Research on Genetic Algorithm-Based Solution Method for Variable Cycle Engine Model

Abstract

Variable cycle engine is a complex system, which is usually mathematically modeled as a series of multi-dimensional nonlinear implicit equations. Processes for solution of these equations are often complicated; therefore, a genetic algorithm-based method was presented in this paper for the solution of the mathematical model. The method was also evaluated by such parameters as initial value sensitivity, computation efficiency, convergence and stability; and compared with Newton-Raphson method. It shows that genetic algorithm-based method is less sensitive to initial values, more capable in convergent and computing stability than Newton-Raphson method, however more time consuming.