Hung-Ching Lu

Design of arbitrary FIR log filters by genetic algorithm approach

Abstract An efficient method is proposed for the design of finite-duration impulse response (FIR) digital filters with arbitrary log magnitude and phase responses. The method is based on an iterative weighted genetic algorithm (GA) approach to obtain a least-squares approximation to the given log magnitude function and the weighting function at each iteration is updated using the result of the previous iteration, which leads to the optimal approximation in the log Chebyshev sense. The method can be easily modified to approximate arbitrary log complex-valued frequency response such as complex FIR log filters. Several numerical design examples are shown to demonstrate the effectiveness of this proposed approach.

A PI-like fuzzy controller implementation for the inverted pendulum system

A PI-like fuzzy controller is designed and implemented for the inverted pendulum system. First, the proposed fuzzy controller adopts the step response phase trajectories to derive the corresponding linguistic control rules of such a system, then applies the relationship between input and output signals in a proportional integral controller to convert the relationship of linguistic control rules to a decision table. At the same time, the membership functions shape is mapped and tuned with constant proportional gain and constant integral gain. Finally, the fuzzy controller of the inverted pendulum system is accomplished by the designed decision table and the determined membership function. By using the Single Board Fuzzy Controller, made by Togai InfraLogic Inc., the proposed methods are proved.

2-D FIR filters design using least square error with scaling-free McClellan transformation

The optimal coefficients of the scaling-free McClellan transformation and the 1-D prototype filters for the design of 2-D digital filters are designed in this brief. These 2-D filters include fan filters with arbitrary inclination, elliptically symmetric filters, elliptically symmetric filters with arbitrary orientation, circular filters and diamond-shaped filters.

Genetic algorithm approach for designing higher-order digital differentiators

Abstract An effect approach is proposed for designing higher-order digital differentiators by the genetic algorithm (GA) method. By minimizing a quadratic measure of the error in the frequency band, appropriate crossover, mutation, and selection operations are used to get the filter coefficients. This method is not only simple and fast but also optimal in the least-squares sense. Comparison to the well-known McClellan–Parks algorithm for minimax equiripple filters shows that both are optimal in the sense of different minimum norms of the error function, but much better performance is obtained with the proposed approach in most of the frequency band except in the narrow-band region near the cutoff edge.

Adaptive self-constructing fuzzy neural network controller for hardware implementation of an inverted pendulum system

A tracking control of a real inverted pendulum system is implemented in this paper via an adaptive self-constructing fuzzy neural network (ASCFNN) controller. The linear induction motor (LIM) has many excellent performances, such as the silence, high-speed operation and high-starting thrust force, fewer losses and size of motion devices. Therefore, the experiment is implemented by integrating the LIM and an inverted pendulum (IP) system. The ASCFNN controller is composed of an ASCFNN identifier, a computation controller and a robust controller. The ASCFNN identifier is used to estimate parameters of the real IP system and the computational controller is used to sum up the outputs of the ASCFNN identifier. In order to compensate the uncertainties of the system parameters and achieve robust stability of the considered system, the robust controller is adopted. Furthermore, the structure and parameter learning are designed in the ASCFNN identifier to achieve favorable approximation performance. The Mahalanobis distance (M-distance) method in the structure learning is also employed to determine if the fuzzy rules are generated/eliminated or not. Concurrently, the adaptive laws are derived based on the sense of Lyapunov so that the stability of the system can be guaranteed. Finally, the simulation and the actual experiment are implemented to verify the effectiveness of the proposed ASCFNN controller.

Design of two-dimensional FIR digital filters for sampling structure conversion by genetic algorithm approach

Abstract By minimizing a quadratic measure of the error in the passband and stopband, a two-dimensional (2-D) diamond-shaped finite-duration impulse response (FIR) digital filter is designed for sampling structure conversion by genetic algorithm (GA) approach. This method is very easy to incorporate the frequency-domain constraints such that the design difficulty inherent to sampling conversion filters can be effectively solved. Several numerical design examples and simulation with test pictures are presented to demonstrate the effectiveness of this proposed approach.

Parameter estimation of fuzzy neural network controller based on a modified differential evolution

A tracking control of a nonlinear system is proposed in this paper via a fuzzy neural network (FNN) controller based on a modified differential evolution (MDE). The proposed modified differential evolution fuzzy neural network controller (MDEFNN) is composed of an FNN identifier, a hitting controller, a computation controller and a MDE estimator. First, the FNN identifier is used to estimate parameters of the nonlinear system. In order to compensate the uncertainties of the system parameters and achieve robust stability of the considered system, the hitting controller is adopted. The computation controller is used to sum up the outputs of the FNN identifier and hitting controller. Furthermore, there are two main learning phases in MDEFNN controller - the training phase and the online phase. In training phase, the mutation operation of the proposed MDE estimator according to fitness function effective produces a mutation vector. The MDE estimator is adopted to estimate the parameters of the MDEFNN controller. Therefore, there are several parameters such as the learning rates of the back-propagation (BP) algorithm, the parameters of error terms which are used in BP algorithm. The initial values of the FNN identifier and some preset parameters of MDEFNN controller can also be estimated by MDE estimator. After the best preset parameters are obtained, the nonlinear system is controlled by using MDEFNN controller. Further, the online parameter learning of the FNN identifier is based on the BP algorithm using error terms in the online phase. Finally, the simulation results are provided to demonstrate robustness, effectiveness and accurate tracking performance of the proposed MDEFNN controller under the conditions of external disturbance.

Robust CMAC control schemes for dynamic trajectory following

Abstract Improved robust CMAC control schemes are proposed for tracing dynamic trajectories in this paper. There are two main structures in the proposed control schemes: one is the robust controller and the other is the improved CMAC network. The robust controller technique can achieve a certain goal without concern for instability of the controlled system in the presence of significant plant uncertainties if the nominal parameter is roughly estimated. Next, in order to reduce the tracing error, a suitable nominal parameter needs to be chosen. Thus, the improved CMAC learning approach under the robust control structure, using the concept of credit assignment, will be employed to determine control variables that can trace other states repeatedly during control processes. Finally, simulation results demonstrate the capability of the proposed control schemes to trace dynamic trajectories.

Design and analysis of direct-action CMAC PID controller

This paper is to propose a direct-action (DA) cerebellar model articulation controller (CMAC) proportional-integral-derivative (PID) controller. The proposed controller, termed the DAC-PID controller, can generate four simple types of the nonlinear functions and then determine a control effort from those functions to control the process. In addition, the real-coded genetic algorithm is used to tune the parameters of the DAC-PID controller such that we can optimize those parameters. The performance of the proposed controller is also discussed in the sense of quantitative analysis. Simulation results demonstrate that the DAC-PID controller is superior to the conventional PID controller tuned by Ziegler-Nichols method and, moreover, as better as the optimal PID controller and the optimal fuzzy-PID controller.

Design and Implementation of a DSP-Based Grey-Fuzzy Controller for Induction Motor Drive

A DSP-based magnetic flux controlled induction motor drive with a grey-fuzzy controller is designed and implemented. This paper investigates the possibility of applying grey theory and fuzzy algorithm in a DSP-based induction motor controller, which requires faster and more accurate response, compared with a traditional fuzzy controller. The proposed controller is composed of two parts. The first one is a grey prediction controller, and the other is a general fuzzy logic controller. The experiments are performed by controlling the induction machines, and the whole system is executed by a digital signal processor (TMS320C30). The experimental results show that the better dynamic speed performances and torque regulation characteristics are obtained by the proposed controller.