| Abstrakt: |
In the process of anaerobic digestion of wastewater, effluent chemical oxygen demand (COD) and gas production are important parameters to measure the effect of anaerobic biological treatment, and are also important indicators for evaluating the performance of water treatment. At present, most of these values in anaerobic biological treatment systems for wastewater are often obtained through manual tests. The disadvantage of manual assays is the long detection time and poor stability. Therefore, the prediction of water COD and gas production based on back propagation neural network (BPNN) is proposed in this paper. Then, aiming at the problems of speed sluggishness and lopsided one-sided minimization in traditional BP neural networks, an improved BP neural network prediction model based on genetic algorithm (GA-BPNN) is proposed. Experimental results show that the performance of GA-BPNN is better than traditional BPNN. In effluent COD prediction, the mean absolute percent error (MAPE) of BP neural network prediction is 60.7234%, while the MAPE of GA-BPNN algorithm is only 20.9854%. In the prediction of gas production, the MAPE of BP neural network prediction is 10.5521%, while the MAPE of GA-BPNN algorithm is only 7.5677%. Moreover, both the effluent COD prediction and the gas production forecasting, GA-BPNN algorithm’s mean square error (MSE), root mean square error (RMSE) and Pearson’s correlation coefficient are all better than BP neural network. [ABSTRACT FROM AUTHOR] |
