|
With the reliability design theory, mechanical design technology, large-scale agricultural machinery operating capability is growing. But with the structural complexity of agricultural machinery increased reliability problems have become increasingly prominent. Therefore, how to allow economic and technical conditions, a reasonable assessment of the reliability of the mechanical system, so as to improve the mechanical system reliability, maintainability and maintenance strategies to provide a reference, is an urgent and significant issue. Poor reliability, failure and more are currently in the process of agricultural machinery in operation a common problem. In this paper, recent production of crawler tractors DFH 1002 as the research object to reliability engineering methods, artificial neural networks, fuzzy mathematics and other subjects combined method, the use of a tractor reliability theory and its application. In the theory of fuzzy neural network research, based on the tractor reliability evaluation conducted in-depth research, the specific findings are as follows: 1. Article describes the significance of the study as well as the reliability reliability evaluation studies on the importance of reliability. In this paper, the existing domestic and international evaluation method based on fuzzy theory are reviewed. 2 pairs of fuzzy theory and artificial neural network technology has been systematically studied, focused on the integration of the two issues, combined with the reliability of the evaluation questions, to build a five-feedforward fuzzy neural network model. The evaluation model includes an input layer, membership function layer, the rule layer, in the final layer and output layer. Application interpolation algorithm to generate the fuzzy evaluation sample processing, and then use feedforward network learning algorithm for BP network training, network training, learning can be obtained after completion of a tractor reliability evaluation results. 3 Use the fault tracing test data obtained reliability, maintainability, availability characteristic quantity of observations. Tractor conducted on the reliability of fuzzy comprehensive evaluation and analysis of four would be based on fuzzy neural network reliability evaluation model used tractor reliability evaluation, select six tractors actual work for the evaluation of the reliability index data objects, through the design of the Matlab language program, and ultimately get a level of reliability of the model tractor, to better ensure the objectivity of the evaluation results. Mechanical system reliability evaluation is a more integrated areas of research, involving many related disciplines, its engineering prospects are broad, theoretical value is significant. With the deepening of theoretical research and engineering application of the urgent needs of research in this area will also further further.
|