Multi-HMM Risk Prediction Assessment Method and Testing on a Ship Integrated Electric Propulsion System

摘要

A new risk prediction assessment method of the integrated electric propulsion system is proposed in the article. Firstly, this paper puts forward the Multi-HMM (Hidden Markov Model) algorithm, which combines the traditional HMM algorithm and the statistical theory, it can predict thesystem state probabilities in the continuous time points in the future. Secondly, according the system operation status and the impact of external environment, the working states of the components of the system are predicted by using the Multi-HMM algorithm. Thirdly, the risk assessment modelof network topology of the system is built, and power flow analysis under the abnormal condition are repeatedly calculated by using the Monte Carlo simulation. Finally the risk prediction assessment results of the integrated electric propulsion system is given. This article also takes a shipintegrated electric propulsion system as an example and the risk prediction assessment method is verified.