The Calculation Of Non-probabilistic Reliability Index Based On Intelligent Algorithm And Its Application

Probabilistic and fuzzy reliability models rely on probability theory and fuzzy mathematics as the theoretical basis. The two models need a lot of experimental data as the basis for the process of the realization. When the experimental data is not sufficient, the non-probabilistic reliability theory is a good supplement to the first two methods. The non-probabilistic reliability index can be determined when only knows the upper and lower bounds of random variables.In the calculation of non-probabilistic reliability index, the optimization method or the one-dimensional optimization algorithm is generally used. But the optimization method can bring the error due to the "interval expansion", and the one-dimensional optimization algorithm has exponential growth in the number of random variables. In order to overcome the disadvantages of the two methods, the one-dimensional optimization algorithm and particle swarm optimization algorithm has been combined, and the optimization model is established to calculate the non-probability reliability index by the way of optimizing iteration. In the non-probabilistic reliability optimization design, reliability constraints is obtained by optimization method. Because of the error of the optimization method, the traditional non-probabilistic reliability optimization design is not accurate. In this paper, the optimization model is introduced to the non-probabilistic reliability optimization design, and a double layer nesting optimization model is established. Outer layer optimization is used to optimize the objective function, and the inner layer optimization makes the design variables satisfy the requirement of the non-probabilistic reliability index. Through a series of numerical examples and simple engineering examples, the method is proved to be simple and accurate, and it has a certain feasibility.The crane as a large common mechanical equipment, occupies an important position in the process of industrialization in the country. However, due to the small volume of the crane equipment production, the design of crane in the same working has very few data for consultation. It is not suitable to use the traditional methods of reliability design. Therefore, this paper applies the new non-probabilistic reliability optimization design model to design the main girder of the crane, and the combination of the theoretical method and the actual production is successful.