A Fuzzy Multi-Objective Optimization Model for Fertilizer Allocation: Zone Partitioning via CryStAl and MILP-Based Bandwidth-Constrained Routing

Fertilizer recommendation plays a pivotal role in maximizing crop yield while minimizing environmental impact and nutrient loss. Traditional practices such as excessive fertilizer use and poor soil assessment have led to soil degradation, structural damage, and nutrient imbalances. To address these challenges, this study introduces a multi-objective, AI-powered framework within the domain of Precision Agriculture (PA). By leveraging zone-specific soil analysis and real-time data from strategically placed agro-sensors (Slave Nodes), the system delivers targeted recommendations. Sensor and UAV deployment are optimized using the Crystal Structure Optimization Algorithm (CryStAl), while a Bandwidth-aware Routing Protocol (BRP) ensures efficient and reliable data transmission. Data preprocessing integrates advanced techniques like the Versatile Loss Pass Weiner (VLPW) filter and Boosted U-Net (BU-Net) for image enhancement, along with outlier detection and dynamic interpolation for sensor data. An Advanced Fuzzy Inference System processes factors such as soil type, leaf disease, and climate conditions to suggest optimal fertilizer types and dosages. The proposed approach enhances resource efficiency, supports environmental sustainability, and improves crop productivity through intelligent, adaptive decision-making.