Study On Kernel Parameter Optimization In Rotating Machinery Fault Diagnosis Based On Improved BFA

Rotating machinery is the core equipment in modern industrial production, carrying out rotating machinery fault diagnosis technology research has important significance for economic and social to ensure such equipment running safe and efficient and to avoid the huge economic loss and catastrophic accidents, meanwhile, it can enrich and develop the technology of mechanical equipment condition monitoring and fault diagnosis.Thesis takes rotating machinery as the objects, carry out rotating machinery fault diagnosis research based on the typical kernel methods—kernel principal component analysis (KPCA) and support vector machine (SVM), for the problems that kernel function and its parameters have a great influence on KPCA and SVM but the optimal parameters are difficult to select, this thesis proposed an improved bacterial foraging algorithm (BFA) to carry out the research of kernel parameters optimal selection for the rotating machinery fault diagnosis. The main contents are as follows:1.The problems exist in standard BFA were analyzed, such as population size, step length, the number of iterations depend on the maximum number of the various of operations, because not introduced convergence criteria, it is difficult to guarantee the accuracy and increase the unnecessary iterative process, an improved BFA was proposed, the simulation experiment of two-dimensional continuous functions prove the improved BFA can not only improve the optimal speed, but also improve the accuracy of solution.2. The influence of kernel parameters on KPCA feature reduction was analyzed, then the algorithm of KPCA feature extraction based on the improved bacteria foraging algorithm was designed, the rotating machinery fault feature example shows that the method can rapidly and accurately to optimize KPCA kernel parameters.3. SVM parameter optimization algorithms based on improved BFA was proposed, by analysis and comparison the optimization performance of improved BFA, standard BFA and other optimization algorithms including genetic algorithm (GA), particle swarm optimization (PSO) and cross validation (CV), the results show that the improved algorithm is superior to others.4. The KPCA based on improved BFA and the SVM based on improved BFA were applied to fault diagnosis of rotating machinery, and achieved the rolling bearing fault diagnosis which based on base multi-sensor information fusion and the gear fault diagnosis which based on multi-sensor information fusion. The results proved the superiority of the method which proposed in this thesis, at the same time, the way which the sensor was installed on the base can overcome the problems of inconvenience and others, and provides a useful reference proposal that has great application and promotion prospects.