Optimal Sensor Placement for High-Rise Building via Genetic Algorithms and Improved Information Matrix Criterion

Abstract

To improve the accuracy of parameter identification and to reduce the test cost in the high-rise structure health monitoring systems , the paper proposes an optimal placement method combines genetic algorithms with improved information matrix criterion (GA-IIM). With traditional improved effective independent method (EFI-DPR), it is easily trapped into bureau of optimum solution, and the proposed method can resolve this problem. The solutions were coded by a two-dimensional integer array, each line stored a feasible solution and compulsory mutation was adopted to avoid the solution reappeared in the same position. To improve the convergence speed, each generation of optimal group was kept to the next one to participate in the competition. With the improved information matrix criterion embedded in the genetic algorithms, the placement design was produced. This technique was used to optimize the sensor placement of a nineteenth layer frame structure. Comparing the optimized sensor placement of the proposed method with that of improved effective independent method by the modal assurance criterion and the mean square error, the proposed method performs better.