Object-oriented Software Architecture Research Articles

Lamarckian Platform: Pushing the Boundaries of Evolutionary Reinforcement Learning Toward Asynchronous Commercial Games

Despite the emerging progress of integrating evolutionary computation into reinforcement learning, the absence of a high-performance platform endowing composability and massive parallelism causes non-trivial difficulties for research and applications related to asynchronous commercial games. Here we introduce Lamarckian <xref ref-type="fn" rid="fn1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">¹</xref> <fn id="fn1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><label>¹</label> The code and demonstrational setup of Lamarckian are publicly available at <uri>https://github. com/lamarckian/lamarckian</uri>. </fn> – an open-source platform featuring support for evolutionary reinforcement learning scalable to distributed computing resources. To improve the training speed and data efficiency, Lamarckian adopts optimized communication methods and an asynchronous evolutionary reinforcement learning workflow. To meet the demand for an asynchronous interface by commercial games and various methods, Lamarckian tailors an asynchronous Markov Decision Process interface and designs an object-oriented software architecture with decoupled modules. In comparison with the state-of-the-art RLlib, we empirically demonstrate the unique advantages of Lamarckian on benchmark tests with up to 6000 CPU cores: i) both the sampling efficiency and training speed are doubled when running PPO on Google football game; ii) the training speed is 13 times faster when running PBT+PPO on Pong game. Moreover, we also present two use cases: i) how Lamarckian is applied to generating behavior-diverse game AI; ii) how Lamarckian is applied to game balancing tests for an asynchronous commercial game.

DOOML: A New Database &amp; Object-Oriented Modeling Language for Database-Driven Web Application Design and Development

A database driven web application is a very common software solution to rising business problems. Modeling the database and the software architecture can be challenging, hence there not being one combined modeling language for database and software architecture, specifically suited for web application development. In this paper we present Database object-oriented Modeling Language (DooML) and its primary Archetype Diagram: a notation for specifying the design of a database schema and corresponding object-oriented software architecture. It combines the syntax for drawing Entity Relationship Diagrams, the Relational Model and Universal Modeling Language Class Diagrams as well to create a mixed diagram, stating database design as well as software design specifications. By default, DooML ensures that the approach of advanced web application development is model-driven and both database-oriented as well as object-oriented.

SHyFTOO, an object-oriented Monte Carlo simulation library for the modeling of Stochastic Hybrid Fault Tree Automaton

Dependability assessment is a crucial activity to ensure the correct operation of complex systems. The output of dependability assessment activities include the quantification of reliability, availability, maintenance and safety related metrics. These metrics can assist in the identification of the system weak points or in the conception of mitigation strategies to increase the system dependability level. The development of advanced computer-aided methodologies to support dependability assessment activities is essential to automate and reduce the efforts implied by this process and similarly, the development of accurate dependability assessment methods is very important to increase the quality of the results. In this context, it is possible to identify different contributions that improve the dependability assessment through general-purpose modeling methodologies. However, existing solutions are ad-hoc applications specified with low-level stochastic formalisms and this complicates their adoption in the industry. Accordingly, this paper presents Stochastic Hybrid Fault Tree Automaton (SHyFTA) based simulation algorithm that allows the accurate dependability analysis of repairable multi-state systems. SHyFTA integrates the stochastic and deterministic operation of the system under study as well as their interactions. The algorithm is formalized through an object-oriented software architecture, which is developed as a software library for the modeling and simulation of repairable SHyFTA models. Following the proposed architecture, a Matlab® implementation of this library, SHyFTOO, has been developed and validated with a thorough test campaign. In order to provide a guideline to the end-users and show the potential of the SHyFTOO library, the case study of a feed-water pumping system is implemented in detail and it is used to evaluate different preventive maintenance policies. The SHyFTOO library can open the way to further investigations that address the interactions between the failure behavior and the functional operation of a system and their combined effect on system dependability.

