基于组合聚类和马尔科夫链的矿井无轨胶轮柴油车发动机瞬态工况构建OA

To accurately reproduce the emission characteristics of trackless rubber-tyred diesel vehicles in underground mines,a transient operating condition construction method based on combined clustering and Markov chain was proposed to address the insufficient adaptability between standard driving cycles and complex underground traffic scenarios.Data preprocessing and kinematic fragment division were performed on data acquired by controller area network(CAN)bus.Subsequently,the Silhouette coefficient and principal component analysis(PCA)were introduced into the traditional K-means clustering framework,and the operating conditions were synthesized by the Markov chain method.The synthesized operating conditions were compared with the published road driving conditions and the Euclidean distance synthesized operating conditions.The results showed that the mean error of the pedal position under the Markov chain algorithm was less than 1％.The innovatively constructed comprehensive parameter value(CPV)synthesized by the Markov method is 12.81 percentage points lower than that of the Euclidean distance synthesized operating condition,which can more effectively reflect the comprehensiveness of the underground driving conditions.