Development of a representative operation cycle characterized by dual time series for bulldozers

Abstract

Driving cycles have been developed for various types of vehicle by different nations and in different areas, as they have a substantial effect on analysis of the fuel economy and the emissions. As the concern about the fuel consumption and the emissions of engineering machinery increases continuously, it has become necessary to develop corresponding operation cycles for engineering machinery. However, a typical operation cycle for bulldozers and the methods for its development is still lacking. Therefore, a representative operation cycle for bulldozers was developed in this study. By taking advantage of readily available data from the Controller Area Network (CAN), large amounts of cycle experimental data were acquired in a typical bulldozing process. Two parameters, namely the bulldozing resistance and the speed, were employed to represent the operation cycle. The values of these parameters were calculated on the basis of the dynamic model and the kinematic model combined with system identification methods. Experimental cycles were divided into operation segments according to the respective operating processes, and characteristic parameters for the operation segments were chosen and calculated accordingly. The optimal representative operation cycle was finally selected on the basis of the smallest Mahalanobis distance. The fuel consumption and the probability distributions of the representative operation cycle were also compared with the average fuel consumption and probability distributions of all the operation cycles and analysed. The average correlation coefficient of the probability distributions was 0.936, whereas the difference in the fuel consumptions was only 1.786%. This indicates that the developed cycle is indeed appropriate for representing the operating process of the bulldozer.