The combination of increasing global food demand with increased food security risks associated with climate change amid a decreasing number of skilled growers necessitates innovative solutions in green- house horticulture. Autonomous growing offers a solution based on greenhouse climate forecasting and (optimal) control. However, current theoretical models developed for greenhouse climate forecast- ing face limitations due to the in-depth physics knowledge required for their use and their dependence on intricate system parameters that are difficult to estimate. Conversely, machine learning models struggle with generalisation to unseen conditions and adherence to physical laws, leading to unrealis- tic predictions in greenhouse environments. This study addresses these challenges by exploring the use of Physics-Informed Machine Learning (PIML) techniques to enhance greenhouse climate fore- casting. A simple differentiable theoretical model for simulating the greenhouse climate is proposed to serve as prior physical knowledge of the system. Subsequently, a novel PIML model is introduced in the form of Controlled Aphynity (CA), which integrates insights from neural controlled differential equations and Aphynity, and is the first such model that allows for the augmentation of incomplete prior knowledge of a dynamical system with data-driven machine learning models while being adaptable to changing dynamics due to forces acting on the system. Experimental results show that CA is capable of successfully augmenting incorrect physics under changing dynamics on the task of humidity deficit prediction in the greenhouse. Furthermore, three ensemble methods combining CA with traditional ma- chine learning techniques are explored and demonstrate promising synergies. A detailed case study evaluates CA and the best-performing ensemble approach on the task of humidity deficit prediction under realistic greenhouse scenarios over a complete crop cycle. Both CA and the ensemble methods exhibit superior adherence to physical laws, lower data requirements, and improved performance on outliers compared to conventional machine learning methods. This research contributes to advancing greenhouse climate modelling, underscoring PIML’s potential in optimising the greenhouse climate.F

Weather forecasting has always been a critical issue across various fields, and traditionally, weather predictions have relied on solving complex atmospheric equations using supercomputers. However, the rise of machine learning has introduced a more efficient method that utilizes historical data to forecast weather patterns. Nonetheless, there is a problem with auto-correlation, where values at a specific time point correlate with those at a previous time. Machine learning models assume independence between data points, which can lead to overfitting if the data is not split properly. This happens because training data might not be independent of test data and models can learn the noise rather than the underlying pattern in data. Also, auto-correlation can apply when having multiple locations with data (named spatial data), because nearby locations share similar characteristics.

This thesis investigates temporal and spatial data splitting strategies to address this challenge to determine which methods provide the most reliable performance estimates. Temporal splitting strategies involve dividing the data using time intervals, while spatial splitting strategies involve dividing based on the geographical location of weather stations. The goal is to identify strategies that minimize bias and variance in the error estimates of weather forecasting models. This will ensure less overfitting and robust and generalisable predictions across different climatic conditions and data regimes.

By systematically evaluating various data splitting strategies, this thesis aims to provide insights into the best approaches for preparing meteorological data for Machine Learning-based weather prediction models, thereby contributing to the advancement of more accurate and efficient weather forecasting techniques. Four different strategies for temporal data splitting were evaluated: random splitting, testing at the end of the dataset, testing in the middle of the dataset, and splitting the test year into multiple parts. Three different strategies for splitting spatial data were evaluated: random splitting, the use of one cluster of cities as the test data with the other clusters as training data, and lastly, choosing only the non-neighbouring clusters as training data. Some of these strategies, such as the random strategy, do not take into account auto-correlation when splitting the data.

Results indicate that the random strategy in both scenarios yields a relatively modest error rate. However, it is advisable to employ a strategy that organizes the test data into a cohesive block and positions it in the centre of the dataset time. Hence, this thesis contributes to the development of more refined weather prediction by shedding light on how data splitting strategies influence outcomes.

This thesis investigates the application of alternating maximisation for active wake control in wind farms, focusing on both numerical static yaw optimisation and multi-agent deep reinforcement learning for dynamic yaw control. As the size and number of offshore wind farms continue to grow, effectively mitigating wake effects ---~where the airflow behind a turbine is disturbed, reducing the efficiency of downstream turbines~--- becomes increasingly important to maximise power output and ensure economic profitability.

Current numerical strategies are either sub-optimal, as they do not leverage the shape of the optimisation surface well, or expensive to optimise due to the computational demands of high-fidelity simulations. Similarly, current reinforcement learning approaches are too sample inefficient on contemporary wind farms of 40+ turbines to learn high-quality policies and likely suffer from credit assignment problems, other agents performing exploration, and relative over-generalisation. Alternating maximisation emerges as a promising candidate to address these challenges by providing an efficient and potentially more effective optimisation method.

