A kind of O＆M pre-alarm and prevention system architecture based on multi-sensor information fusionTechnical field
The present invention relates to a kind of O＆M pre-alarm and prevention system architecture based on multi-sensor information fusion
Background technology
With cloud computing and virtualization, the features such as showing " extensive ", " high density ", " high energy consumption ", " complication ", buildIf with developing New Generation of IDC, lifting data center infrastructure management will become increasingly important, the basis of data centerThe new trend that framework fusion management will develop with intelligence as data center.
At present, O＆M lacks automation means, and passive O＆M, efficiency is low, and extensive IT facilities bring government pressure.NeedThe automatically-monitored of data center is realized, improves the timely alarm ability of system and ambient parameter, system is improved and environment is differentThe response speed and monitoring level often changed.Using the various means perception informations such as sensor (including camera), can be achieved withUnified service management software platform.
The information of fusion multisensor can obtain the performance that single sensor is difficult to obtain, and be mainly reflected in and improve letterCease redundancy and complementarity.The performance advantage that information fusion has is mainly reflected in：Improve the confidence level and target of informationDetectivity, the hunting zone in time and space is expanded, reduce reasoning fog-level, improve detection performance, addThe dimension of target feature vector, improves spatial resolution, enhances system survivability and adaptivity.
The conventional method of sensing data identification has cluster analysis, artificial neural network etc..In Fusion Features, first willThe data of each sensor carry out characteristic vector pickup, are carried out these characteristic vectors using artificial neural network or clustering algorithmFusion, obtain merging target reporting.Further application decision integration technology, such as classical reasoning, Bayesian inference, D-S evidencesReasoning, the processing of broad sense evidence etc., the target reporting that each sensor provides is merged, completes Target state estimator.
Polycaryon processor refers to integrate two or more complete computing engines (kernel) in one piece of processor, now locatedReason device can support multiple processors on system bus, and all bus control signals and command signal are provided by bus control unit.Polycaryon processor provides technical guarantee with for multi-sensor information fusion.
The invention provides a kind of O＆M pre-alarm and prevention system architecture based on multi-sensor information fusion, in association detectionOn the basis of, estimate the state of target, feature extraction is carried out to sensing data；The data that each sensor obtains are meltedClose, extract characteristic vector, obtain merging target reporting, complete the estimation of associated objects situation；Associated objects situation is assessedAnd feedback, so as to build O＆M environment comprehensive situation map；The degree and seriousness of O＆M environmental accident appearance are estimated, and accordinglyInstruction and warning are made in ground, to realize that accident automatic early-warning is taken precautions against.
The content of the invention
It is an object of the invention to provide a kind of O＆M pre-alarm and prevention system architecture based on multi-sensor information fusion.ThisInvention includes following characteristics：
Inventive technique scheme
1. a kind of O＆M pre-alarm and prevention system architecture based on multi-sensor information fusion, it is comprised the following steps that：
1) on the basis of association detection, estimate the state of target, feature extraction is carried out to sensing data；
2) data that each sensor obtains are merged, extracts characteristic vector, obtain merging target reporting, complete to closeJoin target condition estimation；
3) associated objects situation is assessed and fed back, so as to build O＆M environment comprehensive situation map；
4) degree and seriousness of O＆M environmental accident appearance are estimated, and correspondingly makes instruction and warning.
Brief description of the drawings
Fig. 1 is the O＆M pre-alarm and prevention system architecture diagram based on multi-sensor information fusion.
Embodiment
This O＆M pre-alarm and prevention system architecture based on multi-sensor information fusion, comprises the following steps：
1) on the basis of association detection, estimate the state of target, feature extraction is carried out to sensing data；
2) data that each sensor obtains are merged, extracts characteristic vector, obtain merging target reporting, complete to closeJoin target condition estimation；
3) associated objects situation is assessed and fed back, so as to build O＆M environment comprehensive situation map；
4) degree and seriousness of O＆M environmental accident appearance are estimated, and correspondingly makes instruction and warning.