multiobjective optimization

Hirmand River is the main drainage system in the southern part of Afghanistan and has a significant impact on the socio-economic life of Sistan region. The assessment of water potential of Hirmand River is essential for long term period, which requires the use of rainfall-runoff model. The present study focuses on simulating the daily discharge of the upper Helmand basin of Afghanistan using the conceptual model FLEX in combination with multi-objective optimization, which has not been considered in previous studies. At first, the discharge was evaluated integrally by lumped model considering one objection function. Then it was simulated using semi-distributed model by adding the logarithm of the flow to the objection function. The results of this study indicates that the model is efficient in the semi-distributed mode by using two objection functions for calibration of parameters. The coefficients of NS, NS-log in the calibration and validation period were estimated to be 0.86, 0.92, 0.76 and 0.81, respectively. Pareto front analysis gave the best optimum set for each parameter used in model simulation. The model gives satisfactory results based on monthly calibration and can be used to investigate the drought and climate change research of studied basin.

In this research, the capability of a multicolony Ant Colony Optimization algorithm is applied to multiobjective waste load allocation problem. In order to derive nondominated solutions, three different models are used. Two of them are biobjective and the remaining one is a three-objective model. In the first model, minimization of cost and DO violation along the stream flow is considered as multi objective optimization problem, while for the second case, minimization of the cost and equity is investigated. For the third optimization problem, minimization of cost along with equity and DO violation are considered. For the all case studies, the Pareto front is derived which enhances the decision maker to choose one which more suitable for him/her according to the priorities. The case study is the Wilmate river in Oregon State of US. The following research shows the capability of NAACO in multiobjective optimization of waste load allocation problem. According to the discrete pattern of decision variables in the ACO algorithm, it can be easily map to practical waste load allocation problems.