摘要
近年来,现有的可穿戴心理监测装置多源处理技术仍存在数据识别异常和不稳定等问题。为了解决该问题,提出了一种融合卷积神经网络与长短期记忆网络的可穿戴心理监测装置多源处理技术。首先,研究对可穿戴心理监测装置多源处理技术进行分析,然后,研究对可穿戴心理监测装置多源处理模型进行性能分析。实验结果表明,随着迭代次数的增加,当训练集的迭代次数达到4时,模型的损失变化曲线开始趋于平稳,此时的损失值为0.05。同时,当训练集的迭代次数达到6时,模型的准确率曲线也开始趋于平稳。此时,模型的准确率高达98.82%,由此可知,研究模型可以提高心理状态的监测准确性和实时性。
In recent years,the existing multi-source processing technology of wearable psychological monitoring devices still has problems such as abnormal and unstable data recognition.In order to solve this problem,a multi-source processing technology of wearable psychological monitoring device is proposed,which integrates convolutional neural network and long and short term memory network.Firstly,the research analyzes the multi-source processing technology of wearable psychological monitoring device,and then the performance of the multi-source processing model of wearable psychological monitoring device is analyzed.The experimental results show that with the increase of the number of iterations,when the number of iterations of the training set reaches 4,the loss change curve of the model begins to stabilize,and the loss value is 0.05.At the same time,when the number of iterations of the training set reaches 6,the accuracy curve of the model also begins to stabilize.At this time,the accuracy of the model is as high as 98.82%,which shows that the research model can improve the accuracy and real-time monitoring of mental states.
作者
张红梅
ZHANG Hongmei(Xian Medical College,Xi’an 710000,China)
出处
《自动化与仪器仪表》
2024年第8期130-134,共5页
Automation & Instrumentation
基金
医学院校优良学风培育路径研究(2023FDY04)。
关键词
心理监测装置
多源处理技术
CNN-LSTM
心理状态预警
psychological monitoring device
multi-source processing technology
CNN-LSTM
mental state warning
