The 10 most recently published documents
The installed energy capacity of renewable energy generation systems is increasing globally due to the implementation of decarbonization policies. Due to the unpredictable nature of renewable energy sources, there is frequently a mismatch between load demand and energy supply. Stationary energy storage systems act as a buffer that stores excessive energy to balance future energy shortcomings. For residential applications, battery energy storage systems (BESS) are an attractive solution to realize the self-sufficiency of a household equipped with photovoltaics. Despite the decrease in prices, battery costs are still the most significant part of the investment cost of battery energy storage systems. Therefore, estimating battery lifetime and developing operation strategies to hinder aging is essential for improving the feasibility of BESS.
Lithium iron phosphate (LFP) lithium-ion batteries are widely used for residential BESS because of their low cost, long life, and safety. Despite extensive research in the laboratory of small-capacity cells, there are few full-scale field investigations on the lifetime and aging characteristics of commercial BESS equipped with LFP cells. This Ph.D. thesis investigates the realistic aging behavior of a residential-scale BESS equipped with large-format (180 Ah) LFP cells. We aim to create and implement a method to investigate the practical aging on cell, stack, and system levels. The experimental data is also processed by degradation modes analysis to identify the underlying mechanisms of capacity loss.
Each cell underwent primarily detailed electrical characterization, which consists of measuring characteristic charge/discharge curves and internal resistances at different current rates and temperatures. After electrical characterization, exemplary cells were opened in an inert atmosphere glove-box for structural investigation, consisting of size and weight measurements of cell components. The morphology and chemical composition of electrode samples from opened cells were investigated with light microscopy (LM) and scanning electron microscopy (SEM). The results of the detailed initial characterization of single cells were used to create a complete and self-consistent parameter set for each cell. The initial dataset was also used to compare periodical performance test results with the initial aging state of single cells throughout aging experiments.
The realistic aging experiment was carried out by investigating for 1000 days the changes in aging indicators of two commercial, residential scale BESS integrated into a microgrid. The battery stacks of both systems were built with LFP cells from the same batch of detailed characterization cells but installed with different (serial and parallel) configurations. Thus, we could investigate the effect of stack architecture on aging comparatively. At the end of the measurements, it was observed that no stack architecture is superior to another despite different operating voltages and current levels. In parallel, two LFP cells (identical to the battery stack cells) were tested at constant ambient temperature (20 °C) with continuous complete charge/discharge cycles at a constant current higher than the maximum current exhibited by BESS cells. Regarding capacity retention by equivalent full cycles, both cells outperformed BESS stacks. The comparative cell, stack, and system-level aging investigations indicate that good thermal management can provide a better lifetime even under harsher operating conditions.
The individual effects of temperature and load profile on aging were investigated via single-cell experiments in controlled ambient temperature. For this purpose, six test groups of single cells were tested to represent three realistic aging scenarios (continuous cycling, fully charged storage, and partially charged storage) at two different ambient temperatures (35 °C and 50 °C). All cells tested at 50 °C aged faster than those tested at 35 °C according to periodical performance diagnostics. Continuous cycling increased capacity loss among the cells tested at the same ambient temperature compared to fully or partially charged storage.
Single-cell experiment data was analyzed using degradation mode analysis algorithms. The results demonstrate that the loss of lithium inventory, attributed to the irreversible loss of lithium due to continuous growth of the solid electrolyte interface (SEI) layer, is the primary aging mode in all cases.
The peak-to-average power ratio (PAPR), commonly used to describe the amplitude variations of an OFDM (orthogonal frequency-division multiplex) signal, does not accurately reflect its impact on the system performance. This paper applies the mutual information as a metric to assess the effects of nonlinear PAPR reduction schemes on the performance of OFDM systems. Evaluation of the achieved mutual information shows that a significant capacity loss from clipping occurs only at high SNR (signal-to-noise ratio) and a simple compression/expansion technique is proposed to achieve close to optimal performance in this regime. The effectiveness of this method is validated through WER (word error rate) simulations with several modulation and coding schemes.
Der Tanz der Bienen beginnt
(2005)
IEEE802.15.4 und ZigBee sind offene Funknetzstandards, die aufeinander aufbauen. Sie sollen verlustleistungsarme drahtlose Netzwerke über kurze Distanzen ermöglichen, zum Beispiel in der Automatisierungstechnik und Heimautomatisierung. Die mit ihnen einhergehende Standardisierung und vergleichsweise geringen Kosten könnten helfen, Funkschnittstellen verstärkt in Massenmärkten einzusetzen. Der folgende Beitrag bietet eine Zusammenfassung der wichtigsten Eigenschaften der Standards und den aktuellen Stand bei Spezifizierung und Produkten.
This study aims to analyze a novel indirect photoacoustic sensor (PAS) using Machine Learning techniques. The studies focus on understanding the sensor’s repeatability, the influence of temperature and humidity on microphone output voltage, and the applicability of Machine Learning models to accurately describe the sensor’s behavior. To describe the sensor behavior, two studies are carried out in a controlled setting. With a R2 score of 0.964 between the microphone voltage and the gas concentration in ppm, the first study illustrates the sensor’s repeatability for concentration measurements. The second study looks at how temperature and humidity affect microphone output voltage. For this study, R2 score of 0.948 is obtained. The studies underscore the necessity for further investigation of the sensor under diverse testing conditions. The findings demonstrate that the sensor exhibits consistent behavior and can be effectively modeled using Machine Learning techniques.
