Algorithmic Simplification of ANN-Based Adaptive Load Prediction and Control for Photovoltaic Heating Systems

This study investigates the extent to which demand-load prediction and adaptive load control in photovoltaic heating systems depend on computationally intensive artificial neural network (ANN) models. We propose a load-prediction algorithm and a corresponding control strategy designed to streamline the computational workflow and reduce resource requirements. By optimizing the algorithmic structure, the proposed method reduces the computational burden during operation while preserving predictive accuracy. The streamlined workflow enables efficient load prediction and adaptive control in real-time operation. Compared with conventional ANN models, the proposed method achieves an overall (all-interval) mean deviation of −0.2%. Moreover, the proposed adaptive load-control strategy requires only 3.9% of the computational resources needed by conventional approaches. The method rapidly reallocates computational resources in response to changes in real-time input data, thereby minimizing redundant computations. Overall, the results indicate that the proposed algorithm substantially reduces computational complexity while maintaining high predictive accuracy. This approach offers an effective alternative to traditional ANN-based methods and facilitates the practical deployment of photovoltaic heating systems.