Theoretical extension of universal forward and backward Monte Carlo radiative transfer modeling for passive and active polarization observation simulations

This paper extends the theoretical framework of universal forward and backward Monte Carlo radiative transfer modeling for passive and active polarization observation simulations. The framework is built upon newly derived forward and backward vector scattering order-dependent integral radiative transfer equations (IRTEs). The unified mathematical formalism not only establishes the design of object-oriented software architecture, but also speeds up software development via the reuse of developed scalar code. The polarization simulations are implemented with the addition of the Stokes vector and the Mueller matrix weight tracing schemes. To improve numerical performance order by order, the extension implementation of hybrid scattering order-dependent variance reduction techniques are given detailedly, with a focus on the formulation of scattering phase matrix truncation on the basis of remodeling vector IRTEs. To enrich polarization simulation features, not only scattering operators but also source vector function and detector response matrix function are formulated in unified mathematical forms for the hierarchies organization of various atmosphere, surface, source and detector classes.The framework was fully implemented in Multiple-Scaling-based Cloudy Atmospheric Radiative Transfer (MSCART) model. For passive simulation, it not only can handle all one- and three-dimensional (1D and 3D) test cases in the Phase A and B of the International Polarized Radiative Transfer (IPRT) intercomparison project by simulating area-averaged polarized radiance using forward and backward algorithm, but also can simulate polarized radiance in a specified direction at a specified position using backward algorithm; For active simulation, lidar backscatter range-resolved signals can be simulated using forward algorithm. The source package is freely available for research purpose from the corresponding author, with online documentation from https://intersharp.gitlab.io/mscart-docs or http://mscart.nuist.edu.cn.

Redundancy modelling and resolution for robotic mobile manipulators: a general approach

In this work the topic of kinematic redundancy modelling and resolution for robotic mobile manipulators is considered. A set of redundancy parameters is introduced to define a general inverse kinematic procedure for mobile manipulators. Then, redundancy is treated as a non-linear optimization problem with the purpose of finding robot configurations that maximize the designed metric measures. Some strategies to design the optimization objective function are introduced in order to achieve desirable redundant behaviours, such as obstacles avoidance, mobile base motions reductions and dexterity optimization. Moreover, the robot controller has been developed following an object-oriented software architecture principle that allows to keep it general and robot independent. As a prove of reliability and generality of our approach, the same controller has been used to control several different mobile manipulators in a simulation environment, as well as a real KUKA youBot robot.

Evaluating Dependency based Package-level Metrics for Multi-objective Maintenance Tasks

Role of packages in organization and maintenance of software systems has acquired vital importance in recent research of software quality. With an advancement in modularization approaches of object oriented software, packages are widely considered as re-usable and maintainable entities of objectoriented software architectures, specially to avoid complicated dependencies and insure software design of well identified services. In this context, recently research study of H. Abdeen on automatic optimization of package dependencies provide composite frame of metrics for package quality and overall source code modularization. There is an opportunity to conduct comprehensive empirical analysis over proposed metrics for assessing their usefulness and application for fault-prediction, design flaw detection, identification of source code anomalies and architectural erosion. In this paper, we examine impact of these dependency optimization based metrics in wide spectrum of software quality for single package and entire software modularization. Our experimental work is conducted over open source software systems through statistical methodology based on cross validation fault-prediction and correlation.We conclude with empirical evidence that dependency based package modularization metrics provide more accurate view for predicting fault-prone packages and improvement of overall software structure. Thus, application of these metrics can help the developers and software practitioners to insure proactive management of the source code dependencies and avoid design flaws during software development.

OpenGL Based Intuitive Interaction Technology for 3D Graphical System by 2D Devices

We address the problem of how to intuitive interacts with 3D graphic system by 2D devices, and present a general interactive framework for developing human computer interaction interface for graphical system. We consider the subclass of interaction situations in which 2D devices such as mouse and windows intuitively control scene observation, select and manipulate 3D object. Based above, an object-oriented software architecture is presented to construct the virtual world, OpenGL is used to render and control 3D graphic scene and Visual C++ 6.0 platform and Windows system are combined to build the underlying structure of the software. First, compare to OpenGL select mechanism, a novel select mechanism is presented. Second, with common 2D devices, we introduced an intuitive and precise mechanism for scene observation and object manipulation. Experimental examples prove that our interactive technology is feasible and practical.

