Photovoltaic (PV) power production is characterized by high variability due to short-term meteorological effects such as cloud movements. These effects have a significant impact on the incident solar irradiance in PV parks. In order to control PV park performance, researchers have focused on Computer Vision and Deep Learning approaches to perform short-term irradiance forecasting using sky images. Motivated by the task of improving PV park control, the current work introduces the Image Regression Module, which produces irradiance values from sky images using image processing methods and Convolutional Neural Networks (CNNs). With the objective of enhancing the performance of CNN models on the task of irradiance estimation and forecasting, we propose an image processing method based on sun localization. Our findings show that the proposed method can consistently improve the accuracy of irradiance values produced by all the CNN models of our study, reducing the Root Mean Square Error by up to 10.44 W/m${}^2$ for the MobileNetV2 model. These findings indicate that future applications which utilize CNNs for irradiance forecasting should identify the position of the sun in the image in order to produce more accurate irradiance values. Moreover, the integration of the proposed models on an edge-oriented Field-Programmable Gate Array (FPGA) towards a smart PV park for the real-time control of PV production emphasizes their advantages.

Aiming at effectively improving photovoltaic (PV) park operation and the stability of the electricity grid, the current paper addresses the design and development of a novel system achieving the short-term irradiance forecasting for the PV park area, which is the key factor for controlling the variations in the PV power production. First, it introduces the Xception long short-term memory (XceptionLSTM) cell tailored for recurrent neural networks (RNN). Second, it presents the novel irradiance forecasting model that consists of a sequence-to-sequence image regression NNs in the form of a spatio-temporal encoder–decoder including Xception layers in the spatial encoder, the novel XceptionLSTM in the temporal encoder and decoder and a multilayer perceptron in the spatial decoder. The proposed model achieves a forecast skill of 16.57% for a horizon of 5 min when compared to the persistence model. Moreover, the proposed model is designed for execution on edge computing devices and the real-time application of the inference on the Raspberry Pi 4 Model B 8 GB and the Raspberry Pi Zero 2W validates the results.

During the past decades the reports produced by the satellites regarding the clouds became important for the weather prediction, the climate and lately for the generation and the distribution of the solar power. Image segmentation techniques produce these reports and they exploit convolutional neural networks (CNNs) to improve the accuracy results at the expense though of the requirements for execution time and hardware resources. Aiming at improving this trade-off the current work studies the design of a CNN model targeting the placement and the execution on a SoC FPGA while keeping the high accuracy. Moreover, it studies how to optimize the performance by distributing the execution of the CNN’s functions among the SoC’s IP cores and the SoC’s logic. The results are validated by using the Vitis-AI framework to implement the proposed CNN on a Xilinx Zynq UltraScale+ MPSoC ZCU104 board.

The emerging Kolmogorov–Arnold networks (KANs) have set a new standard in machine learning (ML) tasks by prevailing over traditionally deployed multilayer perceptrons (MLPs) thanks to their enhanced interpretability through activation function learning, while they require increased computational complexity and memory footprint. Radial-basis function (RBF)-based KAN models maintain high performance over other variants of KANs with considerable size reduction and consequently more efficient execution. Aiming at effectively supporting the inference of RBF-KANs on Internet-of-Things (IoT) devices, this paper focuses on edge-oriented computing and introduces a soft intellectual property (IP) core, written in hardware description language (HDL), targeting the execution of such networks on all-programmable systems-on-chip (APSoC). The proposed design is fully pipelined and runtime configurable, allowing for real-time inference and latency-sensitive neural network deployment on-the-fly. A testbench reveals up to $43.6\times$ speedup when compared with a commercial edge central processing unit (CPU) and consumes considerably less power. The core’s adaptable design enables efficient allocation of resources and meets diverse throughput demands, making it well-suited for a broad range of IoT applications.