Even with the consistent growth in global photovoltaic (PV) capacity, the necessity for fault detection in PV systems has not been widely addressed regardless of its importance. With International Electrotechnical Commission, Institute of Electrical and Electronics Engineers, and National Electric Code protection standard recommendations for PV systems being vulnerable to fault occurrences, advanced detection techniques are inevitable in PV systems to avail guaranteed protection from electric shocks and fire hazards. Even though numerous fault detection techniques have been conceptualized in last few years, these techniques are yet to be classified and quantified in a common platform. Hence, this paper takes up an initiative to study, classify and analyze the advanced fault detection approaches available in literature. Each fault detection technique is segregated based on the detection approach and are reviewed with respect to: types of faults detected; detection time; sensor requirement; procedural complexity; detection variables; and level of protection achieved. Furthermore, a compatibility study is conducted to evaluate the effectiveness of advanced fault detection techniques for protection against line-line, line-ground, and arc faults that are most common in PV systems. Overall, this investigation serves as a valuable reference for researchers to improve fault detection possibilities in PV systems.

Non-homogeneous irradiation conditions due to environmental changes introduce multiple peaks in non-linear P-V characteristics. Hence, to operate PV at the global power point, numerous algorithms have been proposed in the literature. However, due to the insufficient exploitation of control variables, all the MPPT methods presented in literature fail to guarantee Global Maximum Power Point (GMPP) operation. In this paper, a new detection technology to identify global MPP zones using hybrid Enhanced Leader Particle Swarm Optimization (ELPSO) assisted by a conventional Perturb and Observe (P&O) algorithm is proposed. With inherent mutations, ELPSO applied to MPPT excels in exploring global regions at initial stages to determine the global best leader; whilst, P&O is reverted back soon after global solution space is detected. The transition from ELPSO to P&O is mathematically verified and allowed only when ELPSO finds the global optimal zone. Adapting this hybrid strategy, the proposed method has produced interesting results under partial shaded conditions. For further validation, the results of the proposed hybrid ELPSO-P&O are compared with conventional ELPSO and the hybrid PSO-P&O methods. Experimental results along with energy evaluations confirmed the superiority of the ELPSO-P&O method in obtaining the maximum available power under all shaded conditions.

Optimizing the operational performance of Maximum Power Point trackers (MPPTs) during multiple peak partial shading Conditions (PSCs) prevails as a major challenge in PV power generation. Though many of the newly evolved soft computing and heuristic methods are compatible during PSCs, the need for such techniques in uniform irradiation levels is uncertain because most of these algorithms produce extensive oscillations before converging to the global maximum power point (GMPP). On the other hand, adopting conventional Perturb and Observe (P&O) algorithm is meritorious to reduce the transient power and voltage oscillations. Therefore, this works presents a proficient hybrid tracking technology that provides an adequate trade-off between conventional P&O and advanced soft computing techniques by accurately detecting shade occurrences. For which, the uniqueness in the operating point conductance of P&O at the left most power peak (LPP) in the P-V curve is utilized. Subsequently, the proposed convention utilizes P&O for two major purposes: 1) to track MPP in uniform irradiance and 2) to detect PSCs. More importantly, even during PSCs, the detection algorithm is well designed to operate at GMPP with conventional P&O method itself, and, only for extreme shade cases, Flower Pollination Algorithm (FPA) is introduced to track GMPP.

The traditional protection devices installed in Photovoltaic (PV) arrays generally detect Line-Line (LL) and Line-Ground (LG) faults, when the fault current magnitude exceeds its threshold value defined by various international standards. However, the magnitude of fault current is greatly reduced, due to low irradiance levels, active maximum power point tracker, location of fault, minimal fault mismatch and presence of blocking diodes. Consequently, majority of such faults remain obscured even when the irradiance reaches to a higher level and thereby constitute to reliability issues and severe fire risks. Therefore, both timely fault detection and localization become highly obligatory for sustainable power generation and safety. Thus, this paper proposes a new, robust and efficient fault localization method based on the principle of differential voltage measurement between PV modules of adjacent strings. For accomplishing this task, a new optimized voltage sensor arrangement with minimal number of sensors is followed. Moreover, the proposed convention i) detect LL/LG faults independent of its detection challenges ii) suits grounded/floating PV systems and iii) compatible with/without blocking diodes. For a realistic validation, testing has been performed on a small-scale grid-connected PV system and the efficacy in detecting various array faults are demonstrated via extensive investigations.