Real Measurement Data Research Articles

Physics-Informed Neural Networks for Modeling Water Flows in a River Channel

The impacts incurred by floods regularly affect the planet&#x0027;s population, inflicting social and economic problems. Optimal control strategies based on reservoir management may aid in controlling floods and mitigating the resulting damage. To this end, an accurate dynamic representation of water systems is needed. In practice, flood control strategies rely on hydrological forecasting models obtained from conceptual or data-driven methods. Encouraged by recent works, this research proposes a novel surrogate model for water flow in a river channel based on Physics-Informed Neural Networks (PINNs). This approach achieved promising results regarding the assimilation of real-data measurements and parameter identification of differential equations that govern the underlying dynamics. This work investigates PINN performance in a simulated environment built directly from a configuration of the Saint-Venant equations. The objective is to create a suitable model with high prediction accuracy and scientifically consistent behavior for use in real-time applications. The experiments revealed promising results for hydrological modeling and presented alternatives to solve the main challenges found in conventional methods while assisting in synthesizing real-world representations. <i>Impact Statement</i>&#x2014;The research seeks to contribute to the hydrological modeling area with a surrogate model based on Physics-Informed Neural Networks to water flow in a watershed. In practice, these models use conceptual or data-driven methods. Conceptual models to reach the precision provided by the methodology use large numbers of physical parameters. These parameters can demand deep knowledge about the environment and are possibly hard to identify in a complex basin. On the other hand, while data-driven methods do not require such knowledge about the dynamic system, they depend on a reliable and useful database to guarantee the accuracy of system behavior. We introduce PINNs as a viable solution for training neural networks with a few training data and estimating the PDE parameters that govern the underlying dynamics. In addition, we present a novel strategy for training PINNs inspired by Bayesian inference for parameter estimation problems. The research aims to bring forward opportunities to the nontrivial task of hydrological modeling and tools for balancing learning from both physics and data. Consequently, to develop concepts for real-time applications.

Confronting combined cyber-attacks on collaborative DC microgrids using a communication protocols-based distributed unscented Kalman filter

The functioning of microgrids (MGs), which represent a contemporary arrangement of distribution networks featuring local control over generation and consumption, necessitates telecommunication infrastructure for transmitting and receiving measurement data and control commands. This reliance on telecommunication infrastructure renders MGs vulnerable to the threat of cyber-attacks. The existing methods for identifying cyber-attacks have been restricted solely to MG modeling. However, identifying and restoring a cyber-attack lack dependability and feasibility in a practical application without the concurrent modeling of MGs and their communication protocols. Therefore, this study aims to identify hybrid cyber-attacks (simultaneous false data injection and time delay attacks) to flow sensors and cyber links in direct flow MGs by considering the communication protocols between them, as well as extracting and reconstructing real measurement data. In order to achieve this objective, a generalized unscented Kalman filter (UKF) is introduced in conjunction with communication protocols incorporating an augmented state space to estimate the power of nonlinear loads. For this purpose, two measurement models have been developed to prevent data collisions and reduce the communication load. These models are based on the round-robin scheduling protocol and the try-once-discard weighted protocol. They aim to establish a specific order for data transfer from sensors to the filter. The article concludes by presenting a stability analysis of the protocol-based generalized UKF, which aims to ensure the convergence and limitation of the estimation error. The study presents simulation results on the resilient reconstruction of hybrid cyber-attacks involving unknown constant power loads and noise. The proposed approach utilizes the protocol-based generalized UKF and exhibits superior performance.

