Frank H. F. Leung

Tuning of the structure and parameters of a neural network using an improved genetic algorithm

This paper presents the tuning of the structure and parameters of a neural network using an improved genetic algorithm (GA). It is also shown that the improved GA performs better than the standard GA based on some benchmark test functions. A neural network with switches introduced to its links is proposed. By doing this, the proposed neural network can learn both the input-output relationships of an application and the network structure using the improved GA. The number of hidden nodes is chosen manually by increasing it from a small number until the learning performance in terms of fitness value is good enough. Application examples on sunspot forecasting and associative memory are given to show the merits of the improved GA and the proposed neural network.

Hybrid Particle Swarm Optimization With Wavelet Mutation and Its Industrial Applications

A new hybrid particle swarm optimization (PSO) that incorporates a wavelet-theory-based mutation operation is proposed. It applies the wavelet theory to enhance the PSO in exploring the solution space more effectively for a better solution. A suite of benchmark test functions and three industrial applications (solving the load flow problems, modeling the development of fluid dispensing for electronic packaging, and designing a neural-network-based controller) are employed to evaluate the performance and the applicability of the proposed method. Experimental results empirically show that the proposed method significantly outperforms the existing methods in terms of convergence speed, solution quality, and solution stability.

Improved Hybrid Particle Swarm Optimized Wavelet Neural Network for Modeling the Development of Fluid Dispensing for Electronic Packaging

An improved hybrid particle swarm optimization (PSO)-based wavelet neural network (WNN) for modeling the development of fluid dispensing for electronic packaging (MFD-EP) is presented in this paper. In modeling the fluid dispensing process, it is important to understand the process behavior as well as determine the optimum operating conditions of the process for a high-yield, low-cost, and robust operation. Modeling the fluid dispensing process is a complex nonlinear problem. This kind of problem is suitable to be solved by applying a neural network. Among the different kinds of neural networks, the WNN is a good choice to solve the problem. In the proposed WNN, the translation parameters are variables depending on the network inputs. Due to the variable translation parameters, the network becomes an adaptive one that provides better performance and increased learning ability than conventional WNNs. An improved hybrid PSO is applied to train the parameters of the proposed WNN. The proposed hybrid PSO incorporates a wavelet-theory-based mutation operation. It applies the wavelet theory to enhance the PSO in more effectively exploring the solution space to reach a better solution. A case study of MFD-EP is employed to demonstrate the effectiveness of the proposed method.

A fuzzy sliding controller for nonlinear systems

It is well known that sliding-mode control can give good transient performance and system robustness. However, the presence of chattering may introduce problems to the actuators. Many chattering elimination methods use a finite DC gain controller which leads to a finite steady-state error. One method to ensure zero steady-state error is using a proportional plus integral (PI) controller. This paper proposes a fuzzy logic controller which combines a sliding-mode controller (SMC) and a PI controller. The advantages of the SMC and the PI controller can be combined and their disadvantages can be removed. The system stability is proved, although there is one more state variable to be considered in the PI subsystem. An illustrative example shows that good transient and steady-state responses can be obtained by applying the proposed controller.

Stability analysis of fuzzy control systems subject to uncertain grades of membership

This paper presents relaxed stability conditions for fuzzy control systems subject to parameter uncertainties. As the parameter uncertainties introduce uncertain grades of membership to the fuzzy control systems, the favorable property offered by sharing the same premises in the fuzzy plant models and fuzzy controllers cannot be employed to enhance the stabilization ability of the fuzzy control systems. To widen the applicability of the fuzzy control approach, fuzzy control systems subject to uncertain grades of membership will be investigated. New relaxed stability conditions will be derived to guarantee the stability of this class of fuzzy control systems. A numerical example will be given to show the effectiveness of the proposed approach.

Sampled-Data Fuzzy Controller for Time-Delay Nonlinear Systems: Fuzzy-Model-Based LMI Approach

This paper presents the stability analysis and performance design for a sampled-data fuzzy control system with time delay, which is formed by a nonlinear plant with time delay and a sampled-data fuzzy controller connected in a closed loop. As the sampled-data fuzzy controller can be implemented by a microcontroller or a digital computer, the implementation time and cost can be reduced. However, the sampling activity and time delay, which are potential causes of system instability, will complicate the system dynamics and make the stability analysis much more difficult than that for a pure continuous-time fuzzy control system. In this paper, a sampled-data fuzzy controller with enhanced nonlinearity compensation ability is proposed. Based on the fuzzy-model-based control approach, linear matrix inequality (LMI)-based stability conditions are derived to guarantee the system stability. By using a descriptor representation, the complexity of the sampled-data fuzzy control system with time delay can be reduced to ease the stability analysis, which effectively leads to a smaller number of LMI-stability conditions. Information of the membership functions of both the fuzzy plant model and fuzzy controller are considered, which allows arbitrary matrices to be introduced, to ease the satisfaction of the stability conditions. An application example will be given to show the merits and design procedure of the proposed approach. Furthermore, LMI-based performance conditions are derived to aid the design of a well-performed sampled-data fuzzy controller

The control of switching DC-DC converters-a general LWR problem

The control of switching DC-DC converters is reviewed. It is regarded as a general linear quadratic regulator (LQR) problem, and an innovative optimal and robust digital controller is proposed. The control strategy adopted can achieve good regulation, rejection of modest disturbances, and the ability to cater to switching converters with RHP zeros. This controller design is a general approach that is applicable to all PWM-type DC-DC converters with their circuit topologies known or unknown. Modern CAD techniques are used to reach the final control law. Application to a published Cuk converter is used as an example, and the performance is evaluated. >

A novel genetic-algorithm-based neural network for short-term load forecasting

This paper presents a neural network with a novel neuron model. In this model, the neuron has two activation functions and exhibits a node-to-node relationship in the hidden layer. This neural network provides better performance than a traditional feedforward neural network, and fewer hidden nodes are needed. The parameters of the proposed neural network are tuned by a genetic algorithm with arithmetic crossover and nonuniform mutation. Some applications are given to show the merits of the proposed neural network.

An Improved Genetic Algorithm with Average-bound Crossover and Wavelet Mutation Operations

This paper presents a real-coded genetic algorithm (RCGA) with new genetic operations (crossover and mutation). They are called the average-bound crossover and wavelet mutation. By introducing the proposed genetic operations, both the solution quality and stability are better than the RCGA with conventional genetic operations. A suite of benchmark test functions are used to evaluate the performance of the proposed algorithm. Application examples on economic load dispatch and tuning an associative-memory neural network are used to show the performance of the proposed RCGA.

Lyapunov-function-based design of fuzzy logic controllers and its application on combining controllers

This paper presents the design of fuzzy logic controllers (FLCs) for nonlinear systems with guaranteed closed-loop stability and its application on combining controllers. The design is based on heuristic fuzzy rules. Although each rule in the FLC refers to a stable closed-loop subsystem, the overall system stability cannot be guaranteed when all these rules are applied together. In this paper, it is proved that if each subsystem is stable in the sense of Lyapunov (ISL) under a common Lyapunov function, the overall system is also stable ISL. Since no fuzzy plant model is involved, the number of subsystems generated is relatively small, and the common Lyapunov function can be found more easily. To probe further, an application of this design approach to an inverted pendulum system that combines a sliding-mode controller (SMC) and a state feedback controller (SFC) is reported. Each rule in this FLC has an SMC or an SFC in the consequent part. The role of the FLC is to schedule the final control under different antecedents. The stability of the whole system is guaranteed by the proposed design approach. More importantly, the controller thus designed can keep the advantages and remove the disadvantages of the two conventional controllers.