Michio Sugeno’s research while affiliated with Tokyo Institute of Technology and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (234)


Piecewise Modeling Algorithms Using Numerical Data
  • Chapter

February 2024

·

10 Reads

·

Michio Sugeno

In this study, a novel piecewise modeling method was devised using numerical data. The piecewise model was represented as a rectangular region divided into a state-space. The vertex values of the rectangular region were determined using the learning algorithm based on a simplified fuzzy inference model because the piecewise model was represented by a fuzzy if-then rule with singleton consequents. The proposed algorithm can be used to determine optimal vertex values and positions of segmented regions and with minimum modeling errors. Three examples were considered to demonstrate the effectiveness of the proposed method using numerical simulations.



Convex Stability Analysis of Mamdani-Like Fuzzy Systems With Singleton Consequents
  • Article
  • Full-text available

April 2023

·

267 Reads

·

5 Citations

IEEE Transactions on Fuzzy Systems

We study the stability of a class of discrete-time fuzzy systems with singleton consequents, called Mamdani-like fuzzy systems. The parametric expressions, specific to this class of fuzzy systems, are leveraged to derive stability analysis conditions via Finsler's lemma and Lyapunov stability tools. This allows avoiding the major challenge in dealing with high-dimensional cases, encountered in the related literature when using the classical state-space representation. Moreover, the information of the piecewise region partition can be fully taken into account in the stability analysis with the well-known SS- procedure to further reduce the stability conservatism. The stability of Mamdani-like fuzzy systems can be checked by solving a set of linear matrix inequalities (LMIs), that is numerically tractable with a suitable semidefinite programming software. Several numerical and physically motivated examples are provided to illustrate the effectiveness of the proposed stability analysis results.

Download






Piecewise Multi-linear Model Based Lookup Table Controller for Nonlinear Systems with Input Constraints

January 2021

·

14 Reads

·

2 Citations

Studies in Fuzziness and Soft Computing

This paper proposes a piecewise multi-linear model based lookup table controller for nonlinear systems with input constraints. We construct a piecewise multi-linear model of a nonlinear system. The model is a nonlinear approximation and it can be derived from fuzzy if-then rules with singleton consequents. The model can be expressed as a lookup table. The dynamics is described by multi-linear interpolation of the lookup table elements. We design a piecewise multi-linear controller for the piecewise model via feedback linearization. The controller can be also represented as a lookup table. We propose the lookup table controller for nonlinear systems with input constraints and apply the controllers to TORA (Translational Oscillator with Rotating Actuator) system. Examples are shown to confirm the feasibility of our proposals by computer simulations.


The Fuzzy Theoretic Turn

January 2021

·

3 Reads

·

1 Citation

Studies in Fuzziness and Soft Computing

This short essay characterizes Zadeh’s contribution as the Fuzzy Theoretic Turn, recalling the past two turns in the human history: Kopernikanishe Wende (turn) by Kopernik in astronomy and the Linguistic Turn by Wittgenstein in philosophy.


Citations (45)


... This is the author's version which has not been fully edited and content may change prior to final publication. fuzzy rules are used in fuzzy system applications, such as the Mamdani-type [41], Takagi-Sugeno [42], and Piecewise Multiaffine [43] fuzzy rules. The fuzzy inference machine, which simulates the decisionmaking procedure of human beings based on fuzzy concepts, is a decision-making logic that employs fuzzy rules to determine the output of a fuzzy system. ...

Reference:

Influencing Factor-Decoupled Battery Ageing Assessment for Real-world Electric Vehicles based on Fusion of Fuzzy Logic and Neural Network
Convex Stability Analysis of Mamdani-Like Fuzzy Systems With Singleton Consequents

IEEE Transactions on Fuzzy Systems

... However, for the PEM-based algorithm, fuzzy measures of criteria representing their relative significance are determined by teachers or educators and can be directly applied to derive the evaluation results. Moreover, compared with the classic fuzzy measurement, fuzzy measures take the importance of the combination of decision a ributes into account rather than only considering the importance of the a ributes themselves [22]. For example, teachers may consider that the criteria of online engagement and sustained learning in student performance assessment are both important, while the combined importance of the two could be greater or less than their sum. ...

