The topological analysis OF INVENTIVE (INITIAL) SITUATION

 

Salamatov Y.

 

1. The purposes and difficulties of formalization of ARIZ

Difficulties of transformation of many tools of TRIZ (which represent today " textual (paper) algorithms ") into automatic systems AI (artificial intellect) consist in insufficient completeness and clearness of structure of bases of knowledge of these tools.

There are three levels of development of bases of knowledge:

Questionnaires and simple heuristics (such as - effects, principles of a solution of engineering contradictions); this level cannot be automated;

Rules of modelling and acceptance of decisions, for example, system of standards, - it is possible to automate some fragments;

The algebra of the contradictions, exact models, formalization of qualitative concepts - is possible full automation.

ARIZ there is a most advanced tool for the decision of difficult problems. But also ARIZ is "mix" of two levels - of questionnaires and of rules, it suffers of conceptual incompleteness and uncertainty. For example, in 1-st part of ARIZ:

There is no subroutine of synthesis and processing Inventive situations (IS);

There is no mechanism of synthesis of the TS (if there is no prototype for a mini-problem (mini-task));

There are no precise criteria of a choice of a problem from some set;

The full set of schemes of conflicts in models of problems is unknown;

Regarding problems is difficultly precise to formulate the purpose of changes (the main useful function, the main process of production).

Therefore ARIZ in a present kind it is impossible to automate.

Development of each of these questions up to a level of formalization of steps of algorithm would allow to create elements of automatic processing of problems. Formalization here is understood as a method of revealing and an explanation of investigated laws at which the same idea is always expressed by means of the same words and the marks, located by the same way.

I.e. formation of IS is an active process of penetration into a problem. This process demands expenses of time and at person which solves a problem should be skills of processing of IS. Frequently it is creative process of "extraction" Inventive problems from real complex technical system. In this process the person simultaneously is a source of the information, expert, critic.

In without-machine use ARIZ the model of IS arises and is processed in a brain of the person and formulated at any moment on a step 1.1. as a mini-problem. This process of processing of IS till now was not structured and formalized. Conclusions were very approximate. Not everyone the person were sure in correctness of the formulation of a mini-problem.

Instead of natural language it is necessary to use other type of representation - graphic or graphic-language   (semantic networks).

In the scheme of IS there should be all elements because of which there is a conflict. These elements should be submitted without redundancy and without a multivalence. Then of IS it will be possible to present in the formalized aspect. The obtained statement of IS then is necessary for reducing (having entered restrictions) and to receive in outcome a mini-problem.