High Performance Optimization Algorithms And Model Updating Of Structures

The Finite Element Model (FEM) updating technique has been successfully applied in mechanical, aerospace and automotive industry. Although being based on the same theoretical beckground and using similar technological procedure, the FEM in civil engineering field has faced more challenges comparing with its counterpart in other fields. These challenges include the super size and nonlinearlity of model updating optimization problem due to the complexity of civil structures, the uncertainty of updating parameters and even the objective function due to the uncertainties and nonhomogenousness associated with structures, materials and environments. The optimization techniques, treatment of uncertainty involved in model udating and large-scale model updating of civil engineering structure are investigated in this paper, and the contents are outlined below.Three high efficient optimization methods for FEM updating are developed. (1) A hybrid optimization technique is presented, which draws together the global searching capability of chaos-based optimization technique and high searching efficiency of Trust-Region Newton Method. This hybrid approach is demonstrated to be more efficient and prone to the global minimum than conventional gradient search methods and random search methods; (2) An adaptive Coupled Local Minimizers (CLM) method is proposed by improving the original CLM in three aspects: the selection of initial population, the variable-step searching technique and the adaptive population. These amendments give an adaptive capability of CLM and enable the adaptive CLM to deal with large scale complex optimization problems; (3) An efficient multi-objective optimization technique for FEM updating is proposed based on Taguchi method. Improved Taguchi method has the ability to solve the multi-objective optimization problem by introducing the classification of searching population and new signal to noise ratio.We develop a hybrid finite element model to consider both the rotational and shear flexibility of joints in structures, which is parameterized by a degree of flexibility factor. A parameterizing scheme for boundary conditions updating is presented, by which the boundary conditions can be updated.An updating method based on Component Mode Synthesis (CMS) is proposed to improve the efficiency of updating FEM of complex structures. A condensed structural model is obtained by CMS technique, and then an optimal objective function is obtained, which consists of frequencies of condensed model and modal shapes transformed from condensed coordinates to initial physical coordinates. The updated model is acquired by solving the optimization problem. These method has made the FEM updating of large scale and complex structure possible.Considering the uncertainties involved in model updating, an FEM updating method using fuzzy interval analysis is proposed. The modal parameters and design variables are modeled as fuzzy variables to take into account their uncertainties. Therefore, the most sensitive parameters are selected as updating parameters by the interval eigenvalues sensitivity analysis, and further the interval of updating parameters are determined by interval arithmetic. With the fuzzification of objective function, numerous updated schemes are obtained associated with different degree of fuzziness of the objective function, and then the most physically-compatible solution is chosen by insights to the structures.Finally, the proposed methods for FEM updating are applied to Dongying Yellow River Bridge to demonstrate their validity. By applying the fuzzy updating method, the model updating problem is reduced, and the most meaningful updated model is obtained. The updated results obtained by three optimization techniques are consistent, and the updating approach based on CMS has an advantage of high efficiency and yet sufficient accuracy.