Type-2 Fuzzy Neural Network Research Articles

Robust control for vibration control systems with dead-zone band and time delay under severe disturbance using adaptive fuzzy neural network

This study proposes a novel adaptive control method to deal with the dead-zone and time delay issues in actuators of vibration control systems. The controller is formulated based on a type-2 fuzzy neural network integrating with a new modification of Riccati-like equation. The developed new type Riccati-like equation is significant as it reduces energy consumption of control inputs to minimum. Two approaches are suggested to improve performance of the system using the basic elements of Riccati equation. In addition, a fuzzy neural network is applied to approximate the unmodeled dynamics and a sliding mode controller is developed to enhance the robustness of the system against uncertainties and disturbances. After proving the stability of the proposed controller via Lyapunov criterion, the effectiveness of the proposed approach is validated based on computer simulation for vibration control of a vehicle seat suspension. It is demonstrated that the unwanted vibrations due to external excitations are well controlled despite of the presence of dead-zone and time delay in actuators. Furthermore, when comparing with other two state-of-the-art robust controllers [23,33], the proposed controller provides better vibration suppression capacity and requires less energy consumption.

Improved local fisher discriminant analysis based dimensionality reduction for cancer disease prediction

A good dimensional reduction technique is needed to apply and improve the effectiveness of dimensionality reduction for medical data. High-dimensional data brings great challenges in terms of computational complexity and classification efficiency. It is necessary to compress effectively from high dimensional space to low dimensional space to design a learning curve with good performance. Therefore, dimensional reduction is necessary to study and understand the mechanism of the practical problems of medical data. In this paper, a hybrid local fisher discriminant analysis (HLFDA) method is proposed for the dimension reduction of the medical data. LFDA is a localized variant of Fisher discriminant analysis and it is popular for supervised dimensionality reduction method. The proposed HLFDA is a combination of Locality-preserving projection and LFDA. After the dimensionality reduction process, the data are given to the Type2fuzzy neural network classifier to classify a given data as normal or abnormal. The paper focused on improving performance in terms of prediction accuracy. Three types of UCI cancer dataset is used for analyzing the performance of the proposed method.

DC-link voltage control of three-phase PWM rectifier by using artificial bee colony based type-2 fuzzy neural network

The pulse width modulation (PWM) rectifiers are nonlinear systems due to semiconductor switches in their structure. Therefore, these rectifiers draw a distorted current from AC supply. Many different improvements have been proposed to overcome problems caused by PWM rectifiers. In this paper, DC-link voltage of three-phase PWM rectifier is regulated by using a Type-2 Fuzzy Neural Network (T2FNN) controller that parameters are optimized by using Artificial Bee Colony (ABC) optimization method. The parameters in antecedent and consequent parts of T2FNN are optimized by ABC optimization method. The performance of ABC-T2FNN controller is analyzed under different operating conditions through simulation model based on MATLAB. The operating conditions are considered as constant input, set point, a step DC load change, unbalanced AC supply and regenerative mode. The simulation results obtained from the proposed controller are verified by comparing with the results of the classical T2FNN. When the results of PWM rectifiers are investigated, it is seen that PWM rectifier based on the proposed controller has better dynamic response for all operating conditions than conventional T2FNN controller.

A low-cost integrated MEMS-based INS/GPS vehicle navigation system with challenging conditions based on an optimized IT2FNN in occluded environments

