МОДЕЛИРОВАНИЕ ТРАЕКТОРИИ ДВИЖЕНИЯ АВТОМОБИЛЯ ПРИ ИССЛЕДОВАНИИ ДОРОЖНО-ТРАНСПОРТНЫХ ПРОИСШЕСТВИЙ

Abstract

The article proposes one of the ways to evaluate the trajectory of the cars movement in the study of the causes of the occurrence and reconstruction of the traffic accidents circumstances. The aim is to propose, based on the theory of fuzzy sets and experimental studies, a method for reducing the uncertainty of constructing the calculated trajectory of a vehicle during braking under inaccurate initial data. The basis of the proposed method is a mathematical model for evaluating the trajectory of a vehicle during braking, which allows for the redistribution of vertical reactions not only along the axes, but along the sides of the car, and makes it possible to determine the instant of occurrence of the skid of each wheel by the criterion j/g. A feature of the proposed method is the use of previously developed expert systems for evaluating the braking torques on the wheels of the car and determining the coefficient of adhesion of the car with the road. The use of fuzzy expert information on the importance of factors that affect brake moments and the coefficient of adhesion allows to reduce the volume of experimental studies and significantly shorten the time to reach an objective decision on the causes of emergencies. The discrepancy between the results of the expert system forecast and the experimental data is 5% and 3%, respectively. Presented are the results of experimental studies of the behavior of the car during braking in road conditions, taking into account the magnitude of the braking torques that were applied to the wheels of the car, the magnitude of the coefficient of adhesion along the sides of the car, as well as the transverse displacement of the center of mass of the car due to uneven distribution of cargo. The maximum error in the discrepancy between the simulation results and the experimental data does not exceed 8%. A reference algorithm for estimating the trajectory of a vehicle during braking is presented, which allows one to take into account the stochastic and indistinct uncertainty of the initial data and reduce the range of possible modeling error by 39%. Investigation of road accidents related to changing the trajectory of the car during braking, confirmed the possibility of applying the proposed method in the auto-technical examination.