Multinodal load forecasting for distribution systems using a fuzzy-artmap neural network

Abstract

This work proposes a predictor system (multinodal forecasting) considering several points of an electrical network, such as substations, transformers, and feeders, based on an adaptive resonance theory (ART) neural network family. It is a problem similar to global forecasting, with the main difference being the strategy to align the input and output of the data with several parallel neural modules. Considering that multinodal prediction is more complex compared to global prediction, the multinodal prediction will use a fuzzy-ARTMAP neural network and a global load participation factor. The advantages of this approach are as follows: (1) the processing time is equivalent to the processing required for global forecasting (i.e., the additional time processing is quite low); and (2) Fuzzy-ARTMAP neural networks converge significantly faster than backpropagation neural networks (improved benchmark in precision). The preference for neural networks of the ART family is due to the characteristic stability and plasticity that these architectures have to provide results in a fast and precise way. To test the proposed forecast system, the results are presented for nine substations from the database of an electrical company.

Combined forecasting system for short‐term bus load forecasting based on clustering and neural networks

“…Inputs to the model are only historical load values. A fuzzy-ART neural network is applied to a set of nine substations in New Zealand [19]. The test case involves one day of January 2008.…”

“…(ii) None of the aforementioned studies utilise a full year of hourly load test data. The following periods are used as test sets: [15]: 1 month, [16]: total monthly loads for 12 months, [17]: 2 weeks, [18]: 9 months, [19]: 1 day, [20]: 5 weeks and [21]: 3 days. (iii) The clustering tool is not implemented to support the main forecasting system.…”

Fuzzy Load Forecast with Optimized Parametric Adjustment Using Jaya Optimization Algorithm

“…In addition to the abovementioned soft computing-based methods, there are hybrid intelligent methods that combine at least two of the methods abovementioned to improve the performance of load forecast results. Abreu et al introduced a multi-nodal load forecasting for distribution systems using a hybrid fuzzy art-map neural network [26]. Luy et al proposed a short-term fuzzy load forecasting model using genetic-fuzzy and ant colony-fuzzy optimization approaches [27].…”

Fuzzy ridgelet neural network prediction model trained by improved particle swarm algorithm for maintenance decision of polypropylene plant

“…It set the initial value of neural network close to the global extreme point, and the learning performance of neural network can be improved. 11,12 Therefore, the fuzzy neural network can be combined with the ridgelet neural network to construct the fuzzy ridgelet neural network to predict the maintenance cost of polypropylene plant.The particle swarm algorithm is a common method of training neural network because it does not require the gradient information, it has few adjustable parameters, it is easy to achieve, and it has high running efficiency; therefore, it can also be applied to train the fuzzy ridgelet neural network. Amr M. Ibrahim and Noha H. El-Amary proposed a recurrent neural network trained with particle swarm optimization to predict the voltage instability, and effectiveness of the proposed method is verified through comparing particle swarm algorithm and backpropagation algorithm.…”

“…It set the initial value of neural network close to the global extreme point, and the learning performance of neural network can be improved. 11,12 Therefore, the fuzzy neural network can be combined with the ridgelet neural network to construct the fuzzy ridgelet neural network to predict the maintenance cost of polypropylene plant.…”