Application of particle swarm optimization (PSO) is demonstrated through design of a water distribution pipeline network. PSO is an evolutionary algorithm that utilizes the swarm intelligence to achieve the goal of optimizing a specified objective function. This algorithm uses the cognition of individuals and social behaviour in the optimization process. For the optimization of water distribution system, a simulation – optimization model, called PSONET is developed and used in which the optimization is by PSO. This formulation is applied to two benchmark optimization design problems. The results are compared with the results obtained by other optimization methods. The results show that the PSO is more efficient than other optimization methods as it requires fewer objective function evaluations.

There have been several attempts to use combined simulation-optimization models to solve reservoir operation problems efficiently. In many cases, complex simulation models are available, but direct incorporation of them into an optimization framework is computationally prohibitive. To overcome this problem, in this study, a backpropagation neural network is trained to approximate the simulation model developed for the Chennai city water supply problem. The neural network is used as a submodel in a Hooke and Jeeves nonlinear programming model to find “near optimal policies.” The results are further refined using the conventional simulation-optimization model.

Failures in operation of water supply reservoir systems are often unavoidable during critical hydrologic periods. The failure characteristics of such systems can be represented by performance indicators such as reliability, resilience, and vulnerability. A mixed-integer programming model for the operation of a water supply reservoir during critical periods has been presented in the literature that incorporates these performance indicators. An improved formulation of this model that represents resilience more completely is discussed herein. In addition, a set of constraints with binary integer variables are included to account for reservoir spills. The improvements achieved with the modified model is demonstrated using the same example as presented with the original model.