基于改进LSTM的生鲜牛肉新鲜度预测模型研究

牛肉作为我国消费量较大的肉质食品,其新鲜度受到广泛关注。研究的主要目的是通过改进的循环神经网络模型对牛肉新鲜度进行预测研究。利用红外光谱技术采集了影响牛肉新鲜度的实验数据,即牛肉的系水率、pH值和TVB-N的含量。在利用改进的循环神经网络进行预测的同时,采用了循环神经网络做对比,通过计算均方根误差来评价模型的优势。实验结果表明,改进的循环神经网络均方根误差低于循环神经网络。同时这也证明了上述模型在食品预测方面具有较高的正确性,可用于对食品安全或者新鲜度进行预测。

Improved Beef Quantitative Prediction Based on the Freshness of LSTM

As a meaty food with a large consumption in China,beef has received extensive attention.The main purpose of this study is to quantitatively predict beef freshness through an improved Recurrent Neural Network model.In this paper,the experimental data affecting the freshness of beef was collected by infrared spectroscopy,namely the water rate,pH value and TVB-N content of beef.Predictions are made using an improved Recurrent Neural Net-work.This study also uses a Recurrent Neural Network for comparison,and evaluates the advantages of the model by calculating the root mean square error.The experimental results show that the root mean square error of the improved Recurrent Neural Networks is lower than that of the Recurrent Neural Network.At the same time,it also proves that the model has high correctness in food prediction.The model to predict the safety or freshness of food in advance also promotes the safe development of food.