Hybrid modeling of a circulating fluidized bed boiler for development of a prediction and prescription system for power plant operation

A novel hybrid modeling system of a circulating fluidized bed boiler was developed in this study. The proposed approach combines a computational fluid dynamics model, practice-oriented macro-scale models, and a data-driven reduced order model to capture both the physical and empirical behaviors of the boiler. The application of various modeling strategies allowed for capturing features of the simulated process at various scales and for developing a fast and accurate prediction system. The use of the computational fluid dynamics model improved the predictions of the practice-oriented macro-scale models providing more detailed flow information and the use of the reduced order model allowed for reducing the computation time by a factor of 1350. The hybrid model and its submodels were validated against real measurement data. The validation confirmed good accuracy in predicting key process variables. The developed model is used as a prediction and prescription system to enable operators to optimize the boiler operation and prevent abnormal events. The hybrid system is used for a supercritical, once-through boiler in Łagisza power plant in Poland, however, the presented modeling strategy can be applied in other power plants, which can improve efficiency, reduce emissions, enhance boiler availability, and ultimately contribute to more sustainable energy systems.

Efficient bead-on-plate weld model for parameter estimation towards effective wire arc additive manufacturing simulation

AbstractDespite the advances in hardware and software techniques, standard numerical methods fail in providing real-time simulations, especially for complex processes such as additive manufacturing applications. A real-time simulation enables process control through the combination of process monitoring and automated feedback, which increases the flexibility and quality of a process. Typically, before producing a whole additive manufacturing structure, a simplified experiment in the form of a bead-on-plate experiment is performed to get a first insight into the process and to set parameters suitably. In this work, a reduced order model for the transient thermal problem of the bead-on-plate weld simulation is developed, allowing an efficient model calibration and control of the process. The proposed approach applies the proper generalized decomposition (PGD) method, a popular model order reduction technique, to decrease the computational effort of each model evaluation required multiple times in parameter estimation, control, and optimization. The welding torch is modeled by a moving heat source, which leads to difficulties separating space and time, a key ingredient in PGD simulations. A novel approach for separating space and time is applied and extended to 3D problems allowing the derivation of an efficient separated representation of the temperature. The results are verified against a standard finite element model showing excellent agreement. The reduced order model is also leveraged in a Bayesian model parameter estimation setup, speeding up calibrations and ultimately leading to an optimized real-time simulation approach for welding experiment using synthetic as well as real measurement data.

A FE Model for the Efficient Simulation of the Thermo-Elastic Machine Tool Behavior

In modern machine tools, the thermo-elastic positioning error has a large share on the overall machine tool error. A multitude of different approaches can be found in the literature to correct this error in order to improve the accuracy of the machine tool. However, an FE-based approach to model the thermo-elastic error of 5 axis machine tools is still an active field of research. Therefore, the goal of this work is the improvement of the machine tool accuracy by modelling the machine tool error using a FE model. The model uses real time measurement data of the thermo-elastic error using an R-Test, as well as real time temperature measurements of the machine structure.

Calculation of phase angle diversity for time-varying harmonic currents from grid measurement

Phase angle diversity of harmonic currents emitted by different equipment, like electronic power supplies, modern lamps with electronic ballast, chargers for electrical cars or photovoltaic inverters, results in cancellation effects which represent an efficient way of managing harmonic levels in LV grids. Different methods have been presented in the last decades to quantify and evaluate the summation of harmonic currents based on grid measurements. There is not a general consensus about which method or index should be preferred, especially if time-variation of harmonic currents is considered. The paper discusses the impact of measurement accuracy and aggregation parameters on the calculation of the two main diversity indices, namely diversity factor and summation exponent. In case of time varying harmonics an additional statistical post-processing is required which also depends on several parameters and calculation methods. Based on the results of the application of different methods to real grid measurement data, some guidance on suitability and limitations of the different methods are given.

