An improved Dynamic Programming Method for Solving the Problem of Nonlinear Programming

This paper, which based on the conventional dynamic programming solution, using the method that the decision variables of various stages are fully discrete in their feasible region to solve the optimal target function value under the various state variables. The method can be generic in solving the maximum and minimum objective function value, while avoiding the problem of the different discrete step lengths of the state variables lead to lower the precision of the target value. So, the method will make the solution process of the various stages more specific image, contributing to combining with the practical problems and understanding the connotation of the practical problems (e.g. water resource optimization allocation).