arXiv subjects

Sjoerd Boersma

Publications and source records attributed to Sjoerd Boersma.

7 recordsLinked to original sources

arXiv2026

LettuceVisSim: A Simulator That Generates Lettuce Image Time-series for Vision-Based Reinforcement Learning

Vision-based reinforcement learning holds strong potential for decision-making in controlled environment agriculture (CEA). However, its development is hindered by the scarcity of labelled crop images. To address this gap, LettuceVisSim, a lettuce growth simulator that generates labelled time series of crop images, was developed and validated. The simulator contains a process-based model (PBM) for shoot dry weight dynamics, a canopy layout algorithm for deriving canopy layout representations from shoot dry weight, and a Unity rendering engine for image generation. Five findings support the simulator. First, the PBM reproduced shoot dry weight under dynamic plant-density management with $\mathrm{R}^{2}=0.84$. Second, a piecewise cubic regression mapped shoot dry weight to potential projected area with $\mathrm{R}^{2}=0.94$. Third, the canopy layout representation was validated using 12 experimental datasets each having different dynamic environmental and spacing conditions. It reproduced the ground coverage ratio dynamics observed in measured images, achieving $\mathrm{R}^{2}=0.84$ when driven by measured shoot dry weight and $\mathrm{R}^{2}=0.40$ (0.76 excluding one outlier) when driven by PBM-simulated values. Fourth, the Unity rendering engine converted canopy layout representations into RGB and segmentation images at less than 10~ms. Fifth, a demonstration showed that a lighting-control policy can be learned and applied by observing only crop images that were generated with LettuceVisSim, providing a proof of concept of vision-based reinforcement learning in CEA using LettuceVisSim.

cs.CV↗

arXiv2026

Improving greenhouse fruit-production control by integrating reinforcement learning into short-horizon model predictive control

Greenhouse fruit-production control aims to maximize the economic performance (fruit revenue minus operating costs) while operating within system constraints under external weather disturbances. Control methods need to balance the delayed economic benefit of fruit yield with current operating costs. For such problems, model predictive control (MPC) can explicitly handle system constraints under future weather disturbances, but can become computationally demanding when using sufficiently long prediction horizons for (relatively large) nonlinear greenhouse fruit production models. In contrast, reinforcement learning (RL) can learn control policies offline while considering longer-term economic performance, but struggles to enforce system constraints, and performance may degrade under unseen weather trajectories. This work proposes trajectory-selection RL-MPC, a framework that incorporates longer-term economic information of fruit yield into a short-horizon MPC optimization problem. The framework uses an RL rollout trajectory to define a terminal region constraint and terminal cost. Next, a nonlinear MPC solves a short-horizon optimization problem with these terminal ingredients to find a local optimum. Finally, the framework selects and executes the first input from the trajectory with the better objective value, either from the MPC-predicted or the RL rollout trajectory. The method is applied to GreenLight, a large-scale greenhouse tomato production model that exhibits stiff dynamics. The simulation results show that trajectory-selection RL-MPC with a one-hour prediction horizon matches the closed-loop performance of a high-performing guiding policy while significantly improving over standalone MPC with the same horizon.

math.OC↗

arXiv2025

GreenLight-Gym: Reinforcement learning benchmark environment for control of greenhouse production systems

This study presents GreenLight-Gym, a new, fast, open-source benchmark environment for developing reinforcement learning (RL) methods in greenhouse crop production control. Built on the state-of-the-art GreenLight model, it features a differentiable C++ implementation leveraging the CasADi framework for efficient numerical integration. GreenLight-Gym improves simulation speed by a factor of 17 over the original GreenLight implementation. A modular Python environment wrapper enables flexible configuration of control tasks and RL-based controllers. This flexibility is demonstrated by learning controllers under parametric uncertainty using two well-known RL algorithms. GreenLight-Gym provides a standardized benchmark for advancing RL methodologies and evaluating greenhouse control solutions under diverse conditions. The greenhouse control community is encouraged to use and extend this benchmark to accelerate innovation in greenhouse crop production.