Population-Based Simulation for Public Health: Generic Software Infrastructure and Its Application to Osteoporosis

Policy-making in public health has great socio-economical consequences and must be done using the best available knowledge on the possible options. These processes are often too complex to be evaluated through analytical methods, such that computer simulations are often the best way to produce quantitative evaluations of their performances. For that purpose, we are proposing a complete software infrastructure for the simulation of public health processes. This software stack includes a generic population-based simulator called SynCHroNous Agent- and Population-based Simulator, which has a modern object-oriented software architecture, and is completely configured through eXtensible Markup Language files. These configuration files can themselves be produced by a graphical user interface that allows modeling of public health simulation by nonprogrammers. This software infrastructure has been illustrated with the real-life case study of osteoporosis prevention in adult women populations. This example, which is of great interest for Quebec health decision makers, provides insightful results for comparing several prevention strategies on a realistic population.

A dependency constraint language to manage object‐oriented software architectures

AbstractThis paper presents a domain‐specific dependency constraint language that allows software architects to restrict the spectrum of structural dependencies, which can be established in object‐oriented systems. The ultimate goal is to provide architects with means to define acceptable and unacceptable dependencies according to the planned architecture of their systems. Once defined, such restrictions are statically enforced by a tool, thus avoiding silent erosions in the architecture. The paper also presents results from applying the proposed approach to different versions of a real‐world human resource management system. Copyright © 2009 John Wiley &amp; Sons, Ltd.

Reliability-based optimal design software for earthquake engineering applications

This paper describes a functional tool for engineers to make rational design decisions by balancing cost and safety. Focus is on seismic design, in which nonlinear structural response must be considered. For this purpose, we implement and apply a state-of-the-art algorithm for reliability-based design optimization. The work extends the OpenSees software, which is rapidly gaining users in the earthquake engineering community. Consequently, design optimization with sophisticated nonlinear finite element models of real structures is possible. An object-oriented software architecture is employed that focuses on maintainability and extensibility of the software. This approach also offers flexibility in the choice of optimization and reliability methods for each specific problem, supported by the decoupled nature of the optimization algorithm. Our work utilizes and extends the existing tools for structural reliability analysis in OpenSees. In particular, we employ response sensitivities that are computed within the finite element code by direct differentiation. The implementation is tested through case studies with nonlinear structural response. Discontinuous response gradients are overcome by use of fibre cross sections and smoothed material models. The numerical examples include the seismic design optimization of a six-storey, three-bay, reinforced concrete building. Key words: reliability-based design optimization, nonlinear finite elements, earthquake engineering, object-oriented software development, OpenSees.

A framework for the design and automated implementation of communication aspects in multi-agent systems

This paper presents REFCON, a framework for the automated development of Agent Communication Contexts (ACCs) in multi-agent systems (MASs). ACCs are intended to capture the interaction requirements of a MAS. A formal specification framework is first presented, aimed at modelling an ACC as a set of rules for filtering and filling messages, based on their contents, and the names and roles of the exchanging agents. A XML-based specification language is then introduced, which encodes the specification formalism for the sake of its computer processing. Finally, an object-oriented software architecture capable of supporting ACC-based MAS development is presented. REFCON key characteristic is that it allows a seamless integration of ACC support (even) into an existing MAS, at run-time, independently of the agent platform used for the implementation. This is made possible by a layered software architecture based on computational reflection, a technology that allows transparent evolution and adaptation of existing systems. The REFCON framework is also dynamic, in the two-fold sense that it is capable of both adding new rules and handling multiple contexts, which it can easily switch among, at run-time. The ACC-based design of an example MAS for document sharing is briefly discussed, as a demonstration of the principles put forward.

