Aim The distribution simulation of river contamination is a problem of solving partial differential equation

a new method used for seeking approximate analytic solution of one-dimensional contamination source is studied. Methods Approximate analytic solution is solved by integration of neural network method with mathematical physics method and a three-layer feedforward neural network is utilized to approximate the control equation of one-dimensional contamination. Results and Conclusion The approximate analytic solutions are successfully obtained under three different conditions

the feasibility and rationality of applying neural network to river contamination simulation is hereby proved.