Computer Generated Forces Research Articles

Intelligent agent-based simulation for supporting operational planning in country reconstruction

This paper proposes the development of conceptual models for supporting operational planning in complex scenarios characterised by critical issues such as natural disasters, crisis, asymmetric warfare, etc. In this context, it is fundamental to take into consideration social and psychological aspects; so the author developed a special kind of intelligent agents (IAs) driving computer generated forces (CGF), specific for normalisation and stabilisation phase with particular focus on CIMIC and PSYOPs operations, in order to consider the human aspects as well as the impact of social networks on this context as further development of their IA-CGF. The research provides an overview about conceptual models to be implemented in a simulator for supporting CIMIC and PSYOPs operations planning during stabilisation and normalisation phases in country reconstruction. The author proposes the general architecture of the simulator, focusing on objects attributes and features and verification, validation and accreditation (VV&A) process.

워게임에서 가상군의 자율적 행위를 위한 자동계획 기법

This paper proposes a novel planning method for computer generated forces (CGFs) in war games that plans the behaviors of CGFs according to a given mission and situations. CGFs which are received their missions first plan their tasks for accomplishing the mission and then plan their behaviors for accomplishing each task. After that, they execute their planned behaviors considering the conditions of environments (in other words situations). The tasks and behaviors are hierarchically composed and include start conditions for beginning those and termination conditions for stopping those. CGFs first check whether the start condition of the planned behavior for accomplishing a task is satisfied or not in some degree and perform the behavior if satisfied continuously until the termination condition of the behavior will be met. If the termination condition is satisfied, then they check the start condition of the next planned behavior. This process will be repeated for accomplishing the mission. If the situations of CGFs are different by changing the environments from those of planning time, it may cause the start condition of the planned behavior to be dissatisfied. In this case, CGFs can decide a new behavior using fuzzy rule base. We realized our planning system and tested CGFs with a scenario. Experimental results showed that our system worked well and actively coped with situation changes. It will be possible to make CGFs that can do more autonomous behaviors if we continually develop our method.

Modeling of an airborne torpedo attack and implementation on a simulation environment

Airborne torpedoes are effective weapons that are used against surface and subsurface entities. A carrying platform, which may be a helicopter or plane, drops the weapon with or without a parachute. After a travelling period in the air, the torpedo hits the water and penetrates through it, forming a splash and a cavity. The initial conditions for the movement of the torpedo in the water are produced as a result of velocity at the water entrance and pitch angle and resistance force generated by the water itself. Under the water, the weapon transforms through its internal phases with the aim of a possible target hit. In this work, a certain part of the developed military training simulator is given in order to investigate an airborne torpedo attack scenario. For this purpose, aircraft and torpedo dynamic models, as well as guidance and control laws, are derived; torpedo fall and water entry analyses are also performed. Simulation results are given for each model and/or analysis separately. Modules and libraries for these model simulations are integrated into the component architecture of a commercial computer-generated forces toolkit simulation engine, namely VR-Forces. An original control architecture that is able to manage a wide variety of simulation components with different infrastructures has also been realized.

가상군(Computer-Generated Forces)의 자율지능화 방안 연구

최근 군은 과거 냉전시대와 다른 상황에 능동적으로 대처하고 비용 대비 훈련의 효과를 극대화하기 위한 방안으로 모델링 시뮬레이션 기법에 많은 관심을 보이고 있다. 이 모델링 시뮬레이션을 이용한 훈련이 효과적이기 위해서는 좀 더 사실적인 전장 환경과 전투상황을 모의해야 하는데, 최근에 이를 위해서 컴퓨터상에서 인간 행위와 부대의 집단논리를 표현하는 CGF(Computer Generated Forces)에 대한 기술 개발이 주목받고 있다. 인간 행위를 모델링하는 CGF는 자동화된 병력을 모의함으로써 적군을 모의하거나 미래 전투실험, 새로운 전투개념 개발 등 중요하면서도 다양한 역할을 수행 할 수 있다. 이번 연구에서는 CGF 가상군인의 자율지능화의 방안을 고찰해보고자 한다. 군사시뮬레이션상의 가상군인의 자율지능화의 첫걸음인 과업행위 목록의 적용을 과업행위의 기본요소인 METT+T를 기반으로 그 적용과정을 설명한다. 또한 군사모의 논리는 참 진의 이진값 표현이 외에도 '보유 무기가 충분하면 급속 공격을 빠르게 하라'와 같은 모의 규칙상에서는 군사 전문가에 의해서 결정될 수밖에 없는 '충분하다', '빠르다'와 같은 퍼지 팩트가 존재한다. 이러한 주관성이 가미된 많은 군사모의 논리를 완벽히 표현하기 위해서는 퍼지 추론을 이용하는 것이 효과적인데, 이를 이번 소부대 전투 시뮬레이션에 적용하여 보다 사실적인 모의가 가능하도록 하였다. Modeling and Simulation(M&S) technology gets an attention from various parts such as industry and military. Especially, military uses the technology to cope with a different situation from the one in the Cold War and maximize the effect of training against the cost in the new environment. In order for the training based on M&S technology to be effective, the situations of a battlefield and a combat must be more realistically simulated. For this, a technique development on Computer-Generated Forces(CGF) which represents a unit's simulation logic and a human's simulated behaviors is focused. The CGF simulating a human's behaviors can be used in representing an enemy force, experimenting behaviors in a future war, and developing a new combat idea. This paper describes a methodology to accomplish Computer-Generated Forces' autonomous intelligence. It explains the process of applying a task behavior list based on the METT+T element onto CGFs. On the other hand, in the domain knowledge of military field manual, fuzzy facts such as "fast" and "sufficient" whose real values should be decided by domain experts can be easily found. In order to efficiently implement military simulation logics involved with such subjectivity, using a fuzzy inference methodology can be effective. In this study, a fuzzy inference methodology is also applied.

