Expert Systems for Process Control

ArchCam: Real time expert system for suspicious behaviour detection in ATM site

The highest level of control and supervision of industrial plants, for instance fossile or nuclear power generating systems or chemical production units, still has to be done by human operators. To support these operators, there is an increasing demand for real time and on-line expert systems.

Integration of a real time expert system and a modern graphic display system in a Swedish nuclear power plant control room

Fault prediction and fault compensation are beneficial for the production technology and give a new dimension to fault diagnosis in technical systems. The overall goal of this paper is the presentation of fault prediction and fault compensation procedures as they are studied, implemented and embedded in a real time expert system. This expert system detects and diagnoses faults in hydraulic systems. For this purpose dynamic modelling information, on-line sensor information, special features of the domain of hydraulic systems and expert systems technology are used co-operatively.

We propose a medical network based on a state-of-the-art medical kiosk that addresses the problems of providing preventive and diagnostic healthcare. This paper presents a web-centric diabetic expert system kiosk that is similar to a bank ATM. The patients can directly enter and can be screened for diabetes, with all the necessary tests that are fully automated. The tele diabetic expert system runs on a telehealth server and is connected to the kiosk through the world wide web. Based on the symptoms, tests taken and previous history of the patient, a prescription is generated by the expert system, which is also sent to a hospital where the doctor is online. Any changes required can be made by the physician in the diagnosis and prescription generated by the expert system and it is sent to the patient at the kiosk end as a printout. The whole system has been implemented and tested.

This paper introduces a new approach to the power system restoration problem. The continuous restoration means an online function that supports all the dispatcher actions. It is active and alerts not only in the restorative network state, but also in normal state. This real time expert system is strongly based on the cooperation with the SCADA-EMS functions and the different offline network calculations. The idea is proposed to be implemented in the Hungarian National Control Centre.

The object of the study is the process of operation of traction and external power supply systems as objects of inextricable interconnection while reducing energy costs in the cost of railway transportation in real time. One of the most problematic areas is the technology for choosing energy-efficient power supply schemes for railway traction networks in real time. The methods of forming and transforming graphs of complex schemes of traction and external power supply systems and building expert control systems for the implementation of energy-saving technologies of electrified railways were further developed in the work. In the course of the study, to increase the efficiency of simulation modeling of electric traction networks, the statistical characteristics of the loads of feeders that supply the final boundary sections, stations, depot access tracks, railway junctions and idle voltages on the traction substation tires were obtained. Methods of calculation and modeling of traction power supply have been developed, which take into account the inseparable relationship with power systems and allow choosing rational modes with minimal power flows and energy losses. Proposed methods of managing the modes of operation of the traction power supply system based on a vague description of their states and an expert system that allow solving new problems. Including, the choice of energy-saving power supply schemes in the case of power flows, economic modes of network operation in case of intensification of the transportation process. Thanks to this, ways of reducing power consumption and minimizing energy losses of traction power supply systems are proposed, which allow minimizing power flows and power losses by adjusting load flow parameters and voltage levels of traction substations. And also to increase the energy efficiency of electrified railway lines. The technique of technical and economic feasibility of power supply schemes of traction networks and evaluation of the possibility of switching to cantilever or loop power supply schemes with parallel connection points has been developed. The implemented recommendations save about 25 thousand kWh per 1 km of two-track section.

Automatic diagnostics and prognostics of energy conversion processes via knowledge-based systems

The issue addressed in this research is how to design and build an expert systemwhich capable of diagnosing dog diseases that are caused by parasites, and how this expertsystem later can be useful for dog owners.This expert system is developed based on website technology in order to ease theaccessibility for the users. The finished product is expected to become useful for dog owners,to help them diagnose their dog disease based on the symptoms in order to give the propertreatment. The software development methodology used to develop this expert system is theWaterfall methodology, while the design is done using Unified Modeling Language (UML).The programming language used to build the expert system is PHP for the logic and MySQLfor the database, and the finished product is then tested using black box testing method. Totest system accuracy, disease diagnosis results from the expert system is then compared withdiagnosis results from the real human expert (veterinarian).This expert system is using forward chaining method and Bayes theorem. The forwardchaining concept, in this system, starting with the symptoms that are set into a reasoning andrules to reach some conclusions which in this case are the possible diseases. While the Bayestheorem is used to draw a conclusion from the possible diseases by calculating the probabilityof what kind of disease suffered based on the symptoms, and the weight of the probabilityitself. This expert system, however, still can be developed not only to diagnose diseases thatare caused by parasites, but also to diagnose all kind of dog diseases.

This paper looks at expert systems in management, by using a business game as an experimental vehicle. An expert system called EXGAME was developed to play a business game, which is normally played by students, with minimal human intervention. This paper concentrates on the effectiveness of EXGAME as compared with human players for tasks at different levels. EXGAME was able to replace human players in decision making at the operational level, and indeed outperform them. However, it proved to be impractical to replace human input at the strategic level. The paper also sheds some light on the problems of trying to build an expert system when there is no real expert. A combination of a modular knowledge-base structure and a process of ‘learning by experimentation’ was effective in this case; it is suggested that this may be an appropriate development strategy in other similar situations. © 1998 John Wiley & Sons, Ltd.

Rapid development and deployment of bi-directional expert systems using machine translation technology

Preliminary results are presented of ongoing research to develop visualization paradigms for expert systems reasoning processes. The results are based in part on a prototype implementation that is being developed to visualize the reasoning processes of a rule-based forward chaining expert system. The research is based on the premise that the presentation of information at the highest applicable conceptual level will enhance the assimilation of that information. Hierarchical levels in both the syntactic and semantic levels of reasoning in expert systems are described in detail.

The general task of an expert system for real time monitoring and supervision consists of evaluating the quality of the current situation of a dynamic system and acting consequently, or proposing an action upon the system to an operator, so that its future behaviour leads to an acceptable situation. The particular functions of such an expert system are reduced to the monitoring and supervision processes. While some software packages allow the monitoring function to be made practical, real time supervision of dynamic systems arises as an attempt to aid the human operator in developing its functions, constituting the main way to guarantee the quality of the system response. Guaranteeing the constraints about the response time of the expert system demands adequate information in an optimal state of processing about the functions of the control system. A generic expert system model to implement the monitoring and supervision functions based on inductive learning techniques, together with a data structure to store information from the expert and the control systems, called evolutionary database for supervision, are presented.

Three previously proposed machine learning techniques (a technique based on applying gradient descent training to a fractional value expert system, the fractional value expert system technique’s real time hardware implementation, and a Boolean expert system hardware implementation) are evaluated for application to defense-relevant real time processing challenges. These techniques are defensible, meaning that their decisions are constrained by logical pathways that can be reviewed by a human prior to a decision being made and acted upon by the system (as opposed to simply explained afterwards). These techniques and several conventional techniques are evaluated under multiple defense-relevant real time processing scenarios.

Computer-aided software engineering of expert systems