A Power Factors Full-Dimensionally Hunted Power Distribution Strategy for Railway Power Flow Controller Considering Regenerative Braking

Railway power flow controller (RPFC), as a promising technical route for railway power regulation, its rating determinization and the related flexible power regulation perplex designers and hinder large-scale applications, especially when the regenerative braking energy (RBE) is non-negligible. In this paper, a general power factors full-dimensionally hunted power distribution strategy (PFFDH-PDS), applicable for both off-line design & real-time control, is proposed. Two key methods are developed: 1) Full-dimensionally hunting the power factors (PFs) under the developed mathematic framework, from which the global optimal compensating power and RPFC's rating can be worked out; 2) Precisely partitioning two-phase loads on their PF-P and PF-PF distributing panels; in both methods, RBE is considered. By adopting PFFDH-PDS, RPFC's rating can be reduced by almost 60% compared to the conventional method, while the system has satisfactory grid-connection compatibility. Finally, the effectiveness and performance of the proposal are verified & conducted by real-measured data.

A Comparison of Physics- and Data-based Modeling of Rural Drainage Systems

Rural drainage systems are often large and complex systems, which are difficult to model. In this work, a physical model based on the hydrology and hydraulic laws and a data-driven modeling approach with an ARX structure are created for the rural drainage system in Eiderstedt, Germany. Thereupon, the prediction accuracy for real measurement data of both models is investigated and the respective advantages and disadvantages are discussed. It is shown that both models are able to predict the water level quite well, with the physical and data-based model being more accurate for long- and short-term predictions, respectively.

Challenges and Limitations of Artificial Intelligence Implementation in Modern Power Grid

Ongoing global decarbonization and energy transition have led to significant changes in the traditional power grid, resulting in new challenges for grid operators. These challenges, which are exacerbated by the increased penetration of distributed energy resources (DER), include power system stability, protection and control system malfunction, and supply and demand balance. To better understand the implications of high DER penetration, we proposed a classification of modern grids based on the DER integration level and identified the specific challenges faced by each category. We defined eight key challenges posed by DER on the grid and examined recent academic research on the application of AI techniques to modern power systems. Despite the potential benefits of AI in terms of power system stability and protection, there are currently no commercial applications of this research. Our study analyzed the limitations and barriers to practical AI implementation in power system protection and stability, including the use of synthetic data, insufficient real measurement data, protection selectivity, and the black-box nature of the models. We emphasize the importance of further research and development in the fields of Explainability AI and physics-informed ML to overcome these challenges and enable the successful implementation of AI in modern power grid stability and protection.

Application of Artificial Neural Network in Wildfire Early Prediction Systems

The preservation of forest ecosystems is of vital importance to life on our planet. The increased losses of forests due to fires make the task of forest fire prevention of crucial significance. The present paper describes the development of an artificial neural network (ANN) for forest fire early prediction. The ANN predictor consists of two layers with 5 neurons in the hidden layer. It is trained through backpropagation of an error learning algorithm and is validated to provide prediction with a high degree of accuracy. An additional advantage of the designed predictor is the use of a limited number of input data based on weather and moisture conditions and of an output of a prior computed probability for fire. The training and validation datasets consist of 82 records of real measurement data. The developed and validated ANN can contribute to improvement of the current forest fire prediction systems.

Optimization of an off-grid PV/biogas/battery hybrid energy system for electrification: A case study in a commercial platform in Morocco

The use of hybrid renewable energy systems is growing as a viable option for clean power generation, fueled by the increasing demand for sustainable energy sources and the need to reduce carbon emissions. In this context, this paper evaluates the optimal configuration, as well as the economic and environmental performances of a hybrid solar PV/biogas/battery energy system designed to provide electricity to a commercial platform in Berkane- Morocco. The optimization model aims to determine the optimal capacity of renewable energy systems achieving the most cost-effective levelized cost of electricity, reducing greenhouse gas emissions, and utilizing locally available renewable energy resources. The model was developed using HOMER software incorporating real-measured data for electricity demand, solar irradiance, and biogas availability. It was found that the PV/biogas/battery combination is very optimal in terms of cost and emissions savings in comparison with the use of only one source of power generation. The optimal design of the energy system results in 231 kW of PV modules, 170 kW biogas generator, a 140-kW converter, and a 201 kWh Li-Ion battery park. The optimization results in an LCOE of 0.280 $/kWh; Moreover, the proposed system would save almost 40 % of carbon dioxide (CO2) emissions in comparison with only biogas system. A sensitivity test has been performed, showing that the proposed hybrid system is sensitive to capital subsidies and discount rates. This study demonstrates the economic viability and environmental benefits driven by the integration of the hybrid of PV/Biogas/Battery system in Morocco, making it an attractive alternative for future sustainable development.

