Abstract: Anaerobic fermentation, as a highly efficient organic waste treatment technology, is able to convert agricultural waste into biogas for resource recycling and sustainable energy supply. The anaerobic fermentation process is affected by factors such as the carbon to nitrogen ratio of the reaction substrate, pH, volatile fatty acids, ammonia nitrogen concentration and chemical oxygen demand. Traditional experimental methods often require a large investment of time and resources, so exploring more efficient optimisation methods has become a hot research topic. In this paper, the gas production experiment of anaerobic fermentation with mixed raw materials was carried out, and the experiments were carried out with the ratios of cow dung and corn stover in the reaction  substrate in accordance with 1:1, 2:1, and 3:1, and three sets of parallel experiments were set up to ensure that the results of the experiments were reliable and reproducible. The experimental results showed that the highest methane production was achieved with a 3:1 ratio of cow dung to corn stover, followed by a 1:1 ratio experimental group, and the lowest with a 2:1 ratio experimental group. In this study, we created the Sparrow Search Algorithm (SCSSA) optimised with positive cosine and Cauchy's variation strategy, and optimised its hyperparameters for convolutional bi-directional memory neural network (CNN-BiLSTM), which solved the problem of the model's tendency to converge prematurely into local optimums and improved the global search capability. The experimental data were measured, and the reaction time, ratio of cow dung to corn stover, pH, volatile fatty acid, ammonia nitrogen concentration, and chemical oxygen demand were selected as the input parameters of the model, and the daily gas production and daily methane production were selected as the output parameters. The obtained gas production prediction model based on SCSSA-CNN-BiLSTM mixed feedstock anaerobic fermentation gas production model had an accuracy of 95.29% for daily gas production and 95.87% for daily methane production, and the goodness-of-fit R² reached 0.972.