基于试验设计-遗传算法的船用柴油机冷却系统多目标优化

Multi-Objective Optimization of Marine Diesel Engine Cooling System Based on DOE-GA

关键词:  遗传算法  冷却系统  船用柴油机  耗油率  有效功率

Key words:  Genetic algorithms  Cooling system  Marine diesel engine  Fuel consumption rate  Effective power

摘要:

为了获取船用柴油机冷却系统的最佳运行参数，在柴油机可靠运行的前提下，尽量提升动力性和经济性。建立了柴油机工作过程-燃烧室-冷却系统耦合仿真模型，提出试验设计-遗传算法优化算法，该算法具备节省试验成本和多目标全局寻优的优势。选取柴油机的6个典型推进工况点进行研究，采用该算法对冷却系统运行参数开展多目标优化设计，优化设计以海、淡水泵转速为变量因子，以提升动力性参数和经济性参数为目标，以涡前排温、淡水出机温度和峰值缸压为约束条件，优化后，在低负荷点，淡、海水泵节省功耗79.4%和71.73%，扭矩提升7.2%，有效功率提升7.6%，热效率提升8%，耗油率降低6.73%，摩擦功降低9.71%，峰值缸压降低5%。研究结果表明：耦合仿真模型与实机更加贴近；试验设计-遗传算法在解决强耦合、非线性系统的多目标优化问题具备成本低、效率高、精度高的优点。

Abstract:

In order to obtain the best operation parameters of the cooling system of a marine diesel engine， achieving the goal of improving the power and economy under the premise of reliable operation of diesel engine， diesel engine working process-combustion chamber-cooling system coupling simulation model was established， and DOE-GA optimization algorithm was proposed which had the advantages of saving test costs and multi-objective global optimization. Six typical propulsion conditions were selected for research， DOE-GA optimization algorithm was used to carry out multi-objective optimization design on the operating parameters of the cooling system. The optimal design took the speed of sea and fresh water pumps as variable factors， and the promotion of dynamic and economic parameters as optimization goal， with turbine inlet exhaust temperature， fresh water outlet temperature， and peak cylinder pressure as constraints. After optimization， at the low load point， the fresh water and sea water pumps saved power consumption by 79.4% and 71.73%， torque increased by 7.2%， effective power increased by 7.6%， thermal efficiency increased by 8%， fuel consumption rate decreased by 6.73%， and friction work decreased by 9.71%， peak cylinder pressure decreased by 5%. The research results show that the coupled simulation model is closer to the real machine. The DOE-GA has the advantages of low cost， high efficiency， and high accuracy in solving multi-objective optimization problems of strongly coupled nonlinear systems.