Analysis and Optimization of Maximum Power Point Tracking Algorithms in the Presence of Noise

This paper analyzes the effect of noise on several maximum power point tracking (MPPT) algorithms for photovoltaic systems. Noise is an essential consideration for optimization of MPPT algorithms. For example, for a perturb and observe algorithm, the perturbation size must be big enough to produce a signal larger than the measurement noise; however, the larger the perturbation, the farther the system will deviate from the maximum power point in steady state. There is also a tradeoff with respect to tracking speed, as slower tracking allows for more averaging of the signals, mitigating the noise. The effect of noise and other parameters on tracking performance is quantified, leading to an optimization of the system parameters to provide the best tracking accuracy for a specified tracking speed. The performance of the various algorithms is compared, and the analysis is verified by both simulations and experiments.