Search results for "Reinforcement"
showing 10 items of 230 documents
Deep Q-Learning With Q-Matrix Transfer Learning for Novel Fire Evacuation Environment
2021
We focus on the important problem of emergency evacuation, which clearly could benefit from reinforcement learning that has been largely unaddressed. Emergency evacuation is a complex task which is difficult to solve with reinforcement learning, since an emergency situation is highly dynamic, with a lot of changing variables and complex constraints that makes it difficult to train on. In this paper, we propose the first fire evacuation environment to train reinforcement learning agents for evacuation planning. The environment is modelled as a graph capturing the building structure. It consists of realistic features like fire spread, uncertainty and bottlenecks. We have implemented the envir…
Deep RTS: A Game Environment for Deep Reinforcement Learning in Real-Time Strategy Games
2018
Reinforcement learning (RL) is an area of research that has blossomed tremendously in recent years and has shown remarkable potential for artificial intelligence based opponents in computer games. This success is primarily due to the vast capabilities of convolutional neural networks, that can extract useful features from noisy and complex data. Games are excellent tools to test and push the boundaries of novel RL algorithms because they give valuable insight into how well an algorithm can perform in isolated environments without the real-life consequences. Real-time strategy games (RTS) is a genre that has tremendous complexity and challenges the player in short and long-term planning. The…
Expanding the Active Inference Landscape: More Intrinsic Motivations in the Perception-Action Loop
2018
Active inference is an ambitious theory that treats perception, inference and action selection of autonomous agents under the heading of a single principle. It suggests biologically plausible explanations for many cognitive phenomena, including consciousness. In active inference, action selection is driven by an objective function that evaluates possible future actions with respect to current, inferred beliefs about the world. Active inference at its core is independent from extrinsic rewards, resulting in a high level of robustness across e.g.\ different environments or agent morphologies. In the literature, paradigms that share this independence have been summarised under the notion of in…
Expert Q-learning: Deep Reinforcement Learning with Coarse State Values from Offline Expert Examples
2022
In this article, we propose a novel algorithm for deep reinforcement learning named Expert Q-learning. Expert Q-learning is inspired by Dueling Q-learning and aims at incorporating semi-supervised learning into reinforcement learning through splitting Q-values into state values and action advantages. We require that an offline expert assesses the value of a state in a coarse manner using three discrete values. An expert network is designed in addition to the Q-network, which updates each time following the regular offline minibatch update whenever the expert example buffer is not empty. Using the board game Othello, we compare our algorithm with the baseline Q-learning algorithm, which is a…
Effect of FRP strengthening on the flexural behaviour of calcarenite masonry walls
2017
The use of fiber-reinforced polymers (FRP) for structural strengthening has become increasingly popular in recent years. Several applications of FRP have been proposed and applied, depending on the target of the technique, kind and/or material of the structural member. In particular, because of their great tensile strength, FRP materials are commonly used to enhance the out-of-plane behaviour of masonry walls, allowing to increase their strength, ductility and improving safety against overturning. For these reasons, FRP laminates are often applied in vulnerable ancient buildings in seismic areas to reinforce façades and walls with poor structural features. However, some issues arise when a…
Dependence of fracture toughness of composite laminates on interface ply orientations and delamination growth direction
2004
A critical review has been performed of the published experimental research concerning delamination onset and growth in composite laminate interfaces of different lay-ups under single-mode loadings. It was found that, depending on the loading mode and interface lay-up, the traditional fracture toughness characterization by unidirectionally reinforced composite tests can lead to marked under- or overestimation of material resistance to crack growth. Empirical models of fracture toughness as a function of delamination front orientation with respect to reinforcement directions of the adjacent laminae have been validated and their applicability range established.
Local Buckling of reinforcing steel bars in RC members under compression forces
2018
Buckling of longitudinal bars is a brittle failure mechanism, often recorded in reinforced concrete (RC) structures after an earthquake. Studies in the literature highlights that it often occurs when steel is in the post elastic range, by inducing a modification of the engineered stress-strain law of steel in compression. A proper evaluation of this effect is of fundamental importance for correctly evaluating capacity and ductility of structures. Significant errors can be obtained in terms of ultimate bending moment and curvature ductility of an RC section if these effects are not accounted, as well as incorrect evaluations are achieved by non-linear static analyses. This paper presents a n…
On the use of Deep Reinforcement Learning for Visual Tracking: a Survey
2021
This paper aims at highlighting cutting-edge research results in the field of visual tracking by deep reinforcement learning. Deep reinforcement learning (DRL) is an emerging area combining recent progress in deep and reinforcement learning. It is showing interesting results in the computer vision field and, recently, it has been applied to the visual tracking problem yielding to the rapid development of novel tracking strategies. After providing an introduction to reinforcement learning, this paper compares recent visual tracking approaches based on deep reinforcement learning. Analysis of the state-of-the-art suggests that reinforcement learning allows modeling varying parts of the tracki…
Interpretable Option Discovery Using Deep Q-Learning and Variational Autoencoders
2021
Deep Reinforcement Learning (RL) is unquestionably a robust framework to train autonomous agents in a wide variety of disciplines. However, traditional deep and shallow model-free RL algorithms suffer from low sample efficiency and inadequate generalization for sparse state spaces. The options framework with temporal abstractions [18] is perhaps the most promising method to solve these problems, but it still has noticeable shortcomings. It only guarantees local convergence, and it is challenging to automate initiation and termination conditions, which in practice are commonly hand-crafted.
Shear design of high strength concrete beams in MRFs
2019
This paper presents the criteria for the shear design of high strength concrete (HSC) beams in moment resisting frames (MRFs). The formulation of an analytical model is provided for the case of beams with longitudinal reinforcement in the presence of transverse stirrups. The model is of additive type, in the meaning that the shear resistance of the beam is evaluated as the sum of several contributions. In particular, the contribution of concrete, longitudinal rebars, and transversal reinforcement are taken into account. Furthermore, for assessing the concrete contribution, a classical approach is followed, according to which two effects arise in the shear mechanism: the arc and the beam eff…