基于AGA优化BP网络的焦炭质量模型

Coke Quality Model Based on BP Neural Network Optimized by Adaptive Genetic Algorithms

焦炭生产过程是一个复杂的工业生产过程,为了对焦炭质量进行准确的预测,需要建立有效可靠的焦炭质量模型。针对在对焦炭质量影响因素进行分析的基础上,采用主成分分析方法提取了影响焦炭质量的原料煤的成分,建立测试模型。根据改进的自适应遗传算法优化BP神经网络进行分析,提出了基于AGA优化BP神经网络的焦炭质量模型进行仿真。仿真结果表明,提出的质量模型建模方法是有效的和可靠的,模型预测精度高,对进行焦炭生产过程质量优化、提高焦炭质量具有重要意义。

Coke production process is a complex industrial production process.To carry on the accurate prediction to the coke quality,the effective and reliable coke quality model needs to be built.First,principal component analysis（PCA） is used to extract the main factors on the base of the analysis for effecting coke quality factors.Then through the analysis for improved adaptive genetic algorithm optimization BP neural network,the modeling method of the coke quality model based on BP neural network optimized by adaptive genetic algorithms is put forward.The simulation results show that this modeling method of quality model is effective and reliable.The model prediction accuracy is high.And it has great significance for coke production process quality optimization and improving the coke quality.