Research On Aeroengine Fault Diagnosis Expert System Based On Endoscopic Image Analysis

The new aircrafts continue to arm forces and the training intensity increases year by year.The environment of airport is unfamiliar and harsh.These objective conditions cause a great challenge to the aeroengine maintenance.As an important means of aeroengine damage detection,borescope technology has been very mature.In the outside, the detection of aeroengine mainly takes borescope technology which is one of nondestructive testing methods.You can get some image of borescope inspection, but the guidance of maintenance often can not get directly by these images.The analysis of data and expert diagnosis is must. To solve these problems, analysis of borescope image and expert system diagnosis are combined. Aeroengine blade damage around the feature point extraction and internal crack damage measurement automatically are investigated.The investigation eliminates the influence of human factors on the measurement results and diagnosis results. Accordingly, the efficiency of diagnosis is improved.(1) Multiple diagnostic methods are used in combination to achieve engine maintenance decisions. Conversion the "engine diagnostic testing procedures" and "engine technology uses rules" and experience extracted by endoscopic detection engine expert into a knowledge rule base, inference rules are applied to replace manual inspection, saving the time of diagnosis. Moreover, engine fault case base which is made of fault description and fault solution is build.The case searching and maintenance methods are build.(2) A automatic measurement method of damage leaf is proposed based on cubic spline interpolation.Extract the damage blade profile curve by image preprocessing methods, then use cubic spline interpolation method locate the feature points and calculate size.This method can effectively identify and measure damage.To a certain extent, the measurement error problem is solved which due to differences of borescope personnel,and fault diagnosis efficiency is improved.(3) A method of aeroengine internal cracks automatic measurement is presented. Length of the crack is extracted and calculated by crack split, crack identification and crack measurement, and search rules in the knowledge base to diagnose. To a certain extent, this method solves the problems that crack inspection rely on experience, measurement deviation and low efficiency.(4) An aeroengine internal damage intelligent diagnosis expert system(AIDES) with thei View PC is developed by using the development tools of Microsoft Visual C++6.0 language andMicrosoft Access 2003 database.It achieved intelligent diagnosis, automatic and semi-automatic.