Zdenek Machacek

Accurate forecasting of daily solar radiation is essential for efficient renewable energy management, photovoltaic system optimization, and power grid stability. This study presents a comparative evaluation of six forecasting models: SARIMA, LSTM, NARX, Random Forest, XGBoost, and a stacking ensemble. The models were evaluated using daily solar radiation and meteorological data retrieved from the NASA Prediction of Worldwide Energy Resources (NASA POWER) platform for Poruba, Ostrava, Czech Republic. A rigorous chronological data partitioning strategy was adopted, with data from 2001–2022 used for training, 2023 for validation, and 2024–2025 reserved for prospective testing. The modeling framework incorporated meteorological exogenous variables, cyclic calendar encodings, lagged radiation observations, and rolling statistical features to capture short-term fluctuations and long-term seasonal dynamics. Among the individual models, XGBoost achieved the strongest test performance (R² = 0.7704), followed by Random Forest (R² = 0.7565), NARX (R² = 0.7524), and LSTM, while SARIMA exhibited the weakest predictive accuracy (R² = 0.7265), reflecting its limitations in modeling nonlinear temporal dependencies. The proposed stacking ensemble, combining Linear Regression, Random Forest, and K-Nearest Neighbors as base learners with Ridge regression as the meta-learner, achieved the highest overall performance (R² = 0.9997), substantially outperforming all individual models on the prospective test set. Statistical significance testing using the Friedman omnibus test, the Wilcoxon signed-rank test, and the Diebold–Mariano test with Holm–Bonferroni correction indicated statistically significant differences in predictive performance (p < 0.001). These findings suggest that ensemble learning, supported by structured feature engineering and temporally consistent model evaluation, offers a promising framework for daily solar radiation forecasting.

Accurate forecasting of daily solar radiation is essential for renewable energy management, optimization of photovoltaic systems, and power-grid stability. This study presents an evaluation of six forecasting models: SARIMA, LSTM, NARX, Random Forest, XGBoost, and a stacking ensemble. Daily solar radiation and meteorological data were retrieved from the NASA Prediction of Worldwide Energy Resources (NASA POWER) platform for Poruba, Ostrava, Czech Republic. A chronological data-partitioning strategy was adopted, with data from 2001–2020 used for training, 2023 for validation, and 2024–2025 reserved for prospective testing. The modeling framework incorporated meteorological variables, cyclic calendar encodings, lagged radiation observations, and rolling statistical features to capture short-term fluctuations and long-term seasonal dynamics. Among the individual models, XGBoost achieved the strongest test performance (R² = 0.7704), followed by Random Forest (R² = 0.7565), NARX (R² = 0.7524), and LSTM. SARIMA exhibited the weakest predictive accuracy (R² = 0.7265), reflecting its limitations in modeling nonlinear temporal dependencies. The stacking ensemble, combining Linear Regression, Random Forest, and K-Nearest Neighbors as base learners and Ridge regression as the meta-learner, achieved the highest performance (R² = 0.9997), outperforming all individual models on the prospective test set. Statistical significance testing using the Friedman omnibus test, Wilcoxon signed-rank test, and Diebold–Mariano test with Holm–Bonferroni correction confirmed that the ensemble’s superiority was statistically robust across all pairwise comparisons (p < 0.001). These findings demonstrate that ensemble learning, supported by structured feature engineering and temporally consistent model evaluation, provides an effective and generalizable framework for long-term solar radiation forecasting.

Non-contact laser triangulation is widely used for high-precision measurement and inspection across diverse industrial applications. Although these systems offer speed, accuracy, and non-invasiveness, their reliability is highly sensitive to multiple influencing factors. These include surface characteristics, material transparency, environmental conditions, and measurement configurations such as roughness, illumination, and incidence angle. This study systematically examines the impact of these factors on measurement reliability through repeated experiments conducted in a highly configurable test environment. The results identify key sources of variability and provide practical recommendations for selecting suitable measurement configurations under different operational conditions. By applying these insights, engineers and practitioners can optimize laser triangulation systems for enhanced reliability in demanding real-world applications.