Theory and techniques of data mining in CGF behavior modeling

For the large quantity of data, rules and models generated in the course of computer generated forces (CGFs) behavior modeling, the common analytical methods are statistical methods based on the tactical rules, tactical doctrine and empirical knowledge. However, from the viewpoint of data mining, we can find many of these analytical methods are also each-and-every different data mining methods. In this paper, we survey the data mining theory and techniques that have appeared in the course of CGF behavior modeling from the viewpoint of data mining. Further, we redefine and reclassify these theories and techniques according to the research requirements of data mining (the knowledge structure of data mining). We aim to help the CGF investigators to learn the key points and the research methods of the data intelligent processing in the CGF behavior modeling, and thus we can provide the references for the colleagues who are engaged in the research and practice of the CGF behavior modeling by this article. We also introduce our ongoing work on the CGF behavior modeling based on data mining.

Assessing critical success factors for military decision support

This paper outlines the application of case-based reasoning and Bayesian belief networks to critical success factor (CSF) assessment for parsimonious military decision making. An important factor for successful military missions is information superiority (IS). However, IS is not solely about minimising information related needs to avoid information overload and the reduction of bandwidth but it is also concerned with creating information related capabilities that are aligned with achieving operational effects and raising operational tempo. Moreover, good military decision making, should take into account the uncertainty inherent in operational situations. Herein, we illustrate the development and evaluation of a smart decision support system (SDSS) that dynamically identifies and assesses CSFs in military scenarios and as such de-clutters the decision making process. The second contribution of this work is an automated configuration of conditional probability tables from hard data generated from simulations of military operational scenarios using a computer generated forces (CGF) synthetic environment.

Representing Uncertainty in Computer-Generated Forces

Representing Uncertainty in Computer-Generated Forces

Intelligent Maneuvering Decision System for Computer Generated Forces Using Predictive Fuzzy Inference System

The purpose of this paper is to develop an intelligent maneuvering decision system (IMDS) for computer generated forces (CGF). The proposed CGF can take actions similar to a human pilot to gain an advantageous status over the enemy target using the IMDS. The IMDS will produce the best control command from the control alternatives for the CGF in an air combat environment. In this paper, a predictive fuzzy inference system (PFIS) is proposed as the IMDS for CGF, which incorporates and mimics human thinking capability and the maximum capacity of CGF. Before PFIS executes the fuzzy inference system (FIS) process, it will generate the control alternatives from CGF’ s decision space, and allow CGF to predict its future posture. This study assumes that CGF can accurately predict an enemy target’s future position, and then PFIS applies the predicted data to generate the best control command. In this paper, the proposed algorithm is verified with two types of fighter flying data that are used as the enemy target’s flying trajectories. The simulation and discussion of the proposed algorithm shows that PFIS will enable CGF to obtain the best status in an air combat environment and the performance of the proposed algorithm will be affected by the CGF’s prediction ability for enemy target.

