Font Size: a A A

Multi-sensor Fusion Technology Under Limited Source Conditions

Posted on:2011-02-01Degree:MasterType:Thesis
Country:ChinaCandidate:Y F BiFull Text:PDF
GTID:2178360308455288Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Data source scheduling, the construction of fusion system and evaluation of results of data fusion has been the key area of multi-sensor fusion technology, the hotspot and difficulty of the fusion as well. And it will also be the focus of future development. The main purpose of this paper is hoping to make some progress in three directions under the condition of limited data resources, by means of multi-sensor fusion theory analysis.First of all, Agent is a computing entity which has autonomy, continuity, sociality and agency in a certain environment, it has its own knowledge base and inference mechanism, can respond to environment actively. Multi-Agent System is working like a group by mutual consultation and cooperation. In the multi-Agent, the individual Agent's ability is limited, but through cooperation, a lot of complex tasks can be completed. Therefore, through understanding of the different type and characteristics of data sources, analysis of usual scheduling algorithm, using advanced theory of the multi-Agent, we can set up a distributed multi-sensor data source scheduling model.Secondly, the limitations of sensor resources of fusion system are in two major aspects: ability and link. Ability limitations include processing and observation, link limitations include the time of communication and the bandwidth. Therefore, in consideration of the limitations, by improving existing model which based Dasarathy model, get a new theoretic model of multi-level data fusion. On this basis, according to specific limited (limited ability or link) situation, we can get an improvement program, and applied to real cases, so that we can verify its effectiveness.Finally, by summarizing the evaluation of existing data fusion method, comprehensive analyzing fusion system, we can set up an index set of system effectiveness which includes MOPS (Measure of Performance Set), MOES (Measure of Efficacy Set), MODES (Measure of Decision Efficacy Set), supported by fuzzy theory. At the end, we are hoping to find a quick and effective way to evaluate a fusion system.
Keywords/Search Tags:Multi-Agent theory, Intelligent scheduling, Limited sources, Fusion model, Fusion evaluation
PDF Full Text Request
Related items