scholarly journals Accelerating Learning of Route Choices With C2I: A Preliminary Investigation

2020 ◽  
Author(s):  
Guilherme Santos ◽  
Ana Bazzan
Keyword(s):  
Reinforcement Learning ◽  
Iterative Method ◽  
Learning Process ◽  
Preliminary Investigation ◽  
Experimental Results ◽  
Choice Problem ◽  
The Road ◽  
Road Users ◽  
Infrastructure Network ◽  
Effectiveness And Efficiency

How to choose a route that takes you from A to B? This is an issue that is turning more and more important in modern societies. One way to address this agenda is through the use of communication between the infrastructure (network), and the demand (vehicles). In this paper, we use car-to-infrastructure (C2I) communication to investigate whether the road users (agents) can accelerate their learning process regarding route choice problem, via reinforcement learning (RL). We employ a microscopic simulator in order to compare our method with two others: RL without communication and an iterative method. Experimental results show that our method outperforms both methods in terms of effectiveness and efficiency.

scholarly journals High-Accuracy Tracking Using Ultrawideband Signals for Enhanced Safety of Cyclists

2017 ◽  
Vol 2017 ◽  
pp. 1-13 ◽  
Author(s):  
Davide Dardari ◽  
Nicoló Decarli ◽  
Anna Guerra ◽  
Ashraf Al-Rimawi ◽  
Víctor Marín Puchades ◽  
...  
Keyword(s):  
Risk Assessment ◽  
Real Time ◽  
High Accuracy ◽  
Experimental Results ◽  
Localization Error ◽  
Localization System ◽  
Dynamic Conditions ◽  
The Road ◽  
Road Users

In this paper, an ultrawideband localization system to improve the cyclists’ safety is presented. The architectural solutions proposed consist of tags placed on bikes, whose positions have to be estimated, and anchors, acting as reference nodes, located at intersections and/or on vehicles. The peculiarities of the localization system in terms of accuracy and cost enable its adoption with enhanced risk assessment units situated on the infrastructure/vehicle, depending on the architecture chosen, as well as real-time warning to the road users. Experimental results reveal that the localization error, in both static and dynamic conditions, is below 50 cm in most of the cases.

scholarly journals Using Car to Infrastructure Communication to Accelerate Learning in Route Choice

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Guilherme D. dos Santos ◽  
Ana L. C. Bazzan ◽  
Arthur Prochnow Baumgardt
Keyword(s):  
Autonomous Vehicles ◽  
Metropolitan Areas ◽  
Route Choice ◽  
Learning Processes ◽  
Traffic Situation ◽  
The Road ◽  
Road Users ◽  
Learning Tasks ◽  
Infrastructure Network ◽  
Near Future

The task of choosing a route to move from A to B is not trivial, as road networks in metropolitan areas tend to be over crowded. It is important to adapt on the fly to the traffic situation. One way to help road users (driver or autonomous vehicles for that matter) is by using modern communication technologies.In particular, there are reasons to believe that the use of communication between the infrastructure (network), and the demand (vehicles) will be a reality in the near future. In this paper, we use car-to-infrastructure (C2I) communication to investigate whether the road users can accelerate their learning processes regarding route choice by using reinforcement learning (RL). The kernel of our method is a two way communication, where road users communicate their rewards to the infrastructure, which, in turn, aggregate this information locally and pass it to other users, in order to accelerate their learning tasks. We employ a microscopic simulator in order to compare this method with two others (one based on RL without communication and a classical iterative method for traffic assignment). Experimental results using a grid and a simplification of a real-world network show that our method outperforms both.

scholarly journals Deep Reinforcement Learning via Past-Success Directed Exploration

2019 ◽  
Vol 33 ◽  
pp. 9979-9980
Author(s):  
Xiaoming Liu ◽  
Zhixiong Xu ◽  
Lei Cao ◽  
Xiliang Chen ◽  
Kai Kang
Keyword(s):  
Online Learning ◽  
Adaptive Control ◽  
Reinforcement Learning ◽  
Learning Process ◽  
Control Method ◽  
Learning Algorithms ◽  
Action Selection ◽  
Experimental Results ◽  
Continuous Control ◽  
Exploration And Exploitation

The balance between exploration and exploitation has always been a core challenge in reinforcement learning. This paper proposes “past-success exploration strategy combined with Softmax action selection”(PSE-Softmax) as an adaptive control method for taking advantage of the characteristics of the online learning process of the agent to adapt exploration parameters dynamically. The proposed strategy is tested on OpenAI Gym with discrete and continuous control tasks, and the experimental results show that PSE-Softmax strategy delivers better performance than deep reinforcement learning algorithms with basic exploration strategies.

