Search results for: reinforcement

Authors: Lei Zhang, Tingting Xu, Jun He, Zhenyu Yan, Roger Brooks

Abstract:

Email marketing is one of the most important segments of online marketing. Email content is vital to customers. Different customers may have different familiarity with a product, so a successful marketing strategy must personalize email content based on individual customers’ product affinity. In this study, we build our personalized email marketing strategy with three types of emails: nurture, promotion, and conversion. Each type of emails has a different influence on customers. We investigate this difference by analyzing customers’ open rates, click rates and opt-out rates. Feature importance from response models is also analyzed. The goal of the marketing strategy is to improve the click rate on conversion-type emails. To build the personalized strategy, we formulate the problem as a reinforcement learning problem and adopt a Q-learning algorithm with variations. The simulation results show that our model-based strategy outperforms the current marketer’s strategy.

Keywords: Email marketing, email content, reinforcement learning, machine learning, Q-learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 605

Authors: Amit Gil, Helman Stern, Yael Edan

Abstract:

A cognitive collaborative reinforcement learning algorithm (CCRL) that incorporates an advisor into the learning process is developed to improve supervised learning. An autonomous learner is enabled with a self awareness cognitive skill to decide when to solicit instructions from the advisor. The learner can also assess the value of advice, and accept or reject it. The method is evaluated for robotic motion planning using simulation. Tests are conducted for advisors with skill levels from expert to novice. The CCRL algorithm and a combined method integrating its logic with Clouse-s Introspection Approach, outperformed a base-line fully autonomous learner, and demonstrated robust performance when dealing with various advisor skill levels, learning to accept advice received from an expert, while rejecting that of less skilled collaborators. Although the CCRL algorithm is based on RL, it fits other machine learning methods, since advisor-s actions are only added to the outer layer.

Keywords: Robot learning, human-robot collaboration, motion planning, reinforcement learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1669

Authors: Balwinder Singh

Abstract:

The present research is a paper on the characterization of aluminum alloy-6063 hybrid metal matrix composites using three different reinforcement materials (SiC, red mud, and fly ash) through stir casting method. The red mud was used in solid form, and particle size range varies between 103-150 µm. During this investigation, fly ash is received from Guru Nanak Dev Thermal Plant (GNDTP), Bathinda. The study has been done by using Taguchi’s L9 orthogonal array by taking fraction wt.% (SiC 5%, 7.5%, and 10% and Red Mud and Fly Ash 2%, 4%, and 6%) as input parameters with their respective levels. The study of the mechanical properties (tensile strength, impact strength, and microhardness) has been done by using Analysis of Variance (ANOVA) with the help of MINITAB 17 software. It is revealed that silicon carbide is the most significant parameter followed by red mud and fly ash affecting the mechanical properties, respectively. The fractured surface morphology of the composites using Field Emission Scanning Electron Microscope (FESEM) shows that there is a good mixing of reinforcement particles in the matrix. Energy-dispersive X-ray spectroscopy (EDS) was performed to know the presence of the phases of the reinforced material.

Keywords: Reinforcement, silicon carbide, fly ash, red mud.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 682

Authors: Nitin Singh, Meng Ling, Talha Ahmed, Tianxia Zhao, Reinier van de Pol

Abstract:

We propose the use of Reinforcement Learning (RL) as a viable alternative for optimizing supply chain management, particularly in scenarios with stochasticity in product demands. RL’s adaptability to changing conditions and its demonstrated success in diverse fields of sequential decision-making make it a promising candidate for addressing supply chain problems. We investigate the impact of demand fluctuations in a multi-product supply chain system and develop RL agents with learned generalizable policies. We provide experimentation details for training RL agents and a statistical analysis of the results. We study generalization ability of RL agents for different demand uncertainty scenarios and observe superior performance compared to the agents trained with fixed demand curves. The proposed methodology has the potential to lead to cost reduction and increased profit for companies dealing with frequent inventory movement between supply and demand nodes.

