农机化研究2024，Vol.46Issue(7)：241-248,8.
基于FA-BP神经网络的生姜干燥含水率预测
Prediction of Water Content of Ginger by Optimizing BP Neural Network Based on Firefly Algorithm
摘要
Abstract
In order to explore the drying characteristics of ginger and realize the moisture content prediction of ginger dr-ying,different drying temperatures(50℃,55℃,60℃)and drying wind speeds(1.0m/s,2.0m/s,3.0m/s)were studied.The effect of slice length(30mm,35mm,40mm)on the drying time and drying rate of ginger.Combined with the characteristics of BP neural network adaptive ability,generalization ability,strong learning ability and firefly algo-rithm(FA)with few parameters,strong optimization ability and fast convergence speed.Taking the drying temperature,drying wind speed,slice length and drying time as the input layer,the number of hidden layers is 10,and the output layer is the moisture content of ginger,to build a FA-BP neural network with a topology of 4-10-1 Model.The results showed that drying temperature,drying air speed and slice length were all key factors affecting the moisture content of ginger.In-creasing the drying air speed,increasing the drying temperature and reducing the slice length could effectively shorten the drying time of ginger and improve the drying efficiency.The firefly algorithm is used to optimize the weights and thresholds of the BP neural network,which reduces the training time of the neural network and improves the accuracy.The moisture content prediction results are accurate and rapid,which can provide a scientific basis for the online prediction of moisture content in the drying process of ginge.关键词
生姜/热泵干燥/含水率预测/萤火虫算法/BP神经网络Key words
ginger/heat pump drying/moisture content prediction/firefly algorithm/BP neural network分类
农业科技引用本文复制引用
王雷,胡书旭,钟康生,康宏彬,肖波..基于FA-BP神经网络的生姜干燥含水率预测[J].农机化研究,2024,46(7):241-248,8.基金项目
广东省重点领域研发计划项目(2018B020241003) （2018B020241003）
广东省省级乡村振兴战略专项(2022KJ101) （2022KJ101）
广东省科技创新战略资金项目(粤科资字[2021]133号) （粤科资字[2021]133号）
