An IoT– and Mobile Network–Based Architecture for Monitoring of a Solar PV System

In the context of the gradual depletion of traditional energy sources, solar energy has emerged as a sustainable and environmentally friendly alternative. However, harnessing this renewable resource effectively requires low-cost smart monitoring systems. This paper introduces an Internet of Things (IoT)- and mobile network-based architecture for photovoltaic (PV) power plants, leveraging a cost-effective and scalable IoT framework. The proposed real-time data acquisition and monitoring solution is implemented through two interconnected subsystems. The first subsystem is responsible for gathering all electrical parameters of the PV system, including active/reactive power, power factor, energy yield, DC/AC voltage and current, frequency, and inverter temperature. The second subsystem acquires environmental parameters related to the PV system, such as solar radiation, ambient temperature, humidity, and panel/cell temperature. These subsystems are connected via a 4G mobile router operating on a cellular network. In this configuration, a Raspberry Pi serves as a local server that processes data, stores it locally as a precaution against cloud connectivity loss, and transmits it to the Node-RED platform using message queuing telemetry transport protocol. An internet-connected personal computer is used to collect, store, analyze, and monitor the entire solar power system in real time via the human-machine interface (HMI). Through the HMI, multiple graphs representing electrical, thermal, and environmental parameters are displayed intuitively, enabling users to easily recognize the system's status in real time. In addition, the proposed IoT solution provides key indicators for evaluating the performance of the inverter's DC-to-AC energy conversion process and its operating temperature. Finally, the proposed solution was successfully validated on a 127.4 kWp solar PV system, located in a mountainous district of Da Nang, Vietnam. The results of this study not only offer a simple and reliable tool for the effective remote management of medium- and large-scale PV power systems, but also provide valuable information to support the development of fault diagnosis methods and real-time performance evaluation of the entire system.