SISTEMA WEB PARA AUXÍLIO NO PLANEJAMENTO DE ROTAÇÕES DE CULTURAS

Abstract

Crop rotation is one of the most relevant agronomic practices for the sustainability of production systems, yet its efficient planning involves multiple interdependent variables — agronomic effects between preceding and succeeding crops, operational constraints, price volatility and income risk — which make manual decision-making prone to simplifications and suboptimal outcomes. This work presents the development of a web system to support crop rotation planning, implementing the Target MOTAD mathematical model proposed by Salassi, Deliberto and Guidry (2013), which couples a transportation-type network flow formulation with parametric income-risk control. The system was calibrated with ten historical crop seasons (2015/16 to 2024/25) from the state of Goi´as, considering soybean, maize and cotton, based on yield, price and production-cost data from CONAB and on agronomic yield-adjustment factors grounded in Embrapa technical bulletins. The architecture follows Domain-Driven Design principles, with a C#/.NET 10 backend, an Angular 21 frontend, a MariaDB database, and the linear programming model solved by Google OR-Tools’ GLOP solver. As a result, the system interactively generates the risk-return efficient frontier, allowing the producer to compare Pareto-optimal rotation plans and select the one that best balances expected inco