Modal filters may be obtained by a properly designed weighted sum of the output signals ofan array of sensors distributed on the host structure. Although several researchgroups have been interested in techniques for designing and implementing modalfilters based on a given array of sensors, the effect of the array topology on theeffectiveness of the modal filter has received much less attention. In particular, it isknown that some parameters, such as size, shape and location of a sensor, are veryimportant in determining the observability of a vibration mode. Hence, this paperpresents a methodology for the topological optimization of an array of sensorsin order to maximize the effectiveness of a set of selected modal filters. This isdone using a genetic algorithm optimization technique for the selection of 12piezoceramic sensors from an array of 36 piezoceramic sensors regularly distributedon an aluminum plate, which maximize the filtering performance, over a givenfrequency range, of a set of modal filters, each one aiming to isolate one of the firstvibration modes. The vectors of the weighting coefficients for each modal filter areevaluated using QR decomposition of the complex frequency response functionmatrix. Results show that the array topology is not very important for lowerfrequencies but it greatly affects the filter effectiveness for higher frequencies.Therefore, it is possible to improve the effectiveness and frequency range of a set ofmodal filters by optimizing the topology of an array of sensors. Indeed, using 12properly located piezoceramic sensors bonded on an aluminum plate it is shownthat the frequency range of a set of modal filters may be enlarged by 25–50%.