Construction of Ship Main Engine Performance Evaluation System Based on Principal Component Analysis Method

Abstract

By monitoring changes in thermal parameters, the relationship between the operating status of ship diesel engines and thermal parameters can be reflected, thereby determining the current state of ship diesel engines. Nevertheless, the high complexity of thermal parameters renders their evaluation a challenging endeavor. Therefore, to accurately evaluate the performance of ship engines and reduce the difficulty of data processing, this paper proposes to construct a ship engine performance evaluation system through Principal Component Analysis (PCA). The system uses PCA to check its thermal parameters, thereby elucidating the substantial intrinsic correlation between different thermal parameters of diesel engine performance. The results showed that compared with the fuzzy entropy weight method, PCA was more accurate in evaluating the performance of the propagation host, with a maximum relative error of only 4.2%. For the testing host, PCA accurately detected issues such as high cylinder cooling water temperature, high cylinder liner temperature, high exhaust temperature, and low steam compressor speed. The fuzzy entropy weight rule made it difficult to reflect these issues accurately. In addition, PCA could accurately reflect the severity of the above-mentioned faults through outliers and deviation rates. Meanwhile, compared with the information entropy method, PCA had smaller errors, with an average error of only 2.8%. The above results indicate that PCA can accurately evaluate the performance of ship engines, providing a strong reference for ensuring the performance of ship engines and stable operation of ships.