摘要
采用生物柴油工业化生产经济成本分析的相关论文数据,对论文的数据进行了收集整理,从原料、催化剂、工厂产能、甘油置信度等方面进行了详细分析。引入了人工神经网络对收集到的数据进行优化预测,将59组数据的85%作为训练集,15%作为预测集,采用相关系数和均方误差作为预测结果的评价指标,使用BP网络里的L-M(莱文贝格-马夸特)算法对数据进行研究表征。研究结果表明:使用L-M算法相关系数为0.9741,均方误差为0.0045;预测结果的平均相对误差为5.85%。L-M神经网络算法有效验证了原料成本、工厂产能、甘油置信度对生物柴油最终成本的重要影响。
Based on the relevant papers on the economic cost analysis of biodiesel industrial production,the data of the paper were collected and sorted out,and the detailed analysis was carried out from the aspects of raw materials,catalysts,plant capacity,glycerol confidence,etc.Artificial neural network is introduced to optimize the prediction of the collected data.85%of the 59 groups of data are used as the training set and 15%as the prediction set.Correlation coefficient and mean square error are used as the evaluation indexes of the prediction results.L-M(Levenberg Marquardt)algorithm in BP network is used to study and characterize the data.The results show that:the correlation coefficient of L-M algorithm is 0.9741,the mean square error is 0.0045;the average relative error of prediction results is 5.85%.L-M Neural network algorithm effectively verified the important influence of raw material cost,plant capacity and glycerol confidence on the final cost of biodiesel.
作者
马驰
文振中
MA Chi;WEN Zhen-zhong(College of Energy and Power Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《广州化学》
CAS
2021年第5期70-76,81,共8页
Guangzhou Chemistry
