Intelligent information system for resource planning in grain crops delivery projects on the basis of machine learning

The article proposes an algorithm for resource planning in projects of grain crops delivery. The algorithm proposes to select resources based on the criteria of the minimum work cost. A random forest model is used to forecast the specific fuel consumption. The architecture of an intelligent information system for resource planning in grain crops delivery projects is proposed, this system consists of two subsystems: machine learning for forecasting fuel consumption and decision support for resource planning and user interaction. To accelerate and improve the quality of management decision-making, an intelligent system has been developed based on the proposed architecture. The user interaction is implemented using a graphical interface for PCs in Python 3.9 language. The presented intelligent information system has been tested for adequacy. As a result of using the proposed IIS, fuel consumption is saved in the amount of 6296.5 litres per month, the time for making a management decision is reduced by 1.43 minutes or 95.6%, and the costs of the motor transport enterprise are reduced by UAH 333.6K and 6.3%.