A Self-Adaptive Online Brain–Machine Interface of a Humanoid Robot Through a General Type-2 Fuzzy Inference System

Abstract

Abstract-This paper presents a self-adaptive general type-2 fuzzy autonomous learning system (GT2 FS) for online motor imagery (MI) decoding to build a brain-machine interface (BMI) and navigate a bi-pedal humanoid robot in a real experiment, using EEG brain recordings only. GT2 FSs are applied to BMI for the first time in this study. We also account for several constraints commonly associated with BMI in real practice: 1) maximum number of electroencephalography (EEG) channels is limited and fixed, 2) no possibility of performing repeated user training sessions, and 3) desirable use of unsupervised and low complexity features extraction methods. The novel online autonomous learning method presented in this paper consists of a self-adaptive GT2 FS that can both autonomously adapt its parameters and its structure via creation, fusion and scaling of the fuzzy system rules in an online BMI experiment with a real robot. The structure identification is based on an online GT2 Gath-Geva algorithm where every MI decoding class can be represented by multiple fuzzy rules (models), which are learnt in a continuous (trial-bytrial) non-iterative basis. The effectiveness of the proposed method is demonstrated in a detailed BMI experiment where 15 untrained users can accurately interface with a humanoid robot, in a single thirty-minute experiment, using signals from six EEG electrodes only.Index Terms-General type-2 fuzzy systems, online brain machine interfaces, motor-imagery brain machine interfaces, autonomous learning systems, adaptive learning, phase synchrony features, non-iterative learning.

Robust Interval Type-2 Fuzzy Sliding Mode Control Design for Robot Manipulators

“…sign(s)dt and u µ = −µω of the control law defined in (16) are substituted by their AIT2-FLSs, respectively:û…”

“…However, conventional type-1 fuzzy logic system (T1-FLS) cannot directly handle rule and measurement uncertainties because it uses T1-fuzzy sets (T1-FSs) that are certain. Therefore, these last years, an advanced form of FLS, called type-2 FLS (T2-FLS), has attracted considerable attention and becomes more and more imposed in designing robust controllers for uncertain complex processes, including robot systems [15][16][17][18]. One reason is that a T2-FS is characterized by a membership function (MF) that includes a footprint of uncertainty (FOU), which makes it possible to handle linguistic uncertainties more effectively than T1-FS [19][20][21].…”

General Type-2 Radial Basis Function Neural Network: A Data-Driven Fuzzy Model

“…Therefore, a GT2 Fuzzy Logic System has the potential to outperform not only the use of FLSs of T1, but also to provide a performance than an FLS with IT2 FSs cannot achieve [7]. Although GT2 FLSs are still in their infancy, the number of aplications of higher order fuzzy systems has experienced an important increase during the past five years [15], in particular in areas such as Pattern Recognition [12], [13], Automatic Control [2], [16], Image Processing [17] and Robotics [1], [3], [18]. In this applied context, the usage of GT2 FSs usually increases the computational complexity with respect to T1 and IT2 FLSs.…”

Adaptive Riemannian BCI for Enhanced Motor Imagery Training Protocols

“…This adaptive BCI was based off of quadratic discriminant analysis (QDA), updating the inverse covariance matrix for each class using data from a selected interval of each trial [3]. Subsequent attempts focused on adaptive feature selection, including frequency bands [4], covariate shift estimation [5], [6], and channel and spatial filter selection [7]- [9]. A more recent attempt used positive and negative feedback to change fuzzy rules for classification [10].…”