Analyzing the Goal-Finding Process of Human Learning With the Reflection Subtask

2018 ◽  
pp. 442-459
Author(s):  
Tomohiro Yamaguchi ◽  
Yuki Tamai ◽  
Keiki Takadama
Keyword(s):  
Reinforcement Learning ◽  
Learning Process ◽  
Human Learning ◽  
Learning Processes ◽  
Experimental Results ◽  
New Method ◽  
Main Task ◽  
Continuous Learning ◽  
Learning Framework ◽  
Environmental Mastery

This chapter reports the authors' experimental results on analyzing the human goal-finding process in continuous learning. The objective of this research is to make clear the mechanism of continuous learning. To fill in the missing piece of reinforcement learning framework for the learning robot, the authors focus on two human mental learning processes, awareness as pre-learning process and reflection as post-learning process. To observe mental learning processes of a human, the authors propose a new method for visualizing them by the reflection subtask for human to be aware of the goal-finding process in continuous learning with invisible mazes. The two-layered task is introduced. The first layer is the main task of continuous learning designing the environmental mastery task to accomplish the goal for any environment. The second layer is the reflection subtask to make clear the goal-finding process in continuous learning. The reflection cost is evaluated to analyze it.

URGENSI EVALUASI PEMBELAJARAN DALAM PENDIDIKAN

2018 ◽  
Vol 2 (1) ◽  
pp. 65-69
Author(s):  
Moh Fachri
Keyword(s):  
Learning Outcomes ◽  
Learning Process ◽  
Learning Objectives ◽  
Validity And Reliability ◽  
Evaluation Tool ◽  
Professional Educators ◽  
Institutional Approach ◽  
Learning Evaluation ◽  
Effectiveness And Efficiency ◽  
Evaluation Of Learning

Learning process is the most important part in education as an effort to mature learners, The success of the learning process becomes a benchmark achievement of learning objectives. To know the achievement of the success of learning objectives, it must be done evaluation / assessment. In particular the purpose of evaluation to determine the progress of learning outcomes of learners after following the learning, as well as to determine the level of effectiveness and efficiency of methods, strategies that teachers use in learning. Evaluation of learning has an important and strategic meaning in education, because the learning process becomes meaningful, as well as its evaluation results can be used as a basis to determine the next step, for teachers, principals, institutions, parents, and government. The importance of learning evaluation can be seen from the approach of the learning process, the characteristics of professional educators, and the institutional approach, but it can also be seen from its purpose, function and principles and the validity and reliability of its evaluation tool.

An Implicit to Explicit FEA Solving of Tire F&M with Detailed Tread Blocks

2012 ◽  
Vol 40 (2) ◽  
pp. 83-107 ◽  
Author(s):  
Zhao Li ◽  
Ziran R. Li ◽  
Yuanming M. Xia
Keyword(s):  
Test System ◽  
Mapping Method ◽  
Friction Model ◽  
Experimental Results ◽  
Slip Angle ◽  
Slip Ratio ◽  
The Road ◽  
Numerical Noise ◽  
Nonlinear Stress ◽  
Solving Strategy

ABSTRACT A detailed tire-rolling model (185/75R14), using the implicit to explicit FEA solving strategy, was constructed to provide a reliable, dynamic simulation with several modeling features, including mesh, material modeling, and a solving strategy that could contribute to the consideration of the serious numerical noises. High-quality hexahedral meshes of tread blocks were obtained with a combined mapping method. The actual rubber distributing and nonlinear, stress-strain relationship of the rubber and bilinear elastic reinforcement were modeled for realism. In addition, a tread-rubber friction model obtained from the Laboratory Abrasion and Skid Tester (LAT 100) was applied to simulate the interaction of the tire with the road. The force and moment (F&) behaviors of tire cornering when subjected to a slip-angle sweep of −10 to 10° were studied with that model. To demonstrate the efficiency of the proposed simulation, the computed F&M were compared with experimental results from an MTS Flat-Trac Tire Test System. The computed cornering F&M agreed well with the experimental results, so the footprint shape and contact pressure distribution of several cornering conditions were investigated. Furthermore, the longitudinal forces in response to braking/driving torque application in a slip-ratio range of −100% to 100% were computed. The proposed FEA solution confines the numerical noise within a smaller range and can serve as a valid tool in tire design.

scholarly journals Modeling of Emergency Supply Scheduling Problem Based on Reliability and Its Solution Algorithm under Variable Road Network after Sudden-Onset Disasters

