Application of neural network to automobile engine failure detecting

Automobile industry has become the supporting industry of the main industrial countries by now. With automobilepsilas increasing repairing, perfecting, complicating and automatizing, the traditional failure detecting and repairing method can not meet current requirement for automobile shipment and repairing. By using intelligent neural network technique, people input yearspsila of repairing experience into computer which will have analysis and decision ability in automobile failure detecting similar to human beingpsilas brain. The technique is rapid, exact, reliable, and an important application in the domain of intelligent transportation, a new branch. The paper studies automobile engine failure detecting deeply, adopts improved BP neural network, sets up mathematics model of failures and effecting factors. In this way, failures can be forecasted. Simulation result indicates that the model has stronger self-study ability and better constringency characteristic than traditional algorithm. The forecasting result is exact and practical.