Shamshul Bahar Yaakob

Structural learning of the Boltzmann machine and its application to life cycle management

The objective of this research is to realise structural learning within a Boltzmann machine (BM), which enables the effective solution of problems defined as mixed integer quadratic programming. Simulation results show that computation time is reduced by up to one-fifth compared to conventional BMs. The computational efficiency of the resulting double-layer BM is approximately expressed as the ratio n divided by N, where n is the number of selected units (neurons/nodes) and N is the total number of units. The double-layer BM was applied to efficiently solve a mean-variance problem using mathematical programming with two objectives: the minimisation of risk and the maximisation of expected return. Finally, the effectiveness of our method is illustrated by way of a life cycle management example. The double-layer BM was able to more effectively select results with lower computational overhead. The results also enable us to more easily understand the internal structure of the BM. Using our proposed model, decision makers are able to select the best solution based on their risk preference from the alternative solutions provided by the proposed method.

A hybrid intelligent algorithm for solving the bilevel programming models

In this paper, genetic algorithm (GA) and neural network (NN) are integrated to produce a hybrid intelligent algorithm for solving the bilevel programming models. GA is used to select a set of potential combination from the entire generated combination. Then a meta-controlled Boltzmann machine which is formulated by comprising a Hopfield network and a Boltzmann machine (BM) is used to effectively and efficiently determine the optimal solution. The proposed method is used to solve the examples of bilevel programming for two- level investment in a new facility. The upper layer will decide the optimal company investment. The lower layer is used to decide the optimal department investment. Typical application examples are provided to illustrate the effectiveness and practicability of the hybrid intelligent algorithm.

Particle Swarm Optimization for Multi-function Worker Assignment Problem

A problem of worker assignment in cellular manufacturing (CM) environment is studied in this paper. The worker assignment problem is an NP-complete problem. In this paper, worker assignment method is modeled based on the principles of particle swarm optimization (PSO). PSO applies a collaborative population-based search, which models over the social behavior of fish schooling and bird flocking. PSO system combines local search method through self-experience with global search methods through neighboring experience, attempting to balance the exploration-exploitation trade-off which determines the efficiency and accuracy of an optimization. An effect of velocity controlled for the PSOs is newly included in this paper. We applied the adaptation and implementation of the PSO search strategy to the worker assignment problem. Typical application examples are also presented: the results demonstrate that the velocity information is an important factor for searching best solution and our method is a viable approach for the worker assignment problem.

Power System Equipments Investment Decision-Making under Uncertainty: A Real Options Approach

Power supply failures have caused major social losses in the information society in the present age. Such a loss is estimated up to approximately two trillion yen when a large power failure happens in a big city such as Metropolitan Tokyo. Therefore, it is necessary to provide some remedies such as a diagnosis of the power system equipments not only for preventing the system accident of the equipment beforehand from its failures but also for guarding the social cost from increasing. The objective of the paper is to provide the preliminary research on life cycle management. In this paper, net present value (NPV) analysis and real options approach (ROA) are employed in life cycle management in investment and maintenance of power supply systems in order to keep the continuous normal operation under uncertainty.

A genetic algorithm based double layer neural network for solving quadratic bilevel programming problem

In this paper, an intelligent genetic algorithm (IGA) and a double layer neural network (NN) are integrated into a hybrid intelligent algorithm for solving the quadratic bilevel programming problem. The intelligent genetic algorithm is used to select a set of potential solution combinations from the entire generated combinations of the upper level. Then a meta-controlled Boltzmann machine, which is formulated by comprising the Hopfield model (HM) and the Boltzmann machine (BM), is used to effectively and efficiently determine the optimal solution of the lower level. Numerical experiments on examples show that the genetic algorithm based double layer neural network enables us to efficiently and effectively solve quadratic bilevel programming problems.

