An ALNS-based heuristic for the Crop Planting Schedule Problem

In the Syngenta Crop Analytics Challenge, we have been faced with the problem of crop planting schedule optimization. By analysing the problem structure and comparing it with other similar optimization problems, we have concluded that the given problem is very likely to be an NP-hard problem. Therefore, given the size of the problem instances to be solved, we have developed a heuristic algorithm based on the Adaptive Large Neighborhood Search metaheuristic. In order to determine the harvest date, we also needed to predict the
time of accumulating the enough amount growing degree units. For this purpose, we utilized an ARIMA model. The models, solution algorithm and obtained results are presented in the report.