Abstract

In this study, artificial neural network was used to predict pentachlorophenol (PCP) degradation in aqueous solution by catalytic ozonation process in a laboratory-scale semi-batch reactor. The catalyst used in this process was the alumina (γ-Al₂O₃). Results indicated that after 60 min optimal condition: 0.5 g/L of (γ-Al₂O₃), 0.5 L/min the flow rate of ozone, pH 8 and 100 mg/L PCP initial concentration, 96% of target pollutant was degraded in catalytic ozonation process. In artificial neural network evaluation, a comparison between the model data and laboratory results revealed a high degree of correlation that indicated the model was capable of defining the PCP elimination efficiency with high accuracy. Artificial neural network predicted results are very close to the experimental results with correlation coefficient (R²) of 0.989 and mean square error of 0.000421. The sensitivity analysis indicated that all studied variables (pH, dosage of catalyst and initial concentration of PCP) have strong influence on PCP degradation.

中文翻译：

---

人工神经网络预测和优化非均相催化臭氧化过程中五氯苯酚的降解和矿化作用

摘要

在这项研究中，使用人工神经网络通过实验室规模的半间歇反应器中的催化臭氧化过程来预测水溶液中的五氯苯酚（PCP）降解。在这个过程中使用的催化剂是氧化铝（γ-Al系₂ ö ₃）。结果表明，60分钟后的最佳条件：0.5克/升的（γ-Al系₂ ö ₃），0.5 L / min的臭氧流速，pH 8和100 mg / L PCP初始浓度，在催化臭氧化过程中降解了96％的目标污染物。在人工神经网络评估中，模型数据与实验室结果之间的比较显示出高度的相关性，表明该模型能够高精度地定义PCP消除效率。人工神经网络的预测结果与实验结果非常接近，相关系数（R ²）为0.989，均方误差为0.000421。敏感性分析表明，所有研究变量（pH，催化剂用量和五氯苯酚的初始浓度）对五氯苯酚的降解都有很大影响。