Predictive methods of generalized quantum neural networks and its application

A complex system of coal and methane outbursts has the characteristics coupled, randomized and abrupt change for the system variant, which is a difficult problem to predict coal and methane outbursts using the accurate and effective approach. We proposed a novel generalized quantum neural network called GQNN which can be used to predict coal and methane outbursts. First, we give the influence factors of coal and methane outbursts, and adopt classical neural networks to construct the model for the system. Second, we can determine a proper metrical distances after making linear by constructing the model using the neural networks for the complex nonlinear problems, and then to construct the stats of quantum superposition. Finally, the predictive model can be constructed using GQNN, to forecast the outbursts of coal and gas. The results of simulation shows that generalized neural network predicting coal and methane outbursts is effective and accurate.