Foun-Yuan Liu

Adaptive TSK fuzzy sliding mode control design for switched reluctance motor DTC drive systems with torque sensorless strategy

Abstract This study proposes a novel Adaptive Takagi-Sugeno-Kang (TSK) Fuzzy Sliding Mode controller (abbreviated as AFSC) and investigates the application for a switched reluctance motor (SRM) direct torque control (DTC) drive system without a torque sensor. The sliding mode controller (SMC) is used for reducing the influence of uncertainties and external disturbances, and it performs fast responses. The parameters of the adaptive TSK fuzzy controller (AFC) are adjusted online for further reducing error residues after applying the SMC. Lyapunov stability theory is used for deriving the stability condition of the SMC and the adaptive update law of the AFC. The stability of the overall closed-loop system is also analyzed. To verify the performance and practicality of the controller developed for this study, the AFSC is employed as the speed controller in a SRM DTC drive system. The experimental results reveal that the steady-state speed error is maintained between ±2 rpm when the motor load torque is 1 Nm, and the motor operates at low, medium, high, and variable speeds. Comparing with the conventional SMC, a faster and smoother speed and torque responses are achieved. Moreover, the proposed control strategy is superior to the conventional SMC with respect to robustness for external disturbances.

A Hop-Count Localization Method with Boundary Improvement for Wireless Sensor Networks

WSN localization methods play important roles in various applications, because many applications involve the need of localization. To propose an accurate localization method, this paper focuses on calculating the number of hops and utilizes the multidimensional scaling analysis to estimate the coordinates of unknown nodes. Then, the method of multi-power transmission is used to refine the coordinates of nodes that might have larger errors. Incorporating the above concepts and taking the degree of irregularity (DOI) level in different environments into consideration, we propose a method, called the boundary-improved amorphous with multi-power multidimensional scaling (BIA-MMS) localization method. Comparing the simulation results of BIA-MMS with the results using other localization methods, the BIA-MMS leads to a better improvement in localization accuracy.

Design on Sliding Mode Controller with Adaptive Fuzzy Compensation for Switched Reluctance Motor Drive Systems

In this study, the sliding mode controller (SMC) and the Takagi-Sugeno-Kang (TSK) fuzzy system were employed to design the proposed sliding mode controller with adaptive fuzzy compensation (AFSC). This design can improve the speed control performance of switched reluctance motor drive systems. Lyapunov function was derived to ensure the stability of the controller in the motor drive systems. To verify the performance and feasibility of the controller developed in this study, the AFSC was employed as the speed controller in the switched reluctance motor drive systems. The experiment results verify that the proposed AFSC control strategy can effectively improve the system dynamic response and achieve satisfactory robustness against external disturbances.

An intelligent motor rotary fault diagnosis system using Taguchi method

This paper applies the Taguchi method to filter out the number of input neurons and increases the training efficiency of the dynamic structural neural networks. In order to avoid that omitting the harmonics may affect the fault diagnosis result, this work establishes an index for the fault identification which is based on the features of the first and second harmonics. Together with the identification results of dynamic structural neural network, the diagnosis can be done. The experimental results indicate the proposed method can reduce the iterations dramatically.

An adaptive sliding self-organizing fuzzy controller for switched reluctance motor drive systems

This paper presents an adaptive sliding self-organizing fuzzy controller (ASSOFC) designed using fuzzy theory and a self-organizing algorithm. Composed of a conventional fuzzy controller (FC) and self-organizing algorithm, the ASSOFC adopts the sliding surface signal as an input, uses the algorithm to adjust the central position of the output consequent membership function of the FC, and, by fuzzy control, regulates the learning rate and fuzzy rules in real time to improve control performance. The ASSOFC is embedded into the direct torque control system of a switched reluctance motor (SRM) as a speed controller, and the performance and feasibility of the controller were validated. The experimental results indicate that the root mean square error values for the ASSOFC at various speed ranges are lower than those for a conventional FC, indicating that the proposed controller provides a superior speed response for SRMs.

Applications on Adaptive Recurrent Cerebellar Model Articulation Controller for Switched Reluctance Motor Drive Systems

This study proposed an adaptive recurrent cerebellar model articulation controller (ARCMAC) which consisted of a recurrent Gaussian cerebellar model articulation controller (RCMAC) and a compensation controller. The ARCMAC was used as the speed controller for a switched reluctance motor (SRM), and the contained compensation controller was used to offset the error between the RCMAC and a theoretically idea control. Lyapunov theory was used in this study to derive the update laws of the weights of the ARCMAC, the weights of recurrence, and the parameters of Gaussian functions to ensure the stability of controlled system.

Fuzzy Inference of Excitation Angle for Direct Torque-Controlled Switched Reluctance Motor Drives

This study proposed a fuzzy excitation controller to reduce noise and torque ripples for switched reluctance motors. The design of the controller is simple and can generate appropriate turn-on and turn-off angles according to the speed of torque error in order to improve the torque response. At the motor speed of 300 rpm with one Nm load, the experimental results showed that the implantation of a fuzzy excitation controller in the driver system substantially improved the steady state torque ripples generated when compared to those generated by traditional controllers (at fixed excitation angles), especially at low speeds.

Design of adaptive Takagi-Sugeno-Kang fuzzy estimators for induction motor direct torque control systems

By referencing the adaptive stator flux estimator (ASFE) framework in the model reference adaptive system (MRAS), this study designed an adaptive rotor speed estimator and a stator resistance estimator, and applied the Takagi-Sugeno-Kang (TSK) fuzzy system and projection algorithms to the estimators to establish an induction motor direct torque controlled system without a speed sensor and possessing stator resistance adjustment abilities. In addition, the adaptive TSK fuzzy controller (ATSKFC) was adopted as the speed controller of the system, and was capable of online learning. The transient response was improved by the integration of a refined compensation controller. Induction motor controlled drive system was implemented in this study by using direct torque control (DTC) technology, which had the advantages of a rapid dynamic response, simple system structure, and low computational complexity. In addition, the application of the voltage space vector pulse width modulation (VSVPWM) technique reduced the torque ripples and noise, which are common in a traditional DTC system. The simulation and experimental results demonstrated that, with the proposed adaptive TSK fuzzy rotor speed estimator (ATSKFRSE), adaptive TSK fuzzy stator resistance estimator (ATSKFSRE), and an ATSKFC implanted into the induction motor DTC system, the system provided an excellent speed dynamic response and was able to estimate the rotor speed and stator resistance accurately at an 8-Nm load torque and a wide speed range of 36-2000 rpm.