Design And Application Of IoT Cloud Platform Based On Abnormal Data Detection Algorithm

With the vigorous development of communication technology and cloud computing,the Internet of Things(Io T)has been widely used in production and daily life,and the Io T device has also shown explosive growth.More and more device need to be connected to the network.In the Io T system,there are many kinds of device and complex interaction.Io T devices usually need to run in an environment where the network is not reliable,and there are many limitations in power consumption,volume and computing resources.Therefore,the communication protocols between Io T devices and cloud platforms cannot form a unified standard,and the commonly used Io T communication protocols also have their own characteristics.A large number of Io T devices also produce a large amount of data.It is particularly important to apply machine learning algorithm training model to detect abnormal data in order to observe the operation status of devices.The main work of this paper is as follows:(1)A cloud platform design scheme which can flexibly adapt to a variety of Io T communication protocols is proposed.The platform integrates and encapsulates various common network protocols such as TCP,UDP,MQTT and Co AP in the form of components,so as to realize the functions of unified monitoring,management,online debugging,online stop and start of equipment,which greatly reduces the complexity of communication between equipment and cloud.(2)The Io T cloud platform based on B/S architecture is implemented for equipment access and management to the cloud.The server is based on Java 8 and adopts the high-performance network programming framework Netty,which is used to implement a variety of Io T communication protocols,defines the thing model to describe the representation of the device in the cloud,encapsulates the message sent by the device into a topic message uniformly designed by the Io T cloud platform,and forwards the message to the message bus for the flow between different modules in the platform,Completed the complete access process from the device to the cloud platform.(3)The algorithm model of abnormal data detection is studied and trained,which can detect the data generated by Io T devices,and complete the monitoring of equipment operation status by Io T cloud platform.The traditional abnormal data detection methods can't detect various types of abnormal data comprehensively and accurately.This paper studies the machine learning algorithms such as Random Forest,Neural Network and Isolated Forest,completes the training and optimization of the model,and determines the abnormal data detection algorithm model combined with the characteristics of the algorithm and the analysis and comparison of the experimental results.According to the design scheme of Io T cloud platform with multi-protocol adaptation,the Io T cloud platform is implemented by using many development technologies in Java language ecology,and the algorithm model of abnormal data detection is applied to the Io T cloud platform.