Abstract
A global maximum power point tracking (GMPPT) process must be applied for detecting the position of the GMPP operating point in the minimum possible search time in order to maximize the energy production of a photovoltaic (PV) system when its PV array operates under partial shading conditions. This paper presents a novel GMPPT method which is based on the application of a machine-learning algorithm. Compared to the existing GMPPT techniques, the proposed method has the advantage that it does not require knowledge of the operational characteristics of the PV modules comprising the PV system, or the PV array structure. Additionally, due to its inherent learning capability, it is capable of detecting the GMPP in significantly fewer search steps and, therefore, it is suitable for employment in PV applications, where the shading pattern may change quickly (e.g., wearable PV systems, building-integrated PV systems etc.). The numerical results presented in the paper demonstrate that the time required for detecting the global MPP, when unknown partial shading patterns are applied, is reduced by 80.5%–98.3% by executing the proposed Q-learning-based GMPPT algorithm, compared to the convergence time required by a GMPPT process based on the particle swarm optimization (PSO) algorithm.
Highlights
The economic competitiveness of photovoltaic (PV) systems, compared to conventional power production technologies, has continuously improved during the last years
During the execution of the PV global maximum power point tracking (GMPPT) process, each state depends on the current value of the duty cycle of the DC/DC power converter (Figure 2), the power generated by the PV array and the duty cycle during the previous time-step [28]: n o
The variations of duty cycle and PV array output power versus time for shading the agent had previously acquired enough knowledge from the learning process performed during the GMPPT execution for shading patterns 1–5
Summary
The economic competitiveness of photovoltaic (PV) systems, compared to conventional power production technologies, has continuously improved during the last years. MPPT method [e.g., perturbation and observation (P&O), incremental-conductance (INC) etc.], which is executed after the evolution algorithm execution has been finished, in order to: (i) fine-tune the PV source operating point to the GMPP and (ii) maintain operation at the GMPP during short-term changes of solar irradiation or shading-pattern, which do not alter significantly the shape (e.g., relative position of the local and global MPPs) of the power–voltage curve (e.g., [15]) This enables to avoid frequent re-executions of the evolutionary algorithm, which would result in power loss due to operation away from the GMPP during the search process. The learning rate determines how much the new knowledge acquired at + 1 by the agent will affect the existing estimate in the update of Q(St , at )
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have