Reliability enhancement of power systems through a mean–variance approach

Recently, power-supply failures have caused major social losses. Therefore, power-supply systems need to be highly reliable. The objective of this study is to present a significant and effective method of determining a productive investment to protect a power-supply system from damage. In this study, the reliability and risks of each of the units are evaluated with a variance–covariance matrix, and the effects and expenses of replacement are analyzed. The mean–variance analysis is formulated as a mathematical program with the following two objectives: (1) to minimize the risk and (2) to maximize the expected return. Finally, a structural learning model of a mutual connection neural network is proposed to solve problems defined by mixed-integer quadratic programming and is employed in the mean–variance analysis. Our method is applied to a power system network in the Tokyo Metropolitan area. This method enables us to select results more effectively and enhance decision making. In other words, decision-makers can select the investment rate and risk of each ward within a given total budget.

Double-Layered Hybrid Neural Network Approach for Solving Mixed Integer Quadratic Bilevel Problems

In this paper we build a double-layered hybrid neural network method to solve mixed integer quadratic bilevel programming problems. Bilevel programming problems arise when one optimization problem, the upper problem, is constrained by another optimization, the lower problem. In this paper, mixed integer quadratic bilevel programming problem is transformed into a double-layered hybrid neural network. We propose an efficient method for solving bilevel programming problems which employs a double-layered hybrid neural network. A two-layered neural network is formulate by comprising a Hopfield network, genetic algorithm, and a Boltzmann machine in order to effectively and efficiently select the limited number of units from those available. The Hopfield network and genetic algorithm are employed in the upper layer to select the limited number of units, and the Boltzmann machine is employed in the lower layer to decide the optimal solution/units from the limited number of units selected by the upper layer.The proposed method leads the mixed integer quadratic bilevel programming problem to a global optimal solution. To illustrate this approach, several numerical examples are solved and compared.

Structural learning approach to replacing unreliable units in a power system

Power supply failure will cause major social loss. Therefore, power supply systems have been required to be highly reliable. This research deals with a significant and effective method to decide the investment in a power system damage of the power supply on the society. In CMD2007 Korea, we proposed mean-variance approach to replacing unreliable units in a power system. This paper extends the method so as to solve a problem efficiently. In this research, we will propose the structural learning of a mutual connective neural network. The proposed method enables us to solve the problem defined in terms of mixed integer quadratic programming. In this research, an analysis is performed by using the concepts of the reliability and risks of units evaluated using a variance-covariance matrix and also the effect and expenses of replacement are measured. Mean-variance analysis is formulated as a mathematical programming with two objectives to minimize the risk and maximize the expected return. Finally, we employ a Boltzmann machine to solve the meanvariance analysis efficiently. The result of our method is exemplified using a power network system in Tokyo Metropolitan. By using this method, a more effective selection of results is obtained. In other words, the decision makers can select the expected investment rate and risk of each ward depending on the given total budget. For this reason, the effectiveness of the decision making process can be enhanced.

Analysis the Condition of RazakSAT Based on Attitude Sensor of Atttitude Determination System Data

Attitude determination system (ADS) is a process to control the orientation of the satellite to make sure that the orientation is relative to inertial references frame such as Earth. ADS is consists of mathematical algorithm and sensor as a references measurement to determine attitude of satellite. Since RazakSAT in the orbit, the sensor such as sun sensor, magnetometer and gyroscopes are used to control angular rotation and prevent high spin rates that can damage the satellite. This paper presents the analysis on sensor of attitude determination system based on RazakSAT data. Matlab was used to analyze the X,Y and Z axis of the sensor to determine the condition of the satellite.

Diagnosis System Based on Rough Sets Analysis

Nowadays, power systems play an important role in the whole electric industry. Failures of such systems should result in serious social and economical damages. Therefore, the power systems should be highly reliable. This paper presents the method to build a new type of failure diagnosis system based on rough set theory. The testing data of power systems for their failure conditions are based on experts’ evaluations with uncertainty, especially little knowledge and human experiences are available on power system failure diagnosis. The rough set theory plays a vital role in handling them.