Methods for Estimating Coasting Resistance of Heavy-Duty Commercial Vehicles under Various Loads

This research primarily addresses the issue of resistance model setting for chassis dynamometers or EIL (engine-hardware-in-the-loop) systems under various loads. Based on the data available from the heavy-duty commercial vehicle coast-down test reports, this article proposes three methods for estimating coasting resistance. For heavy-duty commercial vehicles that have not undergone the coast-down test, this article proposes the GA-GRNN (AC) model to predict coasting resistance. Compared to the GA-BPNN model proposed by previous studies, the new model, which achieves 93% prediction accuracy, demonstrates higher estimation accuracy. For heavy-duty commercial vehicles that have undergone the coast-down test, the coasting equal power method proposed in this article can estimate the coasting resistance under various loads. The accuracy and stability of the new method are verified by several coast-down tests. Compared to the existing method proposed by existing scholars, the new method has a higher estimation accuracy, thus compensating for the limitations of the coast-down test in measuring the coasting resistance. When neural network is combined with the coasting equal power method, they not only overcome the limitations of neural network predictions for coasting resistance but also compensate for the limitations of the coasting equal power method in estimating coasting resistance. Ultimately, the methods proposed in this article provide a feasible solution for setting resistance models of chassis dynamometers or EIL systems under various loads, without relying on coast-down tests.