Satellite ACS Health Assessment Based On Bayesian Network And Cloud Model

With the development demand of long life and high reliability of spacecraft,the research on satellite health management technology has become a hot spot.As one of the most significant and complex subsystems of satellite,attitude control system(ACS)has a high incidence of on-orbit faults.This paper studies the health assessment of satellite ACS,constructs the index set of system health assessment.In order to objectively evaluate the health status level of each component of the satellite attitude control system,the health assessment model was established by introducing the cloud model theory,and the health assessment research based on the cloud model was carried out In order to study the influence of working state in the process of health assessment,a real-time health assessment model based on Bayesian network was established by combining knowledge model with current state information.Finally the changing trend of satellite health status time series is predicted,and the simulation analysis verifies the rationality and effectiveness of the proposed health assessment and prediction method.According to the structure and characteristics of the satellite attitude control system,aiming at the complex composition of the subsystem and components of the satellite attitude control system,the health index set of the satellite attitude control system was studied and constructed by using the analytic hierarchy process(AHP),and the health assessment index set of the satellite attitude control system was established.The entropy weight method was used to calculate the weight of health index and measure the importance of each index.The health assessment method based on cloud model is studied.Through the analysis and processing of system data information,cloud model parameters are obtained,and the health index and weight of satellite attitude control system are combined to realize the health assessment of satellite attitude control system.The satellite data reflects the current state information of the system.In this paper,the advantages and feasibility of the health assessment method combining the data information with the knowledge model is discussed,and the health assessment method combining the satellite telemetry data with the Bayesian network is put forward.In order to meet the technical requirements of on-orbit autonomous health management,the structure and parameters of Bayesian health assessment network are learned from on-orbit satellite data,and the health assessment model based on Bayesian network is established.The proposed Bayesian network is used to deduce the health status of the satellite attitude control system.Aiming at the problem of different system health status under various working modes of satellite in orbit,the system health assessment under different working modes is realized by setting the working mode label and connecting it to the Bayesian network.In view of the health state of time series trend prediction,fixed prediction rules for long-term prediction accuracy problem,health assessment by using the value of time series forecast cloud model is established,using both direct and iterative prediction method for satellite attitude control system simulation to evaluate health status of short-term and long-term prediction,realizes the satellite attitude control system healthy trend forecast.The proposed method provides theoretical support and basis for autonomous health management of satellite system,and illustrates the feasibility and effectiveness of the model through simulation experiments.