Miki Sirola

Using Self-organizing Map in a Computerized Decision Support System

Modern computerized decision support systems have developed to their current status during many decades. The variety of methodologies and application areas has increased during this development. In this paper neural method Self-Organizing Map (SOM) is combined with knowledge-based methodologies in a rule-based decision support system prototype. This system, which may be applied for instance in fault diagnosis, is based on an earlier study including compatibility analysis. A Matlab-based tool can be used for example in fault detection and identification. We show with an example how SOM analysis can help decision making in a computerized decision support system. An error state model made in Simulink programming environment is used to produce data for the analysis. Quantisation error between normal data and error data is one significant tool in the analysis. This kind of decision making is necessary for instance in state monitoring in control room of a safety critical process in industry.

SOM Based Decision Support in Failure Management

Computerized decision support system field covers many methodologies and application areas. In this paper self- organizing map (SOM) and knowledge-based techniques are used in combination to reason problematic situations in failure management. A process model that consists of individual connected process components has been developed. A primary circuit of a boiling water nuclear power plant including two branches has been composed. A failure management scenario is thoroughly analyzed and solved with the SOM based decision support system. The structure and reasoning of the computerized decision support system (CDSS) is also shortly discussed. The process model is demonstrated together with the CDSS and shown to be useful. The tool helps operators decision making with various visualizations, and by giving concrete recommendations for possible control actions or other acts.

Self-organizing map in process visualization

Our research group is studying data-analysis based techniques in decision support and visualization. We have had co-operation with a Finnish nuclear power plant Olkiluoto within a long industrial research project. We have developed many decision support schemes and visualizations based on self-organizing map (SOM) method combined with other methodologies. In this paper, we discuss about SOM method in the process visualization of dynamic systems. With a case example produced with the Olkiluoto plant data we show the information value of this method. Some comparisons to other methodologies are made and the assessment of the information value and the definition of the assessment criteria are discussed. The measurement of the information value is a challenging task.

Process state and progress visualization using self-organizing map

The self-organizing map (SOM) [1] is used in data analysis for resolving and visualizing nonlinear relationships in complex data. This paper presents an application of the SOM for depicting state and progress of a real-time process. A self-organizing map is used as a visual regression model for estimating the state configuration and progress of an observation in process data. The proposed technique is used for examining full-scope nuclear power plant simulator data. One aim is to depict only the most relevant information of the process so that interpretating process behaviour would become easier for plant operators. In our experiments, the method was able to detect a leakage situation in an early stage and it was possible to observe how the system changed its state as time went on.

Perspectives in science

This article discusses perspectives in science with a somewhat limited scope. The areas of science included reflect the interest and experience of the author with an interdisciplinary view. Scientific domains, applications and methodologies constitute a unique entity. The three main pillars of the model are psychology, civics and energy production. The psychology here is mainly adapted in work environment. The psychology has two main branches one concentrating on individual behaviour and the other on group behaviour. The latter connected to civics converges towards sociology. The energy production represents the industrial view in the model, and it can be broaden wider in the general model, although here we concentrate mostly on the mentioned application area. From psychology we get assistance in user interfaces and visualizations. One branch in civics is politics. From the energy production and civics we get into production models. From psychology and politics we derive a path to decision making. Also from user interfaces, visualizations and even production models we get into decision making. System technics has a connection to decision making as well as to the other fields mentioned here. Data analysis can be used as a tool in all these fields. In addition scientific visualization is discussed. The use of data analysis in political decision making is presented as an example. For instance in predicting the voting behaviour in a society data from pre-election gallup poll reviews and voting advice applications can be used. This political example shows the opportunity to cross fields in science to derive interesting results.

