An Efficient Constraint Mapping-based Fault-causing Process Localization Method for Intelligent Dispatching Systems of Smart Grid: Case Study

An Efficient Constraint Mapping-based Fault-causing Process Localization Method for Intelligent Dispatching Systems of Smart Grid: Case Study

As the next generation power grid, smart grid, utilizing modern information technologies, is able to deliver power more efficiently and response to wide-ranging conditions and events. Intelligent dispatch system (IDS) is regarded as the core link of the smart grid to control all the systems, and then its smooth running is vital to the smart grid. With numerous functionalities including data aggregation, data analysis, and decision making, the IDS platform is of high complexity, and then the reliable online fault diagnosis of the IDS platform becomes a big challenge when the IDS platform becomes too complicated to avoid faults such as functionality logic fault (FLF), software bug (SB), and hardware fault (HF). In order to decouple the services of IDS for easing understanding, transplanting, and debugging, the software design and implementation are under the principle of modular design and independent service calling. However, due to the sharing of the software and hardware resources of the platform, the running services are coupled tightly, which calls for efficient and reliable fault-causing process localization methods. In this paper, a constraint mappingbased method is proposed for efficient localization of the fault-causing process. Firstly, the complicated interactions among processes are analyzed and categorized into interservice and intra-service constraints, where inter-service constraints include database operation constraint (DOC), software resource constraint (SRC), and hardware resource constraint (HRC), wihle the intra-service constraint is summarized as functional logic constraint (FLC). Then corresponding constraint mapping tables can be constructed for the IDS platform. Originating from the abnormal alerts of processes and fault messages of services, the fault-causing process localization is decomposed into the track of source abnormal process within each service and the localization of fault-causing process among the services regarding intra- and interconstraint mapping tables, respectively. At last, the fault and alarming logics are reproduced in order to achieve both verification of the result of the fault-causing process localization and acquisition of new knowledge from the fault-causing logics. Case studies are presented, and the results show the efficiency of the proposed method.