INVERSE NEURAL NETWORK CONTROLLER FOR CAMERA GIMBAL STABILIZATION

Abstract

In each regulation process, independently of the object of interest, we are trying to save as much energy as possible, and to achieve the highest possible quality of the control process. In order to decrease the time for the control algorithms proposal and tune, to increase quality, precision and to make control algorithms „smart“, several new approaches in cybernetics like fuzzy inference logic, artificial neural networks, hybrids systems, genetic algorithms, etc were invented. This article deals with the controlling of a camera gimbal aimed for Unmanned Aerial Vehicles (UAVs) using artificial neural networks and points out the main advantages of neural networks, like the simplicity of their application after training and their ability to control complex dynamic systems.