Krystian Łapa

A new method for designing neuro-fuzzy systems for nonlinear modelling with interpretability aspects

In this paper we propose a new approach to nonlinear modelling. It uses capabilities ofthe so-called flexible neuro-fuzzy systems and evolutionary algorithms. The aim of our method is not only to achieve appropriate accuracy of the model, but also to ensure the possibility of interpretability of the knowledge within it. The proposed approach was achieved by, among others, appropriate selection of operational criteria applied to evolutionary model creation. It allows to extract interpretable fuzzy rules in the cases which use the learning data e.g. from identification. The possibility of interpretation of knowledge accumulated in the model seems to be important in practice, because it guarantees operation predictability and facilitates production of efficient and accurate control methods. Our method was tested with the use of well-known simulation problems from the literature. HighlightsWe propose a new algorithm to nonlinear modelling.Our method uses flexible neuro-fuzzy systems and population based algorithms.Our method includes various criteria of the interpretability of the fuzzy rules.Our method uses Wards clustering method to initial population generation.

A New Approach to Designing Interpretable Models of Dynamic Systems

In the process of designing automatic control system it is very important to have an accurate model of the controlled process. Approaches to modelling dynamic systems presented in the literature are often approximate, uninterpretable (acting as a black box), not appropriate to work in real-time, so it is not possible to create a hardware emulator on the basis of these approaches. The paper presents a new method to create model of nonlinear dynamic systems which gives a real opportunity for the interpretation of accumulated knowledge. By combining methods of control theory with fuzzy logic rules a good accuracy of the model can be achieved with use of a small number of fuzzy rules. Our method is based on the evolutionary strategy \(\left({\mu,\lambda} \right)\).

New Algorithm for Evolutionary Selection of the Dynamic Signature Global Features

Methods using dynamic signature for identity verification may be divided into three main categories: global methods, local function based methods and regional function based methods. Global methods base on a set of global parametric features, which are extracted from signature of user. Global feature extraction methods have been often presented in the literature. Another interesting task is selection of a features group which will be considered individually for each user during training and verification process. In this paper we propose a new approach to automatic evolutionary selection of the dynamic signature global features. Our method was tested with use of the SVC2004 public on-line signature database.

A New Method for Designing and Complexity Reduction of Neuro-fuzzy Systems for Nonlinear Modelling

In this paper we propose a new method for evolutionary selection of parameters and structure of neuro-fuzzy system for nonlinear modelling. This method allows maintain the correct proportions between accuracy, complexity and interpretability of the system. Our algorithm has been tested using well-known benchmarks.

A new approach to design of control systems using genetic programming

In this paper a new approach to automatic design of control systems is proposed. It is based on a knowledge about modelling object and capabilities of the genetic programming. In particular, a new type of the problem encoding, new evolutionary operators (tuning operator and mutation operator) and new initialization method are proposed. Moreover, we present a modified block schema of genetic algorithm and modification of genetic operators: insertion, pruning, crossover were introduced. Combination of mentioned elements allows us to simplify a design of control systems. It also provides a lot of possibilities in the selection of the control system parameters and its structure. Our method was tested on the model of quarter car active suspension system. DOI: http://dx.doi.org/10.5755/j01.itc.44.4.10214

New Method for Design of Fuzzy Systems for Nonlinear Modelling Using Different Criteria of Interpretability

In this paper a new method for designing neuro-fuzzy systems for nonlinear modelling is proposed. This method contains a complex weighted fitness function with interpretability criteria and new enhanced tuning process for selecting parameters and structure of the system based on a hybrid population-based algorithm (composed of evolutionary strategy, genetic algorithm and bees algorithm). To evaluate this method, we used a well-known dynamic nonlinear modelling problem.

Aspects of Structure and Parameters Selection of Control Systems Using Selected Multi-Population Algorithms

In this paper a new approach for automatic design of control systems is presented. It is based on multi-population algorithms and allows to select not only parameters of control systems, but also its structure. Proposed approach was tested on a problem of stabilization of double spring-mass-damp object.

Aspects of the Selection of the Structure and Parameters of Controllers Using Selected Population Based Algorithms

In this paper we propose a new approach for selection of the structure and parameters of the control system. Proposed approach is based on the selected population-based algorithms. In this approach we considered a combination of the genetic algorithm (it is used for selection of structure of the control system) fused with one of the following algorithms: evolutionary algorithm, firefly algorithm, gravitational search algorithm, bat algorithm and imperialist competitive algorithm (they are all used for the selection of parameters of the control system). In experimental simulations a typical problem of the control process was used.

A New Method for Generating of Fuzzy Rules for the Nonlinear Modelling Based on Semantic Genetic Programming

In this paper we propose a new approach for nonlinear modelling. It uses capabilities of the Takagi-Sugeno neuro-fuzzy systems and population based algorithms. The aim of our method is to ensure that created model achieves appropriate accuracy and is as compact as possible. In order to obtain this aim we incorporate semantic information about created fuzzy rules into process of evolution. Our method was tested with the use of well-known benchmarks from the literature.

A New Interpretability Criteria for Neuro-Fuzzy Systems for Nonlinear Classification

In this paper a new approach for construction of neuro-fuzzy systems for nonlinear classification is introduced. In particular, we concentrate on the flexible neuro-fuzzy systems which allow us to extend notation of rules with weights of fuzzy sets. The proposed approach uses possibilities of hybrid evolutionary algorithm and interpretability criteria of expert knowledge. These criteria include not only complexity of the system, but also semantics of the rules. The approach presented in our paper was tested on typical nonlinear classification simulation problems.