OUR PROGRAMS

For our partners we offer :
Partnership Programs for the universities, research institutes R&D


Quantum optimizer bases of knowledge

The use of soft calculations technologies based on genetic algorithms and fuzzy neural networks, expanded the field of effective application of the fuzzy controller by adding new functions as learning and adaptation. However, it is very difficult to design a global "good" , robust structure of the intelligent system management. This limitation is especially true for contingency management, facility management when operating in a dramatically changing environment (failure of sensors or noise in the measurement system, the presence of time delay control signals or measurement, a sharp change in the structure of the object or control parameters, etc.).

In some practical cases, such conditions can be predicted, but difficult to implement robust control in unexpected situations with projected (for a fixed situation) a knowledge base of fuzzy controller (even on the whole set) predicted by the random situations.

One of the existing solutions is the formation of a finite number of the knowledge base of fuzzy controller for a variety of fixed-control situations.

The question arises: how to identify which of the bases of knowledge to be used at a particular time? In this case, special importance of the choice of the generalized strategy that would give the ability to switch the stream of control signals from the output of the different bases of knowledge of fuzzy control, and (if necessary) to modify their output signal under the current conditions of operation of the facility management. A simple solution to this is to use the method of suspended weights and aggregated output signals from each of the independent fuzzy controller. Unfortunately, this method has limitations, since the distribution of weight factors is often necessary to determine the dynamics of real-time, , search procedure has a combinatorial nature.

Solution of such problems can be found by introducing the principle of self-organization in the process of designing KB of fuzzy controller, which is implemented by means of quantum optimizer bases of knowledge.