Optimization Of 1OT Based On Optical Sensor And Machine Learning

The Internet of things has become a realistic starting point for industrial and other information industries under the background of the country's promotion of industrialization and information integration.Technically,the Internet of things is an intelligent network that transmits data to the specified information processing center through the smart sensor devices and networks,realizing the integration of human society and physical systems.It is widely used in urban construction,transportation and logistics,structural detection,medical and health care,military and other fields.The emergence of artificial intelligence will greatly expand the application scope,application depth and intelligence degree of the Internet of things.With the rapid development of national economy and technology,Maritime military and economy are also flourishing,and the equipment widely used in the ocean is in the special environment of the ocean,its structural health detection is particularly important.Especially in Maritime military matters,the equipment is likely to be damaged to varying degrees,and the data retention and transmission of the equipment,and the accuracy of health status detection is particularly important.This paper studies the key technologies of the three-layer architecture of the Internet of things in the context of the boom of machine learning and the Internet of things,proposes the design of the Internet of things system based on optical sensors and machine learning,and puts forward a sio2 based stress sensor which is not affected by wavelength drift,power attenuation,and light source aging by using the photoelastic effect of materials,thus improving the stability of detection.RFID technology is used to design the transmission mechanism of data protection when the data transmission link of Internet of things is damaged,so as to solve the problem of storing and transmitting the collected stress data when the equipment is damaged.The application of artificial intelligence and data mining makes the structural health detection more intelligent and precise.And a method of structural state detection and protection policy recommendation based on machine learning is designed.Finally,the data set of the Internet of things is processed by data cleaning,attribute specification,etc.,for later data classification and prediction.The improved BP neural network which is based on particle swarm optimization algorithm is used to classify and predict the data sets,which can be used for the state detection of equipment and the structure detection of Bridges.Secondly,the use of RFID technology to design data storage protection mechanism,the use of RFID storage data portability and strong environmental adaptability,the data storage protection,design pressure data collection system based on RFID.Identify the electronic tag through the MFRC522 radio frequency chip,and store the pressure data collected by the pressure sensor to the electronic tag.Data protection and storage are realized,which can be applied to bridge structure data collection and marine equipment data protection.Even when the equipment is completely damaged,the data will be stored.Finally,use data cleaning,attribute specification,etc.to process the IoT data set for later data classification and prediction.The BP neural network improved by the particle swarm optimization algorithm is used to classify and predict the data set,which can be used for device status detection,bridge structure inspection,etc.