Research On Performance Evaluation Of Mechanical Products

This thesis applies the methods of evaluation in the field of mechanical engineering and summarizes the procedure of performance evaluation for mechanical products. New evaluation approaches are proposed, which make the evaluation exacter and more dependable. In this thesis, the following key problems are studied.In order to minimize the correlation among all indicators, based on mathematic statistic methods, the synthetic function of every indication is given.Based on the summarization of standardization for indicators, we put the methods of summarization into four classes. In this way, not only the classes of indicators are predigested, but also the factual needs of performance evaluation for mechanical products are meted.The theory and method of application of entropy fuzzy comprehensive evaluation method in performance evaluation of mechanical products is studied. On the basis of entropy, the objective coefficient can be gotten. The compositive coefficient can be gained by combining the subjective coefficient and the objective coefficient. From the results of fuzzy comprehensive evaluation for performance evaluation for mechanical products, we can get the gradation of the mechanical products' performance.According to the actual circumstance of performance evaluation for mechanical products, the general neural network is reformed. Therefore, we construct the fuzzy neural network. This paper expresses membership function in the network by trapezium formulae .The indicators are fuzzed using the membership function .The network is trained using the top points and midpoint of membership function and the improved back propagation algorithm. Then we can get the adaptive network coefficients.The software system of performance evaluation for mechanical products is developed, which can complete the routine and fuzzy performance. Those can help the decisionmakers to get the correct conclusions.Finally, an example for loading machines of model ZL50 is then presented to verify these methods in this thesis.