Contents I. Introduction
The last few years have seen increased deployment of solar photovoltaic generation due to the high energy prices being achieved in the electricity market. In order to reduce the price impact and the increase of energy consumption, the European Commission has deployed a support package to try to reach a 32% renewable energy consumption by 2030, with wind energy in northern Europe and solar energy in southern Europe [1]. It also foresees that by 2050, almost half of European households will be generating energy from renewable sources, and in particular, solar PV due to the ease of deployment compared to wind. Although this generation technology is well advanced, further research is needed in other application areas, such as the introduction of the intelligence necessary to make this technology increasingly reliable, efficient and resilient. This issue can be addressed with DT concept, an integrated system that allows simulating, monitoring, regulating and control the state of the photovoltaic (PV) generation system [2]. In order to implement a DT, the main elements of the plant under study must first be characterized. In the case under study, photovoltaic generation systems can be characterized by the following elements: i) photovoltaic panel; ii) DC/DC converter to extract the maximum power using MPPT algorithms; iii) DC/AC converter to convert the production of the photovoltaic panel from DC to AC, so that the energy can be injected into the grid; iv) protection elements of the installation. Of these elements, and related to DT implementation, the power electronics converters have the most challenging execution time as they have switching devices that process the PV power and track maximum MPPT power (DC/DC) running at high frequency. Therefore, in order for the DT to run at the same speed as the power electronics converter (5kHz-20kHz switching frequency), it is necessary for the PV panel model to run at the same or higher speed than the power electronics converter or to use different sampling times [3]. Some works has been found in the literature that has tried to implement a model of a PV panel or string of PV panels in devices such as an FPGA, Single Digital Controller (DSC) or System on Chip (SoC), but not in a low-cost micro-controller, as needed for a IoT solution. A PV string model based on the Lambert W function on the SoC Zynq-7000 XC7Z020 to obtain the electrical parameters of a string of 30 PV panels, working under different temperature and irradiance conditions, is implemented in [3]. On the other hand, in [4] a new method to obtain the electrical parameters of a string panel under mismatch conditions is proposed. This method, called Model by Zone (MbZ), alternates between the Single Diode Model, linear model and constant voltage model. In addition, they implement the model in an FPGA, Intel Arria 10AS066K3F40E2SG clocked at 50MHz, which has a high cost that is not affordable in kW PV plants. Finally, in [5] a simulator of a PV string, with fast dynamics and high performance is proposed to obtain a model that can be connected to a power electronics converter. The chosen platform to implement the model is eZdspF28335 DSC-based control system, which also has a cost that is not affordable for a PV panel DT.
