Simultaneous management of water and wastewater using ant and artificial neural network (ANN) algorithms

In the present study, simultaneous management of water and wastewater was examined using ant and artificial neural network algorithms. Ant algorithm is one of the most meta-heuristic algorithms to solve optimization problems. The data were collected in a monthly time series from the Regional Water Organization of Khorasan Razavi during 1998–2016. Water flow data were initially estimated in the study. In this regard, different structures of the back propagation network (1, 2, 3, 4 and 5 nodes in the hidden layer) were designed by using logistic activation function in order to evaluate the efficiency of ANN model and its comparison with the ARIMA method in predicting the time series of the flows for the time horizons of next 3, 6, 9 and 12 month. The output data of each network were finally compared with actual data to evaluate the efficiency of the model and its comparison with ARIMA model using the evaluation criteria of models. Water resources were, then, allocated to drinking, agricultural and industrial sectors using the ant algorithm. According to the results of the total consumption of drinking water for the industrial sector, the southeast and northwest of Mashhad have the greatest amount of water. In other words, the northeastern and southwestern parts of Mashhad have the least amount of water for the industrial sector. And so, there must be a higher priority to water scarcity in these two regions than the other areas.