Keywords: Inventory Management, Reinforcement Learning, Supply Chain Optimization, Uncertainty.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 246

Authors: E. Karakulak, R. Yamanoğlu, M. Zeren

Abstract:

In the current study, fly ash obtained from a thermal power plant was used as reinforcement in pure magnesium. The composite materials with different fly ash contents were produced with powder metallurgical methods. Powder mixtures were sintered at 540^oC under 30 MPa pressure for 15 minutes in a vacuum assisted hot press. Results showed that increasing ash content continuously increases hardness of the composite. On the other hand, minimum wear damage was obtained at 2 wt. % ash content. Addition of higher level of fly ash results with formation of cracks in the matrix and increases wear damage of the material.

Keywords: Mg composite, fly ash, wear, powder metallurgy.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1335

Authors: S.Shanmugam, V.G. Srisanthi, S.Ramachandran

Abstract:

Reinforced concrete has good durability and excellent structural performance. But there are cases of early deterioration due to a number of factors, one prominent factor being corrosion of steel reinforcement. The process of corrosion sets in due to ingress of moisture, oxygen and other ingredients into the body of concrete, which is unsound, permeable and absorbent. Cracks due to structural and other causes such as creep, shrinkage, etc also allow ingress of moisture and other harmful ingredients and thus accelerate the rate of corrosion. There are several interactive factors both external and internal, which lead to corrosion of reinforcement and ultimately failure of structures. Suitable addition of mineral admixture like silica fume (SF) in concrete improves the strength and durability of concrete due to considerable improvement in the microstructure of concrete composites, especially at the transition zone. Secondary reinforcement in the form of fibre is added to concrete, which provides three dimensional random reinforcement in the entire mass of concrete. Reinforced concrete beams of size 0.1 m X 0.15 m and length 1m have been cast using M 35 grade of concrete. The beams after curing process were subjected to corrosion process by impressing an external Direct Current (Galvanostatic Method) for a period of 15 days under stressed and unstressed conditions. The corroded beams were tested by applying two point loads to determine the ultimate load carrying capacity and cracking pattern and the results of specimens were compared with that of the companion specimens. Gravimetric method is used to quantify corrosion that has occurred.

Keywords: Carbonation, Corrosion, Cracking, Spalling.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 3045

Authors: S. O. Adeosun, L. O. Osoba, O. O. Taiwo

Abstract:

Aluminum hybrid reinforcement technology is a response to the dynamic ever increasing service requirements of such industries as transportation, aerospace, automobile, marine, etc. It is unique in that it offers a platform of almost unending combinations of materials to produce various hybrid composites. This article reviews the studies carried out on various combinations of aluminum hybrid composite and the effects on mechanical, physical and chemical properties. It is observed that the extent of enhancement of these properties of hybrid composites is strongly dependent on the nature of the reinforcement, its hardness, particle size, volume fraction, uniformity of dispersion within the matrix and the method of hybrid production.

Keywords: Aluminum alloy, hybrid composites, properties, reinforcements.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 5043

Authors: Romisaa Ali

Abstract:

The Policy Gradient approach is a subset of the Deep Reinforcement Learning (DRL) combines Deep Neural Networks (DNN) with Reinforcement Learning (RL). This approach finds the optimal policy of robot movement, based on the experience it gains from interaction with its environment. Unlike previous policy gradient algorithms, which were unable to handle the two types of error variance and bias introduced by the DNN model due to over- or underestimation, this algorithm is capable of handling both types of error variance and bias. This article will discuss the state-of-the-art SOTA policy gradient technique, trust region policy optimization (TRPO), by applying this method in various environments compared to another policy gradient method, the Proximal Policy Optimization (PPO), to explain their robust optimization, using this SOTA to gather experience data during various training phases after observing the impact of hyper-parameters on neural network performance.

