Design And Implementation Of Smart Home System Based On Android Multimedia Gateway

With the rapid development of Internet of Things and computer technology, intelligent home systems has been developed rapidly, with endless variety of solutions, but due to the limitations of traditional wired solutions integrated wiring, installation inconvenience, scalability is not strong so that it cannot spread widely. For these problems, this thesis designed and implemented a smart home system based on Android gateway using the technique of wireless sensor networks and embedded system. Gesture recognition technology is used in this smart home system solution, and also studied the application of air gesture recongnition technology in Smart Home system.My thesis introduces the research background and the current situation at home and abroad about intelligent home systems, then wireless sensor networks, Android system framework, message push mechanism and gesture recognition Techniques are discussed. Moreover, the thesis firstly describes the overall architecture of the intelligent home system and the functional requirements of the system, detailing the hardware design of gateway control module and ZigBee coordinator module of this intelligent embedded system. Secondly, the system software and application software of control module, software of coordinator modules are also designed.And then, the hardware and software design of ZigBee-based wireless smart socket is implemented, especially the design of small PCB antenna of 2.4GHz. The socket communication distance is tested under two scenarios, as it turns out, the communication distance in the open environmrnt jack is at least 50m, through walls at least 30m, to prove that their ability enough to meet the requirements of the home environment. Finally, we proposed a solution of intelligent home control terminal based on embedded inertial measurement unit of smart phone. Firstly, the raw angular rate and acceleration signals are preprocessed. Secondly, adaptive magnitude thresholds are used to detect the motion interval of the filtered acceleration and angular velocity signals, then an orientation error compensation method is used, once we obtain the pose transformation matrix, the filtered acceleration can be transformed from the smartphone coordinate to the navigation coordinates. Finally, in order to remedy the method of elimination the cumulative error by directly to modify the speed of traditional trajectory reconstruction, a method of modifying the acceleration first and then the speed is proposed. Then the final velocity is integrated to obtain the trajectories of the motion.Through simulation and test shows that the proposed algorithm can improve the accuracy of mobile space trajectory reduction better, for the recongnition rate of gesture is up to 94%.Embedded gateway and intelligent outlet are designed and implemented in this thesis. And also proposed an improved error compensation algorithm to improve the rate of gestures recongnition, so as to control the household equipment more accurately. The system provides important reference value for the air gesture recongnition technology using in the application of smart home.