Work

Sequential sparse Bayesian learning of long-term thermomechanical performance of energy piles from limited multi-parameter monitoring data