Acoustic source localization techniques are widely applied in underwater detection,seismic monitoring,structural health monitoring,and environmental acoustic sensing.The existing acoustic source localization methods are mostly established based on the assumption of a homogeneous propagation medium.However,in practical application scenarios,the medium through which acoustic waves propagate has varying degrees of randomness,which limits the applicability of localization methods based on the homogeneity assumption.To address the issue of source localization under medium randomness,the commonly used method currently is the acoustic source localization method based on the Gaussian distribution of time difference of arrival(TDOA).Nevertheless,further verification is needed to determine whether TDOA follows a Gaussian distribution under random medium conditions and under what conditions it follows a Gaussian distribution.Therefore,random medium models with different parameter settings is constructed in this study and the acoustic travel-time field is calculated by using the fast marching method combined with Monte Carlo simulations to analyze the statistical characteristics of TDOA under random medium conditions.Experiment results indicate that TDOA does not generally follow a Gaussian distribution under random medium conditions,while it approximately conforms to a Gaussian distribution under near-field conditions with relatively small medium perturbation intensity or large correlation length.This study can provide a theoretical basis for the selection and improvement of practical acoustic source localization methods in practical applications.
Analytic hierarchy process(AHP)is widely used in the health assessment of complex equipment such as aerospace due to its structural decision-making advantages.However,the traditional AHP usually relies on a fixed scale when constructing the judgment matrix,which is difficult to reflect the dynamic evolution of index weights caused by factors such as performance degradation and operation environment changes in the whole life cycle of equipment,thus affecting the accuracy of the assessment results.Therefore,a multi-level discriminant analysis is proposed based on dynamic update of judgment matrix for health assessment of spacecraft control system.Firstly,compared with the traditional AHP,the influence of degradation rate of performance index and cumulative failure probability to construct the dynamic update model of the judgment matrix is considered in this method.Secondly,in order to enhance the accuracy and stability of the model,the time-varying adjustment parameters are introduced.The optimization problem is solved by establishing the objective function of minimizing the health degree and using the particle swarm optimization algorithm of a dynamic inertia weight strategy,and the adaptive updating of time-varying adjustment parameters is realized.Furthermore,by integrating component-level health information to the system level,the traceability of component health status and overall health status assessment are achieved.Finally,an empirical analysis is conducted using data from a spacecraft control system.The verification results show that the mean square error of the traditional AHP is 0.0438,while the mean square error of the method proposed in this paper is only 0.0081,with an accuracy rate of 95.18%.Compared to AHP,it is improved by 8.14%,effectively enhancing the accuracy of spacecraft control system health assessment and providing a theoretical basis for the health monitoring and management of spacecraft control systems.
