The Design And Implementation Of Alarm Corralation Analysis Module For Power Communication Network

Electric communication network is another physical network coexist with grid in the power system. As the basis for safe and stable operation of the grid system, Electric communication network guarantees the security of electricity production. In the meantime, and it is also a platform for all types of power of information technology to run the business. Power communication network management system is an important way to support the operation and maintenance of electric power communication network, which needs to show maintenance personnel alarm information that is collected from network devices timely, and all is intended for maintenance personnel to analysis and positioning faults. However, due to the complexity of the electric power communication network alarm transmission, one failure often results in a large number of related alarms that bring lots of difficulties to locate the fault.The design and implementation of electric power communication alarm correlation analysis module in this paper is part of the electric power communication network management system. And it uses Rules Engine to depend on the acquisition of alarm correlation and alarm relationship rules in the rule base to match and find the derivative relationship between alarms and to distinguish failure between roots alarm with derivative warning alarm. In the process of matching, firstly, the module will load current processing alarms resource. Then, it will add alarms which are not marked on other related resources to the alarm queue within a certain time. In the end, it will mark the alarm which are met the alarm relevant rules. It is inadequate to describe the correlation between the overall system alarms according to alarm rules provided by the manufacturers. In order to finish alarm correlation rule base and to distinguish roots alarm with derivative warning alarm effectively. The module, which this paper designs, adopts a hierarchical Apriori algorithm from the library of standardized historical alarm to mine alarm association rules. And it is also in conjunction with alarm rule from factories to update maintenance alarm correlation analysis module in the rule base. SDH alarms as an example are given to the standardized design method to provide a standard warning means of alarm correlation analysis. In order to improve the accuracy of the divided alarm transaction set, clustering technique are used to cluster happened alarm according to the occurrence time which is automatically put alarms that own similar happen time together as the same transaction set. Then, Apriori algorithm is used to do association mining with divided standardized alarm.In the first time, this paper describes the background and significance of alarm correlation analysis from the perspective of the operation and maintenance. In the second time, it introduces some related concepts of power communication network management and alarm correlation analysis. Then, the standardized approach and alarm correlation analysis of the alarm system design is given by this paper. And the detailed implementation of alarm rule mining, root alarms matching and other relevant warning process is also given by this paper. In the end, it summarizes the design and implementation of the proposed alarm correlation analysis modules and points out the future direction of improvement according to existence deficiencies in this system.