Robust estimation of Ackerman angles for front-axle steered vehicles

Driving a tractor in the field can be especially hard and tedious due to the repetitive nature of agricultural operations. Harvesting, spraying, fertilizing, and planting are common tasks that require full attention of farm vehicles operators for long periods of time, usually leading to physical and mental fatigue in drivers. In addition to the benefits of avoiding stressful attention-demanding driving between rows of plants or trees, intelligent vehicles (i. e. vehicles endowed with artificial intelligence techniques) with automated steering allow operators concentrate on alternative tasks and vehicle functions.

In general, driving assistance improves efficiency in the use of machinery when overlaps are reduced or eliminated, and it can also increase operation safety and performance of inexperienced drivers. However, the automation of agricultural equipment also induces the occurrence of serious risks that need to be taken into account. Machines that typically are very heavy, oversized, and run by powerful diesel engines must never operate without strict safety rules. The principal parameter providing feedback to the actuation of a front-axle guided vehicle is the measurement of the actual angle that the front wheels turn after executing computer-generated commands, as the vehicle patented by Mailer [1], which implemented a PID controller getting feedback information from a wheel-angle sensor.

These traditional controllers have been successful for driving assistance in the field; even when the four wheels of agricultural platforms are individually driven and steered, the control inputs for the steering motors have been computed using PID control laws [2]. When accurate steering is necessary, the traditional loop may be expanded as the double closed-loop PID control law proposed by Xiaopeng et al. [3] to reduce steering overshoots and keep errors below 1o. In this approach, the wheel steering angle was used as the feedback for the inner loop, and the angular velocity of the wheel as the feedback for the outer loop. The advent of steer-by-wire systems allows the application of variable steering ratios.

The steering ratio is the quotient between the steering wheel angle and the front wheel angle, and it is fixed for traditional steering systems. Jiang-Yun et al. [4] brought an alternative to traditional control methods based on feedback by implementing a steer-by-wire system in which the front wheel angle is controlled directly with the steering ratio. This option resulted in a simpler design. Regardless of the control law implemented, the actual response of the front wheels will be the major responsible for the vehicle position. Therefore, the real-time determination of the true angle turned by the steering wheels is a matter of great importance.

However, conventional auto-steering systems tend to measure the angle of one of both front wheels and use the simplified bicycle model to emulate the dynamics of motion. The approach presented in this paper takes into account the special features of Ackerman steering and independently measures the angle of each wheel, using the properties of this geometry, and the redundancy of wheel angles, to make final estimations of Ackerman angles more robust and immune to sensor failure.

For full text: click here

(Author: Verónica Sáiz-Rubio, Francisco Rovira-Más, Ishani Chatterjee, José María Molina Hidalgo

Published by Sciedu Press)