基于新型BP神经网络的沼气生产预测

Biogas production prediction method based on novel BP neural network modeling

为了提高BP神经网络预测模型对沼气生产预测的准确性,提出了一种基于新型BP神经网络建模的沼气生产预测方法。在原有方法中引入一种新的功能函数,它优化了传统算法,不仅克服数据量少的问题,而且,与传统BP神经网络相比,拟合精度有了一定的提高。混合配比的原料相较于单一原料产气速率高,但是配比的不同也会相对应的影响产气量,目前甲烷生产企业原料的盲目配比导致了低效益生产。仿真结果表明,本文所述方法对沼气生产过程的预测具有精度高、非线性拟和能力强等优点,克服了低效益生产,可实现相同原料种类下不同配比的沼气企业效益提前精确预测,可广泛应用。

In order to improve the accuracy of BP neural network prediction model for biogas production prediction,a biogas production prediction method based on novel BP neural network modeling is proposed.A new function is introduced in the original method,which optimizes the traditional algorithm,not only overcomes the problem of small amount of data,but also has a certain improvement in fitting accuracy compared with the traditional BP neural network.The mixed ratio of raw materials is higher than that of single raw materials,but the difference in ratio will also affect the gas production.At present,the blind ratio of raw materials in methane production enterprises leads to low-efficiency production.The simulation results show that the method described in this paper has the advantages of high precision and strong nonlinear fitting ability and overcomes the low-efficiency production.It can realize the accurate prediction of the efficiency of biogas enterprises with different ratios under the same raw material types.And it also can be widely used.