Faulty PV panel identification using the Design of Experiments (DoE) method

• Any photovoltaic generator dysfunction detected by the Design of Experiment method.
• A mathematical model describes the maximum power versus irradiance and temperature.
• Complete analysis of the measurements performed to optimize the method.
• Non-standard, then standard methods and practical implementation are proposed.
• Results validate the Design of Experiment method for solar energy applications.

This paper presents a novel application of the Design of Experiments (DoE) method as a practical method to detect any dysfunction of an operating photovoltaic (PV) panel. Electrical field measurements data are used to evaluate the PV panel parameters once and for all. A mathematical model describing the variations of the maximum power versus solar irradiance level and temperature was obtained with the Design of Experiments method. This model is a predictive model. It assesses the value of the maximum power delivered anywhere in the experimental domain. Given a known solar irradiance level and temperature, any discrepancy between the calculated and measured maximum power values signals any dysfunctional operation so that appropriate action can be taken such as maintenance or repair.A complete analysis of the measurements is performed to optimize the method. Firstly, the basic elements of the Design of Experiments method are presented. Three models are then considered: the non-standard model with four trials, the non-standard model with 11 trials, and finally a model with 4 points recalculated to be placed at the vertices of the study area. A practical method of implementation is then proposed.