摘要
棉花是中国重要的经济作物之一,回潮率是棉花交易结算时的重要依据,国家标准《GB1103—2007》中明确公定回潮率为8.5%。为了确保棉花交易中的公平性,构建基于Tensorflow的BP神经网络预测模型对籽棉回潮率进行预测,并且从隐含层方面优化模型。将硬件检测的数据分为训练集和测试集,用训练集的数据训练模型,测试集数据验证模型,其均方误差符合要求。结果表明,构建的BP神经网络预测模型对籽棉回潮率数值的预测结果准确可靠,上述方法不仅能够将籽棉回潮率的多个影响因素考虑进去,而且成本较低,具有一定的推广意义。
Cotton is one of the important economic crops in China, and the moisture regain rate is an important basis for cotton transaction settlement. The national standard "GB1103-2007" clearly defines moisture regain rate as 8.5%. In order to ensure fairness in cotton transactions, a Tensorflow-based BP neural network prediction model is constructed to predict the moisture regain of seed cotton, and the model is optimized from the hidden layer. The hardware detection data is divided into a training set and a test set. The model is trained with the training set data, and the test set data is used to verify the model. The mean square error meets the requirements. The results show that the constructed BP neural network prediction model is accurate and reliable in predicting the value of seed cotton moisture regain. This method can not only take into account the multiple influencing factors of seed cotton moisture regain, but also has a low cost, which has certain popularization significance.
作者
王婷婷
张学东
WANG Ting-ting;ZHANG Xue-dong(College of Information Engineering,Tarim University,Alar Xinjiang 843300,China)
出处
《计算机仿真》
北大核心
2022年第6期274-278,共5页
Computer Simulation
基金
研究生创新创业项目(TDGRI202044)。
关键词
籽棉
回潮率
神经网络
预测
Unprocessed cot ton
Moisture regain
Neural network
Prediction
