Artificial Intelligence for Situation Assessment

نویسندگان

  • Örjan Ekeberg
  • Anders Lansner
چکیده

In the first phase of this master’s thesis, an overview of different approaches involving artificial intelligence to support situation assessment is presented, discussing the potential and the limits of each approach. In particular, two key issues in the context of automated sea surveillance are addressed: identification of particular events and scenarios in the situation picture, and detection of anomalous behaviour in general in the situation picture. Characterizing all events and situations, which may be of interest for a supervisor, is a very difficult task. The set of available examples for each particular event or situation is usually very limited, as the events and situations sought for occur relatively rarely and may vary significantly from one case to another. However, turning it the other way round, these rare events and situations can be detected as anomalies in a model of routine behaviour. Usually, large amounts of data corresponding to routine behaviour are available, which motivates the use of Data Mining and clustering techniques for building models of normal behaviour. In the second phase of this project, anomaly detection in sea traffic, based on clustering of real recorded vessel traffic, is further investigated and implemented. The implemented feature models are based on momentary vessel locations and velocities. Unsupervised clustering is done by a combination of two different cluster models and learning algorithms; one based on Mixtures of Gaussians (MoG) densities and Expectation-Maximization, and the other based on Neural Networks and Adaptive Resonance Theory (ART). Qualitative results from evaluating the implemented models show that the most distinguishing anomalies found in the typical routine traffic correspond to vessels that are crossing sea lanes and vessels that are travelling close to and in the opposite direction of sea lanes. Generally, the implemented models detect the same anomalies to rather large extent. The anomalies detected by the implemented systems are of a rather elementary nature; the type of feature model essentially determines the character of the detectable anomalies. Therefore, a more sophisticated feature model, based on manoeuvres in the motion pattern over time, is proposed as future development of the implemented system. However, the generality of the proposed system should be stressed, as it is applicable to other domains, involving generic motion in the two-dimensional plane, requiring minimal adaptation and no specific domain knowledge as the systems are based on unsupervised algorithms. Artificiell intelligens för situationsanalys Sammanfattning I den första delen av detta projekt presenteras en övergripande utvärdering kring olika tekniker och metoder, baserade på artificiell intelligens, för att stödja situationsanalys, där möjligheter och begränsningar för varje metod och teknik diskuteras. Två nyckelproblem i samband med automatiserad sjöövervakning tas upp; identifiering av särskilda händelser och situationer av intresse, samt detektering av avvikande beteenden i allmänhet. Att karaktärisera alla händelser och situationer, som kan vara av intresse för en övervakare, är en mycket svår uppgift. Mängden tillgängliga exempel på varje enskild typ av händelse eller situation är vanligen mycket begränsad, då de eftersökta händelserna och situationerna förekommer relativt sällan och kan variera avsevärt från ett fall till ett annat. Men, vänder man problemet, kan dessa händelser och situationer upptäckas som avvikelser i en modell för typiskt normalbeteende. Stora mängder data motsvarande typisk normaldata finns vanligen tillgängligt, vilket motiverar databrytningsoch klustringstekniker för att bygga upp modeller av normalbeteende. I den andra delen av projektet undersöks och implementeras algoritmer för avvikelsedetektion hos sjötrafik, baserat på klustring av inspelad sjötrafik. De implementerade särdragsmodellerna är baserade på fartygens momentana hastigheter och positioner. Oövervakad klustring görs med en kombination av två olika klustermodeller och klustringsalgoritmer; en baserad på mixturer av Gauss-fördelningar och ExpectationMaximization (EM), den andra baserad på artificiella neuronnät och Adaptive Resonance Theory (ART). Kvalitativa resultat från utvärderingen av de implementerade modellerna visar att de mest framträdande avvikelserna funna hos typisk rutintrafik svarar mot fartyg som korsar farleder och fartyg som färdas i närheten av och i motsatt riktning av farleder. Generellt så upptäcker de olika modellerna samma avvikelser i ganska stor utsträckning. De avvikelser som upptäckts av de implementerade modellerna är ganska elementära till sin natur; typen av särdragsmodell är fundamental för karaktären hos de framträdande avvikelserna. Därför föreslås en mer sofistikerad särdragsmodell, baserad på manövrar hos rörelsemönstret över tiden, som en lämplig utveckling av detta arbete. Allmängiltigheten hos det implementerade systemet skall dock understrykas, då det relativt enkelt kan överföras och anpassas till andra domäner som innefattar rörelse i det två-dimensionella planet.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

A Review on Applications of Haptic Systems, Virtual Reality, and Artificial Intelligence in Medical Training in COVID-19 Pandemic

This paper presents a survey on haptic technology, virtual reality, and artificial intelligence applications in medical training during the COVID-19 pandemic. Over the last few decades, there has been a great deal of interest in using new technologies to establish capable approaches for medical training purposes. These methods are intended to minimize surgerychr('39')s adverse effects, mostly w...

متن کامل

Transportation Applications of Artificial Intelligence

In this extended abstract, I present an overview of IVHS-related research recently initiated at the University of Michigan Artificial Intelligence Laboratory. Our ultimate goals in this effort are to further the technology underlying decision support for both individual drivers and traffic management centers within a variety of decision contexts, including routing, scheduling, and control of si...

متن کامل

Transportation Applications of Artificial Intelligence ( Extended

In this extended abstract, I present an overview of IVHS-related research recently initiated at the University of Michigan Artificial Intelligence Laboratory. Our ultimate goals in this effort are to further the technology underlying decision support for both individual drivers and traffic management centers within a variety of decision contexts, including routing, scheduling, and control of si...

متن کامل

Can Artificial Intelligence meet the Cognitive Networking Challenge ?

Recent developments in Software Defined Radio technology have created the opportunity to develop networks that are, in principle, highly adaptable and effective under a much wider range of operating conditions than currently possible, but few researchers are able to fully exploit this new flexibility. Cognitive radio networks require multiple interacting capabilities for situation assessment, p...

متن کامل

Level I and Level II Target Valuations for Sensor Management

Issues and Challenges in Situation Assessment (Level 2 Fusion) .............................. 122 Erik Blasch, Air Force Research Laboratory, USA Ivan Kadar, Interlink Systems Sciences, Inc., USA John Salerno, Air Force Research Laboratory, USA Mieczslaw M. Kokar, Northeastern University, USA Subrata Das, Charles River Analytics, USA Gerald M. Powell, U.S. Army RDECOM CERDEC I2WD Daniel D. Cork...

متن کامل

Simple Stochastic Temporal Constraint Networks

Many artificial intelligence tasks (e.g., planning, situation assessment, scheduling) require reasoning about events in time. Temporal constraint networks offer an elegant and often computationally efficient framework for formulation and solution of such temporal reasoning tasks. Temporal data and knowledge available in some domains is necessarily imprecise e.g., as a result of measurement erro...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2007