Straw pretreatment enhances the cellulose accessibility and increases the methane yield from anaerobic digestion. This study investigated the effects of alkali pretreatments with different chemical agents (NaOH, KOH, and Na2CO3) on the physicochemical and thermal characteristics of barley straw, as well as methane production from codigestion with sewage sludge. Artificial neural network modeling with a feedforward neural network (FFNN) and slime mold optimization (SMO) techniques were used to predict methane production. NaOH pretreatment was shown to be the best pretreatment for removing hemicellulose and lignin and for increasing the cellulose accessibility. Moreover, there was a 2.57-fold higher level of methane production compared to that from codigestion with untreated straw. The removal ratios for the total solids, volatile solids, and chemical oxygen demand reached 59.3, 67.2, and 73.4%, respectively. The modeling results showed that the FFNN-SMO method can be an effective tool for simulating the methane generation process, since training, validating, and testing produced very high correlation coefficients. The FFNN-SMO accurately predicted the amount of methane produced, with an R2 of 0.998 and a 3.1x10-5 root mean square error (RMSE).