The double diode model for photovoltaic (PV) modules is currently less adopted than one-diode model because of the difficulty in the extraction of its seven unknown parameters by, I-PV, I-01, I-02, R-s, R-p, a(1) and a(2), which is a serious inverse problem. This paper proposes application of the Fireworks Algorithm (FWA) for the accurate identification of these unknown parameters in such a way to solve effectively this modeling problem. In particular, firstly, the FWA has been comprehensively tested with two different technologies of Mono-Crystalline (SM55 & SP70) and Multi-Crystalline (Kyocera200GT) PV modules. In addition, further statistical and error analysis for three different panels are exclusively carried out to validate the suitability of proposed methodology. The results of proposed algorithm are benchmarked with popular Genetic Algorithm and Particle Swarm Optimization (PSO) methods. Fitness convergence curves or FWA method for SM55, SP70 and Kyocera200GT produce very less objective function as 2.2498E-07, 2.85765E-08 and 4.0075E-08 respectively. This illustrates the wise and accurate validation of FWA method. Calculated curve-fit via FWA in agreement to datasheet curve strongly suggest the FWA can constitute the core of suitable optimization code for two diode PV parameter extraction. (C) 2016 Elsevier Ltd. All rights reserved.