N. Kiupel

Fuzzy supervision and application to lean production

Abstract A novel philosophy of process supervision based on functional redundancy, i.e., analytical or knowledge based redundancy which may specifically be used for lean production, is suggested. The key idea is to replace the conventional residual evaluator of the fault diagnosis system based on crisp logic, by both a decision maker with fuzzy logic for residual pre-evaluation and the human operator to make the final decisions using his natural intelligence, experience and common sense. The purpose of the employment of fuzzy logic for residual pre-evaluation is to release only weighted alarms instead of yes-no decisions, so that (by definition) no false alarms can be produced; besides this, the man-machine interaction becomes much easier. In contrast to the conventional expert system approach, the proposed concept leaves the final yes-no decisions to the natural intelligence, capability and responsibility of the human operator which are still superior to the artificial intelligence and decision making ca...

Process supervision with the aid of fuzzy logic

In this paper we suggest a novel philosophy of process supervision based on functional redundancy: analytical or knowledge based redundancy. The key idea is to replace the conventional residual evaluator of the fault diagnosis system based on crisp logic, by both a decision maker with fuzzy logic for residual pre-evaluation and the human operator to make the final decisions using his natural intelligence, experience and common sense. The purpose of the employment of fuzzy logic for residual pre-evaluation is to release only weighted alarms instead of yes-no decisions, so that (by definition) no false alarms can be produced; besides this the man-machine interaction becomes much easier. In contrast to the conventional expert system approach, the proposed concept leaves the final yes-no decisions to the natural intelligence, capability and responsibility of the human operator which are still superior to the artificial intelligence and decision making capabilities of an expert system.<<ETX>>

A Fuzzy FDI Decision Making System for the Support of the Human Operator

Abstract The goal of Fault Detection and Isolation (FDI) is to decide whether and where a fault in the system under consideration has occurred avoiding wrong decisions that cause false alarms. To achieve a fault detection scheme which is robust in the sense of false alarms a combined quantitative/qualitative supervision system will be used to detect and isolate faults. The quantitative part will be used to generate fault symptoms (residuals) using a quantitative (mathematical) model of the process. These residuals contain the information whether a fault has occured or not. The next step in the FDI process is the residual evaluation . There exists a number of different residual evaluation techniques, for example simple threshold logic tests, statistical decision making, pattern recognition, decision making based on fuzzy logic or neural networks. The fundamental difficulty with residual evaluation is that residuals are normally uncertain, corrupted by noise, disturbances and, if the residuals are generated by model-based techniques, by modelling uncertainties. In order to select from the given residual data the important fault information a human support tool for the generation of a knowledge base for fault diagnosis will be presented in this paper.

Improvement of Mold–Level Control using Fuzzy–Logic

Abstract In this paper we report on a feasibility study of an Improvement of conventional mold-level control using fuzzy logic. For an economic steel production nowadays continuous casting is used. Of special interest for continuous casting is the mold, because in this part of the process the liquid steel will be fed to the mold. From the metallurgical point of view it is necessary to keep the mold-level constant. Greater variations of the mold-level cause non-homogeneity in the product. Especially an overflow of the mold as well as an empty mold has to be avoided, because the liquid steel overflows the working area or the bottom parts of the mold respectively. Beside an interrupt in the production, in addition, high maintenance costs will be caused. At the example of a slab-caster with varying mold geometry it will be shown in this article that in case of disturbances an improvement of the conventional control, using fuzzy logic, can be achieved. The basic idea is the use of a parallel control structure with fuzzy logic. This provides the advantage that under normal operation conditions the conventional controller is still in operation, whilst in the case of a disturbance the fuzzy controller supports the conventional control. This leads to a much better disturbance behaviour as without the fuzzy controller, that means, that in opposite to conventional control especially an overflow can be avoided.