Integration of both global positioning system (GPS) and inertial navigation system (INS) assures a continuous and accurate navigation system. In low-cost low-precision micro-electromechanical system (MEMS)-based INS/GPS integration navigation systems, one of the major concerns is high-level stochastic noise and uncertainties existing in INS sensors and complex model of real noisy data. In such uncertainty-oriented environments, an intelligence structure with extra degrees of freedom which can handle and model a high-level of uncertainties in INS sensors, and an efficient denoising technique as a precursor to the intelligence structure can be efficient solutions. Our approach to these problems is taken in different steps. First, a denoising technique based on empirical mode decomposition (EMD) is used to provide more accurate INS sensor outputs and better generalization ability. Second, an optimized interval type-2 fuzzy neural network is used to model and handle a high-level of uncertainties efficiently and estimate the positioning error of INS sensors when GPS signals are blocked, and still meet both accuracy maximization and complexity minimization. Fast learning and convergence of the algorithm and less computational complexity can be achieved by using an extended Kalman filter in the learning of algorithm and an accurate and simple type-reduction, respectively, which can be utilized in real-time applications with significant performance. The results of EMD-based denoising technique, as a preprocessing phase, verify superior performance in comparison with the discrete wavelet transform denoising method in the signal-to-noise ratio improvement for raw and noisy signals of INS sensors. To verify the effectiveness of our proposed model, we applied challenging conditions consisting of low-cost low-precision inertial sensors based on MEMS technology, long-term outages of GPS satellites, a high-speed experimental test vehicle and noisy real-world data in the real-time flight experiments. The achieved experimental accuracies are compared with the results that we have achieved in other methods, and our proposed method verifies significant improvements.

Hierarchical Fused Model With Deep Learning and Type-2 Fuzzy Learning for Breast Cancer Diagnosis

Breast cancer diagnosis based on medical imaging necessitates both fine-grained lesion segmentation and disease grading. Although deep learning (DL) offers an emerging and powerful paradigm of feature learning for these two tasks, it is hampered from popularizing in practical application due to the lack of interpretability, generalization ability, and large labeled training sets. In this article, we propose a hierarchical fused model based on DL and fuzzy learning to overcome the drawbacks for pixelwise segmentation and disease grading of mammography breast images. The proposed system consists of a segmentation model (ResU-segNet) and a hierarchical fuzzy classifier (HFC) that is a fusion of interval type-2 possibilistic fuzzy c-means and fuzzy neural network. The ResU-segNet segments the masks of mass regions from the images through convolutional neural networks, while the HFC encodes the features from mass images and masks to obtain the disease grading through fuzzy representation and rule-based learning. Through the integration of feature extraction aided by domain knowledge and fuzzy learning, the system achieves favorable performance in a few-shot learning manner, and the deterioration of cross-dataset generalization ability is alleviated. In addition, the interpretability is further enhanced. The effectiveness of the proposed system is analyzed on the publicly available mammogram database of INbreast and a private database through cross-validation. Thorough comparative experiments are also conducted and demonstrated.

An experimental study on inverse adaptive neural fuzzy control for nonlinear systems

In this paper, a type-2 fuzzy neural network is used to adaptive inverse control of a class of nonlinear systems. The proposed method has proper performance for real-time control of nonlinear and time-invariant systems and has a quick response to following changes. In this method, the high approxim ation ability of type 2 fuzzy systems is used to reverse system modeling and this inverse model is used immediately to control the system. In order to measure the efficiency of the proposed method, the control of dynamic, highly nonlinear and time varying water both temperature system with certain and uncertain water pressure is experimentally considered. The results show the accuracy of the proposed method.

Energy management in photovoltaic battery hybrid systems: A novel type-2 fuzzy control

In this paper, a new robust fuzzy control approach is presented to power management in the photovoltaic (PV)-battery hybrid system. The stability and robustness of the developed control scheme is analyzed through the Lyapunov technique. The dynamics of the converters, battery and PV system are assumed to be fully unknown. Specifically, the uncertainties are online estimated by the adaptive interval type-2 fuzzy neural networks. It is shown that the implemented control technique results are in good performance in the presence of time-varying radiation, changes of temperature and output load variation in comparison with the other well-known control techniques.

Chaos analysis and stability control of the MEMS resonator via the type-2 sequential FNN

In this paper, a fuzzy stability control algorithm based on the type-2 sequential fuzzy neural network (T2SFNN) is proposed to suppress the chaotic motion of the MEMS resonator. The dynamical behaviors of the MEMS resonator are revealed by the bifurcation diagram, exponent diagram, phase diagram and corresponding time history. In the process of the controller design for backstepping control, the T2SFNN is used to tackle the unknown external disturbance and eliminate the effect of time delay, along with the cosine function to ensure that the output constraint is not violated. In addition, an absorber is employed to avoid the repeated derivative associated with the backstepping. Finally, the effectiveness of the proposed scheme is verified by numerical simulation.

