Greenhouse Climate Control Benchmark for Reinforcement Learning

A 5.5 KB Excel dataset provides greenhouse climate parameters for benchmarking reinforcement learning control algorithms. It was created by Iman Hindi and published on figshare in 2026, supporting research in optimizing crop yield and resource consumption.

Use Cases

• Benchmark reinforcement learning algorithms like TD3 or PPO using the provided greenhouse climate and control features.
• Analyze the trade-off between crop yield metrics (e.g., stem elongation, cumulative trusses) and resource consumption costs (heating, CO2).
• Train predictive models for internal greenhouse climate conditions or daily resource consumption using the simulated environment data.

Strengths

• Data is explicitly designed for benchmarking reinforcement learning algorithms in greenhouse control.
• Includes key crop parameters (stem elongation, thickness) and resource metrics (heating, electricity, CO2, irrigation) for multi-objective optimization.
• Published under a permissive CC BY 4.0 license.

Limitations

• Very small scale at 5.5 KB, indicating a limited sample or synthetic/simulated data rather than extensive real-world measurements.
• Column structure and row count are unknown, complicating assessment of its dimensionality and volume for machine learning.
• Data is simulated from a proposed RL framework, not directly measured from physical greenhouses, limiting real-world validation.

Provenance

Source

figshare, author Iman Hindi.

Collection Method

Simulated data generated from a proposed reinforcement learning and deep learning framework for greenhouse climate control.

Time Range

null

Freshness

Last updated in March 2026.

Geography

null