DEVS for Human Behavior Modeling in CGFs

In recent years, human behavior simulation has been an active field of research, and various approaches have been used. One of the main challenges is to design models that capture more of the subtleties of human nature to reach a higher level of realism. In this paper, we present the premises of a DEVS (Discrete Event System Specification) based framework for the modeling and simulation of human behavior for CGFs (Computer Generated Forces). We introduce a simplified way of specifying models for the simulation of human behavior. In fact, our focus is to give to the modeler the possibility to represent behavior influenced by certain human factors by fine tuning the proposed models or by extending them with new moderator functions. As an example, we present DEVS models of military behavior as defined in the rules of engagement but moderated by human factors such as personality and stress. We hope to obtain eventually more accurate and realistic models.

Using Graphics Processor Units to Accelerate OneSAF: A Case Study in Technology Transition

Ongoing research aims to accelerate the runtime processing speed of the One Semi-Automated Forces (OneSAF) Computer Generated Forces (CGF) simulation by converting and migrating some of the core algorithms from the host central processing unit (CPU) to an onboard auxiliary graphics processor unit (GPU). In this research the GPU chip is regarded as a surrogate stream processor, and appropriate algorithms are designed to map to the GPU architecture. Processing speed gains are realized both through computational capabilities of the GPU as well as through offloading of the host CPU. Technology transfer of this research into the OneSAF baseline is a key requirement of this research. The OneSAF development program focuses on the same issues of scalability and runtime performance that will be directly affected by use of GPUs. As program architects are marshalling conventional approaches for resolving these challenges, the introduction of GPU-based solutions is being realized. This paper examines the challenges, planned approaches, and benchmarked results for using GPUs to accelerate OneSAF simulation.

Automating the Acquisition of Tactical Knowledge for Military Missions

It is widely accepted that acquisition of the knowledge behind military tactics has been one limiting factor in the development of computer generated forces (CGF) for training simulations. This has been addressed by several researchers with varying degrees of success. A system capable of building a knowledge base directly from a dialogue with a subject-matter expert (SME) could significantly reduce the human effort involved in capturing the knowledge and representing it directly in the modeling language. Because of its highly modular and hierarchical nature, the context-based reasoning (CxBR) modeling paradigm lends itself very well to facilitating the knowledge acquisition process for tactical behaviors. This paper describes an investigation into using CxBR as the foundation for a system that creates a (partial) model of tactical behavior through an interactive process with an SME. Through a sequence of queries from the system, the SME is progressively asked to provide details about the contexts that compose the context-based model of the expert's tactical know-how. A prototype was built and evaluated. A comparison to the effort taken to manually develop a knowledge base is reported. We use the simulation of a non-trivial maritime military confrontation as the benchmark for the comparisons.

Modeling Vigilance Performance as a Complex Adaptive System

Current cognitive models not only lack flexibility and realism, they struggle to model individual behavior and reduced performance. We propose that reduced human performance can be best modeled as a complex adaptive system. We built a multi-agent model, “Reduced Human Performance Model (RHPM),” as a proof of principle. The simulation system realistically simulates the reduction of vigilance that individuals experience during such operations as airport screening, radar-screen operation, and other vital tasks in which attention easily flags. The developed multi-agent system generates individual behavior within a reasonable range. Its use for computer-generated forces (i.e., radar screen operator) would improve the realism of simulation systems by adding human-like reduced vigilance performance. The model represents a well suited tool to mediate between vigilance theories such as signal detection theory and experimental data. Using the model as a surrogate generates insights that potentially create likely hypotheses to improve the theories.

Military route planning in battlefield simulation: effectiveness problems and potential solutions

Path searching is challenging problem in many domains such as simulation war games, robotics, military mission planning, computer generated forces (CGF), etc. Effectiveness problems in military route planning are related both with terrain modelling and path planning algorithms. These problems may be considered from the point of view of many criterions. It seems that two criterions are the most important: quality of terrain reflection in the terrain model and computational complexity of the on(off)-line path planning algorithm. The paper deals with two above indicated problems of route planning effectiveness. Comparison of approaches used in route planning is presented. The hybrid, terrain merging-based and partial path planning, approach for route planning in dynamically changed environment during simulation is described. It significantly increase effectiveness of route planning process. The computational complexity of the method is given and some discussion for using the method in the battlefield simulation is conducted. In order to estimate how many times faster we can compute problem for finding shortest path in network with n big squares (b-nodes) with relation to problem for finding shortest path in the network with V small squares (s-nodes) acceleration function is defined and optimized.

Benefits and Consequences of Automated Learning in Computer Generated Forces Systems

Should computer generated forces (CGF) systems include automated learning capabilities? The CGF research literature contains many statements by CGF experts that the ability to learn will be generally valuable, even necessary, in future CGF systems. A variety of significant benefits for CGF systems and military simulation in general are claimed to follow from automated learning. However, upon closer examination, it seems to be not so obvious that learning by CGF systems would necessarily be beneficial for many uses of CGF systems. This paper takes a respectfully skeptical position regarding CGF learning and provides arguments that CGF learning could compromise and confound the utility of CGF systems for the most common CGF applications.