In static yaw optimisation, the research analysed the effects of wind direction, wind speed, turbine layout, and yaw misalignment on the optimisation surface. Key findings indicate that Nash equilibria are transient and significantly influenced by these factors.

The sampling-based approach to alternating maximisation outperformed other methods, proving robust against variations in initial yaw configurations and optimisation orders. This approach required fewer samples and appears suitable for practical applications.

For dynamic yaw control, alternating maximisation with policy gradients demonstrated greater sample efficiency compared to distributed policy gradient methods. However, convergence to sub-optimal policies was common due to limited exploration capabilities. An oscillating exploration strategy enabled effective exploration of yaw angles, leading to high-quality policies.

The study primarily focused on farm power output, neglecting structural loading and leading-edge erosion. Future research should integrate these factors and explore dynamic yaw control under time-varying wind conditions.

Overall, alternating maximisation shows potential in enhancing wind farm performance through efficient optimisation strategies. Future work should refine these methods and address identified limitations to maximise practical applicability.

The increasing demand for electricity has lead to demand for more efficient energy production. One promising option is wind power, which currently provides an estimated 7.8% of the world’s energy production. One of the problems with wind energy is that a small percentage of the energy is lost due to the wake effect. The wake of a wind turbine is an area of low wind speed and high turbulence which is caused by the spinning of the turbine. This wake effect can mitigated by active wake control, which is a process by which the wake from a turbine is redirected away from downwind turbines, by changing the yaw of the turbine head. Calculating a policy for doing this is computationally expensive to do using numerical optimisation. Therefore, multi agent reinforcement learning is proposed to learn a policy which performs active wake control.
The proposed approach makes use of the popular reinforcement learning algorithm REINFORCE, and extends it using a variety of methods. First, a simplified version of the problem is treated, wherein the wind direction is fixed. Then the problem is made more realistic by introducing changing wind directions. The first extension of REINFORCE that is treated is difference rewards, a reward shaping strategy which seeks to solve the credit assignment problem, thereby improving cooperation between turbines. The second method uses training regimes, which train different agents at different times to stabilise the environment as much as possible. Next, role-based reinforcement learning is used to conteract the complexity of the problem by allowing each agent to specialise for a certain role. Finally, since roles cannot be manually determined for larger farms, influence-based abstraction is used to enable agents to learn the roles themselves, by abstracting spacial information and presenting it to the agent as an observation.
The results demonstrate that multi agent reinforcement learning can be used to perform active wake control in wind farms. Furthermore, the extensions proposed are shown to improve learning, and lead to greater energy output. While multi agent reinforcement learning is shown to be a promising way to tackle active wake control in wind farms, research is needed to improve the stability of the learned policies.

Offline model-based reinforcement learning uses a model of the environment, learned from a static dataset of interactions, to guide policy generation. Sub-optimal planning decisions can be made when the agent explores states that are out-of-distribution, as the world model will have more uncertainty. This paper explores the use of pessimism, the tendency to avoid uncertain states, in the planning procedure. We evaluate Lower Confidence Bound, ensembles, and Monte Carlo dropout in the MinAtar breakout environment. Results indicate that ensemble methods yield the highest performance, with a significant performance gain over the baseline, while LCB also shows varying degrees of improvement. MC dropout is generally shown to not yield a performance improvement.

Traditionally, Recurrent Neural Networks (RNNs) are used to predict the sequential dynamics of the environment. With the advancement and breakthroughs of Transformer models, there has been demonstrated improvement in the performance & sample efficiency of Transformers as world models. The focus has been on partially-observable environments where their capabilities can be maximally utilised. In this paper, we sought to investigate the conditions under which transformers outperform RNNs given a fully observable environment where states obey the Markov property. This provides insight into transformers' generalisation and predictive capabilities. Specifically, our experiments explored the impact of model complexity and the size of the dataset. We observed that transformers did not outperform our baseline implementation when given up to 7000 episodes of trajectory data. It was also observed that having shorter sequence lengths had a negligible impact on the performance of the model, leading to our recommendation of avoiding using transformers in these fully observable environments.

