Index

Comparative analysis of word representations in LLMs using BERT and LSTM architecture

Leakage Detection and Modeling in Water Distribution Systems

Abstract:

Leakage within Water Distribution Networks (WDNs) constitutes a primary concern for utilities, significantly impacting water conservation and the ability to meet consumer demand. In addressing this pervasive issue, the present study explores the efficacy of a comprehensive detection and localization strategy across District Meter Areas (DMAs). The research hinges on the development of a calibrated hydraulic model that adheres to stringent standards, minimizing error margins and setting a precedent in the domain of water network analysis. Central to the study is the implementation of an autoencoder-based deep learning framework, which has demonstrated proficiency in the detection of leaks. To transcend the limitations imposed by sparse sensor networks, a novel hybrid approach was adopted, integrating the hydraulic model with data-driven algorithms to enhance the localization of leaks. The methodology involved segmenting the DMA into subzones and applying rolling window simulations to create a diverse dataset for training a Convolutional Neural Network (CNN). The standout performance of the model, particularly with a 6-hour rolling window, was corroborated by its ability to precisely localize leaks in an actual case scenario. Addressing this issue, our research advances a strategy for leak detection and localization that is both economical and effective, utilizing minimal sensor input and also the importance of additional sensors to localize leak even more accurately. The implications of this research extend towards reinforcing the operational integrity of WDNs, fostering water conservation, and ensuring resource availability and sustainability in the face of escalating demand.

Keywords: Water Distribution Networks, Hydraulic Modelling, Data-Driven Methods, Spectral Clustering, Leakage Detection, Leakage Localization.

This thesis is part of the “UtilityTwin” project.

Adaptive Training for Heat Demand Prediction of District Heating Network

In this thesis, we focus on short-term heat demand forecasting based on measured consumer demand. The thesis also studies approaches from Continual Learning, for developing an adaptive forecasting framework to deploy at the energy supplier to optimize supply and demand.

Heat Demand Forecasting with Multi-Resolutional Representation of Heterogeneous Temporal Ensemble

Abstract:

One of the primal challenges faced by utility companies is ensuring efficient supply with minimal greenhouse gas emissions. The advent of smart meters and smart grids provide an unprecedented advantage in realizing an optimized supply of thermal energies through proactive techniques such as load forecasting. In this research thesis, we propose a deep learning-based forecasting framework for heat demand based on neural networks where the time series are represented as scalograms equipped with the capacity of embedding exogenous variables such as weather, and holiday/non-holiday. Subsequently, CNNs are utilized to predict the heat load multi-step ahead. Finally, the proposed framework is compared with other state-of-the-art methods, such as SARIMAX and LSTM, and two variants of the proposed method.

The quantitative results from retrospective experiments show that the proposed framework consistently outperforms the state-of-the-art baseline method with real-world data acquired from Denmark. A minimal mean error of 7.69% for Mean Absolute Percentage Error and 202.80kW for Root Mean Squared Error is achieved with the proposed framework in comparison to all other methods. Moreover, the proposed method exhibits minimal deviation of performance across different climatic seasons and geographical zones compared to the baseline methods.

This thesis is part of the “UtilityTwin” project.

Graph Neural Networks in Pathological Speech

Apply Graph knowledge to applications on pathological speech (e.g., Parkinson’s disease).

Pre-requirements:

Pattern Analysis
Deep Learning
Optional: Speech and Language Understanding
Optional: Seminar on pathological speech

Please send your grades to paula.andrea.perez@fau.de and luis.rivera@fau.de

Intensity Grading Using 3D Feature Classification with fMRI Images using Deep-Learning Approach

Quantum Machine Learning Techniques in Medical Image Classification: Simulation and Hardware

Data Encoding, Parameterization and Generalization of Quantum Machine Learning for Medical Imaging

Geometry-Aware Key-Point / Object Detection and Pose-Estimation

For a wide range of emerging applications an increasing demand for reliable and accurate object detection and pose estimation, using machine learning based systems arises. This is particularly the case for autonomous systems such as autonomous vehicles and robotics but also in the context of augmented reality [1]. These applications require detecting and locating objects in real-time and in various environments, including cluttered scenes and objects with similar appearances.

However, traditional object detection and pose estimation methods often can only partially detect and locate objects in these challenging situations, leading to inaccurate and unreliable results [2]. This is where Geometry-Aware Key-Point / Object Detection and Pose-Estimation comes in, as it aims to explicitly incorporate additional geometric information into these tasks to improve their accuracy and robustness. In object detection, the goal is to identify the presence and location of objects within an image or video. Pose estimation, on the other hand, refers to estimating the position and orientation of objects in 3D space based on 2D images. By including human domain knowledge in the form of geometric constraints, we would like to utilize the knowledge of domain experts to create more robust and accurate solutions by simultaneously reducing the labeling effort associated with training data-driven solutions for novel applications.

There are various approaches to incorporating geometric information into object detection and pose estimation tasks. One common approach is to use geometry-aware convolutional neural networks (Geo-CNNs) [4], which are designed to incorporate geometric information into the model architecture explicitly. Another approach is to use geometry-aware scene graph generation [5], which uses a graph-based representation to model the geometric relationships between objects in a scene. However, our approach depends on the task at hand, object shape and orientation variability, scene complexity, and we would like to utilize the knowledge of domain experts to create more robust and accurate solutions by simultaneously reducing the labeling effort associated with training data-driven solutions for novel applications. An assessment of existing methods according to those requirements is part of the literature review corresponding to the proposed work. Afterwards, a potential adaptation of an existing method or the design and implementation of a novel approach and the corresponding evaluation should be the central task of the work.

Evaluation will be performed on industrial object detection use-case with high requirements on robustness and performance. The use-case considered for evaluating the proposed method is given by the detection of pallets in the context of an autonomous pallet unloading application. For this work it is planned to start from a public data set [7] and afterwards try to transfer results to our use-case and data, which is at least partially already collected. The thesis shall be carried out within a time period of six months including the literature review.

[1] Realtime 3D Object Detection for Automated Driving Using Stereo Vision and Semantic Information
[2] Viewpoint-Independent Object Class Detection using 3D Feature Maps
[3] Unsupervised 3D Pose Estimation With Geometric Self-Supervision
[4] Modeling Local Geometric Structure of 3D Point Clouds using Geo-CNN
[5] A Comprehensive Survey of Scene Graphs: Generation and Application
[6] From Points to Parts
[7] GitHub – tum-fml/loco: Home of LOCO, the first scene understanding dataset for logistics.
[8] Nothing But Geometric Constraints
[9] DeepIM: Deep Iterative Matching for 6D Pose Estimation