# A Rprop-Neural-Network-Based PV Maximum Power Point Tracking Algorithm with Short-Circuit Current Limitation

**Authors:** Yao Cui, Zhehan Yi, Jiajun Duan, Di Shi, Zhiwei Wang  
**arXiv:** [1811.12541](https://arxiv.org/abs/1811.12541)  
**Date:** 2018-11-29  
**Summary page:** https://www.summarizepaper.com/en/paper/1811.12541v2/

## Key points

- Resilient-backpropagation-neural-network-based algorithm for Photovoltaic (PV) maximum power point tracking (MPPT)
- Incorporates a supervision mechanism to calibrate the Rprop-NN-MPPT reference and limit short-circuit current caused by incorrect predictions
- Aims to overcome steady-state oscillations and loss of tracking direction under fast changing ambient environments
- Addresses the challenge of partial shading by finding the global maximum power point on a multi-peak characteristic curve
- Multiple case studies conducted to verify its effectiveness
- Improved accuracy in tracking the maximum power point of PV systems
- Increased energy efficiency and performance optimization in PV applications
- Provides valuable insights into addressing common challenges faced in MPPT algorithms
- Promising solution through the use of neural networks

## Summary

This paper presents a resilient-backpropagation-neural-network-based algorithm for Photovoltaic (PV) maximum power point tracking (MPPT). The algorithm incorporates a supervision mechanism to calibrate the Rprop-NN-MPPT reference and limit short-circuit current caused by incorrect predictions. Unlike conventional MPPT algorithms such as perturb and observe (P&O), hill climbing, and incremental conductance (Inc-Cond), which are trial-and-error based, the proposed algorithm aims to overcome steady-state oscillations and loss of tracking direction under fast changing ambient environments. Additionally, the algorithm addresses the challenge of partial shading by finding the global maximum power point on a multi-peak characteristic curve. The novel Rprop-NN MPPT algorithm is developed and elaborated in this work, with multiple case studies conducted to verify its effectiveness. By utilizing neural networks and incorporating a supervision mechanism, the proposed algorithm offers improved accuracy in tracking the maximum power point of PV systems. This can lead to increased energy efficiency and performance optimization in PV applications. The paper provides valuable insights into addressing common challenges faced in MPPT algorithms and presents a promising solution through the use of neural networks. Overall, this research contributes to advancing PV technology by providing an innovative approach for efficient power point tracking in varying environmental conditions.