Dynamic aerodynamic parameter estimation using a dynamic particle swarm optimization algorithm for rolling airframes
- PDF / 2,059,549 Bytes
- 16 Pages / 595.276 x 790.866 pts Page_size
- 4 Downloads / 276 Views
(2020) 42:579
TECHNICAL PAPER
Dynamic aerodynamic parameter estimation using a dynamic particle swarm optimization algorithm for rolling airframes Ayham Mohamad1 · Jalal Karimi1 · Alireza Naderi1 Received: 27 May 2020 / Accepted: 1 October 2020 © The Brazilian Society of Mechanical Sciences and Engineering 2020
Abstract The aerodynamic parameters of each flying vehicle dynamically change along its flight profile, because of aerodynamic parameter relationship with flight conditions, and several flight conditions take place during each flight profile. Therefore, in this research, the concept of dynamic aerodynamic parameter estimation (DAPE) is introduced. A two-step strategy is used: In the first step, the aerodynamic forces and moments are estimated; then, after passing through a designed smoothing filter, in the second step, the DAPE is converted to a dynamic optimization problem and solved by a heuristic optimization algorithm that hybridizes the features of particle swarm optimization in tracking dynamic changes with a new evolutionary procedure. Two new algorithms are developed: DAPE and SDAPE. In DAPE algorithm, all aerodynamic parameters are estimated at once by solving a single optimization problem. In SDAPE algorithm, four separate optimization problems are solved. A rolling airframe is the plant studied in this research. Simulation results indicate that SDAPE is better than DAPE in terms of accuracy. Comparing the performance of the newly proposed algorithms with that of three state-of-the-art static optimization algorithms and extended Kalman filter reveals their less run time and acceptable accuracy. Keywords Dynamic aerodynamic parameter estimation · Dynamic optimization · Estimation after modeling · Particle swarm optimization · Smoothing filter
1 Introduction Derivation an exact aerodynamic model for flying vehicles is extremely needed for several tasks like flight dynamic analysis, flight control system design and flight simulation. The aerodynamic parameters are largely affected by flight conditions. Therefore, a realistic aerodynamic model is a multi-dimensional lookup table that is a function of various dimensionless parameters like Mach number, Reynolds number and so on. The aerodynamic coefficients are derived via several ways. Theoretical–empirical methods and computational fluid dynamic (CFD) method, because of their simplified assumptions, may be used in conceptual design analysis. Experimental approaches like wind tunnel tests and flight tests are used for detailed design applications. Technical Editor: Jader Barbosa. * Jalal Karimi [email protected] 1
Department of Aerospace Engineering, Malekashtar University of Technology, Tehran, Iran
The wind tunnel tests have limitations in simulating vehicle’s whole flight envelope and have uncertainties arising from scaled models and wind tunnel walls effect. The most appropriate approach for aerodynamic model derivation is to estimate aerodynamic parameters via flight test data. During last decades, based on flight test data, several appr
Data Loading...