Keywords: Deep neural networks, deep reinforcement learning, Proximal Policy Optimization, state-of-the-art, trust region policy optimization.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 104

Authors: Simon Thiel, Stavros Dalakakis, Dieter Roller

Abstract:

This paper outlines the development of a learning retrieval agent. Task of this agent is to extract knowledge of the Active Semantic Network in respect to user-requests. Based on a reinforcement learning approach, the agent learns to interpret the user-s intention. Especially, the learning algorithm focuses on the retrieval of complex long distant relations. Increasing its learnt knowledge with every request-result-evaluation sequence, the agent enhances his capability in finding the intended information.

Keywords: Reinforcement learning, learning retrieval agent, search in semantic networks.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1445

Authors: Weifeng Wang, Guoqing Huang, Xianwei Zeng

Abstract:

The nearly 21-year-old Jiujiang Bridge, which is suffering from uneven line shape, constant great downwarping of the main beam and cracking of the box girder, needs reinforcement and cable adjustment. It has undergone cable adjustment for twice with incomplete data. Therefore, the initial internal force state of the Jiujiang Bridge is identified as the key for the cable adjustment project. Based on parameter identification by means of static force test data, this paper suggests determining the initial internal force state of the cable-stayed bridge according to the cable force-displacement relationship parameter identification method. That is, upon measuring the displacement and the change in cable forces for twice, one can identify the parameters concerned by means of optimization. This method is applied to the cable adjustment, replacement and reinforcement project for the Jiujiang Bridge as a guidance for the cable adjustment and reinforcement project of the bridge.

Keywords: Cable-stayed bridge, cable force-displacement, parameter identification, internal force state

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1489

Authors: M. Amirjan

Abstract:

In the present study, the properties of Al-Al2O3 nanocomposite hollow sphere structures were investigated. For this reason, the Al-based nanocomposite hollow spheres with different amounts of nano-alumina reinforcement (0-10wt %) and different ratio of thickness to diameter (t/D: 0.06-0.3) were prepared via a powder metallurgy method. Then, the effect of mentioned parameters was studied on physical and quasi static mechanical properties of their related prepared structures (open/closed cell) such as density, hardness, strength, and energy absorption. It was found that, as the t/D ratio increases the relative density, compressive strength and energy absorption increase. The highest values of strength and energy absorption were obtained from the specimen with 5 wt. % of nanoparticle reinforcement, t/D of 0.3 (t=1 mm, D=400μm) as 22.88 MPa and 13.24 MJ/m3, respectively. The moderate specific strength of prepared composites in the present study showed the good consistency with the properties of others low carbon steel composite with similar structure.

Keywords: Hollow sphere structure foam, nanocomposite, t/D (thickness, diameter), powder metallurgy.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2346

Authors: A. Gherbi, L. Dahmani, A. Boudjemia

Abstract:

This paper presents the Finite Element Method (FEM) for analyzing the failure pattern of rectangular slab with various edge conditions. Non-Linear static analysis is carried out using ANSYS 15 Software. Using SOLID65 solid elements, the compressive crushing of concrete is facilitated using plasticity algorithm, while the concrete cracking in tension zone is accommodated by the nonlinear material model. Smeared reinforcement is used and introduced as a percentage of steel embedded in concrete slab. The behavior of the analyzed concrete slab has been observed in terms of the crack pattern and displacement for various loading and boundary conditions. The finite element results are also compared with the experimental data. One of the other objectives of the present study is to show how similar the crack path found by ANSYS program to those observed for the yield line analysis. The smeared reinforcement method is found to be more practical especially for the layered elements like concrete slabs. The value of this method is that it does not require explicit modeling of the rebar, and thus a much coarser mesh can be defined.

Keywords: ANSYS, cracking pattern, displacements, RC Slab, smeared reinforcement.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1195

Authors: O. Vlcek

Abstract:

The paper deals with current issues in research of advanced methods to increase reliability of traditional timber structural elements. It analyses the issue of strengthening of bent timber beams, such as ceiling beams in old (historical) buildings with additional concrete slab in combination with externally bonded fibre - reinforced polymer. The paper describes experimental testing of composite timber-concrete beam with FRP reinforcement and compares results with FEM analysis.

