particle swarm optimization algorithm

In this research, intermittent water supply is optimized using pressure driven hydraulic analysis and particle swarm optimization algorithm with the aim of maximizing the uniformity of water distribution in the network and reliability. In the following, by calculating the resiliency as an efficiency criterion, system performance is evaluated. Obtain ed results from pressure driven hydraulic analysis and demand driven hydraulic analysis are compared. In this regard, the particle swarm optimization algorithm and the EPANET hydraulic analysis model are linked. Pressure driven hydraulic analysis is performed by applying some modifications on the hydraulic calculation process. The proposed model is evaluated on a sample network in several water shortage scenarios. According to the obtained results, the objective function (uniformity of water distribution) has higher values for the scenarios with lower water shortage, so that the maximum value is relevant to the scenario without water shortage. The values of the objective function for scenarios with pressure driven hydraulic analysis are 20% more than its values for scenarios with demand driven hydraulic analysis. By comparing the values obtained for system efficiency criteria it can be observed that the network resiliency and nodal resiliency for most of the scenarios with demand driven hydraulic analysis are more than their values for scenarios with pressure driven hydraulic analysis. This is because of the independency of nodal discharge from nodal pressure in demand driven hydraulic analysis that leads to unreal values for nodal discharges and therefore, hydraulic failure accrues rarely. The maximum value achieved for resiliency is around 99% which is relevant to efficiency threshold of 70% in nodal form. Uniformity and equity of water distribution between demand nodes and as a result satisfaction of stakeholders can be maximized by using optimization models. Employing pressure driven hydraulic analysis models makes it possible to simulate the behavior of water distribution networks realistically.

In this study, a method has been developed to improve reservoir water through selective withdrawal and release of contaminated water from the lower level of the dam. A two-dimensional model called CE-QUAL-W2 was used to simulate the reservoir quality processes. Then this model is coupled with an optimization algorithm (particle swarm optimization). The proposed simulation-optimization approach is able to find optimal operation policies to improve water quality in the reservoir and downstream river. In order to increase the efficiency and reduce the computational time, an artificial neural network model replaced with the reservoir water quality simulation model. The performance of the model is evaluated for a real case for a 2-year simulation period. Different scenarios of reservoir operation are considered and using the simulation model (meta model), the pollutant concentration of the index is calculated and ultimately optimal operation policy of the reservoir is determined. The results show that the average amount of dissolved oxygen in the released from the reservoir for optimal operation policy is about 25% better than the current operating conditions. So that the eutrophication status of the reservoir is improved in relation to the existing condition, which in some months like June is in high eutrophication state, it has been improved in most months of the year to low eutrophication state. The comparison between dissolved oxygen concentrations of reservoir outflow in optimum operation policy with the existing operational conditions showed high efficiency of the proposed method and significant improvement in reservoir water quality.

One of the main aims of water resource planning and management is to estimate and predict groundwater quality parameters which would be used in decision-making. In this regard, many models have been developed which proposed better managements in order to maintain water quality. Most of these models require input parameters which are hardly available or their measurements are time consuming and expensive. Among them, Artificial Neural Network (ANN) models inspired by human's brain are a better choice. The present studied stimulated the electrical conductivity of water quality parameters of Behbahan Plain, using ANN and ANN+PSO models and in the end compared their results with measured data. Data for NO3-, EC, Ca2+, Mg2+, SO42-, HCO3-, CL-, K+, TH and pH were collected during 2009-2016 as input data. The results indicated that the highest prediction accuracy of quality parameters was related to the ANN + PSO model so that the MAE and RMSE statistics had the minimum and  had the maximum value for the model. Considering the high efficiency of artificial neural network model, by training the Particle Swarm Optimization algorithm, it can be used in order to make managerial decisions and ensure the results of monitoring and reducing costs.

Waste load allocation (WLA) is one of the most important elements when evaluating in water quality management problems. Due to multiple and sometimes conflict objectives in WLA problems, a set of Pareto optimal solutions is derived with evolutionary algorithms in which one of these Pareto fronts could be influenced by conflicts. In this research study, simulation-optimization approach was applied by QUAL2Kw simulation model and particle swarm optimization (PSO) as optimization algorithm to assign cBOD point source pollutions for specific location along Gheshlagh River. To reduce the conflicts between beneficiaries for the optimum operation of water resources in river, the level leader-follower and the Nash bargaining game theory models were applied. The results showed that the construction, maintenance and operation costs of the treatment plants for leader-follower and Nash bargaining game theories were about 192 and 293 billion Rial, respectively. The penalties for violating the environmental regulations set by the Iranian environmental protection agency (EPA) for the above theory models were found to be about 32 and 3.9 billion Rial, respectively. Furthermore, the estimated penalty tariff for each overdischarge of allowed cBOD under Stackelberg and Nash bargaining game theories were about 10.8 and 3 Rial, per environmental violation, respectively. The estimated penalty tariff in Stackelberg game is extremely close to current Iran’s EPA penalty tariff.