NDRL: Cotton Irrigation and Nitrogen Application with Nested Dual-Agent Reinforcement Learning

2025-12-18

AI Summary

Cotton water-nitrogen co-regulation faces challenges in quantifying stress signals and suffers from feedback delays, leading to imprecise dynamic optimization. To address this, we propose a macro-micro nested dual-agent reinforcement learning framework: a parent agent guides long-term strategy selection based on cumulative yield rewards, while a child agent integrates water stress factor (WSF) and nitrogen stress factor (NSF) modeling with a hybrid probabilistic policy for daily-level precise decision-making. Our method couples the DSSAT crop simulation model with quantitative stress characterization, overcoming the limitations of conventional static-threshold-based regulation. Two-year simulation results demonstrate that, compared to the optimal baseline, our approach increases yield by 4.7%, improves irrigation water productivity by 5.6% and 5.1%, and enhances nitrogen partial factor productivity by 6.3% and 1.0%, significantly boosting resource use efficiency.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Intelligent Robots: Learning & Optimization for ROB

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Sustainability of Web economics
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Effective irrigation and nitrogen fertilization have a significant impact on crop yield. However, existing research faces two limitations: (1) the high complexity of optimizing water-nitrogen combinations during crop growth and poor yield optimization results; and (2) the difficulty in quantifying mild stress signals and the delayed feedback, which results in less precise dynamic regulation of water and nitrogen and lower resource utilization efficiency. To address these issues, we propose a Nested Dual-Agent Reinforcement Learning (NDRL) method. The parent agent in NDRL identifies promising macroscopic irrigation and fertilization actions based on projected cumulative yield benefits, reducing ineffective explorationwhile maintaining alignment between objectives and yield. The child agent's reward function incorporates quantified Water Stress Factor (WSF) and Nitrogen Stress Factor (NSF), and uses a mixed probability distribution to dynamically optimize daily strategies, thereby enhancing both yield and resource efficiency. We used field experiment data from 2023 and 2024 to calibrate and validate the Decision Support System for Agrotechnology Transfer (DSSAT) to simulate real-world conditions and interact with NDRL. Experimental results demonstrate that, compared to the best baseline, the simulated yield increased by 4.7% in both 2023 and 2024, the irrigation water productivity increased by 5.6% and 5.1% respectively, and the nitrogen partial factor productivity increased by 6.3% and 1.0% respectively. Our method advances the development of cotton irrigation and nitrogen fertilization, providing new ideas for addressing the complexity and precision issues in agricultural resource management and for sustainable agricultural development.

Problem

• Optimizes water-nitrogen combinations for cotton yield enhancement
• Quantifies stress signals for precise irrigation and fertilization control
• Improves resource efficiency in agricultural management through reinforcement learning

Innovation

• Nested dual-agent reinforcement learning optimizes irrigation and fertilization
• Parent agent selects actions based on projected cumulative yield benefits
• Child agent uses stress factors to dynamically optimize daily strategies