While model-free reinforcement learning (MFRL) approaches have been shown effective at solving a diverse range of environments, recent developments in model-based reinforcement learning (MBRL) have shown that it is possible to leverage its increased sample efficiency and generalisation abilities to solve highly complex tasks with fewer resources and environment interactions. The introduction of discrete latent states through categorical distributions allowed DreamerV2, a MBRL approach, to surpass the state-of-the-art MFRL Rainbow algorithm on the Arcade Learning Environment. Despite the successes of this approach, it is not yet understood why discretization improves performance. This paper investigates how the discretization of the latent space through categorical distribution affects planning performance in a deterministic environment. Further investigations are conducted on the model's generalization abilities and the impact of the latent space's shape on performance. By using a dataset of experiences instead of directly interacting with the environment, the models are trained in an offline setting. Results show that the discrete world model underperforms compared to a continuous latent space model while being significantly harder to train. Further investigations concluded that the number of categorical distributions has a high influence on performance and that in the considered setting the discrete world model can generalize better than the continuous baseline but it does so by sacrificing small gains in important metrics.

This thesis project explores how V2X communication between electric vehicles (EV) and charging stations can be used to reduce en-route charging times. This is done using standardized V2X messages for EV charging, as well as proposing an extension to these messages to include data on the intentions of other vehicles. As individual vehicles can have significant effects on the total waiting time at EV charging stations, knowing the intentions of other vehicles will allow drivers to better avoid congested charging stations and achieve a lower total journey time. The performance of the system is evaluated using state-of-the-art traffic and communications simulators showing that using V2X can reduce journey times by 70%. Aside from demonstrating the performance using simulation the system has also been implemented on a real life test vehicle using functional V2X hardware to show that the system is viable for implementation.

Writing software that follows its specification is important for many applications. One approach to guarantee this is formal verification in a dependently-typed programming language. Formal verification in these dependently-typed languages is based on proof writing. Sadly, while proofs are easy to check for computers, writing proofs can be tedious for developers. One particular proof component that currently requires developers attention in many systems, is associative and commutative (AC) reasoning.
We contribute an extension of dependent type-systems to fully automate AC reasoning. This alleviates developers from this task and allows them to concentrate on other proof components. Our approach works by modifying the conversion checker and doesn't compromise soundness or completeness. Furthermore, our approach reuses existing type-checking components, making it easier to implement. We also implemented our theory as an extension of the Agda type-checker. This allowed us to use this implementation to experiment with some example programs.
This thesis can help language designers decide if they want automatic AC reasoning in their language. For language users it can serve as inspiration on how to use such a type-system and finally for researchers we have ideas for future work.

Deep convolutional neural networks (CNNs) have achieved current state-of-the-art in image denoising, but require large datasets for training. Their performance remains limited on smaller real-noise datasets. In this paper, we investigate robust deep learning denoising using transfer learning. We explore the impact of dataset sizes, CNN parameter updates, and noise distribution similarity. Our findings demonstrate that finetuning the decoder while fixing the encoder of an encoder-decoder network architecture avoids overfitting during transfer learning. Moreover, we introduce the concept of noise similarity in transfer learning, showing that reducing the similarity distance between pre-training and finetuning noise significantly enhances CNN performance in denoising. Our results are demonstrated on multiple datasets and various noise camera models. Finally, we validate the robustness and applicability of our approach on real-noise images. All our results and analyses hold and generalize across different datasets, highlighting our insights into the potential of transfer learning for image denoising.

When multiple wind turbines are positioned close to one another, such as in a wind farm, wind turbines located downwind of other turbines are not 100% efficient due to wakes, negatively affecting the total power output of the wind farm. A way to mitigate the loss of power is to steer the wake away from the next turbine, which lowers the current turbine's power output but increases the turbine's total power output. As the number of wind turbines increases, how complex it is to calculate the optimal steering increases exponentially. Reinforcement learning techniques have been a promising candidate to solve this problem. However, single-agent techniques are still very computationally expensive when the number of turbines is the same as an average wind farm. Therefore, this paper aims to see how the QMIX algorithm, a multi-agent reinforcement learning technique, can be efficiently applied to the problem of active wake control. QMIX will be compared to the FLORIS model and a single agent deep reinforcement learning technique TD3 to see if it achieves a higher average reward and converges faster. Finally, QMIX is tested on a larger wind farm to see if it achieves any results in a reasonable amount of time, showing that multi-agent reinforcement learning techniques are more suitable to the problem. This paper shows QMIX has the potential to outperform TD3; although FLORIS is better for smaller wind farms but more research has to be done in applying QMIX to the problem.

