ARTIFICIAL NEURAL NETWORKS APPROACH IN MICROWAVE FILTER TUNING

This paper presents a novel method of cavity fllter tuning with the usage of an artiflcial neural network (ANN). The proposed method does not require information on the fllter topology, and the fllter is treated as a black box. In order to illustrate the concept, a feed-forward, multi-layer, non-linear artiflcial neural network with back propagation is applied. The method for preparing, learning and testing vectors consisting of sampled detuned scattering characteristics and corresponding tuning screw deviations is proposed. To collect the training vectors, the machine, an intelligent automatic fllter tuning tool integrated with a vector network analyzer, has been built. The ANN was trained on the basis of samples obtained from a properly tuned fllter. It has been proved that the usage of multidimensional approximation ability of an ANN makes it possible to map the characteristic of a detuned fllter re∞ection in individual screw errors. Finally, after the ANN learning process, the tuning experiment on 6 and 11-cavity fllters has been preformed, proving a very high e-ciency of the presented method.