Skip to Main Content
This paper utilizes particle swarm optimization method to train the weights of neural network (NN). Particle swarm optimization (PSO) is a population based stochastic optimization technique. Unlike genetic algorithm (GA), PSO has no evolution operators such as crossover and mutation. Compared to GA, the advantages of PSO are that it is easy to implement and there are few parameters to adjust. In this paper, NN is trained to extract important features from the input current waveform to uniquely identify various types of devices using their distinct harmonic Â¿signatures.Â¿ Such automated, noninvasive device identification will be critical in future power-quality monitoring and enhancement systems. A comparative study is made in different training algorithms - particle swarm optimization (PSO), genetic algorithm (GA), gradient descent (GD) and hybrid of PSO&GD for neural network in the mentioned problem. Simulation results show superiority of hybrid algorithm (a combination of GD&PSO) as optimizer over other techniques.
TENCON 2009 - 2009 IEEE Region 10 Conference
Date of Conference: 23-26 Jan. 2009