粒子群算法优化的神经网络对MBR过滤阻力的预测

Neural Network Optimized by Particle Swarm Algorithm for Prediction of MBR Filtering Resistance

摘要: 在MBR系统中，人们通常用膜通量或过滤阻力与操作参数的函数关系表征一个膜污染模型，但是膜污染是一个复杂的动态过程，经典的数学模型难以精确模拟。针对此问题，在本文中使用过滤阻力表征膜污染，利用BP神经网络对MBR系统中的膜的过滤阻力进行预测。然后利用粒子群算法优化神经网络的初始权值与阈值，提高神经网络获得全局最优解的能力。最后，我们将程序预测结果与实际实验数据做对比，发现该模型准确率较高，达到了预期效果。

Abstract: In the MBR system, people usually use the functional relationship between membrane flux or filtration resistance and operating parameters to characterize a membrane fouling model, but membrane fouling is a complex dynamic process, and classical mathematical models are difficult to accurately simulate. In response to this problem, the filtration resistance is used to characterize membrane fouling in this article, and the BP neural network is used to predict the filtration resistance of the membrane in the MBR system. Then the particle swarm algorithm is used to optimize the initial weights and thresholds of the neural network to improve the ability of the neural network to obtain the global optimal solution. Finally, we compare the program prediction results with the actual experimental data, and find that the model has a higher accuracy rate and achieves the expected effect.