The Research Of Multisensor Data Fusion Based On Vehicle Fault Diagnosis

Data fusion technology is used for how to utilize mufti-source information to overcome the inaccuracy while thoroughly describing objects concerned for correct judgment. Under the current complex conditions, sole sensor in the vehicle fault of monitoring system in the vehicle could offer neither real-time, precision, high-reliability navigation information, such as position quickly, nor the functions of self-etermining, adaptability to the change of conditions, anti-jamming, miniaturizing and good property to price ratio. The running state of its pivotal parts can be monitored in real time and alarmed by using of a mass of the information about running state from the good monitoring system, which is useful for total capability of the vehicle being better known and some corresponding measures to enhance its dependability and security being adjusted by its operators.Therefore, in order to better realize real time monitoring on pivotal parts by using of a mass of the information about running state of the vehicle, the main achievements and innovations of the running state of monitoring system in the vehicle based on multisensor information fusion technology are expressed as:1. Analyze and induce the theories and methods of data fusion based on artificial neural network.2. The model of vehicle fault diagnosis system based on monitoring neural network is proposed, which has the advantages of the monitoring neural network and data fusion. It has the superiority which the sole neural network is unable to compare, and its deficiency may be able to make up of the fuzzy theory. The fuzzy reasoning system processing system can be possible to reduce the system order of complexity which the nerve network brings.3. This paper has analyzed the fuzzy judgment method, and blur judges the method and the fuzzy nerve network unifies, proposed one kind the synthesis judgment weight and the judgment matrix determination will unify in the fuzzy nerve network judgment method.and then, we combine fuzzy neural network with thismethod and give a new method to determine the weights and evaluation matrix.4. The train stylebooks come from fault trees and provide them to the neural network for study. The value consists of one part of the repository. The ES can undertake self-learning to expand knowledge library and raise the system 's adaptability and error-allowablity, and make the fault diagnosis more accurate. The BP(Back-Propagation) network's programming design and arithmetic steps with the sigmoid function is illuminated exactly in this paper, and discuss the fuzzy fault diagnosis based on BP network.