Optimal path planning for field operations
Jan Willem Hofstee, L.E.E.M. Spätjens, H. IJken · 2009
Path planning for field operations becomes more and more important. More work is done by workers not knowing the field from experience. A second reason is that in the future more and more operations will done by autonomous vehicles and they require a path. Path planning for rectangular fields is rather simple but for more complex shaped fields tools are needed to support the planning process. GPS developments also enable that more difficult solutions can be realised in practice. With Matlab a tool is developed that reads the boundary coordinates of a field, determines the real vertices, divides the field into convex subfields if necessary, and calculates the costs for different operating directions to find the direction with the lowest costs incurred. Working time is converted to costs to enable the choice to not operate a part of the field for some reason, for example too small in relation to the effort. The tool is tested on some real fields. The results for simple fields are expected. The most optimal direction is the direction parallel to the longest side of the field. For more complex fields that are divided in two or more subfields the solutions are optimal for the individual subfields but the solution for the whole set of subfields is not necessarily optimal because for this interactions between subfields have to be taken into account too. Also, situations where tramlines are not perpendicular to headlands, resulting in small parts of the field either operated twice or not operated at all, have to be taken into account. The developed tool is a good first start but has to be elaborated more to be able to handle more complex field situations and to deliver for these fields also realistic optimal solutions.