Multi-objective particle swarm optimization for regional crop planning

Indian agriculture is heavily dependent on natural resources and climatic situations etc. Uncertain behaviour of climate and continued depletion of natural resources can cause food security issues due to low production. Optimal crop planning is one of the essential tasks for utilizing the minimum resources to acquire the maximum benefit. A novel crop planning model is proposed here for the optimal allocation of available resources under a water scarred region like Bundelkhand using data for the year 2017–18. This study used an evolutionary algorithm called Multi Objective Particle Swarm Optimization using Crowding Distance, to solve the constrained multi-objective crop planning problems. Maximize net returns and minimize the water requirement were the two objective functions used here with area constraints. The optimized results obtained from the multi-objective model were compared with the single objective PSO and linear programming approach. Overall, optimizing the water requirement instead of taking it as a constraint gives better crop planning strategies by allocating the area to suitable crops. Pareto-optimal solutions obtained from the MOPSOCD shows the linear relationship between net returns and water.