M. Heiss, D. Heiss, S. Kampl Lernen linear interpolierter Kennlinien, Automatisierungstechnik at, Vol. 42, No. 8, 1994 These study theses are only available at the library of the Institute of Control Engineering at the Technical University of Darmstadt.  Home/Search   Document Not in Database   Summary   Related Articles

....update. However, simulations with noisy signals have proved, that a gain fl 1 leads to higher robustness and attenuated convergence of the algorithm. It can be shown that the adaptation rule which is given by equations 6, 7 and 8 is equivalent to the normalized LMS algorithm introduced by Heiss [8] for the adaptation of one dimensional input output maps. Criteria for the convergence of the algorithm are also derived in [8] A.2 The MILL Algorithm The associative datafield allows the insertion of datapoints at arbitrary locations in the input space. The mean (or low pass filtered) values of ....

....of the algorithm. It can be shown that the adaptation rule which is given by equations 6, 7 and 8 is equivalent to the normalized LMS algorithm introduced by Heiss [8] for the adaptation of one dimensional input output maps. Criteria for the convergence of the algorithm are also derived in [8]. A.2 The MILL Algorithm The associative datafield allows the insertion of datapoints at arbitrary locations in the input space. The mean (or low pass filtered) values of the controller output (teacher signal) and the setpoint variables can be considered as a new datapoint that represents the ....

M. Heiss, D. Heiss, S. Kampl Lernen linear interpolierter Kennlinien, Automatisierungstechnik at, Vol. 42, No. 8, 1994 These study theses are only available at the library of the Institute of Control Engineering at the Technical University of Darmstadt.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC