Innovative sensorless dual-axis solar tracking system using particle filter

Renewable energy (RE), particularly solar photovoltaic (PV) systems, is a promising alternative to traditional energy sources for worldwide electricity generation. In recent years, PV generation capacity has increased rapidly due to a considerable reduction in installation costs. Nevertheless, in PV systems, energy generation is significantly less than the installed capacity due to the PV modules' orientation. Hence, this study proposes an innovative solar tracking system to enhance PV energy generation. Previous research classified solar tracking systems into two categories: sensor-based and sensorless solar tracking systems. However, tracking performance inevitably deteriorates when implementing sensor-based solar tracking systems in low-visibility conditions. Conversely, sensorless solar tracking systems require substantial historical meteorological data based on geographic location and complex mathematical models estimating the sun's position, increasing computational complexity. This study aims to overcome these difficulties by proposing a sensorless dual-axis solar tracking system that does not require historical meteorological data, complex mathematical models, and sun position sensors to track the sun's position. Particle filter (PF), a robust sampling-based tracking algorithm, is applied to develop an innovative solar tracking strategy. Initially, a set of samples (also called particles) is generated as inputs to PF corresponding to the proposed tracking system's orientation angles: daily and elevation angles. Next, PV power is captured as a measurement of each particle, and a weight associated with the PV power is estimated to represent each particle's relative significance. Furthermore, each particle is estimated and updated recursively, considering its measurement to determine the sun's probable position. The tracking strategy is terminated once the particles' weights are equal, indicating that the optimal position has been achieved (convergence). A comparative study is conducted over 60 days under various weather conditions to evaluate the proposed tracking system's performance compared with that of a fixed flat-plate system. The experimental results indicate that the proposed tracking system improves energy generation performance (after accounting for operational energy consumption) by 20.1% compared with the fixed flat-plate system.

Automated summary

This study presents an innovative sensorless dual-axis solar tracking system that improves PV energy generation by using a particle filter algorithm. It does not require historical meteorological data or complex mathematical models to track the sun's position. Experimental results show that the proposed tracking system enhances energy generation performance by 20.1% compared to a fixed flat-plate system under various weather conditions.