Non-intrusive load monitoring (NILM) is a well-researched concept that aims to provide insights into individual appliance energy usage without the need for dedicated meters. This paper explores the possibility of applying the NILM concept to disaggregate energy data from a community level to a household level. By doing so, it addresses privacy concerns associated with real-time household data collection through smart meters. The proposed approach leverages sequence-to-point learning, a deep learning method, to perform the disaggregation. A method that showed promising results in the domain of NILM. The results show good performance on the evaluation metrics, but they also show that the model does a poor job of capturing the load signature.

Non-Intrusive Load Monitoring (NILM) is a technique used to disaggregate household power consumption data into individual appliance components without the need for dedicated meters for each appliance. This paper focuses on improving the generalizability of NILM algorithms to unseen households using Convolutional Neural Networks (CNNs) and one-shot transfer learning. The research investigates the effectiveness of one-shot transfer learning in fine-tuning a CNN model to accurately detect the ON/OFF state of appliances in households not seen during the training phase of the CNN. The study utilizes the Pecan Street dataset for training and evaluation, which includes detailed energy consumption records from various locations in the United States. The results suggest that one-shot transfer learning could enhance the performance of the NILM algorithm, particularly when multiple data samples are used for fine-tuning. However, the effectiveness of one-shot transfer learning varies strongly depending on the number of samples and the characteristics of the target household.

The ever-evolving power grid is becoming smarter and smarter. Modern houses come with smart meters and energy conscious consumers will buy additional smart meters to place in their home to help monitor their energy consumption. This new smart technology also opens the door to more accurate power consumption forecasting. In this study we look at utilizing a partial hierarchy, in which one of the appliances in a household is modelled separately from the rest of the house, to help improve household energy consumption forecasting accuracy. This is done in conjunction with Auto Regressive Moving Average (ARMA) based models. Three variants of ARMA based models will be looked at: Auto Regressive Integrated Moving Average (ARIMA), Seasonal Auto Regressive Integrated Moving Average (SARIMA), and Auto Regressive Integrated Moving Average with Exogenous variables (ARIMAX). These methods will then be compared to more baseline approaches such as a persistence method and a seasonal moving average. Our analysis has led us to conclude that the partial hierarchy model offers little to no benefit when applied in the field of household energy consumption forecasting when built upon ARMA based models. ARMA based models in general appeared to be poor performers when it came to household energy consumption forecasting.

Electrical load forecasting, namely short-term load forecasting, is essential to power grids’ safe and efficient operations. The need for accurate short-term load forecasting becomes increasingly pressing with increased renewable energy sources, which are stochastic in their power supply. Most forecasting models are focused on the temporal information for predictions, ignoring the spatial information of neighbouring houses. However, as neighbouring houses are often under similar circumstances, such as weather and holiday conditions, predictions
could benefit from this information. Moreover, aggregating electric loads could further improve the accuracy of predictions, as the loads become less
stochastic when aggregated. This paper looks at Graph WaveNet, which can capture the hidden spatial dependencies of different houses without needing geographic information about the houses. The framework is compared against a baseline on different aggregation levels. The results show that the framework can benefit from aggregating the residential electrical loads and improves over the baseline on all aggregation levels.

Bottom up load forecasting, is a technique where energy consumption forecasts are made on lower spatial levels, after which the resulting forecasts are aggregated to form forecasts of higher spatial levels. With the current move to renewable energy sources and the importance of reducing the strain on an already congested electricity grid, accurately forecasting both the location and time of future energy consumption has become more important than ever. To this end, this paper analyses the impact of applying bottom up load forecasting to different spatial levels on the electricity grid, including appliance, household, community and city spatial levels. Energy consumption data for these spatial levels were gathered from the 15-minute Texas and California datasets from Pecan Street Dataport. The results obtained in this study, suggest that energy consumption volatility at lower spacial levels and forecast difficulty at higher spacial levels, play an important role in the performance of applying the bottom up load forecasting technique to energy consumption forecasting problems.

