A real-time online monitoring system for sensor network security based on parallel filtering

Abstract

The invention discloses a sensor network security real-time online monitoring system based on parallel filtering. According to the invention, real-time estimation is carried out on true values of power values measured by all instruments by using a parallel unscented Kalman filter; when the network has a fault, namely, when a system operation failure, an instrument and meter failure or a network attack occurs, deviation is generated between two filters, the deviation is accumulated by using a generalized CUSUM detection method taking KL divergence as an index to obtain decision statistics of fault detection of the whole system and each instrument and meter, and when the decision statistics exceeds a predefined threshold, alarm information is given to indicate that the fault occurs. According to the monitoring system provided by the invention, the abnormal condition of the electric power information physical system is effectively monitored from the perspective of sensor data, and the system has relatively high embedding property, real-time property and stability, is convenient to develop on an original system, and ensures the safe and economic operation of the electric power system.

Description

technical field [0001] The invention relates to the technical field of power system automation, in particular to a sensor network security real-time online monitoring system based on parallel filtering. Background technique [0002] In the context of the Internet of Things era, the power system is developing towards a high degree of integration of information systems and physical systems, gradually forming a new generation of power cyber-physical fusion systems that deeply integrate computing, communication, and control technologies. The ever-increasing level of automation and the increasingly complex information interaction mechanism make the stable operation of the system more dependent on information data, and the faults of the information system and the physical system can be transmitted to each other. new challenge. [0003] At present, the measurement system widely used in the power system is a data monitoring and acquisition system based on a telecontrol terminal wit...

AI Technical Summary

Problems solved by technology

Among them, the protection-based method will affect the real-time performance of information transmission, reduce the redundancy of measurement data, the cost is high, and cannot be guaranteed to be effective all the time; the detection-based method generally adopts a simplified DC power flow model, using the least squares The multiplication or linear Kalman filter method is used to estimate the system state, and the outlier or the CUSUM detection method based on the logarithmic likelihood ratio is used to detect the change point. However, the actual power system is a complex nonlinear system. The existing DC power flow model is based on The designed detection algorithm exposes many deficiencies, such as the difficulty of simplifying the actual system model, the problem of large system state estimation errors, etc., while the existing detection algorithms based on the AC power flow model design are still relatively small, and it is difficult to realize fault However, with the complexity of system fault manifestations, a more general fault real-time online monitoring mechanism designed based on the AC power flow model is needed as complete as possible

Images