Maximum power point tracking using Hill Climbing and ANFIS techniques for PV applications: A review and a novel hybrid approach

Abstract

The development of Maximum Power Point Tracking (MPPT) techniques is continuing in order to increase the generated energy from photovoltaic (PV) generators. A variety of MPPT techniques have been proposed and classified based on three main categories: offline, online and hybrid techniques. This paper presents a review of the most popular techniques for offline and online tracking of the Maximum Power Point (MPP), which are the Adaptive Neuro-Fuzzy Inference System (ANFIS) and Hill Climbing (HC) techniques, respectively. This is in addition to a review for all hybrid techniques reported in the literature demonstrating their main merits and shortcomings. Moreover, the present paper combines the ANFIS and HC as a hybrid technique for the first time. The proposed technique involves the features of the ANFIS and HC techniques and mitigates their shortcomings in order to increase the generated PV electrical energy. The proposed technique is a combination of two stages to assess the duty ratio (control signal) being applied to a boost converter for MPP tracking. The first stage includes a set point calculation loop to estimate the duty ratio. The second stage involves a fine tuning loop to determine the exact duty ratio corresponding to the MPP. This achieves maximum power transfer to the load even under nonuniform climatic conditions using a relatively simple control system. The proposed technique has been simulated in MATLAB/SIMULINK environment and compared with some other MPPT techniques (the Constant Voltage (CV), ANFIS, HC, Incremental Conductance (IncCond) techniques) for steady state and rapidly changing climatic conditions (Ropp and sine radiation tests) as well as load variations. The results reveal that the proposed hybrid MPPT technique outperforms other MPPT techniques in term of performances indicators, which include the tracking speed, tracking accuracy and energy gain factor.