变循环发动机模型的求解算法研究

采用部件法建立了变循环发动机的多维非线性隐式方程组模型,模型具有隐式性,因而求解过程复杂,收敛困难.针对上述问题,采用遗传算法和牛顿-拉夫逊法进行模型求解,提出了算法的有效性评价指标:初值敏感性、收敛性和稳定性.对两种算法进行有效性评价.针对发动机工作性能最优化问题,建立了多目标模型,用遗传算法进行求解,并分析了相关变量随马赫数变化的规律.结果可为变循环发动机模型求解算法的选择与设计提供参考.

Research on Solution Algorithm for Variable Cycle Engine Model

A component method is proposed to build a variable cycle engine model of multidimensional nonlinear implicit equations that are not convergent,which make it difficult to be solved.Thus,Genetic Algorithm and Newton-Raphson are adopted to solve the model and some evaluation indices such as initial value sensitivity,computation efficiency,convergence and stability of algorithm are proposed to evaluate the two methods in the paper.A multiple object model is constructed for engine working performance optimal problem which is solved by Genetic Algorithm,change rule that relative variables with mach numbers is found.The results and conclusions in this paper provide the reference for modeling and solution of variable cycle engine.