Design And Implementation Of Cloud Real Time Supervision System For Construction Robot

Since the "construction industry 4.0" strategy was put forward,more and more construction robots have appeared in the construction site.However,at present,the information level of construction robots is still insufficient,and there is a lack of remote real-time supervision means.With the rapid development of cloud computing technology,computing power and storage resources can be purchased like commodities,and data on the cloud has become the choice of more and more users.Therefore,this topic designs and implements a cloud real-time monitoring system based on public cloud platform,aiming at providing remote real-time monitoring and management services for construction robots.Firstly,this thesis investigates the related technology and theoretical basis involved in the system.Then,according to the application scenarios and main objectives of the system,the requirements of the system are analyzed,and the physical architecture,hierarchical architecture and functional modules of the system are designed.In the design phase of the system,aiming at the data acquisition of construction robot,the data acquisition mode is designed,the unified data model is defined,and the high concurrency of the data acquisition module server is designed.A hybrid data storage scheme is designed,which takes into account the data reliability and efficient reading and writing.Aiming at the problem of sensor data loss of construction robot,this thesis completes the missing data based on time series prediction technology,and selects the best prediction model through experiments.Finally,the real-time performance of the system is designed.In the implementation of the system.In this thesis,an anomaly detection model based on isolation forest is established,and the sensor data collected by the construction robot is scored through the model.According to the scoring results,the quality of the plastering operation of the construction robot is evaluated.Based on netty framework,the server side of data acquisition is realized,which can support more construction robots to communicate with the server at the same time.The problem of data sticking and unpacking is solved.When dealing with the data sent by the construction robot,the filtering rules are formulated,and the sensor missing data is predicted and completed based on the double exponential smoothing method.In order to ensure the security and reliability of business data,this thesis builds a master-slave MySQL cluster.Finally,it realizes the real-time monitoring and control of the construction robot through the web page.In order to ensure the real-time monitoring and control function,this thesis realizes the two-way real-time communication between the construction robot and the web page based on websocket protocol.Finally,the main functions of the system are tested and demonstrated,and the concurrency and real-time performance of the system are tested.The final results show that the system supports real-time monitoring and management of large-scale construction robot,and meets the overall goal and design requirements of the system.