Latent heat thermal energy storage (LHTES) offers a compact and energy-dense solution for storing heat for residential use. However, its large-scale adoption is still limited by several challenges. These include the low effective thermal conductivity of salt hydrates, uncertainty when heat is released from supercooled states, and the lack of experimentally validated system-level performance at the household scale.This study aims to address these challenges by developing and experimentally evaluating a modular thermal energy storage system based on phase change materials, designed specifically for domestic applications with controllable, on-demand heat release. Building on a review of state-of-the-art storage methods, the work presents a modular experimental storage system that uses controllable supercooling to enable on-demand heat release. The system combines loop calorimetry with a dense network of in-situ temperature sensors to accurately measure heat transfer performance.The experimental demonstrator consists of a modular unit comprising a finned aluminium heat exchanger immersed in 85 kilograms of sodium acetate trihydrate (SAT, CH3COONa·3H2O) inside a transparent tank that allows optical access. The system includes a three-dimensional (3D) network of 117 temperature sensors integrated into a controlled hydronic loop. Repeated charging and discharging tests showed stable operation without phase separation, while thermal maps revealed a heat-transfer front that starts at the heat exchanger and gradually moves toward the outer regions of the storage material (that is, the periphery).In a separate long-duration experiment lasting approximately 21 hours and including an intentional idle period, the overall daily efficiency reached about 65 percent. This combined experimental approach provides interpretable, system-level performance indicators and identifies practical and operational improvements. These include optimized heat-exchanger and fin spacing, adding conductive pathways, improving insulation and thermal breaks, and aligning operation with energy tariffs. Together, these measures outline a clear pathway for adapting the concept to deployment in real residential energy systems.

This paper presents the construction of an innovative high-temperature sensor based on the optical principle. The sensor is designed especially for the measurement of exhaust gases with a temperature range of up to +850 °C. The methodology is based on two principles-luminescence and dark body radiation. The core of this study is the description of sensing element construction together with electronics and the system of photodiode dark current compensation. An advantage of this optical-based system is its immunity to strong magnetic fields. This study also discusses results achieved and further steps. The solution is covered by a European Patent.

The presented research addresses the problem of dependency analysis of the ultrasonic signal measured by a sensor in an engine oil bath. The dependency analysis is performed on a selected ultrasonic signal sensor solution containing its own generator and an ultrasonic signal receiver detecting the level of the oil in which it is immersed. The influence of the resulting amplitude of the received ultrasonic signal is mainly due to the level of the measured oil level and the oil temperature, as shown by the regression analysis and ANOVA (Analysis of Variance) testing performed. The analyzed dependence of the time determination of the length of the generated ultrasonic signal envelope is given by a set threshold value, which can be dynamically adjusted based on the backtracking evaluation. The analysis results in the form of approximation by the dependency algorithm confirm the assumption of possible standardization of the envelope parameters with the achievement of accuracy up to 99.02%. The analyzed parameters approximated by the temperature and oil level dependence algorithms include the amplitude of the measured signal, steepness of the rising edge, duration of the envelope, and the digitally processed amplitude value.

In this study, we present the results of measuring the performance of selected Peltier cells such as thermoelectric Peltier cooler modules (TEC), thermoelectric micro-Peltier cooler modules (TES), and thermoelectric Seebeck generator modules (TEG). The achieved results are presented in the form of graphs of powering system output voltage or power efficiency functions of the load impedance. Moreover, a technical solution is also presented that consists of designing a water consumption power supply system, using a renewable energy source in the form of a Peltier cell. The developed measuring system does not require additional batteries or an external power source. The energy needed to power the system was obtained from the temperature difference between two sides of a thermoelectric cell, caused by the measured medium which was flowing in a copper water pipe. All achieved results were investigated for the temperature difference from 1 to 10 K in relation to the ambient temperature.

This study focused on the design of the power supply and the overall control structure of the electronic system, which is charged only with energy from the environment. It describes the basic finding of all the necessary information for the design and implementation with the measurement and evaluation of characteristic parameters. The paper describes the individual possibilities of obtaining energy from the surroundings with the selected energy from the heat of the pipeline. There are also various arrangements of possible elements mentioned in order to ensure the optimal solution of the whole system. Furthermore, a basic evaluation of the comparison of a possible workable solution is made.

This article deals with the design and implementation of a prototype of an efficient Low-Cost, Low-Power, Low Complexity–hereinafter (L-CPC) an image recognition system for person detection. The developed and presented methods for processing, analyzing and recognition are designed exactly for inbuilt devices (e.g., motion sensor, identification of property and other specific applications), which will comply with the requirements of intelligent building technologies. The paper describes detection methods using a static background, where, during the search for people, the background image field being compared does not change, and a dynamic background, where the background image field is continually adjusted or complemented by objects merging into the background. The results are compared with the output of the Horn-Schunck algorithm applied using the principle of optical flow. The possible objects detected are subsequently stored and evaluated in the actual algorithm described. The detection results, using the change detection methods, are then evaluated using the Saaty method in order to determine the most successful configuration of the entire detection system. Each of the configurations used was also tested on a video sequence divided into a total of 12 story sections, in which the normal activities of people inside the intelligent building were simulated.