Fault delineation and stress orientations from the analysis of background, low magnitude seismicity in Southern Apennines (Italy)

The aim of this work was to show that refined analyses of background, low magnitude seismicity allow to delineate the main active faults and to accurately estimate the directions of the regional tectonic stress that characterize the Southern Apennines (Italy), a structurally complex area with high seismic potential. Thanks the presence in the area of an integrated dense and wide dynamic network, was possible to analyzed an high quality microearthquake data-set consisting of 1312 events that occurred from August 2005 to April 2011 by integrating the data recorded at 42 seismic stations of various networks. The refined seismicity location and focal mechanisms well delineate a system of NW-SE striking normal faults along the Apenninic chain and an approximately E-W oriented, strike-slip fault, transversely cutting the belt. The seismicity along the chain does not occur on a single fault but in a volume, delimited by the faults activated during the 1980 Irpinia M 6.9 earthquake, on sub-parallel predominant normal faults. Results show that the recent low magnitude earthquakes belongs to the background seismicity and they are likely generated along the major fault segments activated during the most recent earthquakes, suggesting that they are still active today thirty years after the mainshock occurrences. In this sense, this study gives a new perspective to the application of the high quality records of low magnitude background seismicity for the identification and characterization of active fault systems. The analysis of the stress tensor inversion provides two equivalent models to explain the microearthquake generation along both the NW-SE striking normal faults and the E- W oriented fault with a dominant dextral strike-slip motion, but having different geological interpretations. We suggest that the NW-SE-striking Africa-Eurasia convergence acts in the background of all these structures, playing a primary and unifying role in the seismotectonics of the whole region.

Toward a Magnitude Scale for Low Frequency Seismicity The LP energy for Campi Flegrei caldera

The energy released during a seismic crisis in volcanic areas is strictly related to the physical processes in the volcanic structure. In particular Long Period seismicity, that seems to be related to the oscillation of a fluid-filled crack (Chouet , 1996, Chouet, 2003, McNutt, 2005), can precedes or accompanies an eruption. The present doctoral thesis is focused on the study of the LP seismicity recorded in the Campi Flegrei volcano (Campania, Italy) during the October 2006 crisis. Campi Flegrei Caldera is an active caldera; the combination of an active magmatic system and a dense populated area make the Campi Flegrei a critical volcano. The source dynamic of LP seismicity is thought to be very different from the other kind of seismicity ( Tectonic or Volcano Tectonic): it’s characterized by a time sustained source and a low content in frequency. This features implies that the duration–magnitude, that is commonly used for VT events and sometimes for LPs as well, is unadapted for LP magnitude evaluation. The main goal of this doctoral work was to develop a method for the determination of the magnitude for the LP seismicity; it’s based on the comparison of the energy of VT event and LP event, linking the energy to the VT moment magnitude. So the magnitude of the LP event would be the moment magnitude of a VT event with the same energy of the LP. We applied this method to the LP data-set recorded at Campi Flegrei caldera in 2006, to an LP data-set of Colima volcano recorded in 2005 – 2006 and for an event recorded at Etna volcano. Experimenting this method to lots of waveforms recorded at different volcanoes we tested its easy applicability and consequently its usefulness in the routinely and in the quasi-real time work of a volcanological observatory.

Adaptive and Maladaptive Implications of Reinforcement Learning Processes: Fronto-Striatal Loops and Behavioural Correlates

That humans and animals learn from interaction with the environment is a foundational idea underlying nearly all theories of learning and intelligence. Learning that certain outcomes are associated with specific actions or stimuli (both internal and external), is at the very core of the capacity to adapt behaviour to environmental changes. In the present work, appetitive and aversive reinforcement learning paradigms have been used to investigate the fronto-striatal loops and behavioural correlates of adaptive and maladaptive reinforcement learning processes, aiming to a deeper understanding of how cortical and subcortical substrates interacts between them and with other brain systems to support learning. By combining a large variety of neuroscientific approaches, including behavioral and psychophysiological methods, EEG and neuroimaging techniques, these studies aim at clarifying and advancing the knowledge of the neural bases and computational mechanisms of reinforcement learning, both in normal and neurologically impaired population.

