Design And Implementation Of Mobile Site Operation And Maintenance System Based On Bayesian Network Mining Algorithm

Communication network construction of operators is the key to the construction of basic network.Therefore,it is very important to guarantee the stability and reliability of the communication facilities of the operators.In order to reduce operational cost,the operators outsource the inspection and maintenance work of the communication facilities to substitute maintaining company.However,the lack of management on behalf of the company,resulting in ineffective operation and maintenance of labor,low efficiency of daily work.Through the analysis of current substitute maintaining management system,we found that the efficiency of inspection is low;the data statistics is complicated due to most of the data recorded in the paper form and stored in the database in the form of pictures;operators can not accurately control the site,the data analysis is difficult,the cost of substitute maintaining is high,the process reduction is hard,and so on.Combined with the relevant development background,this thesis firstly in-depth analyzes the business requirements of resource management and personnel management of substitute maintaining and data mining technology,and then proposes a detailed design scheme of the design and implementation of mobile site operation and maintenance system based on Bayesian network data mining algorithm.The system integrates the environment monitoring system,the substitute maintaining management system and security inspection system,And use SSM(Spring,SpringMVC,Mybatis)framework,Bayesian network data mining algorithm and Android platform to achieve the system.SSM framework achieves the low coupling and high scalability of system,and Android client provides an excellent user experience for data acquisition.The system uses data mining technology based on Bayesian network to achieve the function of site fault level division.In this thesis,the Bayesian network structure learning is based on search scoring method,through the test from K2 algorithm,HillClimber algorithm,RepeatedHillClimber algorithm,TabuSearch algorithm and TAN algorithm,select the better classification results of TAN algorithm.It isconvenient to count high failure rate sites,the statistical results as a key object in the next inspection,so that the inspection work more targeted.Through the actual delivery to prove that the system can achieve the scientific management and the intuitive and efficient supervision management of the inspection data,also can quickly record and feedback of the inspection results,and be able to chart in the form of visual display.