Training neural networks with PSO in dynamic environments.
Anna S. RakitianskaiaAndries P. EngelbrechtPublished in: IEEE Congress on Evolutionary Computation (2009)
Keyphrases
- dynamic environments
- neural network
- training process
- training algorithm
- neural network training
- feedforward neural networks
- mobile robot
- particle swarm optimization
- autonomous agents
- backpropagation algorithm
- particle swarm optimization pso
- particle swarm optimizer
- multi layer perceptron
- back propagation
- path planning
- feed forward neural networks
- fuzzy logic
- neural network structure
- optimization algorithm
- collision avoidance
- potential field
- autonomous systems
- artificial neural networks
- evolutionary algorithm
- agent systems
- single agent
- changing environment
- bp neural network
- neural network model
- training set
- real environment
- highly dynamic environments
- collision free
- multilayer perceptron
- activation function
- hybrid algorithm
- genetic algorithm ga
- fault diagnosis
- fitness function
- multi objective
- genetic algorithm