This thesis investigates the application of Knowledge Based systems and Fuzzy Logic to automatic control. The knowledge used by a human operator is put in a computer usable form and applied to a control problem. The idea is not to attempt to enhance the stability or response of the system but given a basically stable and controllable system we apply human type control methods via a computer controller. A system can never be modelled exactly and therefore a controller design must allow for the uncertainty in the model. With fuzzy logic, the system inputs, outputs, parameters, reactions and cross coupling are represented in fuzzy or inexact variables, knowledge and reasoning. An exact (or nearly exact) model of the system is not necessary. A simple aircraft is the process to which this control method is applied. Knowledge, reasoning and feedback similar to what a human pilot utilizes are applied in the control of the process. / Thesis / Master of Engineering (ME)

International Telemetering Conference Proceedings / October 28-31, 1996 / Town and Country Hotel and Convention Center, San Diego, California / We report on the initial design and development of a prototype computer-mediated response system, the Fault Induced Document Officer (FIDO), at the UC Berkeley Center for EUV Astrophysics (CEA) Extreme Ultraviolet Explorer project (EUVE). Typical 24x7 staffed spacecraft operations use highly skilled expert teams to monitor current ground systems and spacecraft state for responding to anomalous ground system and spacecraft conditions. Response to ground system error messages and spacecraft anomalies is based on knowledge of nominal component behavior and the evaluation of relevant telemetry by the team. This type of human-mediated operation is being replaced by an intelligent software system to reduce costs and to increase performance and reliability. FIDO is a prototype software application that will provide automated retrieval and display of documentation for operations staff. Initially, FIDO will be applied for ground systems. Later implementations of FIDO will target spacecraft systems. FIDO is intended to provide system state summary, links to relevant documentation, and suggestions for operator responses to error messages. FIDO will provide the operator with near realtime expert assistance and access to necessary information. This configuration should allow the resolution of many anomalies without the need for on-site intervention by a skilled controller or expert.

The technology of knowledge-based systems undoubtedly offers potential for educational modelling, yet its practical impact on today's school classrooms is very limited. To an extent this is because the tools presently used in schools are EMYCIN -type expert system shells. The main argument of this thesis is that these shells make knowledge-based modelling unnecessarily difficult and that tools which exploit knowledge acquisition technologies empower learners to build better models. We describe how such tools can be designed. To evaluate their usability a model-building course was conducted in five secondary schools. During the course pupils built hundreds of models in a common range of domains. Some of the models were built with an EMYCIN -type shell whilst others were built with a variety of knowledge acquisition systems. The knowledge acquisition systems emerged as superior in important respects. We offer some explanations for these results and argue that although problems remain, such as in teacher education, design of classroom practice, and assessment of learning outcomes, it is clear that knowledge acquisition systems offer considerable potential to develop improved forms of educational knowledge-based modelling.

The recent downturn in world sugar prices has placed even greater demands upon the Australian sugar industry to reduce the costs of sugar manufacture and increase the consistency of producing high quality sugar. One of the proposed approaches in increasing the consistency of very high quality sugar production and leveraging further avenues for cost saving is in the development of a computer based advisory system. This system is able to provide expert knowledge in the area of pan stage management and best practices in the absence of human experts. This thesis explores the design, key features and outcomes of a knowledge based supervisory support system (KBSSS) framework proposed specifically for providing cooperative decision support in the area of pan stage operations within a sugar mill. To demonstrate the viability of the proposed KBSSS framework a prototype system was developed in accordance with the proposed framework. The KBSSS utilises three core innovative system technologies that form the core components of the proposed KBSSS framework. These technologies are: 1) Dynamic industrial pan stage process models for identifying the dynamic relationships between sections of pan stage operations to allow for future forecasting of pan stage operating conditions, 2) Integration techniques for the merging of the developed pan stage process models into the hybrid fuzzy logic expert system rule base to provide localisation adjustment to match with local real world factory operational conditions, and 3) Explanatory capabilities to provide justification and support of system advice and recommendations. As a result of research and development carried out in this thesis, the KBSSS's test results demonstrated in the thesis indicate the viability of the proposed KBSSS framework and highlight the forecasting capabilities of the developed system resulting in favourable outcomes compared to data from pan stage operations. As a result of the research undertaken in the thesis a prototype KBSSS, for pan stage operations, based upon the three core supporting intelligent system technologies reported in the thesis has been developed.

An Extension to the Composite Rule Induction System Discovering knowledge from data is an important task for knowledge management and development of intelligent systems, which is called knowledge acquisition or data mining. Many techniques have been developed for such purpose. For example, ID3, C4.5 (tree induction techniques) and Artificial Neural Networks are among the popular techniques in ¡§Classification and Prediction¡¨ area. However, these methods often use the same criteria to analyze nominal and non-nominal attributes, which is very likely to produce biased knowledge due to mis-match between data type and their algorithms. In Liang (1992), he proposed a composite approach called CRIS to inducing knowledge that introduces statistical concepts and data mining heuristics and found the composite method outperformed other methods including tree induction, discriminant analysis, and neural networks. However, the paper focuses on the classification of binary objects and did not describe how the approach can be applied to a problem with more than two classes in the dependent variable. In this research, we extend the previous approach to solve the problem with more than two classes. We also enhance the approach by adding steps to prioritizing attributes using their identification power and controlling the growth of generated hypothesis. In order evaluate the extended CRIS method, a prototype system, eCRIS, was developed and compared with a commercial data mining package, XLMiner3 (developed by Cytel Software Corporation) using three existing datasets in data mining research. The results indicate that the extended CRIS outperforms tree induction and backpropagation in neural networks in datasets that include both nominal and non-nominal data and performed equally well with them.

