基于改进粒子群算法的航空发动机状态变量建模

为了克服现有航空发动机状态变量建模过程中的不足,采用了一种改进粒子群算法建立航空发动机状态变量模型。首先改进了粒子群算法,提出一种每个粒子根据自身适应值动态调整其惯性系数方法来平衡搜索性能;对群体最优位置进行实时的代内更新以提高搜索速度;为避免陷入局部最优,在最优个体附近进行随机搜索。其次利用该算法建立航空发动机状态变量模型,根据航空发动机在稳态点处的线性化模型应与在该同一稳态工作点处的非线性模型响应一致的原则构造适应值函数,仿真结果表明所建立的状态变量模型不论是稳态过程还是动态过程都与非线性模型响应基本一致,建模精度较高,建立过程简便。

Aero-Engine State Variable Modeling Based on the Improved Particle Swarm Optimization

For overcoming some insufficiencies of establishing the state variable model of aero-engine,the state variable model of aero-engine was proposed based on improved Particle Swarm Optimization(PSO).Firstly,the improvements of PSO were proposed.Each particle could adjust inertia coefficient dynamically based on own fitness value to balance search performance.The best position of the colony was updated in each generation real time.The random search was carried out near by the best individual to avoid plunging into local optima.Secondly,the improved algorithm was used to establish the aero-engine state variable model.The fitness function of PSO was established according to the principle that the aero-engine responses of the linear model should be in accordance with that of the nonlinear model at the same steady working point.The simulation results show that the established state variable model has the same responses as the nonlinear model for both the steady process and the dynamic process,and has high accuracy.And the established process for the model is simple and convenient.