Cognitive Radar Waveform Design Method under the Joint Constraints of Transmit Energy and Spectrum Bandwidth

The water-filling (WF) algorithm is a widely used design strategy in the radar waveform design field to maximize the signal-to-interference-plus-noise ratio (SINR). To address the problem of the poor resolution performance of the waveform caused by the inability to effectively control the bandwidth, a novel waveform-related optimization model is established in this paper. Specifically, a corrected SINR expression is first derived to construct the objective function in our optimization model. Then, equivalent bandwidth and energy constraints are imposed on the waveform to formulate the waveform-related non-convex optimization model. Next, the optimal frequency spectrum is obtained using the Karush–Kuhn–Tucker condition of our non-convex model. Finally, the transmit waveform in the time domain is synthesized under the constant modulus constraint. Different experiments based on simulated and real-measured data are constructed to demonstrate the superior performance of the designed waveform on the SINR and equivalent bandwidth compared to the linear frequency modulated signal and waveform designed by the WF algorithm. In addition, to further evaluate the effectiveness of the proposed algorithm in the application of cognitive radar (CR), a closed-loop radar system design strategy is introduced based on our waveform design method. The experiments under real-measured data confirm the advantages of CR compared to the traditional open-loop radar structure.

Virtual Dental Articulation Using Computed Tomography Data and Motion Tracking.

Dental articulation holds crucial and fundamental importance in the design of dental restorations and analysis of prosthetic or orthodontic occlusions. However, common traditional and digital articulators are difficult and cumbersome in use to effectively translate the dental cast model to the articulator workspace when using traditional facebows. In this study, we have developed a personalized virtual dental articulator that directly utilizes computed tomography (CT) data to mathematically model the complex jaw movement, providing a more efficient and accurate way of analyzing and designing dental restorations. By utilizing CT data, Frankfurt's horizontal plane was established for the mathematical modeling of virtual articulation, eliminating tedious facebow transfers. After capturing the patients' CT images and tracking their jaw movements prior to dental treatment, the jaw-tracking information was incorporated into the articulation mathematical model. The validation and analysis of the personalized articulation approach were conducted by comparing the jaw movement between simulation data (virtual articulator) and real measurement data. As a result, the proposed virtual articulator achieves two important functions. Firstly, it replaces the traditional facebow transfer process by transferring the digital dental model to the virtual articulator through the anatomical relationship derived from the cranial CT data. Secondly, the jaw movement trajectory provided by optical tracking was incorporated into the mathematical articulation model to create a personalized virtual articulation with a small Fréchet distance of 1.7 mm. This virtual articulator provides a valuable tool that enables dentists to obtain diagnostic information about the temporomandibular joint (TMJ) and configure personalized settings of occlusal analysis for patients.

Road friction coefficient estimation under low longitudinal, lateral and combined excitation

On the way to autonomous driving more and more advanced driver assistance systems (ADASs) are developed to increase safety and comfort. These functions require additional quantified information about the environment, for example information about the current road condition, which a human driver would perceive intuitively. In this paper, a road condition estimation, described by a road friction coefficient, is developed using a second order divided differences filter in combination with estimates of the time varying parameters, namely the dynamic wheel radius and the tire stiffness. Using real measurement data of commercial vehicles, the estimation algorithm is compared to a static and less complex longitudinal friction coefficient estimator. The comparison shows good results for both estimators with differences at low excitation and during cornering maneuvers.

