Iranian Journal of

Backgrounds and Objectives:A number of different technologies have recently been studied todetermine the best use of biogas, however, to choose optimize technologies of using biogas for energy recovery it is necessary to monitor and predict the methane percentage of biogas. In this study, a method is proposed for predicting the methane fraction in landfill gas originating from Labscalelandfill bioreactors, based on neural network.
Materials and Methods: In this study, two different systems were applied, to predict the methane fraction in landfill gas as a final product of anaerobic digestion, we used the leachate specifications as input parameters. In system I (C1), the leachate generated from a fresh-waste reactor was drained to recirculation tank, and recycled. In System II (C2), the leachate generated from a fresh waste landfill reactor was fed through a well-decomposed refuse landfill reactor, and at the same time, the leachate generated from a well-decomposed refuse landfill reactor recycled to a fresh waste landfill reactor.
Results: There is very good agreement in the trends between forecasted and measured data. R valuesare 0.999 and 0.997, and the obtained Root mean square error values are 1.098 and 2.387 for training and test data, respectively
Conclusion: The proposed method can significantly predict the methane fraction in landfill gasoriginating and, consequently, neural network can be use to optimize the dimensions of a plant using biogas for energy (i.e. heat and/or electricity) recovery and monitoring system.

Background and Objectives: Rivers are the most important resources supplying drinking, agricultural, and industrial water demand. Their quality fluctuates frequently due to crossing from different regions and beds as well as their direct relationship with their peripheral environments. Thus, it is essential to be considered the surveying and predicating changes in the water qualitative parameters in a river. In this study, in order to estimate some of the qualitative parameters (Total dissolved solids, electrical conductivity and sodium absorption rate) for Tehran Jajroud and Kermanshah Gharasu rivers, we used wavelet-artificial neural network (W-ANN) hybrid model during a statistical period of 24 years. Methods: We compared W-ANN model with ANN model in order to evaluate its capability in detecting signals and separating error signals for estimating water quality parameters of the abovementioned rivers. The evaluation of both models was performed by the statistical criteria including correlation coefficient, the Nash-Sutcliffe model efficiency coefficient (NS), the root mean square error (RMSE) and the mean absolute error (MAE). Results: The results showed that the optimized W-ANN with correlation coefficient of 0.9 has high capability to estimate SAR parameter in the stations studied. Moreover, we found that W-ANN had less error and higher accuracy in the case of EC and TDS parameters rather than ANN model. Conclusion: W-ANN proved high efficiency in forecasting of the water quality parameters of rivers, therefore, it can be used for decision making and assurance of monitoring results and optimizing the monitoring costs.

Background & Objectives: The prediction and quality control of the Karaj River water, as one of the important needed water supply sources of Tehran, possesses great importance. In this study, performance of artificial neural network (ANN), combined wavelet-neural network (WANN), and multi linear regression (MLR) models were evaluated to predict next month nitrate and dissolved oxygen of “Pole Khab” station located in Karaj River. Materials and Methods: A statistical period of 11 years was used for the input of the models. In combined WANN model, the real monthly-observed time series of river discharge (Q) and the quality parameters (nitrate and dissolved oxygen) were analyzed using wavelet analyzer. Then, their completely effective time series were used as ANN input. In addition, the ability of all three models were investigated in order to predict the peak points of time-series that have great importance. The capability of the models was evaluated by coefficient of efficiency (E) and the root mean square error (RMSE). Results: The research findings indicated that the accuracy and the ability of hybrid model of wavelet neural network with the attitude of elimniations of time series noise had beeb better than the other two modes so that hybrid model of Wavelet artificial neural network wase able the improve the rate of RMSE for Nitrate ions in comparison with neural network and multiple linear regression models respectively, amounting to 35.6% and 75.92%, for Dissolved Oxygen ion as much as 40.57% and 60.13%. Conclusion: owing of the high capability wavelet neural network and the elimination of the time series noises in the prediction of quality parameters of river’s water, this model can be convenient and fast way to be proposed for management of water quality resources and assursnce from water quality monitoring results and reduction its costs.

Background and Objectives: In this work, biosorption of hexavalent chromium from aqueous solution with excess municipal sludge was studied. Moreover, the performance of neural networks to predict the biosorption rate was investigated.

Materials and Methods: The effect of operational parameters including initial metal concentration, initial pH, agitation speed, adsorbent dosage, and agitation time on the biosorption of chromium was assessed in a batch system. A part of the experimental results was modeled using Feed-Forward Back propagation Neural Network (FFBP-ANN). Another part of the test results was simulated to assess the model accuracy. Transfer function in the hidden layers and output layers and the number of neurons in the hidden layers were optimized.

Results: The maximum removal of chromium obtained from batch studies was more than 96% in 90 mg/L initial concentration, pH 2, agitation speed 200 rpm and adsorbent dosage 4 g/L. Maximum biosorption capacity was 41.69 mg/g. Biosorption data of Cr(VI) are described well by Freundlich isotherm model and adsorption kinetic followed pseudo-second order model.  Tangent sigmoid function determined was the most appropriate transfer function in the hidden and output layer. The optimal number of neurons in hidden layers was 13. Predictions of model showed excellent correlation (R=0.984) with the target vector. Simulations performed by the developed neural network model showed good agreement with experimental results.

Conclusion: Overall, it can be concluded that excess municipal sludge performs well for the removal of Cr ions from aqueous solution as a biological and low cost biosorbent. FFBP-ANN is an appropriate technique for modeling, estimating, and prediction of biosorption process If the Levenberg-Marquardt training function, tangent sigmoid transfer function in the hidden and output layers and the number of neurons is between 1.6 to 1.8 times the input data, proper predication results could be achieved.