https://hdl.handle.net/20.500.11984/1148 Thu, 09 Jul 2026 03:12:11 GMT 2026-07-09T03:12:11Z https://hdl.handle.net/20.500.11984/14590 Active damping control strategy to avoid resonance issues in diesel-electric vessels with DC distribution systems Alacano, Argiñe; Abad, Gonzalo; Valera, Juan José This paper presents an active damping control strategy especially designed to mitigate the effects of the resonances of low damped power electronic based DC distribution systems. In those systems, some controlled converters are connected to a common DC bus usually exhibiting negative impedance due to its operation as constant power loads. The active damping control strategy is supported by a model based design approach, which defines the dynamic behavior of the entire system. It is implemented in a complete simulation model which considers the nonlinearities of the electronic power converters such as the delays or discretization, among others. The proposed control strategy improves the power quality in both, the transient and the steady state without compromising the voltage and current controllers' bandwidths. The performance of the proposed active damping control strategy is validated through several simulation results. Fri, 01 Jan 2016 00:00:00 GMT https://hdl.handle.net/20.500.11984/14590 2016-01-01T00:00:00Z https://hdl.handle.net/20.500.11984/14588 A multivariable modeling approach for the design of Power Electronics Based DC Distribution Systems in diesel-electric vessels Alacano, Argiñe; Valera, Juan José; Abad, Gonzalo The benefits of using Power Electronics Based DC Distribution Systems in diesel-electric vessels are well known. However, some aspects must be deeply analyzed to guarantee a safe, robust and stable system by design. A multivariable DC Distribution System mathematical model is presented and described in this work, where all the transmission lines and filters impedances are considered. The model has been tackled under a holistic approach in which the average small signal model of the drives/converters can be easily added and `connected' to the main grid model. The stability and signal power quality analysis, as well as the design of controls and active damping strategies can be conducted through this mathematical model at low computational cost. The usefulness of this model in the early design stages is thus presented in this paper through its application over a realistic design scenario. Fri, 01 Jan 2016 00:00:00 GMT https://hdl.handle.net/20.500.11984/14588 2016-01-01T00:00:00Z https://hdl.handle.net/20.500.11984/14571 Enhancing World Models with Specialized Prediction Networks for Reinforcement Learning Mellado Ibañez, Álvaro; Arana-Arexolaleiba, Nestor; Vázquez, Juan Ignacio Training robots in the real-world using reinforcement learning is both expensive and risky. World Models—a simulated environment that mirrors real-world conditions—have been proved to offer an alternative to real-world training. Such simulation-based training not only reduces costs significantly but also reduces the dependency from real-world testing. While previous studies focus on single-network architectures that predict state, reward, and episode termination as a single output, this research proposes a different approach by creating a structure based on specialized prediction networks for each of the aforementioned elements. During the experiment, several simulated environments were used. The main results obtained showed that the specialized-network World Models were capable of learning the environment’s dynamics adequately, and that the proposed architecture outperformed single-network configurations by more effectively capturing these dynamics. Finally, future directions are included on possible ways to enhance World Models efficiency. Wed, 01 Jan 2025 00:00:00 GMT https://hdl.handle.net/20.500.11984/14571 2025-01-01T00:00:00Z https://hdl.handle.net/20.500.11984/14569 Reality Bites: Assessing the Realism of Driving Scenarios with Large Language Models Wu, Jiahui; Lu, Chengije; Arrieta, Aitor; Yue, Tao; Ali, Shaukat Large Language Models (LLMs) are demonstrating outstanding potential for tasks such as text generation, summarization, and classification. Given that such models are trained on a humongous amount of online knowledge, we hypothesize that LLMs can assess whether driving scenarios generated by autonomous driving testing techniques are realistic, i.e., being aligned with real-world driving conditions. To test this hypothesis, we conducted an empirical evaluation to assess whether LLMs are effective and robust in performing the task. This reality check is an important step towards devising LLM-based autonomous driving testing techniques. For our empirical evaluation, we selected 64 realistic scenarios from DeepScenario-an open driving scenario dataset. Next, by introducing minor changes to them, we created 512 additional realistic scenarios, to form an overall dataset of 576 scenarios. With this dataset, we evaluated three LLMs (GPT-3.5, Llama2-13B, and Mistral-7B) to assess their robustness in assessing the realism of driving scenarios. Our results show that: (1) Overall, GPT-3.5 achieved the highest robustness compared to Llama2-13B and Mistral-7B, consistently throughout almost all scenarios, roads, and weather conditions; (2) Mistral-7B performed the worst consistently; (3) Llama2-13B achieved good results under certain conditions; and (4) roads and weather conditions do influence the robustness of the LLMs. © 2024 is held by the owner/author(s). Publication rights licensed to ACM. Mon, 01 Jan 2024 00:00:00 GMT https://hdl.handle.net/20.500.11984/14569 2024-01-01T00:00:00Z