Prediction of long-term trends in biomass energy development suitability and optimization of feedstock collection layout based on deep learning algorithms

In the context of the global energy crisis, the utilization of biomass energy has garnered increasing attention. Biomass energy development suitability serves as a crucial criterion for assessing whether a region is capable of or suitable for developing biomass energy. Accurate predictions of biomass energy suitability provide a strategic direction for regional biomass energy development and utilization. This study investigates biomass energy development suitability. It focuses on the supply-demand relationship and the spatial distribution of raw material collection. Utilizing Python programming language, we constructed LSTM, LSTM-ARIMA, and Transformer time series prediction models, along with a k-means deep learning clustering analysis model. These models were employed to conduct in-depth learning for data prediction and point clustering analysis. The research area chosen was Tongxu County, Henan Province, China, where six time nodes were established for biomass energy development. ArcGIS was employed to analyze and map the spatial layout of feedstock collection across these six milestones, offering some strategic insights for biomass energy development in Tongxu County. The results indicate that both the supply-demand relationship and the spatial layout of biomass collection can effectively evaluate the suitability of local biomass energy development. Regions with a shortage of biomass supply are suitable only for local development. As surplus increases, they can transition to widespread development. For example, in Tongxu County, the previously tight supply-demand landscape is gradually alleviated, with significant improvement in feedstock collection predicted by 2051. A surplus increase of 1.98 × 10⁹ MJ is expected between 2061 and 2071, with biomass energy development suitability transitioning to widespread development between 2063 and 2068. The paper concludes by proposing a strategy for the establishment of secondary feedstock collection and storage facilities with targeted and mobile collection and transportation mechanisms.