Case-based reasoning assisted explanation of genetic algorithm results

Abstract

This paper describes a system for explaining solutions generated by genetic algorithms (GAs) using tools developed for case-based reasoning (CBR). In addition, this work empirically supports the building block hypothesis (BBH) which states that genetic algorithms work by combining good sub-solutions called building blocks into complete solutions. Since the space of possible building blocks and their combinations is extremely large, solutions found by GAs are often opaque and cannot be easily explained. Ironically, much of the knowledge required to explain such solutions is implicit in the processing done by the GA. Our system extracts and processes historical information from the GA by using knowledge acquisition and analysis tools developed for case-based reasoning. If properly analysed, the resulting knowledge base can be used: to shed light on the nature of the search space; to explain how a solution evolved; to discover its building blocks; and to justify why it works. Such knowledge about the search space can be used to tune the GA in various ways. As well as being a useful explanatory tool for GA researchers, our system serves as an empirical test of the building block hypothesis. The fact that it works so well lends credence to the theory that GAs work by exploiting common genetic building blocks.     

n jobs and m machines job-shop problems with sequence-dependent set-up times

This paper presents a mathematical model based on the branch-and-bound technique to solve static scheduling problems involving n jobs and m machines where the objective is to minimize the cost of setting up the machines. Set-up times are sequence dependent and not included in processing times. There is a finite non-zero cost associated with setting the machines which is different for each machine. It is further assumed that the routing may be different for different jobs and a job may return to a machine more than once.     

D.C. conductivity and spectroscopic studies of polyaniline doped with binary dopant ZrOCl<Subscript>2</Subscript>/AgI

Aqueous binary dopant (ZrOCl₂/AgI) is used in different ratios such as 1:1, 1:2 and 2:1 (w/w) for chemical doping to enhance the conductivity of synthesized polyaniline (PANI). The doping of polyaniline is carried out using tetrahydrofuran as a solvent. Doped samples are characterized using various techniques such as I–V characteristics, UV-visible spectroscopy, X-ray diffractometry (XRD), FTIR and photoluminescence (PL) studies. A significant enhancement in d.c. conductivity has been observed with the introduction of binary dopant. UV-visible study shows that optical parameters change considerably after doping. Interestingly, both direct and indirect bandgaps are observed in the doped samples. XRD patterns show the semi-crystalline nature of doped polyaniline. FTIR study shows structural modifications in functional groups with doping in PANI. Photoluminescence spectra exhibit emission properties of the samples.     

Energy modelling for the national economy — generalized model based on a physical systems theory approach

This paper presents an approach to the study of the energy related issues of the national economy using physical systems theory concepts for economic and real life systems. The main issues are: the efficiency of various energy conversion and utilization processes in terms of the macro-level technological coefficients; the direct and indirect energy wastes as calculated on the basis of technological coefficients; the effect of energy wastes on the unit cost of energy available to the consumers; the consumption of natural resources based on the final and intermediate demands of various sectors; and the amount of environmental pollution due to energy wastes disposed of to nature. A model is developed for a simplified two-sector national economy which is later extended to a generalized multi-sector model in a branchchord framework. This generalization is a step forward in simplifying the solution methodology for the multi-sector models through substitution, thus avoiding the necessity of drawing a complex system graph or solving large matrices required in the solution methodology.     

A ROBUST ARTIFICIAL NEURAL NETWORK TECHNIQUE FOR DYNAMIC STABILITY ASSESSMENT

ABSTRACT

This paper presents a robust artificial neural network technique to assess the on-line dynamic stability of power system. Efforts have been made to minimize the error in dynamic stability assessment by optimising the size and introducing more sensitive information in training vector. The problem has been attempted with Kohonen's self-organising feature map (SOFM) to classify the states of power system. The main reason which inspired the authors to apply Kohonen's SOFM technique is to avoid local minima that saturate the learning process in back-propagation algorithm. Conventional QR algrithm in conjunction with S-matrix method is used to allocate dynamic stability indices to output neurons.     