This study explores the development of a quantitative PCR (qPCR) method for measuring fungal biomass in mixed fermentation of Aspergillus oryzae strains grown on soy okara. The research aimed to address the challenge of differentiating between fungal and plant biomass during the bioconversion of food industry by-products. Key steps included optimizing biomass harvesting techniques, developing a selective qPCR assay, and establishing a correlation between qPCR output and dry mycelial biomass.
A 100 µm strainer method was identified for reliable biomass harvesting. The qPCR assay development involved in-silico and in-vitro primer analysis, with the ITS5 and 58A2R primer pair selected. Despite challenges with high GC content affecting the melting curve, the assay was optimized for selectivity and sensitivity.
Correlation between qPCR copy number and dry mycelial biomass was established for A. oryzae WK and KS strains, with R² values of 0.9795 and 0.9326, respectively. This enabled the quantification of mycelial biomass throughout the fermentation process, revealing that after 48 hours, mycelial biomass accounted for 24.92% and 31.95% of the total mixed biomass for White Koji and Koji Shoyu strains, respectively.
Developed method provides a tool for monitoring fungal growth in mixed biomass systems, supporting research in sustainable food production and the valorization of food industry by-products through fungal fermentation.
This study explores the temporal response of an indirect photoacoustic CO 2 sensor, focusing on combined temperature and humidity variations. The intricate relationship between temporal resolution, environmental conditions, and sensor repeatability is investigated. Through Studies 1 to 4, the impact of feature differences and explicit values on repeatability under varying temporal resolutions is assessed. A temporal resolution of 700 seconds is seen at the threshold R2 score of 0.80 when using changes in temperature and humidity as features, compared to 300 seconds when using explicit values of temperature and humidity as features. Furthermore, using time difference as an extra feature allows for prediction with varying temporal resolution, resulting in R2 score of up to 0.9933. The findings enhance the understanding of the sensor’s behavior in dynamic settings, contributing to its practical applicability and performance optimization.
Time-Sensitive Networking (TSN) is becoming increasingly important. Especially in the field of industrial applications, the demand for uniform, converged real-time networks is continuously increasing. Furthermore, the request to integrate wireless, mobile, and real-time capable network elements is getting more and more relevant to industrial automation use cases. To address these requests, the 3rd Generation Partnership Project (3GPP) has extended their specifications for mobile telecommunication protocols by descriptions to integrate 5G mobile networks into TSN starting from Release 16 onwards. While the specifications provide a good theoretical overview, there is still a lack of real implementations or even proof of concepts. Therefore, we started an implementation of a 5G network that is ready to be integrated into existing TSN. This work gives an overview of the current work in progress, mainly focusing on the implementation of the TSN Application Function (TSN AF) and the time synchronization features within the TSN Translators (DS-TT and NW-TT). It also shows current limitations and difficulties and how we have overcome them with our setup.
A new algorithm for incremental learning in the context of Tiny Machine learning (TinyML) is presented, which is optimized for low-performance and energy efficient embedded devices. TinyML is an emerging field that deploys machine learning models on resource-constrained devices such as microcontrollers, enabling intelligent applications like voice recognition, anomaly detection, predictive maintenance, and sensor data processing in environments where traditional machine learning models are not feasible. The algorithm solve the challenge of catastrophic forgetting through the use of knowledge distillation to create a small, distilled dataset. The novelty of the method is that the size of the model can be adjusted dynamically, so that the complexity of the model can be adapted to the requirements of the task. This offers a solution for incremental learning in resource-constrained environments, where both model size and computational efficiency are critical factors. Results show that the proposed algorithm offers a promising approach for TinyML incremental learning on embedded devices. The algorithm was tested on five datasets including: CIFAR10, MNIST, CORE50, HAR, Speech Commands. The findings indicated that, despite using only 43% of Floating Point Operations (FLOPs) compared to a larger fixed model, the algorithm experienced a negligible accuracy loss of just 1%. In addition, the presented method is memory efficient. While state-of-the-art incremental learning is usually very memory intensive, the method requires only 1% of the original data set.
This study introduces EmbeddedTrain, an innovative algorithm optimized for on-device learning in deep neural networks, specifically designed for low-power microcontroller units. EmbeddedTrain refines sparse backpropagation by dynamically adjusting the level of sparity, including the ability to selectively skip training steps. This feature significantly lowers computational effort without substantially compromising accuracy. Our comprehensive evaluation across diverse datasets—CIFAR 10, CIFAR100, Flower, Food, Speech Command, MNIST, HAR, and DCASE2020—reveals that EmbeddedTrain achieves near-parity with full training methods, with an average accuracy drop of only around 1% in most cases. For instance, against full training, EmbeddedTrain’s accuracy drop is minimal, for example, only 0.82% on CIFAR 10 and 1.07% on CIFAR100. In terms of computational effort, EmbeddedTrain shows a marked reduction, requiring as little as 10% of the computational effort needed for full training in some scenarios, and consistently outperforms other sparse training methodologies. These findings underscore EmbeddedTrain’s capacity to efficiently manage computational resources while maintaining high accuracy, positioning it as an advantageous solution for advanced embedded device applications in the IoT ecosystem.