Study On Stability Control Method Of Ship Power System Based On Fault Prediction

The stability of the ship's power system includes stability under normal conditions of use and should also include relative stability under fault conditions.Ships are sailing on the sea,and the environment is bad.It is easy to cause short circuit faults caused by insulation deterioration,and even lead to more serious cascading failures.Generally such gradual failures are traceable before the failure occurs.If we can monitor the analysis and find fault signs,we can make predictions for possible faults.Then we can make early response,maintain or replace risk lines in time,or make targeted preparations.As far as failure occurs,it is possible to reduce the effect of failure,that is,the so-called predictive failure recovery.It improves the stability and safety of ship power system,and then improves the vitality and competitiveness of the whole ship.This article first briefly analyzes the failure of the ship's power system,focusing on the short-circuit failure caused by the deterioration of the insulation.Then analyze the causes of short-circuit faults and their evolution mechanisms.Set up a low-voltage line Mayr arc mode for ship power systems.And simulate arc faults during early failures.Then,based on the Calman filter,we analyze the early fault currents and voltage signals,and extract the early fault feature signals.The Calman slow filter is used to estimate the current signal.The difference between estimated signal and actual monitoring value is used to extract the characteristic and reference signals.Similarly,the method of fast estimation is used to estimate voltage signals and square wave fitting to extract feature signals.Then we compare the two characteristic signals to determine the health of the circuit.After that,Markov chain method is applied to predict the early failures.According to the line health condition,it is divided into three states: normal availability,light fault and failure.According to the current state and state transition probability matrix,we predict the future trend of the line fault and predict the probability of failure occurrence and the safe available time before the failure.Finally,combined with the annular ship power system,the improved particle swarm optimization algorithm is used to predict the recovery plan.The load correlation matrix is used to describe the power grid structure,and its search mechanism is introduced.The objective function of the power supply and the number of switches and the capacity constraints are determined.The basic particle swarm optimization algorithm is modified to adjust the mutation rate and inertia weight dynamically.When predicting possible failures,we search for predictive recovery schemes suitable for this condition.