Dynamical analysis and anti-oscillation-based adaptive control of the FO arch MEMS with optimality

This paper mainly investigates dynamical analysis and anti-oscillation-based adaptive control issues of the fractional-order (FO) arch microelectromechanical system (MEMS) with optimality. With the aid of phase diagram and Lyapunov exponent, the dynamical analysis reveals that chaotic oscillation originated in non-equilibrium interaction of potential energy can deteriorate the performance of the FO arch MEMS under different FOs and excitation amplitudes. During the design process, a recurrent non-singleton type-2 fuzzy neural network with a transformation is employed to approximate uncertain function and the Nussbaum function is used to tackle the problem of unknown control direction from actuation characteristic in the field of Caputo fractional derivative. Meanwhile, a tracking differentiator based on the hyperbolic sine function is designed to solve the repeated differentiation of virtual control because of complicated FO calculation. Then, an anti-oscillation-based adaptive control scheme fused with the Nussbaum function, neural network, tracking differentiator and optimal control is developed in the framework of backstepping by means of continuous frequency distributed model. The asymptotic stability of the proposed solution is proved based on the Lyapunov stability criterion. Finally, simulation results validate the feasibility of the presented scheme under multi-step methods of numerical solution.

Design and Verification of an Interval Type-2 Fuzzy Neural Network Based on Improved Particle Swarm Optimization

In this study, we proposed an interval type-2 fuzzy neural network (IT2FNN) based on an improved particle swarm optimization (PSO) method for prediction and control applications. The noise-suppressing ability of the proposed IT2FNN was superior to that of the traditional type-1 fuzzy neural network. We proposed dynamic group cooperative particle swarm optimization (DGCPSO) with superior local search ability to overcome the local optimum problem of traditional PSO. The proposed model and related algorithms were verified through the accuracy of prediction and wall-following control of a mobile robot. Supervised learning was used for prediction, and reinforcement learning was used to achieve wall-following control. The experimental results demonstrated that DGCPSO exhibited superior prediction and wall-following control.

EKF-Neural Network Observer Based Type-2 Fuzzy Control of Autonomous Vehicles

This paper proposes a novel robust path-following strategy for autonomous road vehicles based on type-2 fuzzy PID neural network (PIDT2FNN) method coupled to an Extended Kalman Filter-based Fuzzy Neural Network (EKFNN) observer. Uncertain Gaussian membership functions (MFs) are employed to self-adjust the universe of discourse for MFs using the adaptation mechanism derived from Lyapunov stability theory and Barbalat's lemma. External disturbances are significant in autonomous vehicles by changing the driving condition. Furthermore, parametric uncertainties related to the physical limits of tires and the change of the vehicle mass may significantly affect the desired performance of autonomous vehicles. The robustness of the proposed controller against the parametric uncertainties and external disturbances is compared with one active disturbance rejection control (ADRC) algorithm, and a linear-quadratic tracking (LQT) method. The obtained results in terms of the maximum error and root mean square error (RMSE), demonstrate the effectiveness of the proposed control algorithm to reach the minimized path-tracking error.

Sliding-Mode-Control-Theory-Based Adaptive General Type-2 Fuzzy Neural Network Control for Power-line Inspection Robots

In this paper, adaptive general type-2 fuzzy neural network control for motion balance adjusting of a power-line inspection robot is developed. It is used to enhance the anti-interference ability of the controlled plant. General type-2 fuzzy system is adopted because of its ability to more effectively handle uncertainties which may exist as external disturbances and parameter perturbations. The structure of general type-2 fuzzy system is designed by mimicking the neural network. The adaptive laws can be obtained based on the sliding mode control theory. This provides a kind of dynamic general type-2 fuzzy system whose membership functions and consequent parts are changing adaptively. The proposed controller is used to control an under-actuated non-linear power-line inspection (PLI) robot. Different simulation conditions are considered to test the anti-interference ability of the proposed controller. Simulation results indicate that the proposed method can strengthen the PLI robot’s anti-interference ability in a better way as compared to its interval type-2 fuzzy counterpart and PD controller.

