Modeling, Optimization and Real-time Optimal Control of Hybrid Electric Vehicles and Marine Vessels

日益增长的环境问题和降低能源消耗的需要促进了混合动力电力推进系统技术的快速发展。拟议的研究计划旨在解决几个根本性问题，这些问题阻碍了该技术的进一步进步，并将其更广泛地应用于重型车辆和海上船舶。研究将引入先进的建模、设计优化和驾驶员或任务自适应实时最优控制技术，使混合动力技术充分发挥其性能、能源效率、减排和全生命周期成本节约的潜力。这项研究将在三个密切相关的领域进行：a)开发基于驾驶员或船舶的实际操作而不是目前固定的标准驾驶周期进行优化的动态、实时最优控制的知识和技术；b)为混合动力船舶的优化设计和控制开发建立新的方法和建模工具平台；以及c)为下一代插电式混合动力汽车和船舶(PHEV/PHES)的设计和控制优化形成基于元模型的全局优化(MBGO)技术的透彻理解和系统的算法/工具开发。基于出行、驾驶员自适应的智能最优功率控制和能量管理技术，其自学习能力将基于系统采集的运行数据的车船运行模式识别，以及考虑电池性能退化的混合动力系统模型的离线最优控制方案生成。混合动力船舶推进系统的建模研究将填补目前研究和工业实践中存在的空白，为设计和控制优化提供一个具有适当复杂性和保真度的集成顶层系统模型。对降阶船舶阻力和螺旋桨推力模型的改进和验证将成为建模研究的一部分。这些复杂的设计和控制优化问题形成了计算密集型的黑盒全局优化问题，需要更高效、更健壮、更高维的全局优化技术。对先进的MBGO理论和算法的继续研究将满足这一需求。提出的研究将混合动力系统模型与高级优化相结合，形成新的基于模型的设计与优化技术。该研究计划将提高我们对混合动力推进和优化技术的理解和能力，开辟新的研究领域，为大量具有尖端研究和实践经验的HQP提供宝贵的培训，并满足行业的迫切需求。

Growing environmental concerns and the need to reduce energy consumption have given rise to rapid advances in hybrid electric propulsion system technology. The proposed research program is aimed at addressing several fundamental issues that are hindering the further advance of the technology and its broader transportation applications to heavy-duty vehicles and marine vessels. The research will introduce advanced modeling, design optimization, and driver or mission adaptive real-time optimal control techniques to allow the hybrid propulsion technology to reach its full performance, energy efficiency, emissions reduction and life-cycle cost saving potentials. The research will be carried out in three closely related areas: a) developing the knowledge and techniques for dynamic, real-time optimal controls that are optimized based on the actual operations of a driver or a ship, not on the fixed standard driving cycles as present; b) establishing the new methodology and modeling tool platform for the optimal design and control developments of hybrid electric marine vessels; and c) forming a thorough understanding and systematic algorithm/tool development of Metamodel Based Global Optimization (MBGO) techniques for the design and control optimizations of the next generation Plug-in Hybrid Electric Vehicles and Ships (PHEV/PHES). The trip-based, driver adaptive intelligent optimal power control and energy management techniques, and its self-learning capability will be based on vehicle and ship operation pattern identification from systematically acquired operation data, and off-line optimal control plan generation using the hybrid propulsion system model with battery performance degradation consideration. The hybrid electric marine propulsion system modeling research will fulfill the void existed in present research and industrial practice to produce an integrated top-level system model with appropriate complexity and fidelity for design and control optimization. Improvement and validation of reduced-order hydrodynamic ship drag and propeller thrust models will form part of the modeling research. These complex design and control optimization problems form computationally intensive black-box global optimization problems, and calls for more efficient, robust and high dimensional global optimization techniques. The continued study on advanced MBGO theory and algorithms will meet this need. The proposed research will combine the hybrid propulsion system model and advanced optimization to form the new Model Based Design and Optimization technology. The research program will improve our understanding and ability to utilize hybrid propulsion and optimization technologies, open new research areas, provide invaluable training for a large number of HQPs with cutting edge research and hands-on experiences, and address the urgent needs from industry.