Real-Time Systems. Design Principles for Distributed Embedded Applications. Herman Kopetz. Second Edition (811374), страница 16
Текст из файла (страница 16)
276].Example: The basic level concept temperature is more fundamental than the sub-conceptoil-temperature or the encompassing concept sensor data.Studies with children have shown that basic-level concepts are acquired earlier thansub-concepts or encompassing concepts. As a child grows up it continually buildsand adds to its conceptual landscape by observing regularities in the perceptions andutility in grouping properties of perceptions into new categories [Vig62]. These newcategories must be interlinked with the already existing concepts in the child’s mindto form a consistent conceptual landscape. By abstracting not only over perceptions,but also over already existing concepts, new concepts are formed.A new concept requires for its formation a number of experiences that havesomething in common and form the basis for the abstraction. Concept acquisition isnormally a bottom-up process, where sensory experiences or basic concepts are thestarting point.
Examples, prototypes and feature specification play an important rolein concept formation. A more abstract concept is understood best bottom up bygeneralizations from a set of a suitable collection of examples of already acquiredconcepts. Abstract analysis and concrete interpretation and explanation should beintertwined frequently. If one remains only at a low-level of abstraction then theamount of non-essential detail is overwhelming.
If one remains only at a high-level ofabstraction, then relationships to the world as it is experienced are difficult to form.In the real world (in contrast to an artificial world), a precise definition of aconcept is often not possible, since many concepts become fuzzy at their boundaries[Rei10, p. 272].Example: How do you define the concept of dog? What are its characteristic features?Is a dog, which has lost a leg, still a dog?Understanding a new concept is a matter of establishing connections between thenew concept and already familiar concepts that are well embedded in the conceptual landscape.Example: In order to understand the new concept of counterfeit money, one must relatethis new concept to the following already familiar concepts: (1) the concept of money,2.2 The Conceptual Landscape37(2) the concept of a legal system, (3) the concept of a national bank that is legalized to printmoney and (4) the concept of cheating.
A counterfeit money bill looks like an authenticmoney bill. In this situation, examples and prototypes are of limited utility.In the course of cognitive development and language acquisition, words (names)are associated with concepts. The essence of a concept associated with a word canbe assumed to be the same within a natural language community (denotation), butdifferent individuals may associate different shades of meaning with a concept(connotation), dependent on their individual existing conceptual landscape and thediffering personal emotional experiences in the acquisition of the concept.Example: If communicating partners refer to different concepts when using a word orif the concept behind a word is not well established in the (scientific) language community,(i.e., does not have a well-defined denotation), then effective communication amongpartners becomes difficult to impossible.If we change the language community, the names of concepts will be changed,although the essence of the concept, its semantic content, remains the same.
Thenames of concepts are thus relative to the context of discourse, while the semanticcontent remains invariant.Example: The semantic content of the concept speed is precisely defined in the realmof physics. Different language communities give different names to the same concept: inGerman Geschwindigkeit, in French vitesse, in Spanish velocidad.2.2.2Scientific ConceptsIn the world of science, new concepts are introduced in many publications in orderto be able to express new units of thought. Often these concepts are named by amnemonic, leading to, what is often called, scientific jargon. In order to make anexposition understandable, new concepts should be introduced sparingly and withutmost care.
A new scientific concept should have the following properties[Kop08]:lllllUtility. The new concept should serve a useful well-defined purpose.Abstraction and Refinement. The new concept should abstract from lower-levelproperties of the scenario under investigation. It should be clear which propertiesare not parts of the concept. In the case of refinement of a basic-level concept, itshould be clearly stated what additional aspects are considered in the refinedconcept.Precision.
The characteristic properties of the new concept must be preciselydefined.Identity. The new concept should have a distinct identity and should be significantly different from other concepts in the domain.Stability. The new concept should be usable uniformly in many different contexts without any qualification or modification.38l2 SimplicityAnalogy. If there is any concept in the existing conceptual landscape that is, insome respects, analogous to the new concept, this similarity should be pointedout. The analogy helps to establish links to the existing conceptual landscape ofa user and facilitates understanding.
According to [Hal96, p. 5]:Analogical reasoning mechanisms are important to virtually every area of highercognition, including language comprehension, reasoning and creativity. Humanreasoning appears to be based less on an application of formal laws of logic than onmemory retrieval and analogy.The availability of a useful, well defined, and stable set of concepts and associatedterms that are generally accepted and employed by the scientific community is amark for the maturity of a scientific domain.
An ontology is a shared taxonomy thatclassifies terms in a way useful to a specific application domain in which allparticipants share similar levels of understanding of the meaning of the terms[Fis06, p. 23]. Progress in a field of science is intimately connected with conceptformation and the establishment of a well-defined ontology.Example: The main contributions of Newton in the field of mechanics are not only in theformulation of the laws that bear his name, but also in the isolation and conceptualization ofthe abstract notions power, mass, acceleration and energy out of an unstructured reality.Clear concept formation is an essential prerequisite for any formal analysis orformal verification of a given scenario.
The mere replacement of fuzzy conceptsby formal symbols will not improve the understanding.2.2.3The Concept of a MessageWe consider a message as a basic concept in the realm of communication.A message is an atomic unit that captures the value domain and the temporaldomain of a unidirectional information transport at a level of abstraction that isapplicable in many diverse scenarios of human communication [Bou61] andmachine communication. A basic message transport service (BMTS) transports amessage from a sender to one or a set of receivers.
The BMTS can be realized bydifferent means, e.g., biological or electrical.For example, the message concept can be used to express the information flowfrom the human sensory system to the conceptual landscape of an individual. Themessage concept can also model the indirect high-level interactions of a humanwith his environment that are based on the use of language.Example: We can model the sensory perception, e.g. of temperature, by saying that amessage containing the sensed variable (temperature) is sent to the conceptual landscape.A message could also contain verbal information about the temperature at a location that isoutside the realm of direct sensory experience.The message concept is also a basic concept in the domain of distributed embeddedcomputer systems at the architecture level.
If the BMTS between encapsulated subsystems is based on unidirectional temporally predictable multicast messages, then2.2 The Conceptual Landscape39the data aspect, the timing aspect, the synchronization aspect, and the publicationaspect are integrated in a single mechanism. The BMTS can be refined at a lower levelof abstraction by explaining the transport mechanism. The transport mechanism couldbe wired or wireless.
The information can be coded by different signals. Theserefinements are relevant when studying the implementation of the message mechanism at the physical level, but are irrelevant at a level where the only concern is thetimely arrival of the information sent by one partner to another partner.A protocol is an abstraction over a sequence of rule-based message exchangesbetween communicating partners. A protocol can provide additional services, suchas flow control or error detection. A protocol can be understood by breaking it downto the involved messages without the need to elaborate on the concrete transportmechanisms that are used.2.2.4Semantic Content of a VariableThe concept of a variable, a fundamental concept in the domain of computing, is ofsuch importance for the rest of the book that it justifies some special elaboration.A variable can be considered as a language construct that assigns an attribute to aconcept.
If the point in real-time, the instant, when this assignment is valid, is ofrelevance, then we call the variable a state variable. As time progresses, theattribute of a state variable may change, while the concept remains the same.A variable thus consists of two parts, a fixed part, the variable name (or theidentifier), and a variable part called the value of the variable that is assigned tothe variable. The variable name designates the concept that determines what we aretalking about. In a given context, the variable name – which is analogous to thename of a concept in a natural language community – must be unique and point tothe same concept at all communicating partners. The meaning that is conveyed by avariable is called the semantic content of the variable.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
