Quantum-enhanced forecasting and optimal scheduling for energy management in photovoltaic greenhouses

Photovoltaic (PV) greenhouses represent a promising pathway toward sustainable agriculture by enabling clean energy utilization and improving crop productivity. However, their energy management is constrained by two critical challenges: the uncertainty in energy supply and demand, and the complexity of non-convex optimization. This paper proposes a novel energy management method based on variational quantum algorithms. Firstly, a quantum machine learning forecasting model, namely a dual attention mechanism-based quantum long short-term memory (DAM-QLSTM), is developed to improve the prediction accuracy by exploiting its expressive capacity for modeling nonlinear dependencies. Secondly, based on the prediction results, a variational quantum circuit (VQC)-based rolling optimization model is formulated to address the non-convex optimization problem with multi-physics constraints, leveraging quantum interference to improve exploration of the solution space. Finally, the proposed DAM-QLSTM-VQC method is validated using practical data from a PV greenhouse in Beijing, demonstrating a 9.6% improvement in renewable energy utilization compared with particle swarm optimization, along with reductions in total operating cost, carbon and pollutant emissions. Compared with model predictive control, the proposed method reduces cost by 1.5% and emissions by 4.7%. These findings highlight the potential of quantum computing to enhance the energy management efficiency in complex agricultural energy systems.