---

This paper is concerned with the determination of optimal sensor locations for structural modal identification in a strap-braced cold formed steel frame based on an improved genetic algorithm (IGA). Six different optimal sensor placement performance indices have been taken as the fitness functions two based on modal assurance criterion (MAC), two based on maximization of the determinant of a Fisher information matrix (FIM), one aim on the maximization of the modal energy and the last is a combination of two aforementioned indices. The decimal two-dimension array coding method instead of binary coding method is applied to code the solution. Forced mutation operator is applied whenever the identical genes produce via the crossover procedure. An improvement is also introduced to mutation operator of the IGA. A verified computational simulation of a strap-braced cold formed steel frame model has been implemented to demonstrate the effectiveness and application of the proposed method. The obtained optimal sensor placements using IGA are compared with those gained by the conventional methods based on several criteria such as norms of FIM and minimum in off-diagonal terms of MAC. The results showed that the proposed IGA can provide sensor locations as well as the conventional methods. More important, based on the criteria, four of the six fitness functions, can identify the vibration characteristics of the frame model accurately. It is shown through the example that in comparison with the MAC-based performance indices, the use of the FIM-based fitness functions results in more acceptable and reasonable configurations.

---

Presenting structural damage detection problem as an inverse model-updating approach is one of the well-known methods which can reach to informative features of damages. This paper proposes a model-based method for fault prognosis in engineering structures. A new damage-sensitive cost function is suggested by employing the main concepts of the Modal Assurance Criterion (MAC) on the first several modes’ data. Then, Chaotic Imperialist Competitive Algorithm (CICA), a modified version of the original Imperialist Competitive Algorithm (ICA) which has recently been developed for optimal design of complex trusses, is employed for solving the suggested cost function. Finally, the optimal solution of the problem is reported as damage detection results. The efficiency of the proposed method for damage identification is evaluated by studying three numerical examples of structures. Several single and multiple damage patterns are simulated and different number of modal data are utilized as input data (in noise free and noisy states) for damage detection via suggested method. Moreover, different comparative studies are carried out for evaluating the preference of the suggested method. All the obtained results emphasize the high level of accuracy of the suggested method and introduce it as a viable method for identifying not only damage locations, but also damage severities.