Unsupervised reinforcement learning via state entropy maximization

Reinforcement Learning (RL) provides a powerful framework to address sequential decision-making problems in which the transition dynamics is unknown or too complex to be represented. The RL approach is based on speculating what is the best decision to make given sample estimates obtained from previous interactions, a recipe that led to several breakthroughs in various domains, ranging from game playing to robotics. Despite their success, current RL methods hardly generalize from one task to another, and achieving the kind of generalization obtained through unsupervised pre-training in non-sequential problems seems unthinkable. Unsupervised RL has recently emerged as a way to improve generalization of RL methods. Just as its non-sequential counterpart, the unsupervised RL framework comprises two phases: An unsupervised pre-training phase, in which the agent interacts with the environment without external feedback, and a supervised fine-tuning phase, in which the agent aims to efficiently solve a task in the same environment by exploiting the knowledge acquired during pre-training. In this thesis, we study unsupervised RL via state entropy maximization, in which the agent makes use of the unsupervised interactions to pre-train a policy that maximizes the entropy of its induced state distribution. First, we provide a theoretical characterization of the learning problem by considering a convex RL formulation that subsumes state entropy maximization. Our analysis shows that maximizing the state entropy in finite trials is inherently harder than RL. Then, we study the state entropy maximization problem from an optimization perspective. Especially, we show that the primal formulation of the corresponding optimization problem can be (approximately) addressed through tractable linear programs. Finally, we provide the first practical methodologies for state entropy maximization in complex domains, both when the pre-training takes place in a single environment as well as multiple environments.

Thermal and air quality modeling of an electric vehicle cabin with low-cost sensors and reinforcement learning

The design process of any electric vehicle system has to be oriented towards the best energy efficiency, together with the constraint of maintaining comfort in the vehicle cabin. Main aim of this study is to research the best thermal management solution in terms of HVAC efficiency without compromising occupant’s comfort and internal air quality. An Arduino controlled Low Cost System of Sensors was developed and compared against reference instrumentation (average R-squared of 0.92) and then used to characterise the vehicle cabin in real parking and driving conditions trials. Data on the energy use of the HVAC was retrieved from the car On-Board Diagnostic port. Energy savings using recirculation can reach 30 %, but pollutants concentration in the cabin builds up in this operating mode. Moreover, the temperature profile appeared strongly nonuniform with air temperature differences up to 10° C. Optimisation methods often require a high number of runs to find the optimal configuration of the system. Fast models proved to be beneficial for these task, while CFD-1D model are usually slower despite the higher level of detail provided. In this work, the collected dataset was used to train a fast ML model of both cabin and HVAC using linear regression. Average scaled RMSE over all trials is 0.4 %, while computation time is 0.0077 ms for each second of simulated time on a laptop computer. Finally, a reinforcement learning environment was built in OpenAI and Stable-Baselines3 using the built-in Proximal Policy Optimisation algorithm to update the policy and seek for the best compromise between comfort, air quality and energy reward terms. The learning curves show an oscillating behaviour overall, with only 2 experiments behaving as expected even if too slow. This result leaves large room for improvement, ranging from the reward function engineering to the expansion of the ML model.

Directly training spiking neural networks for cyber-physical systems: from supervised to reinforcement learning

Spiking Neural Networks (SNNs) are bio-inspired Artificial Neural Networks (ANNs) utilizing discrete spiking signals, akin to neuron communication in the brain, making them ideal for real-time and energy-efficient Cyber-Physical Systems (CPSs). This thesis explores their potential in Structural Health Monitoring (SHM), leveraging low-cost MEMS accelerometers for early damage detection in motorway bridges. The study focuses on Long Short-Term SNNs (LSNNs), although their complex learning processes pose challenges. Comparing LSNNs with other ANN models and training algorithms for SHM, findings indicate LSNNs' effectiveness in damage identification, comparable to ANNs trained using traditional methods. Additionally, an optimized embedded LSNN implementation demonstrates a 54% reduction in execution time, but with longer pre-processing due to spike-based encoding. Furthermore, SNNs are applied in UAV obstacle avoidance, trained directly using a Reinforcement Learning (RL) algorithm with event-based input from a Dynamic Vision Sensor (DVS). Performance evaluation against Convolutional Neural Networks (CNNs) highlights SNNs' superior energy efficiency, showing a 6x decrease in energy consumption. The study also investigates embedded SNN implementations' latency and throughput in real-world deployments, emphasizing their potential for energy-efficient monitoring systems. This research contributes to advancing SHM and UAV obstacle avoidance through SNNs' efficient information processing and decision-making capabilities within CPS domains.