ELoran signal message prediction algorithm based on Alexnet‐ECA

AbstractEnhanced Loran (eLoran) is the most important backup system for Global Navigation Satellite System (GNSS). However, most current eLoran demodulation methods are based on digital signal processing, which may produce large errors under the influence of strong interference or noise. The authors propose an eLoran signal message prediction algorithm based on deep learning. Using time‐frequency (TF) analysis convert one‐dimensional signal into two‐dimensional high‐resolution image. It improves Alexnet by adjusting the convolutional layer structure, adding an attention module, and reducing the pooling layer window to construct the Alexnet‐ECA network. The network is used to identify TF image features and predict signal message. The experimental results show that novel method for eLoran signal message prediction based on Alexnet‐ECA can achieve high accuracy and robustness in various SNR scenarios. The accuracy is more than 90% at SNR greater than 10db and not less than 80% even in the complex environments. Compared with Resnet34 and VGG16 networks, Alexnet‐ECA demonstrates its efficiency and effectiveness using real measurement data. This method can enhance the performance and reliability of eLoran receivers and applications.

The effects of coolant mass flow rate and atmospheric indicators in a PV/T system with experimental and ANN’s models

In this study, the effects of the mass flow rate of coolant and other atmospheric and positional variables on the performance and efficiency of a PV/T system have been experimentally researched and modeled with ANN’s models. Mass flow rate, irradiance, ambient temperature, wind speed, wind direction, relative humidity and coolant water inlet temperature have been measured while solar azimuth angle and solar elevation angle have been calculated. These parameters have been selected as the input parameters (independent variables), mainly having the preliminary focus on the mass flow rate of the coolant. The performance parameters namely, electrical power production, thermal energy production, total energy production, temperature decrease on the PV module, electrical efficiency increase, surface temperature of the PV module and water outlet temperature for the PV modules have been experimentally measured. They were also computed by ANN models, as the output variables (dependent variables) with the most robust estimation of FBPANN based models depending on the real measurement data set from the experimental works. The relevance of the dependability of each targeted variables were revealed by statistical indicators in order to maintain the sensitivity of the ANN calculations. Among all the dependent parameters that were predicted by the ANN models, the strictest predictions were found for electrical power production by the FFANN1 model with the R value of 99.985% and the R2 value of 99.971%. Also, all of the performance parameters have robustly been predicted by the FFANN models. Besides, the order of importance of the predictors that effects dependent variables has been found. The mass flow rate has been determined to be highly effective on electrical efficiency increase, surface temperature of the PV module, electrical power production, thermal energy production and total energy production. The benefit of the present work is that the effects of daily working parameters such as mass flow rate of coolant with respect to the other atmospheric indicators on instantaneous electrical and thermal performance of a PV or PV/T system can be robustly predicted with their order of importance with the developed ANN models. It will make the investors to easily manage their systems.

Designing a Model of Solar Photovoltaic Water Pumping System for Off-Grid Localities in Tanzania

Modeling of photovoltaic (PV) water pumping system for rural water localities is vital especially in these modern days, in Tanzania. This is due to the rise of high number of solar-powered rural water projects constructed lacking appropriate design. The fate result into poor and substandard projects delivering below expectations. Kikombo Solar Water Project situated in Dodoma; Tanzania is one among poorly designed solar project chosen as a case study to validate proposed model. The main problem of Kikombo project is unsatisfactory amount of water from domestic points to satisfy population demand. This matter has been raised by pumping-station attendants and the Kikombo village community. The Kikombo water project can only deliver 14.21 m3/day while required demand is 50 m3/day. In order to solve the existing problem, solar water pumping model was developed. It consists of PV array, a motor-pump (AC), inverters (DC–AC converter), maximum power point tracker (MPPT) integrated with a controller and a storage water tank. Analysis was done in MATLAB to estimate which parameters should be upgraded to meet the water requirements. Moreover, the input and output variable for each model component were identified and simulated in MATLAB software. Finally, PV model was validated and system capacity improvement to rectify designs limitation for Kikombo is proposed. The model results were validated through real measurement data at Kikombo water pumping station. The differential error of 0.04062 was observed as a system error. PV model displayed an error of 0.0509 for power output, 0.0383 for current and 0.0131 for voltage; inverter displayed error of 0.0686 and 0.0406 for motor pump. In order to acquire sufficient water demand of 50 m3 /day, the PV panel and pump capacities should be increased to 8.82 kW and 7.35 kW respectively from the existing PV panel’s power of 2.5 kW and motor-pump of 4 kW.