Keywords: Timber-concrete composite, strengthening, fibre-reinforced polymer, experimental analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2076

Authors: Ross Jeffry, Muhammad N. S. Hadi

Abstract:

High strength concrete (HSC) provides high strength but lower ductility than normal strength concrete. This low ductility limits the benefit of using HSC in building safe structures. On the other hand, when designing reinforced concrete beams, designers have to limit the amount of tensile reinforcement to prevent the brittle failure of concrete. Therefore the full potential of the use of steel reinforcement can not be achieved. This paper presents the idea of confining concrete in the compression zone so that the HSC will be in a state of triaxial compression, which leads to improvements in strength and ductility. Five beams made of HSC were cast and tested. The cross section of the beams was 200×300 mm, with a length of 4 m and a clear span of 3.6 m subjected to four-point loading, with emphasis placed on the midspan deflection. The first beam served as a reference beam. The remaining beams had different tensile reinforcement and the confinement shapes were changed to gauge their effectiveness in improving the strength and ductility of the beams. The compressive strength of the concrete was 85 MPa and the tensile strength of the steel was 500 MPa and for the stirrups and helixes was 250 MPa. Results of testing the five beams proved that placing helixes with different diameters as a variable parameter in the compression zone of reinforced concrete beams improve their strength and ductility.

Keywords: Confinement, ductility, high strength concrete, reinforced concrete beam.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2176

Authors: M. Sedighizadeh, A. Rezazadeh

Abstract:

A self tuning PID control strategy using reinforcement learning is proposed in this paper to deal with the control of wind energy conversion systems (WECS). Actor-Critic learning is used to tune PID parameters in an adaptive way by taking advantage of the model-free and on-line learning properties of reinforcement learning effectively. In order to reduce the demand of storage space and to improve the learning efficiency, a single RBF neural network is used to approximate the policy function of Actor and the value function of Critic simultaneously. The inputs of RBF network are the system error, as well as the first and the second-order differences of error. The Actor can realize the mapping from the system state to PID parameters, while the Critic evaluates the outputs of the Actor and produces TD error. Based on TD error performance index and gradient descent method, the updating rules of RBF kernel function and network weights were given. Simulation results show that the proposed controller is efficient for WECS and it is perfectly adaptable and strongly robust, which is better than that of a conventional PID controller.

Keywords: Wind energy conversion systems, reinforcementlearning; Actor-Critic learning; adaptive PID control; RBF network.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 4842

Authors: S. Srinivasa Moorthy, K. Manonmani

Abstract:

The new, polymer composites consisting of e-glass fiber reinforcement with titanium oxide filler in the double bonded unsaturated polyester resin matrix were made. The glass fiber and titanium oxide reinforcement composites were made in three different fiber lengths (3cm, 5cm, and 7cm), filler content (2 wt%, 4 wt%, and 6 wt%) and fiber content (20 wt%, 40 wt%, and 60 wt%). 27 different compositions were fabricated and a sequence of experiments were carried out to determine tensile strength and impact strength. The vital influencing factors fiber length, fiber content and filler content were chosen as 3 factors in 3 levels of Taguchi’s L₉ orthogonal array. The influences of parameters were determined for tensile strength and impact strength by Analysis of variance (ANOVA) and S/N ratio. Using Artificial Neural Network (ANN) an expert system was devised to predict the properties of hybrid reinforcement GFRP composites. The predict models were experimentally proved with the maximum coincidence.