eess.SY↗

arXiv2024

A Bayesian Neural ODE for a Lettuce Greenhouse

Greenhouse production systems play a crucial role in modern agriculture, enabling year-round cultivation of crops by providing a controlled environment. However, effectively quantifying uncertainty in modeling greenhouse systems remains a challenging task. In this paper, we apply a novel approach based on sparse Bayesian deep learning for the system identification of lettuce greenhouse models. The method leverages the power of deep neural networks while incorporating Bayesian inference to quantify the uncertainty in the weights of a Neural ODE. The simulation results show that the generated model can capture the intrinsic nonlinear behavior of the greenhouse system with probabilistic estimates of environmental variables and lettuce growth within the greenhouse.

math.OC↗

arXiv2023

Reinforcement Learning Versus Model Predictive Control on Greenhouse Climate Control

Greenhouse is an important protected horticulture system for feeding the world with enough fresh food. However, to maintain an ideal growing climate in a greenhouse requires resources and operational costs. In order to achieve economical and sustainable crop growth, efficient climate control of greenhouse production becomes essential. Model Predictive Control (MPC) is the most commonly used approach in the scientific literature for greenhouse climate control. However, with the developments of sensing and computing techniques, reinforcement learning (RL) is getting increasing attention recently. With each control method having its own way to state the control problem, define control goals, and seek for optimal control actions, MPC and RL are representatives of model-based and learning-based control approaches, respectively. Although researchers have applied certain forms of MPC and RL to control the greenhouse climate, very few effort has been allocated to analyze connections, differences, pros and cons between MPC and RL either from a mathematical or performance perspective. Therefore, this paper will 1) propose MPC and RL approaches for greenhouse climate control in an unified framework; 2) analyze connections and differences between MPC and RL from a mathematical perspective; 3) compare performance of MPC and RL in a simulation study and afterwards present and interpret comparative results into insights for the application of the different control approaches in different scenarios.

math.OC↗

arXiv2021

Feedforward-Feedback wake redirection for wind farm control

This work presents a combined feedforward-feedback wake redirection framework for wind farm control. The FLORIS wake model, a control-oriented steady-state wake model is used to calculate optimal yaw angles for a given wind farm layout and atmospheric condition. The optimal yaw angles, which maximize the total power output, are applied to the wind farm. Further, the lidar-based closed-loop wake redirection concept is used to realize a local feedback on turbine level. The wake center is estimated from lidar measurements \unit[3]{D} downwind of the wind turbines. The dynamical feedback controllers support the feedforward controller and reject disturbances and adapt to model uncertainties. Altogether, the total framework is presented and applied to a nine turbine wind farm test case. In a high fidelity simulation study the concept shows promising results and an increase in total energy production compared to the baseline case and the feedforward-only case.

eess.SY↗

arXiv2020

A Distributed Model Predictive Wind Farm Controller for Active Power Control

Due to the fluctuating nature of the wind and the increasing use of wind energy as a power source, wind power will have an increasing negative influence on the stability of the power grid. In this paper, a model predictive control strategy is introduced that not only stabilizes the power produced by wind farms, but also creates the possibility to perform power reference tracking with wind farms. With power reference tracking, it is possible for grid operators to adapt the power production to a change in the power demand and to counteract fluctuations that are introduced by other power generators. In this way, wind farms can actually contribute to the stabilization of the power grid when this is necessary instead of negatively influencing it. A low-fidelity control-oriented wind farm model is developed and employed in the developed distributed model predictive controller. In this control model, the wake dynamics are taken into account and consequently, the model's order is relatively large. This makes it, from a computational point of view, challenging for a centralized model predictive control to provide real-time control for large wind farms. Therefore, the controller proposed in this paper is a distributed model predictive control. Here, the central control problem is divided into smaller local control problems that are solved in parallel on local controllers, which significantly reduces the computational complexity and brings the application of model predictive control in a wind farm a step closer to practical implementation. The proposed control solution is tested in simulations on a 10 and 64 turbine wind farm.

eess.SY↗