摘要
为更精确地关联预测硫在高含硫气体中的溶解度,提出将遗传算法(GA)和LM-反向传播神经网络(LM-BP ANN)相结合的预测模型。设计了该模型的计算过程,讨论了模型参数的设置。以温度、压力和气体组分作为BP神经网络预测模型的输入变量,利用GA优化了BP神经网络的初始权值和阈值,采用遗传算法优化后的BP神经网络计算了元素硫在高含硫气体中的溶解度。结果表明,该模型训练结果与实测值之间的平均相对误差为5.90%,测试结果与实测值的平均相对误差为5.54%;该方法较BP神经网络模型具有预测精度高、收敛速度快的优点;该模型具有较好的模拟及内推、外推功能。
To associate and predict the solubility of sulfur in high sulfur gas more accurately, a GA-LM-BP ANN based on genetic algorithm （GA） and LM-back propagation （BP） neural network is proposed. Implementation procedure and parameters setting of this model are introduced in detail. The input variables of BP neural network predictive model are temperature, pressure and gas component. Genetic algorithm is used to optimize the weights and bias of BP neural network. The designed GA-LM-BP ANN is used to calculate the solubility of sulfur in high sulfur gas. The simulation results show that the average relative deviation （AARD） between training results and measured values is 5.90% , and the AARD for the test results is 5.54%. The proposed model offers the advantages of high precision and fast convergence in contrast with BP neural network. The model is a better way to simulate, interpolate and extrapolative the solubility of sulfur in high sulfur gas.
出处
《现代化工》
CAS
CSCD
北大核心
2014年第9期142-147,149,共7页
Modern Chemical Industry
基金
国家自然科学基金项目(51174172)
教育部博士点专项科研基金项目(20125121110003)
关键词
硫沉积
遗传算法
LM-BP神经网络
元素硫
高含硫气体
溶解度
sulfur deposition
genetic algorithm
LM-BP neural network
elemental sulfur
high sulfur gas
solubility
