Power System Fault Diagnosis Based on Krill Herd Algorithm

When the power system fails, the alarm signal will be uploaded, if a complex fault occurs, the amount of signal will increase greatly and the error alarm signal will be prone to occur. In order to ensure the accuracy of grid fault diagnosis, this paper introduces the actual status and alarm signal of protection and circuit breaker in the fault hypothesis. Aiming at the difference between the expected action state and the actual action state to build the objective function reflects the correctness of the hypothesis, and the error correctness indicator is used to eliminate the false alarm signals, thus an improved method of power grid fault diagnosis is given. The objective function is computed by the Krill Herd Algorithm (KHA) to obtain the actual state of protection and the circuit breaker. The effectiveness of the proposed model is verified by simulation of IEEE39 nodes, which shows that the method can accurately diagnose faults under complex conditions and obtain reasonable fault diagnosis results.