Optimised Tuning of a PID-Based Flight Controller for a Medium-Scale Rotorcraft
| Autor: | Lindokuhle J. Mpanza, Jimoh O. Pedro |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2021 |
| Předmět: |
0209 industrial biotechnology
Computer science Industrial engineering. Management engineering optimisation PID controller 02 engineering and technology T55.4-60.8 Trim Theoretical Computer Science dynamic modelling 020901 industrial engineering & automation 0203 mechanical engineering Control theory Genetic algorithm genetic algorithm ant colony optimisation Cuckoo search particle swarm optimisation 020301 aerospace & aeronautics Numerical Analysis Computer simulation rotorcraft UAV Particle swarm optimization QA75.5-76.95 cuckoo search Computational Mathematics Nonlinear system Computational Theory and Mathematics Control system Electronic computers. Computer science |
| Zdroj: | Algorithms Volume 14 Issue 6 Algorithms, Vol 14, Iss 178, p 178 (2021) |
| ISSN: | 1999-4893 |
| DOI: | 10.3390/a14060178 |
| Popis: | This paper presents the parameter optimisation of the flight control system of a singlerotor medium-scale rotorcraft. The six degrees-of-freedom (DOF) nonlinear mathematical model of the rotorcraft is developed. This model is then used to develop proportional–integral–derivative (PID)-based controllers. Since the majority of PID controllers installed in industry are poorly tuned, this paper presents a comparison of the optimised tuning of the flight controller parameters using particle swarm optimisation (PSO), genetic algorithm (GA), ant colony optimisation (ACO) and cuckoo search (CS) optimisation algorithms. The aim is to find the best PID parameters that minimise the specified objective function. Two trim conditions are investigated, i.e., hover and 10 m/s forward flight. The four algorithms performed better than manual tuning of the PID controllers. It was found, through numerical simulation, that the ACO algorithm converges the fastest and finds the best gains for the selected objective function in hover trim conditions. However, for 10 m/s forward flight trim, the GA algorithm was found to be the best. Both the tuned flight controllers managed to reject a gust wind of up to 5 m/s in the lateral axis in hover and in forward flight. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|