This paper compares two positioning systems, namely ultra-wideband (UWB) based micro-location technology and dead reckoning and a RPLidar based simultaneous localization and mapping (SLAM) solution. This new approach can be used to improve the quality of the positioning system and increase the functionality of advanced driver assistance systems (ADAS). This is achieved by using stationary nodes and UWB tags on the vehicles. Thus, the redundancy of localization can be achieved by this approach, e.g., as a backup to onboard sensors like RPlidar or radar. Additionally, UWB based micro-location allows additional data channels to be used for communication purposes. Furthermore, it is shown that the regular use of correction data increases UWB and dead reckoning accuracy. These correction data can be based on onboard sensors. This shows that it is promising to develop a system that fuses onboard sensors and micro-localization for safety-critical tasks like the platooning of commercial vehicles.
Shutdowns of photovoltaic installations are a problem that has been increasingly affecting private investors who have built home installations of several to a dozen kWp over the last few years. This problem, most often caused by outdated infrastructure, appears in many countries and impacts energy production. This work focuses on three aspects of the problem. The first one answers the question of how shutdowns of the photovoltaic installation affect production, and how significant the energy loss happens when the PV inverter is not working or is in the restart phase. The second aspect proposes an original, low-cost method that reduces the number of shutdowns. This method relates to the auto-consumption mechanism associated with domestic water heaters and the system for measuring voltage and energy consumption from the electrical network. The solution is based on constant monitoring of the network voltage and the switching of heaters based on a dedicated algorithm. Additionally, continuous analysis also allows for reporting observed irregularities to the electricity supplier. The third and final factor corresponds to the real impact of shutdowns on the long-term aspect of the investment and the extension of its payback period, and to what extent the proposed solution shortens this period. Through a detailed analysis on the issue of shutdowns, the proposed solution allows for a reduction in the number of shutdowns by over 40%. However, due to the fact that it discusses a specific case, this solution requires a calibration and adjustment process, which is discussed in the article.
In response to the growing need for accurate forecasting of electricity generation from PV installations, which is crucial both for enhancing self-consumption and for balancing the power grid, this study presents a comparative analysis of selected machine learning models. The research focuses on the XGBoost algorithm and LSTM neural networks, applied to predict PV energy production based on meteorological data and historical generation records from four medium-sized PV installations (30–50 kWp) located in Poland. Meteorological data were retrieved from open sources and combined with actual production measurements to build the training dataset. This paper discusses the challenges posed by these data at the given latitude, as well as issues related to processing data from newly launched installations. The performance of both approaches was evaluated in short- and medium-term forecasting, with particular attention to prediction accuracy, robustness to noisy data, and the ability to capture nonlinear relationships.
This paper focuses on the different filtering methods of movement around a circle whose acquisition was made using system based on the Ultra-wideband (UWB) technology. The main aim of this study is to get the most accurate way of filtering circular motion for the UWB system for statistical terms. Filtration presented in this paper can be used to locate objects using ultra-wideband technology, and thanks to the use of motion at speed up to 50km/h, it will be possible to use test results to reflect drones or autonomous vehicles, particularly where standard locations systems (e.g. GPS) is not available - such as tunnels, underground car parks, shopping centers. This document proposes an approach based on the filtration points that was assigned to the buckets placed in three-dimensional space on the expected trajectory of the circular object movement. During the data processing the following filters have been used: median, arithmetic mean, Savitzky-Golay in combination with arithmetic mean, and moving average in combination with arithmetic mean. The number of used measurement buckets, as well as their size, was also analyzed. The results were compared with the use of the length of the line, that traces the localized object in correlation with the number of measured buckets used.