Recurrent Interval Type-2 Fuzzy Wavelet Neural Network with Stable Learning Algorithm: Application to Model-Based Predictive Control

Fuzzy neural networks, with suitable learning strategy, have been demonstrated as an effective tool for online data modeling. However, it is a challenging task to construct a model to ensure its quality and stability for non-stationary dynamic systems with some uncertainties. To solve this problem, this paper presents a novel identification model based on recurrent interval type-2 fuzzy wavelet neural network (RIT2FWNN) with new learning algorithm. The model benefits from both advantages of recurrent and wavelet neural networks such as use of temporal data and fast convergence properties. The proposed antecedent and consequent parameters update rules are derived using sliding-mode-control-theory. To evaluate the proposed fuzzy model, it is utilized to design a nonlinear model-based predictive controller and is applied for the synchronization of fractional-order time-delay chaotic systems. Using Lyapunov stability analysis, it is shown that all update rules of the parameters are uniformly ultimately bounded. The adaptation laws obtained in this method are very simple and have closed forms. Some stability conditions are derived to prove learning dynamics and asymptotic stability of the network by using an appropriate Lyapunov function. The efficacy and performance of the proposed method is verified by simulation examples.

Prediction Intervals for Granular Data Streams Based on Evolving Type-2 Fuzzy Granular Neural Network Dynamic Ensemble

Granular data streams (GDSs) are a class of high-level abstract multitime scale description of data streams. Prediction intervals (PIs) for GDSs that provide estimated values as well as their corresponding reliability play an important role for assisting on-site workers to perceive the nonstationary environment in real time. However, constructing reliable PIs for GDSs constitutes a significant challenge. To provide a solution to the problem, an interval type-2 (IT2) fuzzy granular neural network (FGNN) dynamic ensemble approach (IT2FGNNDEnsemble) is proposed in this article. To fully reflect the uncertainty of GDSs, an interval value learning algorithm based IT2FGNN is developed, which can automatically generate, prune, merge, and realize recall in a single-pass learning mode. In addition, an evolving dynamic ensemble method is presented by providing an adaptive structure that considers a tradeoff between coverage and width of PIs, which can dynamically generate and prune the element of an ensemble according to current data tendency. A number of synthetic and industrial data streams experimentally validate the performance of the proposed IT2FGNNDEnsemble by using the state-of-the-art comparative methods. It is demonstrated that the proposed approach exhibits a good performance on PIs for practical applications.

Dynamical analysis of the fractional-order centrifugal flywheel governor system and its accelerated adaptive stabilization with the optimality

This paper investigates the issues of dynamical analysis and accelerated adaptive stabilization of the fractional-order (FO) centrifugal flywheel governor system with optimality. The dynamic behavior of the centrifugal flywheel governor system is revealed and its local stability is discussed in the fractional calculus (FC). A speed function is introduced to accelerate convergence rate within a pre-given time and a hierarchical type-2 fuzzy neural network (HT2FNN) is employed to play an approximating role for unknown nonlinear items. An extended state tracking differentiator which overcomes repetitive differentiation problem is used to approximate the derivative of virtual control input. Then, a stabilization controller is designed by integrating with the speed function, neural network and tracking differentiator in the framework of backstepping. It is proved that the proposed scheme guarantees the boundedness of all signals of the closed-loop system by using the frequency distributed model and makes the predefined cost function smallest. Finally, simulation results verify the effectiveness of the presented scheme.

Commercial Vehicle-Based Robust Control of Seated Whole-Body Vibration Using Adaptive Indirect Type-2 Fuzzy Neural Network

Drivers of commercial vehicles are invariably subject to chronic diseases such as back pain as a result of exposure to severe cabin excitations. In this paper, a six degrees of freedom (6-DOF) coupled human-body and seat suspension system is modeled in order to reduce the vibrations transmitted to the head, seat, and the relative seat and cabin floor displacement. The contributions of the present paper are: 1) two significant but inherently conflicting control objectives are employed, namely the seat acceleration and the relative displacement between seat and cabin floor to account for the effect of seat endstops, 2) A novel learning rate gradient descent based neural network approximator algorithm coupled to an adaptive indirect type-2 fuzzy neural network (T2FNN) controller to converge the controller to the ideal parameters of the uncertain model. 3) The controller model takes into account the seat suspension nonlinearities due to the nonlinear asymmetric piecewise damper and the cubic hardening of the suspension spring. The proposed controller employs the principle of type-2 fuzzy systems with interval membership function and unknown specifications. The effectiveness of the closed-loop system is validated regarding the uncertainties compared to observer-based sliding mode controller (SMC) and a high-fidelity virtual lab MSC.ADAMS-Simulink platform to validate the results in practical scenarios.

