本論文研究「冰水主機自我診斷之Linux嵌入式系統開發」,係以Xscale Linux架構下的嵌入式工業電腦,整合電力監測電錶、變頻器與PLC可程式控制器,以結合MYSQL資料庫,作為運算、控制與狀態追蹤,將結果畫面以PHP網頁即時顯示於網頁瀏覽器,供遠端管理者查詢目前冰水主機運轉狀況,進而修改相關參數,實現節能策略佈局。並提出基於人工類神經網路模型的研發作為冰水主機自我檢測,達到故障預警協助管理保養與即時故障排除,避免系統損壞,保障操作人員安全。另外針對嵌入式電腦進行遠端遙控,即時調整更新系統程式,以因應當操作環境條件改變時確保監控效果最佳化;並裝設防火牆(Firewall)強化網路資料傳輸安全,降低中毒與資料外洩風險。 本研究結果除發揮Embedded Linux System應有功能外,大幅提昇系統安全性、穩定性與實用性。
This paper studies the topic of “A Study of Self- Diagnosis of Chiller System on Linux Embedded Platform “, which is based on the industrial embedded computers of Xscale Linux framework. The study integrates monitor of electricity meters, VSD and PLC programmable controllers to combine with MYSQL database with a view to computing, controlling and tracking the status. The results will display in the form of PHP page on a web browser for remote managers to check the current state of operation and amend the relevant parameters, and then achieve energy-saving strategies layout. Moreover, proposing the invention based on artificial neural network-based model to do self-diagnosis of chiller system, which can assist in the management of maintenance and troubleshooting immediately to prevent the system from damage and to protect the operators’ safety. In addition, the invention supports administer remote control on embedded computers to update the program on the system instantly for the purpose of ensuring the best results of the monitoring when operating environments are changed. Furthermore, the system sets up the firewall to strengthen the safety of network data transmission and reduce the risk of virus attacks and information disclosed. The results of this study not only display the functions of Embedded Linux System, but also promote security, stability and practicality of the system.