جستجوی مقالات مرتبط با کلیدواژه « adaptation » در نشریات گروه « برق »

The surrounding context information of business processes is unpredictable and dynamically changing over time. Therefore, they should dynamically adapt themselves to different changes in context information such as business rules or computational changes in available resources. For example, we may add a particular delivery service for golden customers, or provide new payment services in case of unavailability of service providers, or change a service invocation based on available bandwidth. Unlike other methods which provide a context Meta model or a shallow context taxonomy in a specific limited scope of the business domain, we focused on ontology engineering methods not only to propose a general multipurpose context ontology but also to present our proposal as an underlying ontology for other researchers to extend and customize it for their applications. In this paper, the business process adaptation knowledge is modelled in the form of concepts, relations, and axioms which comprises time, resources, performers, locations, environment, and rules to model and record whole context information of adaptation mechanism. We characterized our work in comparison with related studies to show its completeness and demonstrated it by using an online learning management system and virtual class case studies.

In this paper, the algorithms for adaptive rejection of a radar clutter are synthesized for the case of a priori unknown spectral-correlation characteristics at wobbulation of a repetition period of the radar signal. The synthesis of required algorithms for the non-recursive adaptive rejection filter (ARF) of a given order, which is used in modern radar systems, reduces to determination of the weighting coefficients vector, which realizes the best effectiveness index for radar signal extraction from the moving targets on the received clutter background. As the effectiveness criterion, we consider the averaged (over the Doppler signal phase shift) improvement coefficient for a signal-to-clutter ratio (SCR), which is usually used at describing the modern radars. On the base of extreme properties of the characteristic numbers (eigennumbers) of the matrices, the optimal vector is defined as the eigenvector of the correlation matrix of the clutter, corresponding to its minimal eigenvalue. After transition to the estimate value of the clutter correlation matrix according to the adaptive Bayesian approach, the general type of the vector of optimal ARF weighting coefficients is determined. The specific adaptive algorithms, depending upon the ARF order, are obtained, which in the particular cases can be reduced to the known algorithms confirming its authenticity. The comparative analysis of the synthesized and known algorithms is performed. It is shown that adaptation of offered processing algorithms in combination with its time adjustment leads to significant benefits in clutter rejection effectiveness compared to the known processing algorithms. Application of the algorithms obtained during ARF designing allows increase the effectiveness of radar signal extraction from the moving targets in much more range of its radial velocities on the clutter background in the case of unknown spectral-correlation properties.

An adaptive process consists of dynamic elements, and management rules which govern their run-time behaviors. The WS-CDL describes collaborative business processes between service consumers and providers. Adapting the processes to runtime changes becomes a demanding challenge, because the WS-CDL has static technology-dependent structure, and does not support the separation of concerns. Here, we propose a model-driven approach to transform WS-CDL into UML state machine (behavioral and protocol models), and subsequently into implementation code. Besides separating the business logic from the implementation, the state machine has a dynamic structure which is verifiable and adaptable. As a result, we can easily modify the process flow or change the management rules at run-time, and reflect their effects on the running process instances. We present an ‘itinerary purchase’ case study for prototyping the transformation rules.