An ant colony-BP neural network model has been put forward.Adopting method combining ant colony with back propagation(BP) algorithm to train the neural network

can overcome the drawback of pure BP algorithm to easily converge on local optima

decreasing the sensitivity of algorithm to the initial value.A model for predicting the coal ash fusion point has been established by using ant colony and BP neural network

and verification of prediction performance for the said model being carried out.Results show that the predicted precision of said method is more enhanced than that of a single BP neural network model

the trained network model can be used for prediction of coal ash fusion point.