No description available.

Supply Chain Management (SCM) is becoming increasingly important in the modern business world. In order to effectively manage and integrate a supply chain (SC), a deep understanding of overall SC operation dynamics is needed. This involves understanding how the decisions, actions and interactions between SC members affect each other, and how these relate to SC performance and SC disruptions. Achieving such an understanding is not an easy task, given the complex and dynamic nature of supply chains. Existing simulation approaches do not provide an explanation of simulation results, while related work on SC disruption analysis studies SC disruptions separately from SC operation and performance. This thesis presents a logic-based approach for modelling, simulating and explaining SC operation that fills these gaps. SC members are modelled as logicbased intelligent agents consisting of a reasoning layer, represented through business rules, a process layer, represented through business processes and a communication layer, represented through communicative actions. The SC operation model is declaratively formalised, and a rule-based specification is provided for the execution semantics of the formal model, thus driving the simulation of SC operation. The choice of a logic-based approach enables the automated generation of explanations about simulated behaviours. SC disruptions are included in the SC operation model, and a causal model is defined, capturing relationships between different types of SC disruptions and low SC performance. This way, explanations can be generated on causal relationships between occurred SC disruptions and low SC performance. This approach was analytically and empirically evaluated with the participation of SCM and business experts. The results indicate the following: Firstly, the approach is useful, as it allows for higher efficiency, correctness and certainty about explanations of SC operation compared to the case of no automated explanation support. Secondly, it improves the understanding of the domain for non-SCM experts with respect to their correctness and efficiency; the correctness improvement is significantly higher compared to the case of no prior explanation system use, without loss of efficiency. Thirdly, the logic-based approach allows for maintainability and reusability with respect to the specification of SC operation input models, the developed simulation system and the developed explanation system.

M.A. (Business Information Technology) / Knowledge management has developed greatly over the last few decades, particularly in striving for economic and commercial effectiveness. With the growth of technology-based knowledge management systems and an increase in the number of organisations implementing them, there is concern as to how these systems are being accepted by knowledge workers. The systems are currently a pertinent issue on business agendas, and organisations across all sectors are recognising the critical role that effective ~ ones will play in their future success (Malhotra, 2000:56). This creates a concern where these systems are expected to bring success in organisations or to improve return on investments without a deeper understanding of their utilisation by knowledge workers. The goal of this dissertation is to understand the way in which technology-based knowledge management systems are being utilised and accepted by knowledge workers, and furthermore to establish if knowledge workers have full understanding of the systems they use. The focus of this research is therefore on the human dimensions in relation to the systems, and on how they support organisational intellectual capital. The history of interactive computlnq shows repeatedly that it is the human issues which make or break new methods and tools at work. What are technology-based knowledge' management systems? How are they managed? Moreover, how can managers harness the potential of the knowledge workers to expand the knowledge base of the organisation? In order to answer these questions, this research determines how knowledge workers utilise the systems and their level of acceptance of this technology. It considers the importance of organisations that want their employees to use the systems effectively by contributing ideas and knowledge out of their own goodwill. Employees will do so if the concept of trust (of any technology system) has been imparted to them fully.

Knowledge refinement tools seek to correct faulty rule-based systems by identifying and repairing faults indicated by training examples that provide evidence of faults. This thesis proposes mechanisms that improve the effectiveness and efficiency of refinement tools by the best use and selection of training examples. The refinement task is sufficiently complex that the space of possible refinements demands a heuristic search. Refinement tools typically use hill-climbing search to identify suitable repairs but run the risk of getting caught in local optima. A novel contribution of this thesis is solving the local optima problem by converting the hill-climbing search into a best-first search that can backtrack to previous refinement states. The thesis explores how different backtracking heuristics and training example ordering heuristics affect refinement effectiveness and efficiency. Refinement tools rely on a representative set of training examples to identify faults and influence repair choices. In real environments it is often difficult to obtain a large set of training examples, since each problem-solving task must be labelled with the expert's solution. Another novel aspect introduced in this thesis is informed selection of examples for knowledge refinement, where suitable examples are selected from a set of unlabelled examples, so that only the subset requires to be labelled. Conversely, if a large set of labelled examples is available, it still makes sense to have mechanisms that can select a representative set of examples beneficial for the refinement task, thereby avoiding unnecessary example processing costs. Finally, an experimental evaluation of example utilisation and selection strategies on two artificial domains and one real application are presented. Informed backtracking is able to effectively deal with local optima by moving search to more promising areas, while informed ordering of training examples reduces search effort by ensuring that more pressing faults are dealt with early on in the search. Additionally, example selection methods achieve similar refinement accuracy with significantly fewer examples.

Yes / We present a technique for automatically acquiring operational knowledge on how to adapt assembly systems to new production demands or recover from disruptions. Dealing with changes and disruptions affecting an assembly station is a complex process which requires deep knowledge of the assembly process, the product being assembled and the adopted technologies. Shop-floor operators typically perform a series of adjustments by trial and error until the expected results in terms of performance and quality are achieved. With the proposed approach, such adjustments are captured and their effect on the station is measured. Adaptation knowledge is then derived by generalising from individual cases using a variant of the k-nearest neighbour algorithm. The operator is informed about potential adaptations whenever the station enters a state similar to one contained in the experience base, that is, a state on which adaptation information has been captured. A case study is presented, showing how the technique enables to reduce adaptation times. The general system architecture in which the technique has been implemented is described, including the role of the different software components and their interactions.