Simplification of physical system theory in the modelling of manufacturing,organizational and other socio-economic systems

The number of variables and equations associated with the system model of any large scale system are sufficiently large. In the case of models developed using physical system theory (PST), the number of variables and equations are twice the number of edges in the system graph. The system graph represents the interactions of discrete components united at a finite number of interfaces. The PST model of any real life manufacturing, organizational or socio-economic system has a large number of edges and equations. If, somehow, the number of variables can be reduced, the PST model can be simplified, thus increasing its practicability. An approach is presented to simplify the PST models of manufacturing, organizational and other socio-economic systems that exhibit some peculiar features. For the special case of forward flow systems (which are quite common in real life) the solution procedure is simplified considerably. The proposed methodology of simplification is illustrated with the example of a multi-stage manufacturing sytem.     

Use of the Lyapunov stability method for policy design in a class of system dynamics models

Policy design in system dynamics models is a complex issue. The use of the Lyapunov stability method, a very popular method in modern control theory, is highlighted in synthetically generating policy structures in a class of system dynamics models that aim at designing policies for stabilizing model response. The use of the method is demonstrated by applying it to a well-known production stock problem.     

Generalization of I-O-W model for waste management policy analysis and national planning

The generalization of the input-output-waste (I-O-W) model for different types of resources is presented in this paper as a new paradigm of national consistency modelling. This incorporates waste management (WM) policy analysis in a national planning context. The I-O-W model is an extension of the input-output model that formulates the waste aspects explicitly. The systematic analysis of intersectoral flow of different resources and wastes is important at the macro level for preparing resource plans that can be further integrated to frame overall economic plans. The paper briefly overviews the I-O-W model and generalizes it for five basic resources, i.e. material, energy, manpower, capital and services. The resource balance has been established by treating input to be equal to output plus waste for each resource independently as well as for the overall flow. The inputs required and the possible outputs for the generalized I-O-W model are listed. An illustrative example has been given that works out the direct, indirect and total energy. flows for a hypothetical economy. The scheme of integration is presented for overall WM policy analysis and national planning.     

Application of physical system theory and goal programming to modelling and analysis of waste management in national planning

Conventionally, national planning models have emphasized two major interfaces of the economic system, i.e. the input, and the output. The third important interface, i.e. waste, has not received the attention it deserves in the economic planning process. Attempts to include waste parameters in economic environmental modelling has been limited to tackling certain environmental considerations in national planning. The broader perspective of waste management (WM) has, by and large, been ignored in most of the literature in this context. This paper presents a graph theoretical system model of a multi-sectoral national economy including WM sectors. The system model of the economy has been integrated to a multi-objective goal programming model to fulfil the WM goals in a particular planning period. The model takes care of a variety of factors like resource structure, standard of living, sustained economic growth, waste recycling, environmental quality, etc. The model may be applied to determine the optimal quantity of waste to be generated in various sectors, and to assess optimal waste recycling efforts, along with the type of technology to be employed in various sectors. The application of the model has been illustrated by the example of a two sector economy having one production sector and one WM sector, and the sensitivity analysis has been carried out for alternate objective functions. Statistical analysis and the interpretation of the results are presented. The different variables and goals have been categorized as highly sensitive, moderately sensitive and less sensitive. The discussion of the results has been extended to show how the proposed model can be applied to real life national planning from a WM viewpoint.     

Modelling of quality in physical system theory: an extended framework

Quality is a vital concept that has important bearing on the functioning of any system. In order to control the quality effectively, it has to be systematically planned. Further, the quality in the system has a significant impact on the cost. The cost and flow dimensions of any manufacturing or socioeconomic system can be very well modelled and analysed in the framework of physical system theory (PST), which is based on the concept of a linear graph and deals with two categories of variables; an ‘across’ variable and a ‘through’ variable. The “across’ variables are governed by the law of compatibility, whereas the ‘through’ variables are governed by the law of continuity. The quality variables are not governed by the strict definitions of the “across’ and ‘through’ variables and the constraints imposed on them. Hence, in order to utilize the PST framework for modelling the quality variables in addition to the cost and flow variables, we attempt to develop an extended framework with quality as a third category of variables. An overview of the PST concepts is provided and the need for an extended framework is discussed. Then, the extended framework of the PST is presented with the modelling of basic processes, i.e. transformation, transportation and storage, in the developed framework with additional quality equations and functions. The proposed framework is illustrated with the simple example of a manufacturing system with three stages of production.