Combining neural methods and knowledge-based methods in accident management

Accident management became a popular research issue in the early 1990s. Computerized decision support was studied from many points of view. Early fault detection and information visualization are important key issues in accident management also today. In this paper we make a brief review on this research history mostly from the last two decades including the severe accident management. The authors studies are reflected to the state of the art. The self-organizing map method is combined with other more or less traditional methods. Neural methods used together with knowledge-based methods constitute a methodological base for the presented decision support prototypes. Two application examples with modern decision support visualizations are introduced more in detail. A case example of detecting a pressure drift on the boiling water reactor by multivariate methods including innovative visualizations is studied in detail. Promising results in early fault detection are achieved. The operators are provided by added information value to be able to detect anomalies in an early stage already. We provide the plant staff with a methodological tool set, which can be combined in various ways depending on the special needs in each case.

New visualization techniques and their assessment

Our research group has been studying data-analysis based techniques in decision support and visualization. We had a long industrial research project in co-operation with a Finnish nuclear power plant Olkiluoto. We developed many decision support schemes based on Self-Organizing Map (SOM) method combined with other methodologies. Also several visualizations based on various data-analysis methods were developed. Data from the Olkiluoto plant and training simulator was used in the analysis. In this paper some of these visualizations are presented, analyzed, and assessed with a psychological framework. Measuring the information value of the visualizations is a real challenge. The developed visualizations and visualization techniques are also compared with some existing visualizations and techniques in current plants and research laboratories. The visualizations and the visualization techniques are developed further, and completely new visualizations and techniques are developed. We point out what additional value the new visualization techniques can produce. A detailed test case of using Self-Organizing Map (SOM) method with Olkiluoto plant data is presented. With this practical example the information value of this method is shown, and it is also pointed out how it can be assessed, and what are the most reliable criteria in this assessment.

Using Conceptual Decision Model in a Case Study

Decision making is mostly based on decision concepts and decision models built in decision support systems. Type of decision problem determines application. This paper presents a case study analysed with a conceptual decision model that utilises rule-based methodologies, numerical algorithms and procedures, statistical methodologies including distributions, and visual support. Selection of used decision concepts is based on case-based needs. Fine tuning of the model is done during construction of the computer application and analysis of the case examples. A kind of decision table is built including pre-filtered decision options and carefully chosen decision attributes. Each attribute is weighted, decision table values are given, and finally total score is calculated. This is done with a many-step procedure including various elements. The computer application is built on G2 platform. The case example choice of career is analysed in detail. The developed prototype should be considered mostly as an advisory tool in decision making. More important than the numerical result of the analysis is to learn about the decision problem. Evaluation expertise is needed in the development process. The model constructed is a kind of completed multi-criteria decision analysis concept. This paper is also an example of using a theoretical methodology in solving a practical problem.

Detecting aging of process sensors with noise signal measurement

In this paper, methods for detecting failures in process sensors from the noise measurement due to aging issues are examined. The data are acquired from the water level and pressure measurement transmitters in the Olkiluoto nuclear power plant in Finland: units Olkiluoto 1 and Olkiluoto 2. Methods found from the literature about the failure indicators are presented. Changes in the sensor response time as well as in the resonance peaks in the signal are identified from the power spectrum of the signal. In addition, a new method for fingerprinting the sensors using the Principal Component Analysis (PCA) of the signal spectra is presented. By following the changes in these fingerprints and the variations between parallel measurements of the redundant sensors, symptoms of sensor failures can be detected. In the experiments we were able to produce stable fingerprints for the differential pressure transmitters used in the water level measurement. Potential failure in one differential pressure sensor in unit Olkiluoto 2 is found with the fingerprint method and by analyzing the changes in the spectrum.

Data fusion of pre-election gallups and polls for improved support estimates

In this paper, gallup results and a questionnaire in the context of a voting advice application related to the Finnish presidential election are combined. The main emphasis is on preprocessing phases where raw data is reformed to temporal data sets. We also pay attention to find optimized parameters for a merged recursive model. Aggregated data from a questionnaire was stored frequently and modified by a differential equation. The method presented in this paper allows us to visualize more accurately the daily support of each candidate before the election. The results can be used for further research such as forecasting the results and the success of presidential campaigns.