Expert systems components

Besides these three major components, many expert systems also comprise an explanation subsystem and a knowledge acquisition subsystem,... 

Implementation of advanced performance degradation models, necessitate the inclusion of advanced instrumentation and sensors such as pyrometers for monitoring hot section components, dynamic pressure transducers for detection of surge and other flow instabilities such as combustion especially in the new dry low NO combustors. To fully round out a condition monitoring system the use of expert systems in determining fault and life cycle of various components is a necessity. 

There are some aspects of process design in which decisions are based primarily on past experience rather than on quantitative performance models. Problems of this type include the selection of constraction materials, the selection of appropriate models for evaluating the physical properties of homogeneous and heterogeneous mixtures of components, and the selection of safety systems. Advances in expert systems technology and information management will have a profound impact on expressing the solutions to these problems. 

The typical structure of an expert system is shown in Fig. 43.1. Three basic components are present in all expert systems the knowledge base, the inference engine and the interaction module (user interface). The knowledge base is the heart of the expert system. It contains the necessary expert knowledge and experience to act as a decision support. Only if this is correct and complete enough the expert system can produce meaningful and useful conclusions and advice. The inference... 

Today, analytical evaluation is done on a large scale in a computerized way by means of data bases and expert systems (Sect. 8.3.6). In particular, a library search is a useful tool to identify pure compounds, confirm them and characterize constituents in mixtures. Additionally, unknown new substances may be classified by similarity analysis (Zupan [1986], Hippe [1991], Warr [1993], Hobert [1995]). The library search has its main application in such fields where a large number of components has to be related with large sets of data such as environmental and toxicological analysis (Scott [1995], Pellizarri et al. [1985]). 

Taken together, all of these points suggest that it might be possible to prepare a toolkit consisting of the essential components of an ES, apart from the knowledge base, and then fill it with application-specific data. This is such a useful way to work that in all expert systems there is a clean separation between the information that the ES manipulates and the tools that are required to perform that manipulation. This division between the part of an ES that changes between applications and that which is constant has led to the development of the expert system shell. 

There are three major components to an expert system ... 

The contents of a knowledge base, the facts and rules, or heuristics about a problem will be discussed shortly. The problem-solving and inference engine is the component of the system that allows rules and logic to be applied to facts in the knowledge base. For example, in rule-based expert systems, "IF-THEN" rules (production rules) in a knowledge base may be analyzed in two ways ... 

The facts in a knowledge base include descriptions of objects, their attributes and corresponding data values, in the area to which the expert system is to be applied. In a process control application, for example, the factual knowledge might include a description of a physical plant or a portion thereof, characteristics of individual components, values from sensor data, composition of feedstocks and so forth. 

The success of SYNLMA shows that it is possible to base an expert system on a theorem prover. The advantage of using a theorem prover as deductive component is that it allows us to experiment with a number of different representations for chemical information. The same flexibility makes it easy to add new starting materials and reaction rules from large commercial online databases. 

An integrated GC/IR/MS instrument is a powerful tool for rapid identification of thermally generated aroma compounds. Fourier transform infrared spectroscopy (GC/IR) provides a complementary technique to mass spectrometry (MS) for the characterization of volatile flavor components in complex mixtures. Recent improvements in GC/IR instruments have made it possible to construct an integrated GC/IR/HS system in which the sensitivity of the two spectroscopic detectors is roughly equal. The combined system offers direct correlation of IR and MS chromatograms, functional group analysis, substantial time savings, and the potential for an expert systems approach to identification of flavor components. Performance of the technique is illustrated with applications to the analysis of volatile flavor components in charbroiled chicken. 

Thus, multilinear models were introduced, and then a wide series of tools, such as nonlinear models, including artificial neural networks, fuzzy logic, Bayesian models, and expert systems. A number of reviews deal with the different techniques [4-6]. Mathematical techniques have also been used to keep into account the high number (up to several thousands) of chemical descriptors and fragments that can be used for modeling purposes, with the problem of increase in noise and lack of statistical robustness. Also in this case, linear and nonlinear methods have been used, such as principal component analysis (PCA) and genetic algorithms (GA) [6]. 

In an expert system some essential components can be identified. A first part is a set of facts which specify relations between defined objects. This part is often referred to as the data base. A second part is a collection of rules which specify how to reach conclusions by combining the facts. In most cases a rule is equivalent to a statement about relations between object clauses, or it may be given as a relation between not fully defined objects. Thus, the boundary line between rules and facts is rather illdefined. 

Knowledge-based systems are computer programs which apply knowledge about a specific domain in order to derive new conclusions. These conclusions are on the level of a human expert in this field, but constrained by the field of expertise. Applied to catalysis, the knowledge base contains heuristics about relationships between chemical and physico-chemical properties of solids and their catalytic properties, as well as known properties for such solids which may be used as components of catalysts [2, 18]. An expert system is able to combine this knowledge about different catalyst properties which may be necessary to catalyze the required reaction steps, or which should be avoided because they catalyze side reactions. 

The system ESKA (Expert System for Selection and Optimization of Catalysts [20]) was designed at BASF specifically for hydrogenation reactions. The main component of a catalyst is proposed on the basis of activity patterns which describe the applicability of catalysts for different types of hydrogenations. The system also is able to propose secondary catalyst components and, if necessary, a support material which is stable under reaction conditions and does not have any undesired catalytic properties. Based on heuristics for required as well as undesired side reactions and for different catalytically active components, the system also proposes reaction conditions as temperature, pressure, the solvent or the pH. 

Kito and Hattori et al. have described INCAP (IN-tegration of Catalyst Activity Patterns [21-23]), an expert system which rates the applicability of catalyst components for the desired reaction based on known activity patterns for different catalyst properties. The system was successfully applied for the selection of promoter components for the oxidative dehydrogenation of ethylbenzene to styrene. An improved version INCAP-MUSE (INCAP for MUlti-Componcnt catalyst SElcction [24-26]) selects as many catalyst components until all required catalyst properties are present. Although the system was successfully applied to oxidation reactions, more recently better results have been obtained by neural network methods (Section 2.6.2.2). 

By doping a primary catalyst component with lower-valent metal cations, additional oxygen vacancies will be created which facilitate the incorporation of electrophilic oxygen species chemisorbed on the surface into the bulk where they will not oxidize adsorbed methyl radicals. Also, the promoter oxide should be basic, not be reducible, oxidizablc, or easily volatiz-ablc. It should form a mixed oxide with the main component which may be possible if the ionic radii arc similar. According to these rules, the expert system proposes as potential catalyst components combinations of substances with appropriate chemical and physico-chemical properties (Table 2). Many of these systems already have been described in the literature... 

---