The Prediction on Initial Discharge Capacity of AB₅-Based Alloy with Simulated Annealing

抄録

The prediction for initial discharge capacity of AB₅-based alloy has tested by employing the simulated annealing method in artificial neural network (ANN). The parameters in cooling schedule were determined and their influences on network prediction performance were discussed. A practical cooling schedule was established and the network with better performances was obtained by associating simulated annealing algorithm with gradient descent algorithm.