摘要
人工神经网络具有较强的自组织、自学习、自记忆联想的能力,更适合处理在已知条件和结果之间无明确关系情况复杂的流域系统。利用实测资料,运用BP网络建立了厌氧序批示反应器出水指标预测模型,引入了Levenberg-Marquardt算法对反应器相关运行参数进行预测,并与机理模型进行了精度比较分析,表明前者精度高于后者。
Artificial neural network has capabilities of good self-organization, strong self-learning, self-memory and association. It is more suitable to deal with a complicated fluid area system of unclear relationship between known condition and result. By using experimental reference and BP network, a predictive model of water fluid indication of anaerobic sequential batch reactor is established. Relevant processing parameters of reactor are predicted by introducing Levenberg-Marquardt algorithm and accuracy is compared and analyzed with that of mechanistic model. The experimental results demonstrate the accuracy of the neural network model is more accurate.
出处
《电脑开发与应用》
2008年第12期13-15,共3页
Computer Development & Applications
基金
山西省自然科学基金项目(20051035)资助
