Multi-objective Optimization of Sensor and Excitation Layouts for Frequency Response Function-Based Structural Damage Identification

The accuracy of many damage identification methods depends significantly on the quality of measurements collected by sensors, such as accelerometers, concerning the response characteristics of a structure. Often the number of sensors used to collect measurements is limited due to available funds, equipment, and access. In addition, the excitation location can significantly affect a sensor's ability to collect quality measurement information. Therefore, both the location and number of sensors and the location of the excitation must be optimized to maximize the quality of information collected. A multi‐objective optimization approach is presented that minimizes the number of sensors specified while maximizing the sensitivity of the frequency response functions (FRFs) collected at each specified sensor location with respect to all possible damaged structural elements. The multiple Pareto‐optimal sensor/excitation layouts obtained aid in determining the number of sensors required to obtain an effective level of measurement information. The benefit of using Pareto‐optimal sensor/excitation layouts is investigated by using the optimized layouts to collect measurement information for a FRF‐based structural damage identification method. Trial results confirm that an increase in damage identification accuracy and efficiency is achieved when Pareto‐optimal sensor/excitation layouts are used instead of nonoptimal layouts. In addition, the Pareto‐optimal layouts improved damage identification accuracy in noisy measurement environments due to increasing the quality of measurements collected.