Layout and size optimization of sanitary sewer network using intelligent ants

The incremental solution building capability of ant algorithm is exploited in this paper for the efficient layout and pipe size optimization of sanitary sewer network. Layout and pipe size optimization of sewer networks requires that the pipe locations, pipe diameters and pipe slopes are optimally determined. This problem is a highly constrained Mixed-Integer Nonlinear Programming (MINLP) problem presenting a challenge even to the modern heuristic search methods. In this paper, the Ant Colony Optimization Algorithm (ACOA) is equipped with a Tree Growing Algorithm (TGA) to efficiently solve the sewer network layout and size optimization problem and its performance is compared with the conventional application of the ACOA. The method is based on the assumption that a base layout including all possible links of the network is available. The TGA is used to construct feasible tree-like layouts out of the base layout defined for the sewer network, while the ACOA is used to optimally determine the pipe diameters of the constructed layout. An assumption of sewer flow at maximum allowable relative depth is made allowing for the calculation of the optimal pipe slopes in the absence of any pump and drop in the network. Two different formulations are, therefore, proposed and their performances are tested against hypothetical problems. In the first formulation, ACOA is used in a conventional manner for pipe size optimization while an ad hoc engineering concept for the layout determination. In the second formulation, however. ACOA equipped with TGA is used to simultaneously determine both the layout and pipe sizes of the network. Proposed formulations are used to solve three hypothetical test examples of different scales and the results are presented and compared. The results indicate the ability of the proposed method to optimally solve the problem of layout and size determination of sewer networks. (C) 2012 Elsevier Ltd. All rights reserved.