A fault diagnosis concept for a high-pressure-preheater line

This contribution describes a combined analytical/fuzzy model-based fault diagnosis concept which has been applied to the high-pressure-preheater line of a power plant. The key idea is to divide the whole system under supervision into several subsystems and to employ all available analytical knowledge to generate residuals for each subsystem seperately. Thereby the necessary observers for fault detection are of a handable size. For fault isolation, the residual evaluation is done by applying qualitative knowledge about the fault effects on the residuals and about the interaction of the different subsystems. Here a rule base is evaluated using fuzzy logic. With this method a complex system can be supervised without the need for a complex analytical model of the whole system. Furthermore, the presentation of the fault isolation results leaves the final decision about a fault alarm to the human operator. Results from a power plant prove the successful application of the proposed supervision concept.

Improvement of mold-level control using fuzzy logic

Abstract This paper reports on a feasibility study on an improvement of conventional mold-level control using fuzzy logic. Nowadays, for economic steel production, continuous casting is used. Of special interest for continuous casting is the mold, because in this part of the process the liquid steel will be fed to the mold. From the metallurgical point of view it is necessary to keep the mold-level constant, as large variations of the mold-level cause non-homogeneity in the product. Especially, an overflow of the mold (as well as an empty mold) has to be avoided, because the liquid steel overflows the working area or the bottom parts of the mold respectively. Besides an interruption to the production, high maintenance costs will be caused. Using an example of a slab-caster with varying mold geometry, this paper shows that in the case of disturbances an improvement on conventional control can be achieved by the use of fuzzy logic. The basic idea is the use of a parallel control structure with a complementary fuzzy logic controller. This provides the advantage that under normal operating conditions the conventional controller, in this case a PI controller, is still in operation, whilst in the case of a disturbance the fuzzy controller supports the conventional control. This leads to a much better disturbance control behaviour than without the fuzzy controller, with the result that, in contrast to conventional control, overflows can be avoided.

Fuzzy residual evaluation concept (FREC)

Suggests a novel philosophy of process supervision based on knowledge based redundancy. The key idea is to replace the conventional residual evaluator of the fault diagnosis system based on crisp logic, by both a decision maker with fuzzy logic for residual evaluation and the human operator to make the final decisions using his natural intelligence, experience and common sense. The purpose of the employment of fuzzy logic for residual evaluation is to release only weighted alarms instead of yes-no decisions, such that (by definition) no false alarms can be produced; besides this the man-machine interaction becomes much easier. In contrast to the conventional expert system approach, the proposed concept leaves the final yes-no decisions to the natural intelligence, capability and responsibility of the human operator. In addition this method can be seen as an extension to the quantitative model-based techniques. Nevertheless the huge amount of information, which is normally given by most of the fault diagnosis schemes, should be both, filtered and reduced in the sense of the detectability and reliability of an FDI scheme. As an application example this concept has been applied to a part of a power plant in order to prove this theory.

Fuzzy control of steam turbines

Abstract A feasibility study of a fuzzy logic speed control of a steam turbine is reported. The efficiency of the fuzzy control is compared with conventional PID control. As a basis for this comparison, we use a mathematical model of the turbine which has been well validated at a real turbine in connection with the design of the PID control. For this model, a simple fuzzy controller is designed. To assess the resulting fuzzy control, we study by simulation the step response of the speed at a full load reduction, as well as the robustness with respect to parameter variations of the turbine.

FDI with computer-assisted human intelligence

The goal of fault detection and isolation (FDI) is to decide whether and where a fault in the system under consideration has occurred avoiding wrong decisions that cause false alarms. To achieve a fault detection scheme which is robust in the sense of false alarms a combined quantitative/qualitative supervision system is used to detect and isolate faults. The quantitative part is used to generate fault symptoms (residuals) using a quantitative (mathematical) model of the process. These residuals contain the information about whether a fault has occurred or not. The next step in the FDI process is the residual evaluation. There exists a number of different residual evaluation techniques, for example simple threshold logic tests, statistical decision making, pattern recognition and decision making based on fuzzy logic or neural networks. The fundamental difficulty with residual evaluation is that residuals are normally uncertain, corrupted by noise, disturbances and, if the residuals are generated by model-based techniques, by modelling uncertainties. In order to select from the given residual data the important fault information a human support tool for the generation of a knowledge base for fault diagnosis is presented in the paper.