基于人工神经网络的纸浆浓度实时监测系统  

Real Time Monitoring System of Pulp Concentration Based on Artificial Neural Network

作　　者：孟晓龙 MENG Xiaolong(Shanghai Normal University,School of Tourism,Shanghai,201418,China;Shanghai Institute of Tourism,School of Foundational Education,Shanghai,201418,China)

机构地区：[1]上海师范大学旅游学院,上海201418 [2]上海旅游高等专科学校公共教学部,上海201418

出　　处：《造纸科学与技术》2021年第6期34-39,共6页Paper Science & Technology

基　　金：校(院)级科研项目(KY2020-DL13);校(院)人才队伍建设工程项目(RS2021-CY04)。

摘　　要：纸浆浓度直接关系到纸张生产质量,有效监测纸浆浓度变化值对于提高纸张生产质量具有重要的现实意义。在此背景下,设计一种基于人工神经网络的纸浆浓度实时监测系统。基于B/S三层架构设计系统框架结构,包括采集层、处理层和显示层;选择声速测量装置、测温仪、单片机以及通信设备等作为系统主要硬件设备。在软件部分,以声速测量装置、测温仪以及打浆时间作为输入值,通过人工神经网络推算纸浆浓度值,并对比标准值,判断纸浆浓度是否合格,完成纸浆浓度实时监测。系统功能测试:系统监测功能实现,完成了纸浆浓度监测任务,并与实际情况大致相同,误差较小,说明系统监测精度较高;系统性能测试:响应时间,吞吐量,并发用户数,资源利用率等均满足设定的预期标准,证明系统性能满足要求。Pulp concentration is directly related to paper production quality. Therefore, effective monitoring of pulp concentration change is of great practical significance to improve paper production quality. Under this background, a real-time pulp concentration monitoring system based on artificial neural network is designed. the system framework is design based on B/S three-layer architecture, including acquisition layer, processing layer and display layer;Sound velocity measuring device, thermometer, single chip microcomputer and communication equipment are selected as the main hardware equipment of the system. In the software part, taking the sound velocity measuring device, thermometer and beating time as input values, the pulp concentration value is calculated through artificial neural network. Then, compared with the standard value to judge whether the pulp concentration is qualified, so as to complete the real-time monitoring of pulp concentration. The results show that the system function test: the system monitoring function is realized, the pulp concentration monitoring task is completed, which is roughly the same as the actual situation, and the error is small, indicating that the system monitoring accuracy is high. The system performance test, including response time, throughput, number of concurrent users and resource utilization, meet the set expected standards, proving that the system performance meets the requirements.

关 键 词：人工神经网络 纸浆浓度 硬件设备 软件程序 监测系统 

分 类 号：TP236.51[自动化与计算机技术—检测技术与自动化装置]