Evolving fuzzy inference system based online identification and control of a quadcopter unmanned aerial vehicle

Among different rotary wing unmanned aerial vehicles; quadcopters are used most commonly in both civil and military sector. Comparatively higher portability, smaller size, simple method of assembly and reconstruction and lower expenditure have caused the rapid growth of the quadcopter. Precise mathematical model of quadcopter considering uncertainties is necessary for the better performance of the commonly used first principle based controllers. However, modelling quadcopter incorporating uncertainties is hard to achieve. A solution to the problem is the utilization of modelfree data-driven methods. An evolving intelligent system (EIS) based data-driven technique called evolving Takagi-Sugeno fuzzy inference system (eTS) is employed in this paper for online identification of a quadcopter from realtime experimental data. Besides, an adaptive fuzzy controller is developed using fuzzy c means clustering to control the altitude of the quadcopter.

I. Introduction

Unmanned aerial vehicles (UAVs) are aircraft with no aviator on board. UAV autonomy varies from partial to complete, which begins from the human operator based partial remote control to fully autonomous control by onboard computers. Autonomy enables UAVs to perform some tasks very well where human involvement would be dangerous, or too tedious. Besides, they are light in weight, fuel efficient and their maintenance is easier than their manned counterparts. Therefore, UAVs are becoming popular day by day with enormous applicability in both military and civil sectors. In relation to the wing types, UAVs are usually classified into three subdivisions, and they are namely: 1) fixed wing, 2) rotary wing, and 3) flapping wing. The rotary wing UAVs (RUAVs) can be further classified by the number of rotors like a helicopter, quadcopter, hexacopter, octocopter, etc. Among various RUAVs, the most commonly used is the quadcopter [1]–[3] since hexacopter and octocopters are comparatively expensive and cumbersome.