Theory of Systems and Systems Analysis (Kravets, 2012)

This course is a broad map of systems thinking rather than a single technical theory. Its eighteen lectures move from the idea of a system and ways to describe structure, through state, behavior, information, feedback and modeling, to multi-criteria choice, decomposition, aggregation, and then-current software tooling. The durable contribution is the sequence of questions it asks. The bank uses that sequence to discover topics, then replaces load-bearing claims with current primary sources.

What the course contributes #

The opening lectures insist on three separations that remain valuable:

• the system is not identical to its list of parts; relations and the selected boundary determine what counts as the whole;
• internal description concerns elements, relations, state and transformation, while external description concerns exchanges and observable behavior in context;
• the same real thing supports multiple models because a model is built for a question, not as a complete duplicate of reality.

The middle block connects information and control, introduces static and dynamic models, distinguishes continuous and discrete descriptions, and surveys measurement scales. The final block treats choice as a relation over alternatives, shows why multiple criteria generally produce a set of non-dominated alternatives rather than a unique optimum, and presents decomposition as an iterative modeling operation whose stopping rule depends on purpose.

That material now maps to topic chapters instead of remaining one course summary:

Information-bearing visual audit #

Every one of the 237 pages was rendered and reviewed, with the pages containing diagrams inspected again at readable resolution. The load-bearing visual families are:

• linear, hierarchical, network and matrix structures;
• a system boundary with subsystems and matter, energy and information exchanges;
• directed graphs for informational and causal descriptions;
• input/output model abstraction and model-class trees;
• frame and predicate knowledge representations;
• Pareto regions and preference graphs;
• decomposition trees, stopping logic and aggregation examples.

None is retained as a raster. Most are schematic enough to redraw more clearly, but several embed course-specific labels, unattributed examples, or obsolete software screenshots. The topic entries therefore use independently constructed diagrams tied to current terminology. No visual from the laboratory manuals or student work was used.

Where the source has aged #

The final lecture treats Windows 95, early object-oriented tooling, CASE environments and virtual reality as representative “new technologies.” Those pages are useful only as history. Current model-based systems engineering is anchored instead in ISO/IEC/IEEE 15288:2023 and OMG SysML 2.0, whose formal version was adopted in 2025.

Some terminology is also too loose for modern use. Shannon entropy measures uncertainty in a probability distribution, not semantic meaning or general “amount of knowledge.” A cognitive map records an analyst's asserted causal structure; it does not establish causality. A decomposition tree is not proof that the omitted environment, cross-cutting interaction, or lifecycle concern is irrelevant.

Source claims audit #

How to cite it #

Use the course for its historical synthesis and topic map. For current engineering processes, formal information measures, control design, or deployable modeling technology, cite the relevant standard or primary technical source instead. The archived file preserves the course as received, while the topic chapters carry the current account.