Browsing by Author "Premasiri, KW"

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    item: Thesis-Abstract
    Development of rule based expert system for design guidelines of domestic buildings
    Premasiri, KW; Dias, WPS
    Use of expert systems for Engineering applications is still in very primitive stages in Sri Lanka. However it is being used satisfactorily in many developed countries over several areas of the Engineering applications. In Medical diagnostic applications it had been used successfully over many decades. Later it has been spreaded over some other disciplines such as Engineering fault diagnosis and selecting bad ones from production lines etc. Further it has been reported that attempts are being done to carry out Engineering designs. Algorithmic methods are still prevalent in structural Engineering designs and these are mainly data driven and non-availability of any single data will stop the running of design process. Expert systems are knowledge driven and non-availability of piece of knowledge will not stop the process. Rule based expert system approach is very popular among the system builders due to its flexibility in applications. In this assignment rule based method was utilized in the process of developing a proto type computer model "BUILD GUIDE" on the design guidelines for domestic buildings, operating in the DOS environment. Chapter I of this report is dealing with the background details of the study including objectives of the study and methodology used and aspects of the Artificial Intelligence and Expert Systems. Chapter 2 of this report is on the literature review of the Expert systems and various aspects of them including structure of them, advantages, disadvantages, uncertainty of data and reasoning and methods of developing expert systems etc. Chapter 3 of this report is reserved to discuss about the structuring of design guidelines for domestic buildings, usefulness of design guidelines of domestic buildings and related topics. For this purpose the design entity was considered in three basic stages namely, preliminary design, outline or conceptual design and detailed design stages. Further a description about the expert system development software is available. An output of a sample session for a small domestic building is included in the chapter 3.5 to illustrate about the output report of the BUILD GUIDE system. Finally a comprehensive comparison of design and diagnosis processes is provided, as it is necessary to understand it clearly before developing an expert system on design and diagnosis. The "CONFAULT" system, which has been formulated to diagnose the faults of Reinforced Concrete structures by identifying sub fault types, is used with the BUILD GUIDE system for the comparison. Further, this report consists with a chapter on brief description of the expert systems for design applications.
  • Thumbnail Image
    item: Thesis-Abstract
    Studies on textile waste water treatment
    Premasiri, KW; Rathnayake, N
    In Sri Lanka the waste water discharged from most Textile Processing Factories to water bodies are found to be poorly treated or untreated. This may be due to reasons such as involvement of high capital and operation costs for treatment facilities, difficulties encountered in maintenance of mechanical and electrical equipments and the unawareness about the necessity of treatment of waste water. The non availability of literature regarding textile waste treatment may also be a reason for it. As such studies on treatment methods of textile waste water are very important. In this report the author explains available literature on characteristics of textile waste water and its treatment methods and measurements related to waste water. Laboratory studies were carried out for the removal of colour and C.O.D. from textile waste water. Regarding the removal of colour several experimental methods were carried out based on the principle of coagulation and flocculation. For colour adsorption, powdered activated carbon and "commercially available polyelectrolytes were used. The Conventional jar test was employed in these experiments. Conventional activated sludge process was used in' the removal of C.O.D. This was carried out by varying the influent C.O.D; values. A basic Laboratory scale activated sludge plant model complete with aeration tank, settling tank and other essential components was used for this purpose. By adding powdered activated carbon to the waste water and carrying out flocculation and coagulation with alum, 100% removal of colour was achieved at certain dosages. Similarly removal of colour can be achieved by using polyelectrolytes also. Powdered activated carbon requirement varies from 100 mg. to 1000 mg. for 1 liter of waste water.More than 80% of available C.O.D. could be removed by adopting conventional activated sludge treatment method with proper controls.