Implementation and performance of the high level trigger electron and photon selection for the ATLAS experiment at the LHC

The ATLAS experiment at the Large Hadron Collider (LHC) will face the challenge of efficiently selecting interesting candidate events in pp collisions at 14 TeV center of mass energy, while rejecting the enormous number of background events, stemming from an interaction rate of up to 10/sup 9/ Hz. The First Level trigger will reduce this rate to around O(100 kHz). Subsequently, the High Level Trigger (HLT), which is comprised of the Second Level trigger and the Event Filter, will need to further reduce this rate by a factor of O(10/sup 3/). The HLT selection is software based and will be implemented on commercial CPUs, using a common framework built on the standard ATLAS object oriented software architecture. In this paper an overview of the current implementation of the selection for electrons and photons in the HLT is given. The performance of this implementation has been evaluated using Monte Carlo simulations in terms of the efficiency for the signal channels, rate expected for the selection, data access and manipulation times, and algorithm execution times. Besides the efficiency and rate estimates, some physics examples will be discussed, showing that the triggers are well adapted for the physics programme envisaged at LHC. The electron and photon trigger software is also being exercised at the ATLAS 2004 Combined Test Beam, where components from all ATLAS subdetectors are taking data together along the the H8 SPS extraction line at CERN; from these tests a validation of the selection architecture chosen in a real on-line environment is expected.

E-Learning in der Radiologie - praktischer Einsatz des Content-Management-Systems Ilias

Due to the possibility of using different kinds of visualization, e-learning has the advantage of allowing individualized learning. A check should be performed to determine whether the use of the web-based content management system ILIAS simplifies the writing and production of electronic learning modules in radiology. Internet-based e-learning provides access to existing learning modules regardless of time and location, since fast Internet connections are readily available. Web Content Management Systems (WCMS) are suitable platforms for imparting radiology-related information (visual abilities like the recognition of patterns as well as interdisciplinary specialized knowledge). The open source product ILIAS is a free WCMS. It is used by many universities and is accepted by both students and lecturers. Its modular and object-oriented software architecture makes it easy to adapt and enlarge the platform. The employment of e-learning standards such as LOM and SCORM within ILIAS makes it possible to reuse contents, even if the platform has to be changed. ILIAS renders it possible to provide students with texts, images, or files of any other kind within a learning context which is defined by the lecturer. Students can check their acquired knowledge via online testing and receive direct performance feedback. The significant interest that students have shown in ILIAS proves that e-learning can be a useful addition to conventional learning methods.

Object Oriented Software Architecture Design based on UML/Petri Net approach for deadlock prevention of Real Time Systems

The objective of this article is to present an approach based on UML dynamic diagrams and on Petri Net Objects for the deadlock prevention of Real Time Systems. The main idea consists of translating the sequence diagrams which express the initial specifications of the system in a unique Petri Net model which represents the global behaviour of the entire system and which belongs to the special class of Petri Nets called Simple Sequential Process with Resources. After the detection of possible unmarked siphons in the structure of the global model, a deadlock prevention strategy based on the addition of local control places is applied in order to remove structural deadlocks. The augmented Petri Net model is then translated into a set of collaboration and state-transition diagrams which represent the final deadlock free UML specification. Finally, a software architecture free of deadlocks based on an object which contains the global behaviour of the system and which centralises all the interactions between the software objects is presented. The approach is illustrated through an example of Real Time System used at the global coordination level of a Batch System.

A middleware architecture for distributed systems management

This paper presents a middleware solution for global management of any kind of distributed system, such as networks of PCs/workstations, clusters or server farms. Our approach lies in an object-oriented software architecture that models all kind of management information using the common information model (CIM) developed by the Distributed Management Task Force (DMTF). The classes and attributes obtained after the modeling process are mapped to a Lightweight Directory Access Protocol (LDAP) repository. This paper discusses the key features of our middleware that allows that any element (physical, logical, device, user or system) can be managed using a network-oriented and topology-independent approach. A representative example of management domain illustrates the procedures followed to build management applications using our middleware architecture.

