Study On Intelligent Measurement And Mutation Prediction Method Of Comprehensive Information Network Traffic In Coal Mine

With more and more applications for coal mine big data,internet of things,and cloud computing,the country faces many heterogeneous subsystems in the information and network construction of coal mines.It is difficult to share data in real time,find out the reason of abrupt change in network traffic and know when the traffic change.Therefore,how to carry out scientific network data management,accurately measure and predict the trend of changes in network traffic,and prevent network congestion have become a hot topic in the field of coal mine information network security.Based on the detailed analysis of the coal mine information network architecture and traffic characteristics,this paper conducts in-depth research on the method of intelligent measurement and prediction of coal mine integrated information network traffic.In view of the independence of the coal mine subsystem network,the inconsistency of information protocols,and the difficulty of data interaction,the structure of the existing coal mine information network platform is analyzed.The characteristics of the main six types of coal mine information sources are analyzed,and a three-tier network architecture designed in the B/S model is introduced and then a integrated coal mine integrated information network architecture solution is proposed which have a detailed analysis and design on the overall architecture of the platform,network structure,network topology,subsystem functions,etc.Taking the Shenhuaning coal Meihuajing coal mine network structure as object,the information management topological structure design program for the comprehensive information data transmission of coal mines is put forward.In order to carry out the expert system to enhance the information utilization rate,conduct in-depth analysis and excavation,and provide coal mine companies with solid big data infrastructure support in terms of safety status index analysis and evaluation,production risk analysis and evaluation.Aiming at the problems caused by the possible burst traffic under the high-speed backbone network link in coal mines for traffic measurement,an adaptive adaptive traffic sampling algorithm is proposed.This algorithm solves a series of problems such as single traffic sample measurement algorithm's low fitting of complex traffic sample measurement characteristics,high measurement error,and unsatisfactory sampling traffic measurement ability of abnormal traffic.By analyzing the characteristics of coal mine diversified information service traffic,a multi-segment discrimination screening counter and matching sampling function library design are proposed to calibrate the specific sampling measurement algorithm of multi-sampling and improve the accuracy of multi-sampling measurement under the coal mine information network.In response to the possible change of the traffic sample,an updateable sampling library function is proposed.For a small number of unrecognized traffic sample measurement problems,a measurement scheme is proposed that compares the adaptive traffic length threshold with the adaptive simple random probability and complete sampling modes.The results of data simulation experiments show that:(1)The improved algorithm can have a huge increase on the accuracy of small traffic measurement;(2)The improved algorithm has remarkable performance and accurate judgement in abnormal traffic measurement performance when attacked.In view of the randomness and irregularity of the time series of coal mine network traffic,the chaotic characteristics of the time series in coal mines' information network traffic are analyzed.The SCADA class traffic of coal mines is selected as the index quantity,and combine the chaotic attractor dimension,Lyapunov index with coal mine network traffic data,then the chaotic characteristics of time series are deduced and calculated to obtain the maximum saturated embedding dimension.Through the acquisition of 24-hour traffic data of Meihuajing Coal Mine central switch,the time series of coal mine network traffic is established.With the help of theoretical derivation and simulation experiments,the optimal delay time of the network,the maximum saturated embedding dimension and the maximum Lyapunov exponent are obtained.The coal mine comprehensive information network traffic has chaotic characteristics,which provides support for constructing a traffic forecasting model using chaos theory.The characteristics of traditional traffic forecasting algorithms are analyzed,and a prediction method for coal mines' information traffic is proposed by combining wavelet neural network and chaos algorithm.In order to prevent the system from falling into a local minimum,a nonlinear self-feedback term is designed to introduce dynamical system equations for the interconnection weight space,which helps the system to escape the local extremum during the learning process.In order to improve the convergence speed,a wavelet neural network algorithm model based on simulated annealing strategy is proposed and the function relationship between system connection weights and chaos mechanism probability is established.The control of annealing temperature T is used to improve the convergence of training.In order to improve the learning speed,a new chaotic fast learning algorithm is designed which is related to the feedback item parameters.Through the establishment of the wavelet neural forecasting model of chaotic time series in coal mine network traffic,the chaotic characteristics of the weight space are analyzed.Through the single-step and multi-step multi-algorithm traffic forecasting simulation experiments,the results show that the wavelet neural network variable traffic prediction method based on chaotic fast learning algorithm has excellent prediction performance for coal mine information traffic.