LAUSR

Automated tuning can significantly improve productivity and save the costs of manual operation in the microwave filter manufacturing industry. This article proposes a mathematical model of scattering data optimization to find the accurate coupling matrix for multiple-version microwave filters, a core step of automated microwave filter tuning. For the large-scale problem of coupling coefficient combination, we propose a decision set decomposition strategy that evenly divides the entire frequency interval into several subintervals according to the correlation between scattering data. With this strategy, we design a microscale (small-size subsets of the decomposed decision set) searching algorithm, which solves each suboptimization problem by searching the decision subset instead of the entire decision set. To verify the validity of the proposed algorithm for multiple-version microwave filters, experiments are conducted on three versions of microwave filters from a real-world production line, including the two-port eighth-order, ninth-order, and tenth-order microwave filters. Experimental results show that the proposed model is feasible within the industrial error for the multiversion microwave filter tuning problem. Besides, the proposed algorithm outperforms the state-of-the-art optimization algorithms in the coupling matrix optimization problem.