Keywords: Analysis of variance (ANOVA), Artificial neural network (ANN), Polymer composites, Taguchi’s orthogonal array.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2357

Authors: S. Bahadır Yüksel, Alptuğ Ünal

Abstract:

The Composite Shear Walls (CSW) with steel encased profiles can be used as lateral-load resisting systems for buildings that require considerable large lateral-load capacity. The aim of this work is to propose the experimental work conducted on CSW having L section folded plate (L shape steel made-up sections) as longitudinal reinforcement in boundary regions. The study in this paper present the experimental test conducted on CSW having L section folded plate as longitudinal reinforcement in boundary regions. The tested 1/3 geometric scaled CSW has aspect ratio of 3.2. L-shape structural steel materials with 2L-19x57x7mm dimensions were placed in shear wall boundary zones. The seismic behavior of CSW test specimen was investigated by evaluating and interpreting the hysteresis curves, envelope curves, rigidity and consumed energy graphs of this tested element. In addition to this, the experimental results, deformation and cracking patterns were evaluated, interpreted and suggestions of the design recommendations were proposed.

Keywords: Shear wall, composite shear wall, boundary reinforcement, earthquake resistant structural design, L section.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1766

Authors: S. A. Naeini, H. R. Rahmani, M. Hossein Zade

Abstract:

Many classical bearing capacity theories assume that the natural soil's layers are homogenous for determining the bearing capacity of the soil. But, in many practical projects, we encounter multi-layer soils. Geosynthetic as reinforcement materials have been extensively used in the construction of various structures. In this paper, numerical analysis of the Plate Load Test (PLT) using of ABAQUS software in double-layered soils with different thicknesses of sandy and gravelly layers reinforced with geogrid was considered. The PLT is one of the common filed methods to calculate parameters such as soil bearing capacity, the evaluation of the compressibility and the determination of the Subgrade Reaction module. In fact, the influence of the geogrid layers on the bearing capacity of the layered soils is investigated. Finally, the most appropriate mode for the distance and number of reinforcement layers is determined. Results show that using three layers of geogrid with a distance of 0.3 times the width of the loading plate has the highest efficiency in bearing capacity of double-layer (sand and gravel) soils. Also, the significant increase in bearing capacity between unreinforced and reinforced soil with three layers of geogrid is caused by the condition that the upper layer (gravel) thickness is equal to the loading plate width.

Keywords: Bearing capacity, reinforcement, geogrid, plate load test, layered soils.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 786

Authors: Carlos V. Regueiro, Jose E. Domenech, Roberto Iglesias, Jose L. Correa

Abstract:

In this work a visual and reactive contour following behaviour is learned by reinforcement. With artificial vision the environment is perceived in 3D, and it is possible to avoid obstacles that are invisible to other sensors that are more common in mobile robotics. Reinforcement learning reduces the need for intervention in behaviour design, and simplifies its adjustment to the environment, the robot and the task. In order to facilitate its generalisation to other behaviours and to reduce the role of the designer, we propose a regular image-based codification of states. Even though this is much more difficult, our implementation converges and is robust. Results are presented with a Pioneer 2 AT on a Gazebo 3D simulator.

Keywords: Image-based State Codification, Mobile Robotics, ReinforcementLearning, Visual Behaviour.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1551

Authors: K. S. Sivakumaran, Bo Chen

Abstract:

The floor beams of steel buildings, cold-formed steel floor joists in particular, often require large web openings, which may affect their shear capacities. A cost effective way to mitigate the detrimental effects of such openings is to weld/fasten reinforcements. A difficulty associated with an experimental investigation to establish suitable reinforcement schemes for openings in shear zone is that moment always coexists with the shear, and thus, it is impossible to create pure shear state in experiments, resulting in moment influenced results. However, Finite Element Method (FEM) based analysis can be conveniently used to investigate the pure shear behaviour of webs including webs with reinforced openings. This paper presents the details associated with the finite element analysis of thick/thin-plates (representing the web of hot-rolled steel beam, and the web of a cold-formed steel member) having a large reinforced opening. The study considered simply-supported rectangular plates subjected to in-plane shear loadings until failure (including post-buckling behaviour). The plate was modelled using geometrically non-linear quadrilateral shell elements, and non-linear stress-strain relationship based on experiments. Total Langrangian with large displacement/small strain formulation was used for such analyses. The model also considered the initial geometric imperfections. This study considered three reinforcement schemes, namely, flat, lip, and angle reinforcements. This paper discusses the modelling considerations and presents the results associated with the various reinforcement schemes under consideration.

