Sustainable irrigation planning using meta-heuristic algorithms

The present study proposes a novel, systematic, multi-algorithm, benchmark-based screening framework. Ten Meta-Heuristic Algorithms (MHAs), namely, Sand Cat Swarm Algorithm (SCSA), Chaotic Sand Cat Swarm Algorithm (CSCSA), White Shark Algorithm (WSA), Grey Wolf Optimization (GWO), Gravitational Search Algorithm (GSA), Jaya Algorithm (JA), Chaotic Jaya Algorithm (CJO), Archimedes Optimization (AO), Teaching Learning Based Optimization (TLBO), Elephant Herding Optimization (EHO) covering four distinct inspiration categories (human, animal, physics, and mathematics) are evaluated on thirteen standard benchmark functions using a weighted ranking approach. TLBO, AO, CJO, and EHO are identified as the most promising algorithms after screening. Subsequently, these are used to optimize net annual benefits for the study area of the Sri Ram Sagar Project (SRSP), India. Additionally, the study explicitly models the conjunctive use of surface and groundwater under three scenarios, S1-S3: S1 and S2 represent 75% and 90% inflow dependability levels with groundwater utilization, respectively, while S3 is for a 75% inflow level without groundwater. The study also integrates environmental flow requirements, command-area limitations, groundwater withdrawals, canal capacity restrictions, and others. TLBO performed best, with optimal net annual benefits of ₹3815.08 crores for S1, ₹1766.83 crores for S2, and ₹2731.78 crores for S3, and respective irrigation intensities of 80.85%, 30.04%, 64.48%, indicating improvement in land and water-use efficiency. The structured screening methodology, which selects the best-performing algorithms across diverse categories, is the first of its kind at the global and national levels in irrigation planning. This approach also facilitates an adaptable framework for any large-scale reservoir systems worldwide.