In this paper, we investigate the cycle-based ammonia-coverage-ratio reference generator design for Diesel engine two-cell selective catalytic reduction (SCR) systems. Since the dynamic programming algorithm is time-consuming, the motivation of this research is to develop the ammonia-coverage-ratio reference generator which has the capacity to produce the optimal desired ammonia coverage ratios onboard. With the sizing optimization results obtained from the dynamic programming algorithm, we propose an ammonia-coverage-ratio reference generator by using the adaptive-network-based fuzzy inference system (ANFIS). In the generator, the inputs include the engine-out NOx concentration, the ammonia concentration between the two cans, and the ammonia concentration in the tailpipe. The performance of the generator is comparable to the optimized results using the dynamic programming algorithm. (C) 2015 Elsevier Ltd. All rights reserved.