Prediction method of output varied rules in polymer flooding based on chaotic neural network

The chaotic characteristic of water ratio and oil output was determined with Lyapunov exponent in the situation of polymer flooding, the chaotic attractors in phase spaces were reconstructed, and space embed dimension was calculated, the chaotic neural network model was established. Finally, a new prediction method of water ratio and oil output was formed. Training process indicates the method has powerful approaching ability, classing ability and convergence. Actual experiment results verify validity and veracity of this method.