Operator controlled, vehicle-based highway crack-sealing machine

This paper presents a new heavy vehicle system developed for the automated sealing of cracks in highway pavement. The machine has been developed based on the authors' experience over the last twelve years building such machines. A special sealant applicator that deploys a pressurised reservoir was developed to allow for high-speed, automated sealing operations. A telescoping long-reach robotic arm was developed, which can position the sealing head mounted to its end in a 3.65 m (12 ft) by 4 m (13 ft) large workspace within 0.4 mm (1/64 in) at a traverse speed up to 1 m/s (3 ft/s). To ensure adequate sealant flow, a special electrically heated hose was developed. A control network of both fast 100 Mbps Ethernet and 10 Mbps Wireless allows for distributed real-time motion control, I/O control and machine vision. A fully object-oriented, multi-threaded and Internet-based software architecture allows reliable and efficient system integration and provides an interactive and responsive user interface for feature-rich and easy imaging, path planning and system control. An operator-assisted crack identification method is used and the path planning method is based on simulated annealing in order to optimally seal a network of cracks.

AN OBJECT ORIENTED MODEL FOR SCHEDULING IN AGILE MANUFACTURING

Agility represents a key factor in industry to handle the continuous market changes. Companies must re-organize their activities to be agile and competitive in such a dynamic environment. In particular, production planning and control tools are very important to optimize the manufacturing process responsiveness to sudden changes in customer demand. In this paper, an attempt has been made to develop an object-oriented software architecture that allows the optimal line organization to be determined once a set of parts to be produced has been ordered. An optimization module represented by a simulated annealing algorithm has been interfaced with an object oriented architecture to build up a framework that can be used to select both the optimal schedule, the buffer capacity and the speed of the conveyors in a very short time. The main characteristic of the proposed system is represented by its extremely high flexibility and reconfigurability with respect to the sudden changes in the customer orders that are reflected on the mix to be produced. A test case is provided to demonstrate the potential of using the proposed software architecture. A discussion on the obtained results is also developed to prove how the system is able to cope with lines characterized by different configurations of available resources.

Patterns in scientific software: an introduction

Patterns are well understood methodology for object-oriented software architecture, especially for business applications. Scientific programmers have generally avoided object-oriented approaches because of their heavy computational overhead, but the benefits of using patterns for scientific problems can outweigh their costs. This article introduces the concept of object oriented software patterns and discusses how they can be applied to scientific software problems. After a brief explanation of what patterns are and why they can be relevant to scientific software, the author explores the application of patterns to dynamic-systems simulation, such as molecular dynamics, and identifies four design patterns that emerge in modeling such systems. To illustrate how to reuse a general pattern for a specific problem, he applies one of the dynamic simulation patterns to the different problem of hydrodynamic chemistry tracers.

Entwicklung eines inversen Bestrahlungsplanungssystems mit linearer Optimierung

Our approach shows that inverse planning for intensity-modulated beams in radiotherapy can be solved efficiently by the mathematical method of linear optimization. The completeness property of this method guarantees that calculated treatment plans fulfill the dose constraints given by the oncologist. Techniques developed by our group can also avoid the possible infeasibility caused by a physically impossible dose distribution. In contrast to other optimization methods, the simplex algorithm used for linear optimization allows a very high optimization speed, i.e., very short planning cycles.These preconditions were integrated in the development of the software system MIPART („Munich Inverse Planning And Radiotherapy Treatment“). The object-oriented software architecture of MIPART achieves a maximum of extensibility and flexibility.Our clinical tests show that MIPART can be easily integrated in the clinical routine in spite of the complexity of data in the field of intensity-modulated radiotherapy. Moreover, qualitatively better treatment plans can be generated than in conventional treatment planning, especially in complicated cases.

Framelets—small and loosely coupled frameworks

article Free Access Share on Framelets—small and loosely coupled frameworks Authors: Wolfgang Pree Univ. of Constance, Constance, Germany Univ. of Constance, Constance, GermanyView Profile , Kai Koskimies Nokia Research Center, Helsinki, Finland Nokia Research Center, Helsinki, FinlandView Profile Authors Info & Claims ACM Computing SurveysVolume 32Issue 1esMarch 2000 pp 6–eshttps://doi.org/10.1145/351936.351942Online:01 March 2000Publication History 10citation1,009DownloadsMetricsTotal Citations10Total Downloads1,009Last 12 Months7Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF