CORES-the continuous restoration expert system

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.

Specifying fault tolerance in mission critical intelligent systems

A real-time expert system is being developed using 0PS83 rule based programming system for monitoring and analysis, diagnosis of faults and malfunctions, and on-line optimization of fermentation processes. The main purpose of the expert system is to assist the operators in running and improving fermentation conditions with minimum amount of supervision. The expert system consists of a data base which includes physical attributes of the process (color, and general appearance of the fermentation broth) and includes operator/user given off-line data like microscopic states of the microorganisms. Optimization and estimation algorithms can be interfaced with the expert system. These can be used in conjunction with the fault detection capabilities of the expert system for making intelligent decisions about the abnormalities in fermentation experiments, and rule out contradictory sensor information.

This paper uses two modeling tools to analyze the reliability of real-time expert systems: (1) a stochastic Petri net (SPN) for computing the conditional response time distribution given that a fixed number of expert system match-select-act cycles are executed, and (2) a simulation search tree for computing the distribution of expert system match-select-act cycles for formulating a control strategy in response to external events. By modeling the intrinsic match-select-act cycle of expert systems and associating rewards rates with markings of the SPN, the response time distribution for the expert system to reach a decision can be computed as a function of design parameters, thereby facilitating the assessment of reliability of expert systems in the presence of real-time constraints. The utility of the reliability model is illustrated with an expert system characterized by a set of design conditions under a real-time constraint. This reliability model allows the system designers to: (1) experiment with a range of selected parameter values; and (2) observe their effects on system reliability.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A vehicle management system (VMS) research and development project is underway at McDonnel Aircraft to integrate control of systems on combat aircraft and enhance diagnostics. A VMS demonstrator has been constructed to test VMS methodologies and architectures. Two real-time expert systems are under development to diagnose electrical and fuel system failures on the VMS demonstrator: an electrical expert system, located on a Sun Microsystems SPARCstation 1 that acquires data from a centralized computer; and a fuel expert system, located in the control computers using a shared memory allocation with the control software. Both expert systems used the CLIPS inference engine developed by NASA. Future enhancements to the two expert systems will include more use of probability in determining failures and testing other versions of the CLIPS inference engine.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A real-time expert system has recently been installed at British Steel Sections, Plates and Commercial Steels, Scunthorpe Works. The system is at the bloom and billet mill and has been designed to allow engineers to add and modify rules through a generic graphical rule editor. The mill has 26 soaking pits which reheat the steel prior to rolling. The expert system covers the soaking pits and the gas mixing station. It currently diagnoses operational problems with the soaking pits and generates messages which are displayed to the process operator via an existing message screen. Combustion and heat balance information is also calculated by the system for each pit every five minutes. Prior to installing the expert system, the calculations were performed manually for individual pits on an irregular basis. The on-line calculations are to be used to derive new diagnostic rules and provide supporting evidence to existing rules. (2 pages)

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

In this paper an expert system is used as a part of a feedback controller which consists of algorithms and logic. The logic is often the dominating part, particularly when different control functions are combined. This logic can be implemented in an expert system. The task of the expert system is to orchestrate the application of different numerical algorithms to the controlled plant in real time. This problem contains elements of both monitoring and planning. A prototype implementation is described and experiments with this system are evaluated. A real time expert system architecture is outlined based on experiences from the experiments.

The Gensym real-time expert system G2 integrates dynamic models and rules to combine analytic and heuristic knowledge representation. The inference engine can reason about object connectivity and can use generic forms of knowledge, interpreted for specific instances in the domain. Designed for process control and related applications, G2 allows the process engineer to implement and manage the expert system. A network architecture allows hierarchical or co-operating expert systems in the plant.The first Expert Control System using G2 is described. The Reliable Water Company in Boston has implemented G2 for control and on-line diagnostics for a water purification process. The system was first demonstrated in June 1987, in prototype form, purifying water from the Boston Harbor. Delivery versions are now being deployed.

A Hybrid Expert System, GIS, and Simulation Modeling for Environmental and Technological Risk Management

To meet the ever-increasing demand of the rapidly-growing semiconductor manufacturing industry it is critical to have a comprehensive methodology integrating techniques for process optimization real-time monitoring and adaptive process control. To this end we have accomplished an integrated knowledge-based approach combining latest expert system technology machine learning method and traditional statistical process control (SPC) techniques. This knowledge-based approach is advantageous in that it makes it possible for the task of process optimization and adaptive control to be performed consistently and predictably. Furthermore this approach can be used to construct high-level and qualitative description of processes and thus make the process behavior easy to monitor predict and control. Two software packages RIST (Rule Induction and Statistical Testing) and KARSM (Knowledge Acquisition from Response Surface Methodology) have been developed and incorporated with two commercially available packages G2 (real-time expert system) and ULTRAMAX (a tool for sequential process optimization).© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Algebraic formulae for solving systems of max-min inverse fuzzy relational equations

Use of an expert system shell to develop a power plant simulator for monitoring and fault diagnosis

A method for control employing rule-based search is reviewed, and a Rule-Based Controller achieving economical real-time performance is described. Code optimization, in the form of LISP-to-Pascal knowledge base translation, provides real-time search execution speed and a processing environment enabling highly integrated symbolic and numeric computation. With a multiprocessor software architecture specifying rule-based protocol for control task communication, and a hardware architecture providing concurrent implementation within a multi-microprocessor system, the controller realizes a set of cooperating real-time expert systems. Based on experience gained through the design and implementation of a Rule-Based Flight Control System, the proposed approach appears applicable to a large class of complex control problems.