Auxiliary Decision Of CPS Intelligent Lighting System Based On SVM

Cyber Physical System is a complex network-based system which providesintelligent services by combining computing, network communication and intelligentcontrol. It is a newly put forward research area that can change the style of our life.With the popularity and much further development of technologies on computernetworks such as cloud computing, artificial intelligence and so forth, people have anincreasing need on computing performance for working, information sharing,searching the Internet or games. By and by, people desire of advance in computertechnology and R&D focus on advanced artificial intelligence, massive dataprocessing, better computing performance, more natural human-computer interactionand so on. While smart-house become an important field of computer scientific andtechnological research, researchers started further studies on the combination of CPSand intelligent-house. Scholars from various countries carried out extensiveresearches based on embedded systems, embedding chips on more household goodsforming family-wide sensor networks, introducing more user-friendlyhuman-computer interaction. The embed systems nowadays cannot reunion to be anetwork system because of lacking protocols and used to be designated to achieve thegoal independently. Besides, most of them cannot use high-AI algorithm as thelimitations of hardware. In conclusion, they can’t satisfy needs of users.Major subject of this dissertation is studies the relationship between CPS andsmart-house while CPS technology used in the application of smart-house and itdiscuss about the definition and characteristics of CPS. In the experiment we use themethod of centralized control to make advantage of the devices in the smarthouse.The intelligent lightning system designed based on SVM in the research is for validating thesis in the dissertation.Classifier based on SVM is chosen and makeacceptable experimental result in the process of data mining. The first study ismodeling situation of smart-house using data sets collected by wireless sensornetwork. Then using SVM make auxiliary decision and improve the AI of lightingsystem. The system can conserve energy by controlling lights and sensorsdynamically. As we investigate the deficiency of the lightning system usingnowadays，we did a overall research and making UML models of the system. First，in this experiment we did a research on the achievement of context-aware in theSmart-house with Cyber Physical System. Second, we use SVM to classify the datasetand its accuracy rate is90%.By our strategy for sensor we can decrease four fifth ofthe times we using sensor.