To effectively control the urban pipeline leakage, the study combines minimum night flow with pressure-driven leakage hydraulic model to calculate the leakage coefficient of network nodes and establish an enhanced hydraulic model considering pressure dependent leakage terms. The model uses the genetic algorithm to solve the problem, taking the minimized pipeline leakage loss as the objective function, the maximum installed number of the pressure reducing valve, the node pressure flow equation, and the maximum allowable leakage rate as the constraint conditions, to find out the installation number and locations of the pressure reducing valves. The one which costs the least is chosen from all the feasibilities as the optimal solution, and HJ area partition is taken as an example to illustrate in detail the calculation procedures of the decision support model. Our study possesses the guiding significance for the control of the actual pipeline leakage.