Sliding mode control of a class of nonlinear systems based on recurrent type-2 fuzzy RBFN

In this paper a type-2 fuzzy RBFN-based sliding mode control is used for a class of nonlinear systems. A new recurrent type-2 fuzzy radial basis neural network (we named RT2FRBFN) is used to approximate the conventional sliding mode control law. This method does not require prior information about the system; therefore, simultaneously, RT2FRBFN detects the system's dynamics, as well as the estimated dynamics in the sliding mode controller. Finally, the proposed RT2FRBFN sliding mode control system is used to tracking control design with regard to uncertainty in a class of nonlinear systems. Combination of backstepping method and sliding mode control helps to compensate the control signal and get a better performance. The backstepping method is used to improve the final threshold stability and use the sliding mode control to obtain high speed and unchangeable response to uncertainty. Simulation results show the proposed RT2FRBFN-based sliding mode control has better performance than RBFN-based sliding mode control in the presence of uncertainty.

Using an Interval Type-2 Fuzzy Neural Network and Tool Chips for Flank Wear Prediction

The precision of part machining is influenced by the tool life. Tools gradually wear out during the cutting process, which reduces the machining accuracy. Many studies have used machining parameters and sensor signals to predict flank wear; however, these methods have many limitations related to sensor installation, which is not only time-consuming and costly but also impractical in industry. This paper proposes an interval type-2 fuzzy neural network (IT2FNN) based on the dynamic-group cooperative differential evolution algorithm for flank wear prediction. Moreover, the Taguchi method is used to design cutting experiments for collecting experimental data and reducing the number of experiments. The CIE-xy color chromaticity values, spindle speed, feed per tooth, cutting depth, and cutting time are used as inputs of the IT2FNN, and the output is the flank wear value. The experimental results indicate that the proposed method can effectively predict flank wear with higher efficiency than other algorithms.

Fuzzy forecasting for long-term time series based on time-variant fuzzy information granules

As to the long-term time series forecasting, it is more challenging and practical to obtain the trend information and fluctuation range of sequence data than single-step prediction values. In this article, by means of fuzzy information granules (FIGs) and recurrent fuzzy neural networks, a novel long-term prediction model for time series is proposed. Based on a variable-length division method, generalized zonary time-variant fuzzy information granule (GZT-FIG) is constructed, which can express the variation trend, fluctuation range and dispersion degree of sequence data. Furthermore, in order to improve the anti-noise ability and memory ability, type-2 fuzzy sets and long short-term memory mechanism are introduced into the prediction scheme, based on which a self-evolving interval type-2 LSTM fuzzy neural network (eIT2FNN-LSTM) is provided. Compared with the existing works related to fuzzy-neural models, the involvement of long short-term memory mechanism effectively improves the memory ability to achieve the long-term prediction. In order to verify the validation and effectiveness of the proposed scheme, several groups of experiments, including synthetic sequences and real-life time series, are carried out. The experimental results reveal the better predictive performance and rich semantic information.

A novel fractional-order fuzzy control method based on immersion and invariance approach

In most of industrial applications, the dynamics of the system in hand are perturbed by a number of operational conditions. Also the outputs of the sensors always include noise. To alleviate these common problems, this paper presents a novel fuzzy control approach based on the immersion and invariance (I&I) approach under the conditions of unknown dynamics and measurement errors. The adaptation laws for the parameters of the proposed non-singleton type-2 fuzzy neural network (NT2FNN) are derived through a stability analysis based on I&I method. The effectiveness of the proposed membership function (MF) and non-singleton fuzzification is verified by comparison with the conventional Gaussian MF in the presence of measurement errors. The performance of the proposed control method is compared with other techniques and an experimental study is provided to show the capability of the proposed control scheme in real-time applications.