An Improved PSO Algorithm with Object-Oriented Performance Database for Flight Trajectory Optimization

Abstract

In order to improve accuracy and convergence speed for flight trajectory optimization program in flight management computer and enhance its maintainability, an improved particle swarm optimization (PSO) algorithm with object-oriented performance database is proposed. Firstly, an object-oriented performance database is built by Microsoft Visual C++ and MATLAB/SIMULINK mixed software development environment. Through synthetically use class hierarchy and specialized function library, the flight performance data is retrieved and its data file can be replaced with adapt for different aircraft types. Secondly, the mass point motion mathematical model is built according to mass point dynamics and energy states. Objective functions for trajectory optimization in vertical flight profile are acquired by the Minimum Principle of Pontryagin. Thirdly, adaptive inertia weight is introduced, the equality constraints is processed using the penalty function method. Finally, trajectory in vertical flight profile is optimized through using the improved PSO based on the object-oriented performance database. Meanwhile, the PSO algorithm flow for vertical flight profile trajectory optimization is given. Through using of the improved PSO, trajectory optimization of Boeing 737-800 aircraft in vertical flight profile is carried out. Comparison results between optimization results and flight test data show that the calculated results of proposed algorithm rapidly converges to optimal solution with higher precision.