基于NSGA II算法分布度改进的ATREX发动机性能优化方法研究

Research on Air Turbo Ramjet Expander Performance Optimization Based on Diversity Improved NSGA II Algorithm

关键词:  膨胀式空气涡轮冲压发动机  变比热  化学平衡方法  多目标遗传算法  变工况

Key words:  Air Turbo Ramjet Expander  Variable specific heat  Chemical equilibrium method  Multi-objective genetic algorithm  Off-design condition

摘要:

为使飞行轨迹上膨胀式空气涡轮冲压发动机(Air Turbo Ramjet Expander, ATREX)推力满足飞行器要求，同时比冲为对应推力下的最优值，基于改进分布度的NSGA II算法建立了以推力、比冲为优化目标的ATREX多目标优化模型。本文首先提出了基于个体优化目标间直线距离的筛选函数，改善了NSGA II算法优化结果的分布度；然后基于改进的NSGA II算法建立了ATREX性能多目标优化模型，获得了地面状态发动机推力、比冲最优解。在优化结果分布度接近前提下，与基于原NSGA II算法建立的ATREX性能多目标优化模型对比，基于改进NSGA II算法建立的优化模型所需初始种群个数及迭代时间均下降30%左右。

Abstract:

In order to make the Air Turbo Ramjet Expander （ATREX） meets thrust requirements of the aircraft along the flight path, and at the same time the engine specific impulse is optimal under the corresponding thrust, an ATREX multi-object optimization model is built based on the diversity improved NSGA II algorithm, the engine specific impulse and thrust are selected as the optimization goal. During the modeling process, first, the searching function based on linear distance of individual optimized targets is proposed to improve the distribution of the NSGA II optimization results; then, the ATREX performance multi-object optimization model is built based on the improved NSGA II algorithm, the optimal ATREX thrust and impulse are obtained at sea level static condition. When the distribution of the optimization results is close, compared to the ATREX multi-object optimization model built based on the NSGA II, the initial population number and the convergence speed of the optimization model built based on the improved NSGA II are reduced by about 30%.