System analysis

from Wikipedia, the free encyclopedia

The system analysis is practically applicable method of system theory . The observer of the system first constructs a model of an existing or planned system as a black box and refines it in the further course. The processor has to make a selection with regard to the relevant elements and relationships of the system. The model created is - especially in the case of complex systems - mostly a limited, reduced, abstract image of reality, with the help of which statements can be made about past and future developments and behavior of the system in certain scenarios. The process is applicable to nearly any system including physics , biology , demography , economics , geography , engineering, and computer science .

The opposite process is called system synthesis . It is used in practice for. B. used in the simulation of laws, often with the help of software .

Work steps

  1. Survey and analysis of a given problem
  2. Specification of a general objective
  3. Establishing the system boundaries to differentiate between system and environment.
  4. Determine those system elements that are considered relevant to the question.
  5. Determination of those relationships between the system elements that are considered relevant to the question.
  6. Determination of the system properties at the macro level.
  7. Establishing the relationships of the system to the environment or to other systems when changing from the view of the system as an isolated or closed system to the open system.

presentation

Presentation of the analysis results:

  • qualitative: concept map , flowchart , impact diagrams
  • semi-quantitative: arrow diagram (the-the-relations)
  • quantitative: xy, xt diagrams, among other things, mathematical systems of equations

Formal and graphic methods are used for the system analysis .

Keith Edwards uses the following elements in his work to represent various pattern systems:

He also names the following theoretically possible combinations, which, however, are only useful to a very limited extent in practice:

  • Assignment between data stream display and data storage (for verification).
  • Temporal change in data processing by control signals (for function control).

The derivation of states ("States") through events ("Events") and vice versa is possible. A constant limitation to a set of elements that makes sense for the respective level of detail is necessary in order to arrive at a suitable, i.e. transparent and thus usable result. The representation differentiates between control flows, data flows, instantaneous events and physical flows of matter or energy.

Examples of computer science

System integration

In computer science, system analysis is the first phase in the design process. The system analyst describes the system elements relevant to his question and their relationships to one another (usually with an information model ). The aim of the system analysis is, for example, to describe the environment without a machine (actual state) in order to plan a machine based on this actual model. The target model shows how the machine should look. The differences between the actual and target model make it clear what the machine to be designed should achieve. The system analysis does not examine how the machine is implemented. In this context, a machine is understood to be hardware and software as a unit. Systems analysis can also be used prior to optimizing , migrating and converting systems.

Application system development

When creating application systems in a business context or when adapting standard software (" customizing "), it can make sense to model the relevant business processes (for example with event-driven process chains [EPK]). These models not only serve as a basis for planning organizational measures ( process management ), but are also suitable for determining requirements for application systems in order to efficiently support business processes through IT without media discontinuity .

literature

See also