- 學位論文
應用灰色預測與模糊推論之智慧型家庭系統設計與實作
Design and Implementation for Smart Home Systems Based on Grey Prediction and Fuzzy Reasoning
摘要
近年來，全球天氣有時會發生極端的現象，以及家庭常發生一氧化碳中毒事件，與室內二氧化碳濃度過高時，造成人們不舒服、疲倦或頭痛，甚至出現嘔吐。然而，目前市面上的室內環境監測系統，並沒有對環境監測等級做一個更精細的分類，也沒有預測的功能，且有些感測器是有線傳輸模式，架設與配線工作麻煩也是一大問題。 因此，為了改善這些問題，本論文提出一個將灰色預測與模糊推論應用於環境危險程度評估的智慧型家庭系統之研究。其中智慧型家庭系統分為兩大部分，分別為智能家電控制與室內環境監測。系統結合無線感測網路(wireless sensor network, WSN)改善感測器架設與配線困難的問題，以及提供室內家電無線監控與感測器數據採集，遠端方面是使用智慧型手機結合全球行動通訊系統(global system for mobile communications, GSM)來與遠端家電做控制，並提供簡訊服務方便查詢遠端電器狀態。而電腦監控端會將一氧化碳、二氧化碳、溫度感測器的數據收集後做分析，利用灰色預測將收集到溫度數據做下一時段預測，將預測溫度與目前溫度比較，得到溫度變化量。最後使用模糊推論將溫度變化量、一氧化碳、二氧化碳數據做運算，推論出目前室內的環境危險等級，並給予相關對應措施意見，提供更為安全的家庭生活環境。
並列摘要
In recent years, global weather has caused extreme phenomenon. The family often stays in carbon monoxide poisoning, and the high indoor carbon dioxide concentration causes people uncomfortable, tired or headachy, event vomiting. However, in present market condition, the indoor environmental monitoring system does not do a finer classification for the environmental monitoring level without predict function. In addition, most sensors are wired, and difficult for deployment. Therefore, in order to improve these problems, this paper proposes a grey prediction and fuzzy reasoning application to the degree of environmental risk assessment of the smart home system. In our framework , the smart home system is divided into two parts, the control of smart appliances and indoor environmental monitoring. The system combines wireless sensor networks (WSN) to improve the problems of the sensor erection and wiring difficulties, as well as provides indoor appliance wireless monitoring and sensor data acquisition. The remote user can use smart phones with the global system for mobile communications (GSM) to control the remote home appliance, and use SMS services to look up a remote electrical state simple. While the computer monitor end collects the data of carbon monoxide, carbon dioxide, temperature sensor data to analyze, making use of grey projected to collect temperature data for the next period forecast, comparing with the forecast temperature and the current temperature to get the temperature change. Finally, by using the fuzzy reasoning of temperature change, carbon monoxide, carbon dioxide data to do calculations to infer the indoor environmental risk rating, the proposed system provides corresponding measures to have a more secure family living environment.