TreeContainment is a well-known problem within phylogenetics, which asks whether a binary phylogenetic tree is embedded in a binary phylogenetic network. For this problem, Jones, Weller and van Iersel (2022) have created an algorithm that uses dynamic programming on tree-decompositions to achieve a running time that is exponential in the tree-width parameter instead of in the number of reticulations. However, due to the implicit formulations of two crucial steps in this algorithm, this running time cannot be achieved in practice without finding ways to generate the required structures explicitly. This paper provides two new sub-algorithms that can do that. To further improve the performance of the algorithm, I introduce a number of criteria and other methods that can be used to reduce the number of structures generated. Additionally, I describe a way to manipulate the nice tree-decompositions to create a more favorable order of bags for the dynamic programming. These sub-algorithms and improvements are used by two new algorithms TWITCH and PITCH, whose implementations are compared to a brute force algorithm and a new branching cherry-picking algorithm named BOTCH. The latter has an FPT running time that is exponential in the number of vertices that have only reticulations as children. The comparisons show that the implementations of the dynamic programming algorithms TWITCH and PITCH are slower in practice than the brute force algorithm, despite their numerous improvements. Of the four new implementations, BOTCH has the best test results and it is shown to be fast in practice.

The European Union is on course for introducing a European Digital Identity that will be available to all EU citizens and businesses. This will have a huge impact on how citizens and businesses interact online. Big Tech companies currently dictate how digital identities are used. As a result they have amassed huge amounts of private user data. Movements like Self-Sovereign Identity aim to give users control over their online identity. TrustVault is the first data wallet that gives users back control not only of their identity, but all their data. TrustVault allows users to store all their data on their smartphone and control who they share it with. The user has fine-grained access control based on verifiable user attributes. EBSI connects TrustVault to the European Self-Sovereign Identity Framework allowing users to use Verifiable Credentials from public and private institutions in their access control policies. The system is serverless and has no Trusted Third Parties. TrustVault replaces the for-profit infrastructure of Big Tech with a public and transparent platform for innovation.

Renewable energy generation projects are often measured by their peak capacity. A wind farm rated at 25 MW will generate 25 MW of power under the right circumstances. This peak capacity is reached very little in practice. However, these generators are forced to purchase grid operation infrastructure that can handle this peak generation event. The high voltage grid connections are expensive and increasingly difficult to receive permits for. This work presents a solution in which the high voltage grid connection is undersized in comparison to the renewable energy generator. A battery energy storage system is installed in the local grid to solve the issue of excess energy generation (congestion).

A simulation of the local network has been built that models a battery energy storage system (BESS), the network and uses data from a solar park. A case study in which a 19 MW solar park is connected to the high voltage grid with a transformer of only 14 MW as well as a 14 MW | 30 MWh BESS on the network is investigated in the rest of the work. Furthermore a BESS control strategy for price arbitrage on the TenneT imbalance market is presented and encoded such that it can be optimised.

Four heuristics are presented that time and size the congestion issue in a manner the control strategy of the BESS can prepare for and solve congestion when necessary. These heuristics are tested against strategies optimized for revenue maximisation through price arbitrage. While the most aggressive strategies did not solve all the congestion events in these simulations, we found that the heuristic that takes the average generation of the solar park into account performs the best while remaining appropriately conservative.

A basic evolutionary algorithm is presented that optimizes a BESS control strategy for price arbitrage when the BESS is not needed on the local network to solve congestion. Although the strategies earn ~33% less revenue due to the congestion related limitations, the optimisation surrounding congestion does improve revenue by 2.58%.

The results presented in this work suggest that this setup of a local grid can be economically viable and that the BESS can solve the congestion issue when steered with an appropriate control strategy. We hope to inspire parties that battery energy storage systems can earn substantial revenue aside from solving issues on a (local) grid.

Every day, Intrusion Detection Systems around the world generate huge amounts of data. This data can be used to learn attacker behaviour, such as Techniques, Tactics, and Procedures (TTPs). Attack Graphs (AGs) provide a visual way of describing these attack patterns. They can be generated without expert knowledge and vulnerability reports. The goal of AGs is to reduce alert fatigue, and to give another perspective on attacker behaviour.
SAGE, the state-of-the-art method for generating AGs uses FlexFringe's state merging algorithms to learn the underlying state machine to model these attacks. A big challenge of these state merging algorithms is learning infrequent behaviour. Next to that, the underlying state machines cannot deal with noisy input data.
In this work, a sequence-automaton alignment algorithm is used to align sequences of states to a state machine. Our method iteratively aligns infrequent sequences to the model, effectively learning both frequent and infrequent behaviour.
The method is evaluated on a security competition dataset, where experiments show that the algorithm is able to recover from noise such as added or removed events. The learned models are also reduced to half its original size, while better fitting to the training data.
Last, we show how our method can be used to learn the model of an anomaly detection dataset. The test data is predicted using three anomaly conditions, and results in all anomalous data being labelled correctly. The F1-score is competitive compared to other state of the art methods.