Research On Monitoring And Analysis Model Of Data Transmission Condition In Data Center

Facing the ever-changing environment and the rapid growth of China’s economy,modern society puts forward extremely strict standards for the safe,stable and sustainable operation of electric power.Therefore,ensuring the safety of electric power has become the core of today’s industry,and its importance is far more than in the past.The State Grid Corporation is currently constructing a data full-link monitoring system between provinces,cities,and affiliated units-the cornerstone of the Internet of Things’ ubiquitous power.However,due to the vast array of data links and intricate parameters,pinpointing the source of the data link’s fault is challenging,and the data link monitoring is one-sided and subjective.Due to the increasing scope of massive monitoring data processing,increasingly diverse processing methods,increasingly diversified processing methods,and increasing application requirements,there is an urgent need for a more in-depth analysis of the massive monitoring data.In order to solve the above problems,a fault diagnosis method based on data link monitoring is studied in this dissertation.Based on the data itself,this topic first introduces monitoring data such as data center,data link and data security,historical alarms and logs,and analyzes common types and their characteristics;Select the appropriate algorithm to realize abnormal fault detection and determine the relevant parameters of typical monitoring faults;Based on the results of abnormal fault detection and parameter analysis,a fault diagnosis model for abnormal monitoring data is proposed.First of all,this project makes an in-depth study of the multi-dimensional information in the link,and then uses the recursive multi-dimensional information fusion method to fuse the multi-dimensional information in the link in order to obtain the optimal and accurate feature subset.On this basis,the random forest method is used to measure the importance of each attribute on the optimal attribute subset,and then the importance of each attribute is judged,and then the maximum value of each attribute is obtained.a fault diagnosis method based on self-organizing mapping neural network is proposed.In order to solve this problem,this project intends to use the method based on particle swarm optimization to optimize it.Finally,aiming at the problems of PSO method,a new PSO method is established and applied to PSO method.Finally,modeling is carried out based on data link data,The standard self-organizing map neural network and the standard particle swarm optimization self-organizing map neural network model are trained and tested as the contrast experiment of the improved particle swarm optimization self-organizing map neural network model.The experimental results verify the effectiveness of the improved self-organizing map neural network algorithm in abnormal fault diagnosis.