Online State-of-Health Estimation for NMC Lithium-Ion Batteries Using an Observer Structure

State-of-health (SoH) estimation is one of the key tasks of a battery management system, (BMS) as battery aging results in capacity- and power fade that must be accounted for by the BMS to ensure safe operation over the battery’s lifetime. In this study, an online SoH estimator approach for NMC Li-ion batteries is presented which is suitable for implementation in a BMS. It is based on an observer structure in which the difference between a calculated and expected open-circuit voltage (OCV) is used for online SoH estimation. The estimator is parameterized and evaluated using real measurement data. The data were recorded for more than two years on an electrified bus fleet of 10 buses operated in a mild European climate and used regularly in the urban transport sector. Each bus is equipped with four NMC Li-ion batteries. Every battery has an energy of 30.6 kWh. Additionally, two full-capacity checkup measurements were performed for one of the operated batteries: one directly after production and one after two years of operation. Initial validation results demonstrated a SoH estimation accuracy of ±0.5% compared to the last checkup measurement.

Enhancing heliostat calibration on low data by fusing robotic rigid body kinematics with neural networks

Solar tower power plants rely on very precise alignments of their heliostats for efficient operation. To achieve this, the heliostats have to be calibrated regularly. Open-loop calibration procedures are the most common type due to their cost-effectiveness. Two main approaches to these algorithms exist: geometry-based robotic kinematics and neural network-based models. While the former is reliable and requires little data, it only yields moderate accuracy. The latter, however, promises higher accuracies but is data-hungry and unreliable. We here present a two-layer hybrid model that combines a well-established geometric model for pre-alignment with a neural network disturbance model whose impact is gradually adapted through a regularization sweep. This approach ensures that the prediction accuracy is, in the worst-case, equivalent to that of the rigid-body model. Moreover, it helps to identify deficiencies that may have been overlooked by the physical approach. Especially, it is capable to compute deviation from the geometry models averaged optimum. For testing, real measurement data from daily heliostat calibration at the solar tower in Jülich is used. To report accuracies which hold true throughout the year a special training/validation data sampling is employed, which allows for a conservative performance assumption. The results demonstrate that the Hybrid-model outperforms rigid-body models starting from the first measurement, achieving a top performance below 0.7 milliradians. In conclusion, the proposed hybrid model provides a cost effective in-situ solution for heliostat calibration with highest accuracies on low data in solar tower power plants for all open-loop calibration methods.

Multiagent Q-Learning-Based Mobility Management for Multi-Connectivity in mmWAVE Cellular Systems.

Effective mobility management is crucial for efficient operation of next-generation cellular systems in the millimeter wave (mmWave) band. Massive multiple-input-multiple-output (MIMO) systems are seen as necessary to overcome the significant path losses in this band, but the highly directional beam makes the channels more susceptible to radio link failures due to blockages. To meet stringent capacity and reliability requirements, multi-connectivity has attracted significant attention. This paper proposes a multiagent distributed Q learning-based mobility management scheme for multi-connectivity in mmWave cellular systems. A hierarchical structure is adopted to address the model complexity and speed up the learning process. The performance is assessed using a realistic measurement data set collected from Wireless Insite in an urban area and compared with independent Q learning and a heuristic scheme in terms of handover probability and spectral efficiency.