As the core power unit of complex electromechanical systems, accurate health assessment of diesel engines is essential for safe operation. The Interval Belief Rule Base (IBRB) method integrates observed data with expert knowledge to support system assessment. However, engine operating parameters change over time because of wear and aging. Additionally, traditional optimization methods struggle to balance global search speed with local convergence efficiency. To address these issues, this paper proposes an Interval Belief Rule Base method based on Hybrid Optimization and Adaptive Intervals (IBRB-HOAI). First, an adaptive reference interval is introduced by combining K-means clustering and quantile interval estimation, dynamically generated based on the actual operating state of the engine. The health assessment baseline is optimized. The applicability of the model is enhanced. Second, the global exploration ability of particle swarm optimization is combined with the local refinement ability of the projected covariance matrix adaptation evolution strategy. The model parameters are collaboratively optimized. Finally, experimental verification is conducted on a diesel engine dataset containing 2700 sample points. Compared with the traditional IBRB method, the proposed method achieves a significant reduction in MSE of 97.5%. It outperforms other machine learning methods. The effectiveness of the proposed method is verified.
Zheng et al. (Fri,) studied this question.