论文部分内容阅读
为了降低变循环发动机模型求解时对初始值的依赖性,提升算法的全局收敛性,同时提高模型求解的效率,提出了一种基于改进的混合粒子群算法的变循环发动机模型求解思路。首先建立了变循环发动机的部件级模型,并建立了发动机的共同工作方程组;然后采用Broyden法对牛顿-拉夫森算法中的雅可比矩阵进行更新计算,在经典粒子群算法的基础上引入粒子中心,作为干扰项,并引入限制因子和自适应时变惯性系数;最后,综合了两种改进的算法,提出改进的混合粒子群算法。实验结果表明:该算法不仅继承了牛顿-拉夫森算法的高计算效率,还吸收了改进的粒子群算法的全局收敛优点,可实现模型大范围收敛。
In order to reduce the dependence on the initial value of the variable cycle engine model and improve the global convergence of the algorithm and improve the efficiency of the model solving, a new idea of variable cycle engine model based on improved hybrid particle swarm optimization is proposed. Firstly, a component-level model of a variable cycle engine is established and a common working equation set of the engine is established. Then, the Jacobian matrix in Newton-Raphson algorithm is updated and calculated by Broyden’s method. Based on the classical particle swarm optimization algorithm, Center, as the interference term, and introduces the limiting factor and the adaptive time-varying inertia coefficient. Finally, two improved algorithms are synthesized and an improved hybrid particle swarm optimization algorithm is proposed. The experimental results show that the new algorithm not only inherits the high computational efficiency of Newton-Raphson algorithm, but also absorbs the advantages of global convergence of the improved particle swarm optimization and can achieve large-scale convergence of the model.