Complexity ◽  
2020 ◽  
Vol 2020 ◽  
pp. 1-15 ◽  
Author(s):  
Tinggui Chen ◽  
Shiwen Wu ◽  
Jianjun Yang ◽  
Guodong Cong ◽  
Gongfa Li
Keyword(s):  
Road Network ◽  
Road Networks ◽  
Sudden Onset ◽  
The Road ◽  
Bee Colony ◽  
Proposed Model ◽  
Effectiveness And Efficiency ◽  
In Transit ◽  
Emergency Supply

It is common that many roads in disaster areas are damaged and obstructed after sudden-onset disasters. The phenomenon often comes with escalated traffic deterioration that raises the time and cost of emergency supply scheduling. Fortunately, repairing road network will shorten the time of in-transit distribution. In this paper, according to the characteristics of emergency supplies distribution, an emergency supply scheduling model based on multiple warehouses and stricken locations is constructed to deal with the failure of part of road networks in the early postdisaster phase. The detailed process is as follows. When part of the road networks fail, we firstly determine whether to repair the damaged road networks, and then a model of reliable emergency supply scheduling based on bi-level programming is proposed. Subsequently, an improved artificial bee colony algorithm is presented to solve the problem mentioned above. Finally, through a case study, the effectiveness and efficiency of the proposed model and algorithm are verified.

scholarly journals Public Roads as Places of Interspecies Conflict: A Study of Horse-Human Interactions on UK Roads and Impacts on Equine Exercise

Animals ◽  
2021 ◽  
Vol 11 (4) ◽  
pp. 1072
Author(s):  
Danica Pollard ◽  
Tamzin Furtado
Keyword(s):  
Near Miss ◽  
Online Questionnaire ◽  
Near Misses ◽  
Logistic Regression Models ◽  
Emotional Management ◽  
The Road ◽  
Human Interactions ◽  
Road Users ◽  
Conflict Risk ◽  
The Individual

Real or perceived traffic risk is a significant barrier to walking and cycling. To understand whether similar barriers influence equestrians, this study obtained exercise behaviours, road use and experiences of road-related incidents from UK equestrians (n = 6390) via an online questionnaire. Multivariable logistic regression models were used to identify factors associated with road use and experiencing a near-miss or injury-causing incident in the previous year. Content analysis identified themes around equestrians’ decisions not to use roads. Our results show that most equestrians (84%) use roads at least once weekly, and in the previous year, 67.7% had a near-miss and 6.1% an injury-causing incident. Road use differs regionally, with exercise type and off-road route availability. Road-using equestrians covered greater daily distances and were younger. However, younger equestrians were at higher risk of near-misses. Respondents’ decisions not to use roads were based on individualised risk assessments arising from: the road itself, perceptions of other road users, the individual horse and the handler’s own emotional management. Roads were perceived as extremely dangerous places with potentially high conflict risk. Injury-causing incidents were associated with increasing road-use anxiety or ceasing to use roads, the proximity of off-road routes, having a near-miss and type of road use. Targeted road-safety campaigns and improved off-road access would create safer equestrian spaces.

scholarly journals Goal-driven active learning

2021 ◽  
Vol 35 (2) ◽  
Author(s):  
Nicolas Bougie ◽  
Ryutaro Ichise
Keyword(s):  
Decision Making ◽  
Reinforcement Learning ◽  
Learning Process ◽  
Real World ◽  
Imitation Learning ◽  
Learning Approaches ◽  
Wide Range ◽  
Fixed Set ◽  
Complex Decision Making ◽  
Complex Decision

AbstractDeep reinforcement learning methods have achieved significant successes in complex decision-making problems. In fact, they traditionally rely on well-designed extrinsic rewards, which limits their applicability to many real-world tasks where rewards are naturally sparse. While cloning behaviors provided by an expert is a promising approach to the exploration problem, learning from a fixed set of demonstrations may be impracticable due to lack of state coverage or distribution mismatch—when the learner’s goal deviates from the demonstrated behaviors. Besides, we are interested in learning how to reach a wide range of goals from the same set of demonstrations. In this work we propose a novel goal-conditioned method that leverages very small sets of goal-driven demonstrations to massively accelerate the learning process. Crucially, we introduce the concept of active goal-driven demonstrations to query the demonstrator only in hard-to-learn and uncertain regions of the state space. We further present a strategy for prioritizing sampling of goals where the disagreement between the expert and the policy is maximized. We evaluate our method on a variety of benchmark environments from the Mujoco domain. Experimental results show that our method outperforms prior imitation learning approaches in most of the tasks in terms of exploration efficiency and average scores.

Sign in / Sign up

Export Citation Format

Share Document