Optimization of spatial pipeline with multi-hoop supports for avoiding resonance problem based on genetic algorithm.

Abstract

The reasonable layout of hoops can effectively avoid the excitation frequency of engine rotors and greatly reduce the vibration level of pipeline systems. In this study, a spatial pipeline supported by multi-hoops was taken as the object, the method of using genetic algorithm to efficiently obtain the optimal layout of hoops to avoid resonance was investigated. The finite element model of the pipeline system was created as the basic model of optimization, spring elements were applied to simulate the mechanical characteristics of hoop and the influence of spring element direction on the vibration characteristics of pipeline system were mainly described. In the optimization of avoiding resonance for spatial pipelines, the optimization goal was to maximize the first-order natural frequency, the positions of the hoops were converted into the node number as design variables, and the final optimization model of pipeline to avoid resonance was determined on the premise of reasonably setting of constraint conditions for design variables. The genetic algorithm was utilized to solve the optimization model, and two optimization methods were proposed, which were named as "genetic algorithm calling finite element model" and "genetic algorithm updating stiffness matrix" respectively. Finally, a case study was carried out to display the proposed methods. The maximum deviation between the calculation and the test results is less than 1.5% for the first three order natural frequencies, which proves the rationality of the created finite element model of spatial pipeline. Furthermore, the optimization practices show that the reasonable hoop layout of the pipeline system can be obtained by the two optimization methods, but the efficiency of the optimization performed by "genetic algorithm updating stiffness matrix" is much higher than that of "genetic algorithm calling finite element model".