A fast fixed-point algorithm for independent component analysis
Neural Computation
Fast and robust fixed-point algorithms for independent component analysis
IEEE Transactions on Neural Networks
Software application implement in java for electrical lines dimensioning
ACS'08 Proceedings of the 8th conference on Applied computer scince
WSEAS Transactions on Computers
Particle swarm optimization models applied to neural networks using the R language
WSEAS TRANSACTIONS on SYSTEMS
Hi-index | 0.00 |
Because of increasingly common damage and thefts of the transmission towers, a new function is proposed to enhance the state maintenance of transmission lines. It relies on the monitoring of the high-voltage transmission towers to ensure the reliability and safety of the power grid operation. An approach which combines Independent Component Analysis (ICA) with neural network based on Particle Swarm Optimization (PSO) algorithm is presented. Its purpose is to extract the vibration source signals caused by the theft of the towers. Initially, the proposed algorithm separates the source vibration signals from the observed mixing signals based on the FastICA of negentropy and Exploratory Projection Pursuit (EPP). In order to distinguish the vibration pulse signal from other similar interference pulses, the algorithm of the feed forward neural network (FFNN) is used to identify the vibration pulses. It is trained online with particle swarm optimization (PSO), which incorporates the pulse extraction algorithm based on an adaptive threshold. Numerical results show that the algorithm is effective in extracting and identifying the vibration signals, and it also can suppress the interference signals.