عنوان مقاله English
نویسندگان English
Dynamic model identification and state variables estimation from the corrupted measurement data have been attracted much research efforts during the recent years. In this way, Kalman and H-infinity filters have been increasingly used to estimate the parameters individually. In this paper, a mixed kalman-H_∞ filter is designed in an innovative approach using a multi-objective optimization method. It is desired to simultaneously employ the advantages of both filters to minimize both the root-mean squared errors and the upper bounds limit of estimation errors associated with Kalman and H-infinity filters, respectively. Some Pareto optimum design points are presented for two case studies from which trade-off optimum design points can be simply selected.
کلیدواژهها English
