Real-time Evaluation of Predictive Fuzzy Terminal Sliding Mode Controller for Quadrotors

This paper introduces a practical solution for quadrotors under the effect of time delays. In actual control systems, time delays are inevitable because of the communication delays, heavy algorithm computations, and sensing delays. These delays have the potential to cause instability and deteriorate system performance if they are not appropriately addressed. In this paper, a Smith predictor is introduced to reduce the negative impacts of time delays and maintain closed-loop stability. However, the Smith predictor's performance may be hindered by modeling inaccuracies, and its effectiveness depends on a precise system model. A robust terminal sliding mode controller with a new sliding surface is suggested as a solution to this problem. Additionally, a fuzzy logic-based switching function is used in place of the discontinuous signum function to lessen noise impacts and chattering. Real-time results using a quadrotor platform show that the proposed methods can effectively control the platform with smaller tracking errors and reduced levels of chattering and noise. Also, it achieves at least a 27% improvement in the tracking error relative to other methods for the worst scenario.