Automation of motor grader working body positioning in the formation of the design profile of a highway

Abstract

The accuracy of forming the design profile of a subgrade is regulated by Table 2 of GOST R 59864.1-2022, which became the sole basis for geometric acceptance after Annex A was removed from SP 78.13330.2012 by Amendment № 3. A discrepancy between the rated performance of on-board motor grader automation equipment and the actual geometry of the finished surface persists even with fully serviceable hardware, while the quantitative distribution of the total error among its sources has not been established. The aim of the study is to build a model of blade cutting-edge positioning that combines the measuring, actuating and information channels, and to distribute the elevation error among the sources. The kinematics of the blade suspension were linearised about the working position; the disturbance caused by the micro-relief was represented as the sum of two Markov processes with correlation intervals of 12,0 and 1,2 m; the electrohydraulic loop was described by a first-order lag; Monte Carlo simulation with 10,000 runs was performed for five loop configurations and three working speeds, followed by a factor decomposition of the variance. A planar reference with a 50 m working section and a vertical curve radius of 10,000 m produces a systematic deviation of 31,3 mm and fails to meet the ±30 mm tolerance. A spatial system operating on the current error leaves 24,98% of control points outside the tolerance; feedforward correction from the digital profile model reduces this share to 11,23%, a 2.5-fold reduction of the drive time constant brings it to 10,22%, and the integration of satellite and inertial channels brings it to 1,03%. The share of the reference measuring channel in the total variance rises from 50.8 to 93,2%. The cross slope is held with a standard deviation of 0,66-0,93‰ against a ±10‰ tolerance, because the reference channel error is common to points of the same cross section and cancels in the difference of elevations. The positioning accuracy of the working body is limited successively by the geometric adequacy of the reference model and by the reference measurement error, rather than by drive response; the lower bound of about 12 mm is set by the accuracy of the geodetic setting-out network.