Security monitoring, early warning and alarm Method based on security boundary for regional integrated energy system

Abstract

Effective security early warning and alarm technology can provide comprehensive information such as the security levels, violation components, and the security trend of the system before a fault occurs. The core early warning and alarm information provided by existing methods is insufficient. To fill these gaps, this paper proposes an early warning and alarm method to monitor the security state of regional integrated energy system (RIES). Firstly, the security boundary and fuzzy theory are introduced. Secondly, the assessment methods of geometric security distance, overload and node parameter violation are further given, respectively. Thirdly, a two-stage early warning and alarm method is proposed, which can analyze the system operating state and determine the security level. The first stage can preliminarily classify the operating state of RIES through AC security boundary, including normal state, alert state and emergency state, and then send out signals correspondingly. The second stage can calculate the final security levels based on fuzzy inference and fuzzy comprehensive evaluation. For interconnected RIES, the security levels include I, IIa ~ IId and IIIa ~ IIId. For radial RIES, the security levels include I and IIIa ~ IIId. Fourthly, the comprehensive and detailed early warning and alarm information is given. The overload components and N − 1 components are located by DC security boundary analysis; the violation degree and parameters violation locations are further given based on energy flow calculation; the security trend is also predicted by the average value of the geometric security distance of the time series operating points. Finally, the correctness and effectiveness of proposed method are tested on typical cases. The results show that there are indeed five security levels: III, IIc, Ia, Ic and Id, the violation components and corresponding violation degrees can be obtained accurately, and the downward trend of N-0 and N − 1 security for RIES can be predicted further. Compared with existing methods, this paper provides more complete early warning and alarm information such as detailed security levels, violation components, violation degree and security trend. The time complexity of online early warning and alarm is only O(rt), and the online computational time is reduced by 2 orders of magnitude compared with the existing method. The detailed information provided by the proposed mothed can help RIES dispatchers assess security levels and make decisions timely.