the-loop test and real-world test are carried out. Power matching and parametric design are the fundament of energy management of HEV to explore, an approach of parameter matching of HEV based on Chinese transit bus city driving cycle is developed, so the major power-train components, such as engine, integrated starter generator, traction motor, battery and final drive gear ratio, are selected to meet the requirements on dynamic performance and economic performance from the results of parameter matching method. A forward simulation model of the series-paralle hybrid electric bus is construsted based on Matlab/Simulink software make use of empirical modeling approach and combine with the aid of theoretical modeling. It provides the essential simulated platform for the exploration of energy management control stragety.Analysis the operation mode of the series- parallel hybrid electric city bus power-train system, and according to the character of its configuration, control strategy model of power balancing distribution between engine and battery not only advance the engine operating efficiency but also take the battery efficiency into consideration is proposed. The equivalent fuel model of battery is implemented and combined with the fuel of engine to constitute the objective function that is to minimize the fuel consumption at each sampled time and to coordinate the power distribution in real-time between the engine and battery. There are three algorithms applied to seek for the optimal fuel economy, they are Principle of Pontryagin, quivalent consumption minimization strategy and dynamic programming. The simulation results indicate that the fuel economy of proposed global optimization control strategy is higher than the instantaneous optimization control strategy, and also advance 34% than the prototype city bus.In order to satify the adaptability of diverse driving style, a markov model for driver’s power requirements based on statistical data from several driving cycle is construsted, and an optimal power management control strategy based on stochastic dynamic programming (SPD), this algorithm consists of two successive steps, namely, policy evaluation and policy improvement, repeated iteratively until convergence. For each possible SOC and velocity state, the policy iteration be intuitively interpreted as the expected cost function value averaged over a stochastic distribution of drive cycles starting at that state. The obtained control law is in the form of a stationary full-state feedback and can be directly implemented.Although overcome the limitation of DP algorithm and SDP method will still suffer from“curse of dimensionality”in EMS design.That is,the computation and memory needed by value iteration and policy iteration will increase exponentially with the number of states.This will limit its application in engineering.To deal with this problem, the method ofλestimation is introduced. The method is the Equivalent fuel Consumption Minimization Strategy based on driving pattern recognition in essence. The main idea of adaptive real time control strategy is periodically updating the equivalence factor dependence on the corresponding driving condition. It is assumed that information about the route is available in advance. Using this knowledge, global optimization methods can be used in real-time control to approach optimal fuel consumption while keeping the state of charge (Soc) of the batteries at a desired level. The measured fuel consumption and the obtained battery Soc trajectory demonstrate good performance of the proposed adaptive control.An approach of designing SPHEB real-time energy management strategy was derive from the offline global optimization results. The Macro-distribution rules of SPHEB powertrain energy flow under various optimal control and the powertrain operating mode switching rules and power distribution rules were designed, throuth the method of statistical analysis and multivariate nonlinear regression. From the correlation analysis of regression formula, there is a conclution that: the formula regressed by the Levenberg-Marquardt algorithm, whose calculated values was highly relevant to the optimized values, could be used to make an engine operating map for optimizing the distribution of powertrain energy flow. Under the guidance of the concept of“driver-vehicle-road”, the integrated control strategy which is constitued by the rules of driving pattern recognition and driver intention combined with the energy management contrl strategy designed above. The simulation results show that the driving pattern recognition based control strategy can be adaptive to the various driving cycle and indicate a good application prospect.Last, the dissertation carried out a series of tests to validate the proposed power-balancing control strategy for series-parallel hybrid electric bus. Both the hardware in the loop simulation and real-world test are adopted to collect the information about fuel consumption and drive performance. And some opinions for further improvement of performances of the system were given. The conclutions of the test are given as following: under the control of proposed strategy, the series-parallel hybrid electric bus not only can be satisfied the drive performance but also achieve the improvement of fuel economy by comparing with the prototype bus up to 23.73%.更多还原