Keywords: Cold-formed steel, finite element analysis, opening, reinforcement, shear resistance.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1974

Authors: O. Vlcek

Abstract:

The paper deals with current issues in research of advanced methods to increase reliability of traditional timber structural elements. It analyses the issue of strengthening of bent timber beams, such as ceiling beams in old (historical) buildings with additional concrete slab in combination with externally bonded fiber - reinforced polymer. The study evaluates deflection of a selected group of timber beams with concrete slab and additional CFRP reinforcement using different calculating methods and observes differences in results from different calculating methods. An elastic (EN 1995) calculation method and evaluation with FEM analysis software were used.

Keywords: Timber-concrete composite, strengthening, fibre-reinforced polymer, theoretical analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1749

Authors: C. P. Providakis, G. M. Angeli, M. J. Favvata, N. A. Papadopoulos, C. E. Chalioris, C. G. Karayannis

Abstract:

An effort for the detection of damages in the  reinforcement bars of reinforced concrete members using PZTs is  presented. The damage can be the result of excessive elongation of  the steel bar due to steel yielding or due to local steel corrosion. In  both cases the damage is simulated by considering reduced diameter  of the rebar along the damaged part of its length. An integration  approach based on both electromechanical admittance methodology  and guided wave propagation technique is used to evaluate the  artificial damage on the examined longitudinal steel bar. Two  actuator PZTs and a sensor PZT are considered to be bonded on the  examined steel bar. The admittance of the Sensor PZT is calculated  using COMSOL 3.4a. Fast Furrier Transformation for a better  evaluation of the results is employed. An effort for the quantification  of the damage detection using the root mean square deviation  (RMSD) between the healthy condition and damage state of the  sensor PZT is attempted. The numerical value of the RSMD yields a  level for the difference between the healthy and the damaged  admittance computation indicating this way the presence of damage  in the structure. Experimental measurements are also presented.

 

Keywords: Concrete reinforcement, damage detection, electromechanical admittance, experimental measurements, finite element method, guided waves, PZT.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2607

Authors: Izaz Ahmad, Farhatullah, Shahbaz Ali, Farhad Ali, Faiza, Hazrat Junaid, Farhan Zaid

Abstract:

Current trends in remote health monitoring to monetize on the Internet of Things applications have been raised in efficient and interference free communications in Wireless Body Area Network (WBAN) scenario. Co-existence interference in WBANs have aggravates the over-congested radio bands, thereby requiring efficient Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) strategies and improve interference management. Existing solutions utilize simplistic heuristics to approach interference problems. The scope of this research article is to investigate reinforcement learning for efficient interference management under co-existing scenarios with an emphasis on homogenous interferences. The aim of this paper is to suggest a smart CSMA/CA mechanism based on reinforcement learning called QIM-MAC that effectively uses sense slots with minimal interference. Simulation results are analyzed based on scenarios which show that the proposed approach maximized Average Network Throughput and Packet Delivery Ratio and minimized Packet Loss Ratio, Energy Consumption and Average Delay.

Keywords: WBAN, IEEE 802.15.4 Standard, CAP Super-frame, Q-Learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 559

Authors: Soofiyan Atar, Adil Shaikh, Sahil Rajpurkar, Pragnesh Bhalala, Aniket Desai, Irfan Siddavatam

Abstract:

Deep reinforcement learning (deep RL) algorithms leverage the symbolic power of complex controllers by automating it by mapping sensory inputs to low-level actions. Deep RL eliminates the complex robot dynamics with minimal engineering. Deep RL provides high-risk involvement by directly implementing it in real-world scenarios and also high sensitivity towards hyperparameters. Tuning of hyperparameters on a pneumatic quadruped robot becomes very expensive through trial-and-error learning. This paper presents an automated learning control for a pneumatic quadruped robot using sample efficient deep Q learning, enabling minimal tuning and very few trials to learn the neural network. Long training hours may degrade the pneumatic cylinder due to jerk actions originated through stochastic weights. We applied this method to the pneumatic quadruped robot, which resulted in a hopping gait. In our process, we eliminated the use of a simulator and acquired a stable gait. This approach evolves so that the resultant gait matures more sturdy towards any stochastic changes in the environment. We further show that our algorithm performed very well as compared to programmed gait using robot dynamics.

