Abstract: Most of self-tuning fuzzy systems, which are
automatically constructed from learning data, are based on the
steepest descent method (SDM). However, this approach often
requires a large convergence time and gets stuck into a shallow
local minimum. One of its solutions is to use fuzzy rule modules
with a small number of inputs such as DIRMs (Double-Input Rule
Modules) and SIRMs (Single-Input Rule Modules). In this paper,
we consider a (generalized) DIRMs model composed of double
and single-input rule modules. Further, in order to reduce the
redundant modules for the (generalized) DIRMs model, pruning and
generative learning algorithms for the model are suggested. In order
to show the effectiveness of them, numerical simulations for function
approximation, Box-Jenkins and obstacle avoidance problems are
performed.
Abstract: The paper presents a method in which the expert
knowledge is applied to fuzzy inference model. Even a less
experienced person could benefit from the use of such a system, e.g.
urban planners, officials. The analysis result is obtained in a very
short time, so a large number of the proposed locations can also be
verified in a short time. The proposed method is intended for testing
of locations of car parks in a city. The paper shows selected examples
of locations of the P&R facilities in cities planning to introduce the
P&R. The analyses of existing objects are also shown in the paper
and they are confronted with the opinions of the system users, with
particular emphasis on unpopular locations. The results of the
analyses are compared to expert analysis of the P&R facilities
location that was outsourced by the city and the opinions about
existing facilities users that were expressed on social networking
sites. The obtained results are consistent with actual users’ feedback.
The proposed method proves to be good, but does not require the
involvement of a large experts team and large financial contributions
for complicated research. The method also provides an opportunity to
show the alternative location of P&R facilities. Although the results
of the method are approximate, they are not worse than results of
analysis of employed experts. The advantage of this method is ease of
use, which simplifies the professional expert analysis. The ability of
analyzing a large number of alternative locations gives a broader
view on the problem. It is valuable that the arduous analysis of the
team of people can be replaced by the model's calculation. According
to the authors, the proposed method is also suitable for
implementation on a GIS platform.