Training and Assessing Complex Decision-Making in a Virtual Environment

The current report describes a procedure for training and assessing complex decision-making in a virtual environment. Focusing on small unit leaders, 7 experienced and 7 inexperienced Army platoon leaders performed missions in a combat simulator, where they were required to direct the activities of 3 subordinate leaders and computer-generated forces in 4 different operations in a virtual urban setting. Objective and subjective assessments of the training value of the simulations showed that both experienced an inexperienced platoon leaders improved their decision-making across the four missions, and both groups rated this "virtual environment" training procedure as useful and positive. Lessons learned and implications for designing similar training protocols for other domains are discussed.

TRAINING AND ASSESSING COMPLEX DECISION-MAKING IN A VIRTUAL ENVIRONMENT

The current report describes a procedure for training and assessing complex decision-making in a virtual environment. Focusing on small unit leaders, 7 experienced and 7 inexperienced Army platoon leaders performed missions in a combat simulator, where they were required to direct the activities of 3 subordinate leaders and computer-generated forces in 4 different operations in a virtual urban setting. Objective and subjective assessments of the training value of the simulations showed that both experienced an inexperienced platoon leaders improved their decision-making across the four missions, and both groups rated this “virtual environment” training procedure as useful and positive. Lessons learned and implications for designing similar training protocols for other domains are discussed.

Computational Model of Recognition-Primed Decisions (RPD): Improving Realism in Computer-Generated Forces (CGF)

Modeling an intelligent adversary has provided great challenges to simulated training realism. Traditional approaches to modeling have relied on rule-based and analytical decision-making models in an attempt to optimize the decision making of an intelligent computer-generated adversary. In order to promote realistic transfer of training in the realm of command and control, the trainee must experience realistic decision-making behavior in the enemy. This means that the enemy must make realistic decisions based on environmental constraints, goals, and intent. The enemy's decisions must then be reflected by the simulated agents. The Recognition-Primed Decision (RPD) Model is a descriptive model of expert decision making in real-world settings. We have several related projects where the goal is to translate the conceptual RPD Model into a computer model that can simulate realistic expert decision making. In attempting this feat, we have discovered many valuable lessons about modeling cognition and decision making, and about the assumptions and mechanisms underlying the RPD Model. The purpose of this paper is to report those findings.

The Fully Immersive Team Training (FITT) Research System: Design and Implementation

This paper describes the design and implementation of the Fully Immersive Team Training (FITT) research system. FITT was developed to support research on the use of distributed virtual environments (VEs) for team training. A generic search mission was created and modeled after decision-making and procedural tasks used by military and civilian small teams in conducting missions in emergency situations. The hardware and software functional requirements for the development of FITT were defined by the characteristics of the mission environment to be simulated, the team members' actions and interactions necessary to complete the mission, and instructional features and interventions relevant to training research. These requirements included locomotion, object manipulation and aiming, communication among participants, design of avatars for participants and computer-generated forces, data capture and playback, and a host of networking issues. A brief description of the first experiment conducted with FITT is presented as an example of how the system will be used in VE training research.

Investigating the application of web-based simulation principles within the architecture for a next-generation computer generated forces model

With a heavy emphasis on distribution and reuse, web-based simulation portends a dramatic shift in the application of simulation as a problem-solving technique and decision-support tool. Next-generation simulation systems of all kinds should be evaluated and constructed with an appreciation of the potential paradigm shift that web-based simulation represents. Fiscal constraints indicate that next-generation computer generated forces (CGF) models will be used to support a wide range of missions, unifying — and replacing — a variety of CGF systems currently in existence. We describe the evolution of web-based simulation and derive a collection of modeling principles that characterize our vision of the web-based simulation future. We examine these principles in terms of their implications for next-generation CGF systems.

Modeling Training Effects in Computer Generated Forces

The use of Computer Generated Forces (CGF) in simulations ranges from small human performance models to large scale Advanced Distributed Simulations (ADS). CGF have been used to assess human performance aspects of system designs and to augment man-in-the-loop simulators with friendly and opposing forces to make simulated battlefield training exercises realistic. However, the current human performance models and ADS available to system designers, trainers, and resource allocation analysts do not include the effects of training on the performance of the CGF. For the most part, the CGF entities in these simulations reflect a constant level of training. This paper discusses a methodology for modeling the effects of training in CGF and other human performance models