Data Fusion Technology Based On GA-BP And Its Application In Intelligent Lighting System

Smart home is the most mature industry in internet of things,and it is the development tendency in the future.Smart home contains many subsystems,such as intelligent lighting system,intelligent security system,intelligent audio-visual system,intelligent curtain system etc.Although the development speeds and functions of each subsystem are different,they have the same main structure.They are all composed of the home network,sensor modules,central control modules and terminal actuators.The study of one subsystem can be applied to the whole smart home system.This paper research the intelligent lighting control system.Firstly,the research status of smart home and intelligent lighting control system is studied.The whole design solutions of intelligent lighting control system are brought forward.Secondly,the characteristics of smart home network modes are compared.Bus line and wireless are combined in this system to arrange lines and set up ZigBee star home sensor network.Multi sensors are selected to collect environmental information.Then the information is fused by BP neural network,and optimized by genetic algorithm.The optimized GA-BP algorithm is designed,and it is simulated and verified by MATLAB.Finally,the system platform of intelligent lighting control is designed,including wireless sensor modules(wireless pyroelectric infrared sensor modules,wireless photosensitive sensor modules and wireless pressure sensor modules),central controller,light driver modules and PC monitoring program,and the system is tested to verify its validity and reliability.Two points are important and creative for the design in this dissertation.One is the design of wireless sensors,and the other is the design of ZigBee network.According to a survey,traditional intelligent lighting system only uses pyroelectric infrared sensors to collect environmental information,while these sensors can’t judge the static state of human body.So,traditional intelligent lighting system always results in misjudgment.Through analysis of still behavior,I find people usually sit or lie if they remain static.So pressure sensors are added in this dissertation to assist in collecting information.Photosensitive sensors are also used to coordinate lighting brightness to increase system functions.In this dissertation,ZigBee technology is used to build wireless sensor network,which uses star topology.This network is simple,scalable,and easy to maintain and update.