Prediction model of tractor residual value coefficient considering factor of invisible wear

Abstract: Depreciation takes up a large proportion in the total activity cost of agricultural machinery. As the update speed of agricultural machinery was gradually accelerating, the impact of the invisible wears was strengthened by technological progress on agricultural machinery depreciation. Aiming to this problem, two types of invisible wears, which were caused by increased labor productivity and the emergency of more advanced machines, were analyzed and quantified respectively, and based on which, the comprehensive invisible wear degree was given in this paper. According to the traditional ASABE (American Society of Agricultural and Biological Engineers) residual value coefficient prediction model of tractors, considering the invisible wear factor, an improved prediction model of the residual value coefficient for tractors was established. Based on the auction data of 25 wheeled tractors collected from the agricultural machinery and tools auction results in the agricultural machinery service center, the age, the average annual working hours and the comprehensive degree of the invisible wears were used as the main independent variables to construct the improved model of the tractor residual value coefficient prediction and to analyze the law of the tractor value change. In addition, in order to increase the calculation accuracy, the discount coefficient was introduced to consider the time value of the capital. The results showed when the tractor age was above five years, the prediction value of the improved model was closer to the real residual value coefficient, and compared with the original model, the improved model was with less prediction deviation and higher prediction precision that was improved by more than 27%; but when the tractor age was within four years, the prediction deviation for the improved model was bigger than the original model, and the prediction result of the improved model was not satisfactory. The main reasons causing the results were that the effect of the invisible wears on the tractor was very small due to smaller technological progress in its early working period, but with the increasing of the tractor age and the development of science and technology, on one hand, the reproduction value of the same type of tractor would go down because of the improvement of manufacturing process and the increase of social labor productivity; on the other hand, the new type of machines, which were with more advanced technology, better performance, higher productivity and lower consumptions of raw materials and energy, would occur, so the effect of the invisible wears on the tractor residual value would intensified gradually, which also demonstrated that when predicting the residual value of agricultural machinery used for a relatively long period of time, the influence of the invisible wears must be considered. At the same time, the method of the geometry-based "Fréchet distance" metric was applied to further comprehensively compare and evaluate the prediction accuracies of the original model and the improved model. The results showed that compared with the original model, the improved model was with the higher prediction precision that was improved by 24.9%, and was able to significantly improve the accuracy of residual value prediction of tractors. The research of this paper can provide scientific and accurate decision-making basis for the calculation of depreciation cost and the determination of replacement time of agricultural machinery.