On the f-divergence for discrete non-additive measures
  • Citing Article
  • September 2019

Information Sciences

... To help the reader understand the paper, several important details about this new derivative have been provided along with a variety of examples; for further information, refer to [10]. The Sugeno derivative theory has garnered significant attention and publications; for some of them, see [1,3,[6][7][8]11,12] and the references cited therein. Particularly, Sugeno focused on statistics with distorted probabilities in 2018 [11], utilizing the Sugeno derivative [10] and the Choquet integral. ...

Application of Choquet Calculus to Statistics With Distorted Probabilities
  • Citing Article
  • December 2018

IEEE Transactions on Fuzzy Systems

... Fig. 4 presents the structure of the proposed FLC design done in MATLAB/Simulink ® . The Mamdani Type 1 block refers to the rule based defuzzification system as presented in Nguyen et al. [18]. The selected rules are shown in Fig. 5. ...

Fuzzy Control Systems: Past, Present and Future

IEEE Computational Intelligence Magazine

·

·

·

[...]

·

Michio Sugeno

... Moreover, since linguistic labels may be assigned to singletons of singleton-type fuzzy systems, they can be called Mamdani-like fuzzy systems [14]. These fuzzy systems have been also called piecewise multiaffine systems due to their input-output multiaffine relation with respect to an affine-in-control system [15]. Mamdani-like fuzzy modeling has been found advantageous to study nonlinear systems for many reasons [3]. ...

LMI-Based Stability Analysis Conditions for Piecewise Multiaffine Systems

IEEE Transactions on Fuzzy Systems

... Besides, with such tuned PID they cannot handle real-world uncertainties like wind gust, i.e., their PID controller is not robust against uncertainties. Some other commonly used methods for controlling quadcopter are Lyapunov theory based methods (Zuo 2013), Linear Quadratic (LQ) techniques (Bader et al. 2014;Bouabdallah et al. 2004b;Rinaldi et al. 2013), Sliding Mode Control (SMC) methods (Bouabdallah and Siegwart 2005;Lee et al. 2009;Waslander et al. 2005;Xiong and Zhang 2016;Xu and Ozguner 2006), back-stepping techniques (Bouabdallah and Siegwart 2005;Madani and Benallegue 2006), Feedback Linearization (FBL) techniques (Lee et al. 2009;Taniguchi et al. 2014), H∞ robust control techniques (Chen and Huzmezan 2003;Walker and Postlethwaite 1996) etc. Bouabdallah et al. (2004b) have developed a Linear Quadratic Regulator (LQR) based algorithm for controlling the quadcopter. The efficiency of their controller is compared with PID controller. ...

Tracking control of a four rotors helicopter via dynamic feedback linearization based on piecewise bilinear models
  • Citing Article
  • January 2014

... The concept of capacity and Choquet integral as important tools in non-additive measure theory [4] were initiated by [2], which later became widely used in many fields such as statistics and financial economics [5,8,9], multicriteria decision making [6,11,13], economic modeling [7], risk measuring [7,8] and fuzzy measure theory [15,16]. Recently, the study of numerical Choquet integration on the real line has been studied using the statistics software R by [17,19]. They also proposed several algorithms for some functions with some examples for Hellinger distance between two monotone measures. ...

On the f-divergence for non-additive measures
  • Citing Article
  • July 2015

Fuzzy Sets and Systems

... The concept of capacity and Choquet integral as important tools in non-additive measure theory [4] were initiated by [2], which later became widely used in many fields such as statistics and financial economics [5,8,9], multicriteria decision making [6,11,13], economic modeling [7], risk measuring [7,8] and fuzzy measure theory [15,16]. Recently, the study of numerical Choquet integration on the real line has been studied using the statistics software R by [17,19]. ...

A Way to Choquet Calculus
  • Citing Article
  • October 2015

IEEE Transactions on Fuzzy Systems

... Another possible issue is the feasibility of inversion; an uninvertible singular matrix can block the application of the Feedback linearization though some methods, e.g., PCH ( [8]), dynamic approximation ( [9]), etc. are invented. One typical way to pick the controlled variables is the yaw ( ) and the position ( ) combination. ...

Quadrotor control using dynamic feedback linearization based on piecewise bilinear models
  • Citing Conference Paper
  • December 2014