Automated generation of guidance lines for operational field planning

• Published in: Biosystems engineering 2010-12, Vol.107 (4), p.294-306
• Main Authors: Hameed, I.A., Bochtis, D.D., Sørensen, C.G., Nøremark, M.
• Format: Article
• Language: English
• Subjects: accuracy; Agricultural equipment; agricultural field operations; Agricultural machinery and engineering; agricultural machinery and equipment; Agricultural practices; Agronomy. Soil science and plant productions; Algorithms; area; Biological and medical sciences; Combinatorial analysis; crop management; Curved; fields; Fundamental and applied biological sciences. Psychology; Generalities. Biometrics, experimentation. Remote sensing; global positioning systems; kinematics; Mathematical analysis; Mathematical models; planning; Real time; Representations; satellites; shape; simulation models; steering systems; topography
• ISSN: 1537-5110; 1537-5129
• DOI: 10.1016/j.biosystemseng.2010.09.001

A number of advanced satellite positioning based systems for agricultural machines have been developed and are commercial available for aiding and supporting navigation efforts up to full auto-steering. Furthermore, in terms of the route planning for agricultural field operations, advanced methods based on combinatorial optimisation of fieldwork patterns have recently been introduced. In order to apply and implement these methods in the high-level control system of agricultural machines, an appropriate representation of the field as a geometrical entity made up of discrete geometric primitives, such as points, lines, and polygons is needed. Preferably, such a representation must be generated in real-time providing the input to the whole accuracy range of navigation systems as well as the range of operation types, equipment characteristics, and machinery kinematics. Here, a method for real-time generation of field geometrical representation for operational planning is presented. The representation regards simple or complex fields for both convex and non-convex field boundaries, where generated tracks can be straight or curved. As demonstration cases for the method, 15 different field types were evaluated. According to the experimental results, the computational time of the method was in the range of 0.11–239.73 s for the case of single-block fields and in the range of 2.24–402.59 s for multiple-blocks fields. The tested fields were of different shapes and the area ranged from 0.21 ha to 44.93 ha. ► Algorithmic generation of parallel, straight or curved, field-work tracks. ► Generation of guidance lines for auto-steering supported field vehicles. ► Formalised mathematical description of 2-D field geometrical representation. ► Assessment tool for the effectiveness of alternative driving directions in a field.