# Load Estimation for Electric Power Distribution Networks

**Authors:** Chiebuka Eyisi, Saeed Lotfifard  
**arXiv:** [2109.09873](https://arxiv.org/abs/2109.09873)  
**Date:** 2021-09-20  
**Summary page:** https://www.summarizepaper.com/en/paper/2109.09873v1/

## Key points

- The paper focuses on Distribution Load Estimation (DLE) in Distribution Management Systems (DMS)
- Authors propose a method using Representative Load Curves (RLC) to improve real-time load estimation
- Adaptive Neuro-Fuzzy Inference Systems (ANFIS) are used to estimate the RLC
- ANFIS combines fuzzy logic and neural network techniques to learn from input-output data
- Accurate RLCs obtained through ANFIS provide better pseudo-measurements for load estimation
- Method demonstrated on an 11 kV radial distribution network using MATLAB software
- Performance evaluated based on mean absolute percent error (MAPE) criterion
- Paper presents an innovative approach to load estimation by introducing RLC generated through ANFIS
- Results show improved accuracy in real-time load estimation and enhanced DMS performance.

## Summary

The paper titled "Load Estimation for Electric Power Distribution Networks" by Chiebuka Eyisi and Saeed Lotfifard focuses on Distribution Load Estimation (DLE), which is a crucial function of the Distribution Management System (DMS). The authors propose a novel method for presenting historical load data in the form of Representative Load Curves (RLC) to improve real-time load estimation in distribution networks. To estimate the RLC, the authors employ Adaptive Neuro-Fuzzy Inference Systems (ANFIS). ANFIS combines fuzzy logic and neural network techniques to adaptively learn from input-output data. By utilizing ANFIS, accurate RLCs can be obtained, which in turn provide better pseudo-measurements for load estimation. The proposed method is demonstrated on an 11 kV radial distribution network using MATLAB software. The performance of the estimated RLC is evaluated based on the mean absolute percent error (MAPE) criterion, which quantifies the accuracy of the estimates. Overall, this paper presents an innovative approach to load estimation in distribution networks by introducing Representative Load Curves generated through ANFIS. The results highlight the effectiveness of this method in improving real-time load estimation accuracy, thereby enhancing the performance of Distribution Management Systems.