An object-oriented knowledge-based system for hydroelectric power plant turbine selection
The objective of this work is the development of an Expert System for the selection of the turbines for any hydroelectric power plant. The selection is based upon the calculation of different variables referred to as head, flow, speed, and load. Using essential data of a certain project, the system will produce the mathematical results and suggest the best choice of turbine for it. It was important to present in this work material found in the literature, because hydroelectric turbine data is difficult to find. Its accumulation in one document will be a great benefit for the user and is thus an additional contribution of this thesis. The turbine theory also is an indispensable part of the expert-system code because it became the knowledge base to support the inference engine. The system was developed to be simple and user friendly, requiring no programming skills on the user's part. The utilization of this expert system as a decision tool in the selection process of a turbine enables future modifications of the program to be made easily. It also permits changes and adaptations according to the needs of the power plant. In addition the system will provide, upon request, a short justification for the decision reached. The turbine for a hydroelectric power plant should be selected and designed to suit the specific range of conditions under which the plant is going to operate. Careful studies of streamflow and reservoir operation and accurate field data are necessary for an economical selection to be made. Using an object-oriented knowledge base all the information about certain turbine use is written in a form of production rules. The data available will trigger the rules that match the category of such turbine. Graphics and field data were collected from different sources. All field data were graphically analyzed and the frequency of each turbine in each range was defined. These data were used to write the rules that make the knowledge base complex and realistic. The range of the variables was divided in categories. Using these categories, the examples were graphically analyzed. The graphics were also divided in categories. Each graphic was defined as a Class of the object-oriented knowledge base. The probability of each turbine occurring in any of the classes was calculated. The head and flow ranges are the main attributes of each class. According to the class being used other attributes are added. The system was designed to minimize the calculation time as well as the cost of the project, since the user does not have to be an expert in the field of Hydroelectric Power. If the user is an expert in this field it is estimated that he will spend 10 minutes to run the program and to get results. If the user is not an expert in this field it is estimated that he will spend 30 minutes to run the program and to get the results. In case of an expert doing the selection and all the calculation by hand it is estimated, that he will spend 8 hours for the calculation and selection of the turbine. In case of a nonexpert doing the selection and calculation by hand it is estimated that he will spend at least 5 months, considering that he will have to spend most of the time to become knowledgeable about turbine selection. Another point to be considered, when doing the selection by hand, it is the possibility of making mistakes during calculation and selection.
School Location:USA - Ohio
Source Type:Master's Thesis
Keywords:object oriented knowledge based system hydroelectric power plant turbine selection
Date of Publication:01/01/1992