Keywords: model-based reinforcement learning, gait stability, supervised learning, pneumatic quadruped

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 519

Authors: Zasiah Tafheem, Mahadee Hasan Shourav, Zahidul Islam, Saima Islam Tumpa

Abstract:

The paper aims at investigating the effect of number of story on different structural components of reinforced concrete building due to gravity and lateral loading. For the study, three building models having same building plan of three, six and nine stories are analyzed and designed using software package. All the buildings are residential and are located in Dhaka city of Bangladesh. Lateral load including wind and earthquake loading are applied to the building along both longitudinal and transverse direction as per Bangladesh National Building Code (BNBC, 2006). Equivalent static force method is followed for the applied seismic loading. The present study investigates as well as compares mainly total steel requirement in different structural components for those buildings. It has been found that total longitudinal steel requirement for beams at each floor is 48.57% for three storied building, 61.36% for six storied building when the total percentage is taken as 100% in case of nine storied building. For an exterior column, the steel ratio is 2.1%, 3.06%, 4.55% for three, six and nine storied building respectively for the first three floors. In addition, it has been noted that total weight of longitudinal reinforcement of an interior column is 14.02 % for threestoried building and 43.12% for six storied building when the total reinforcement is considered 100% for nine storied building for the first three floors.

Keywords: Equivalent Static Force Method, longitudinal reinforcement, seismic loading, steel ratio.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1724

Authors: Majid Yazdandoust

Abstract:

This study, tries to suggest a design method based on displacement using finite difference numerical modeling in reinforcing soil retaining wall with steel strip. In this case, dynamic loading characteristics such as duration, frequency, peak ground acceleration, geometrical characteristics of reinforced soil structure and type of the site are considered to correct the pseudo static method and finally introduce the pseudo static coefficient as a function of seismic performance level and peak ground acceleration. For this purpose, the influence of dynamic loading characteristics, reinforcement length, height of reinforced system and type of the site are investigated on seismic behavior of reinforcing soil retaining wall with steel strip. Numerical results illustrate that the seismic response of this type of wall is highly dependent to cumulative absolute velocity, maximum acceleration, and height and reinforcement length so that the reinforcement length can be introduced as the main factor in shape of failure. Considering the loading parameters, geometric parameters of the wall and type of the site showed that the used method in this study leads to efficient designs in comparison with other methods, which are usually based on limit-equilibrium concept. The outputs show the over-estimation of equilibrium design methods in comparison with proposed displacement based methods here.

Keywords: Pseudo static coefficient, seismic performance design, numerical modeling, steel strip reinforcement, retaining walls, cumulative absolute velocity, failure shape.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2101

Authors: J. Novak, A. Kohoutkova, V. Kristek, J. Vodicka, M. Sramek

Abstract:

While the use of cast-in-place concrete for an airfield and highway pavement overlay is very common, the application of precast concrete elements is very limited today. The main reasons consist of high production costs and complex structural behavior. Despite that, several precast concrete systems have been developed and tested with the aim to provide a system with rapid construction. The contribution deals with the reinforcement design of a hexagonal element developed for a proposed airfield pavement system. The sub-base course of the system is composed of compacted recycled concrete aggregates and fiber reinforced concrete with recycled aggregates place on top of it. The selected element belongs to a group of precast concrete elements which are being considered for the construction of a surface course. Both high costs of full-scale experiments and the need to investigate various elements force to simulate their behavior in a numerical analysis software by using finite element method instead of performing expensive experiments. The simulation of the selected element was conducted on a nonlinear model in order to obtain such results which could fully compensate results from experiments. The main objective was to design reinforcement of the precast concrete element subject to quasi-static loading from airplanes with respect to geometrical imperfections, manufacturing imperfections, tensile stress in reinforcement, compressive stress in concrete and crack width. The obtained findings demonstrate that the position and the presence of imperfection in a pavement highly affect the stress distribution in the precast concrete element. The precast concrete element should be heavily reinforced to fulfill all the demands. Using under-reinforced concrete elements would lead to the formation of wide cracks and cracks permanently open.

Keywords: Imperfection, numerical simulation, pavement, precast concrete element, reinforcement design, stress analysis.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 688

Authors: S. I. Durowaye, O. P. Gbenebor, B. O. Bolasodun, I. O. Rufai, V. O. Durowaye

Abstract:

Particles are the most common and cheapest reinforcement producing discontinuous reinforced composites with isotropic properties. Conventional fabrication methods can be used to produce a wide range of product forms, making them relatively inexpensive. Optimising composite development must include consideration of all the fundamental aspect of particles including their size, shape, volume fraction, distribution and mechanical properties. Research has shown that the challenges of low fracture toughness, poor crack growth resistance and low thermal stability can be overcome by reinforcement with particles. The unique properties exhibited by micro particles reinforced ceramic composites have made them to be highly attractive in a vast array of applications.

Keywords: Ceramic composites, Mechanical properties, Microparticles, Thermal stability.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1970

Authors: DucKien Thai, Seung EockKim

Abstract:

A numerical analysis of a reinforced concrete (RC) wall under missile impact loading is presented in this study. The model created by Technical Research Center of Finland was used. The commercial finite element code, LS-DYNA was used to analyze. The structural components of the reinforced concrete wall, missile and their contacts are fully modeled. The material nonlinearity with strain rate effects considering damage and failure is included in the analysis. The results of analysis were verified with other research results. The case-studies with different reinforcement ratios were conducted to investigate the influence of reinforcement on the punching behavior of walls under missile impact.

Keywords: Missile Impact, Reinforced Concrete Walls, LSDYNA, Dynamic Analysis, Punching Behavior.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 2506

Authors: Hadi Firouzi, Majid Nili Ahmadabadi, Babak N. Araabi

Abstract:

Conceptualization strengthens intelligent systems in generalization skill, effective knowledge representation, real-time inference, and managing uncertain and indefinite situations in addition to facilitating knowledge communication for learning agents situated in real world. Concept learning introduces a way of abstraction by which the continuous state is formed as entities called concepts which are connected to the action space and thus, they illustrate somehow the complex action space. Of computational concept learning approaches, action-based conceptualization is favored because of its simplicity and mirror neuron foundations in neuroscience. In this paper, a new biologically inspired concept learning approach based on the probabilistic framework is proposed. This approach exploits and extends the mirror neuron-s role in conceptualization for a reinforcement learning agent in nondeterministic environments. In the proposed method, instead of building a huge numerical knowledge, the concepts are learnt gradually from rewards through interaction with the environment. Moreover the probabilistic formation of the concepts is employed to deal with uncertain and dynamic nature of real problems in addition to the ability of generalization. These characteristics as a whole distinguish the proposed learning algorithm from both a pure classification algorithm and typical reinforcement learning. Simulation results show advantages of the proposed framework in terms of convergence speed as well as generalization and asymptotic behavior because of utilizing both success and failures attempts through received rewards. Experimental results, on the other hand, show the applicability and effectiveness of the proposed method in continuous and noisy environments for a real robotic task such as maze as well as the benefits of implementing an incremental learning scenario in artificial agents.

Keywords: Concept learning, probabilistic decision making, reinforcement learning.

Procedia APA BibTeX Chicago EndNote Harvard JSON MLA RIS XML ISO 690 PDF Downloads 1480