Building a knowledge-based system for production process in a fatty acid plant | |
| Author | Kumar, V. R. |
| Call Number | AIT Thesis no.CS-89-05 |
| Subject(s) | Expert systems (Computer science) Chemical industry |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering |
| Publisher | Asian Institute of Technology |
| Abstract | Sound production methods and customer satisfaction are the most important factors for the success of any industry. The pressures on production processes in Chemical industries are not constant. The processes in Chemical industries vary according to the type of raw materials available and the type of products required by the customers. Hence, planning the operations in a chemical industry should take into account a variety of factors and the changing supply and demand situations. A Fatty acid Chemical Plant located near Bangkok was considered for the study. A study about the Plant with respect to the production processes was carried out. The production processes in the Chemical Plant involves various stages such as splitting, treatment, hydrogenation, distillation processes, etc. The raw materials used by the Chemical Plant vary with their availability, cost and also with changing nature of the demand. Various problems involved in the production processes for which the experts in the plant have difficulty in solving them were identified, especially those relating to material flow. For pedagogical implementation, we have considered the problem of allocation of tanks for raw materials, intermediate materials, finished products and by-products. Certain number of tanks of varying capacity are available to store certain materials, where the number of materials are greater than the number of tanks. Storing and retrieving of materials are done in these tanks. Intermediate raw materials are also stored temporarily between processes. Constraints such as, certain materials can mix with some materials without any condition, certain materials can mix with some materials with certain conditions with respect to the iodine value, some materials cannot mix with any material, some materials require cleaning of the tank before being stored etc., have to be satisfied while storing in tanks. The time of arrival or discharge of materials is not known. The issue is on scheduling the tanks under these conditions so as to effective utilise the storage space keeping in view the overall efficiency of the plant. A Knowledge Based System was designed for the problem and a prototype of the design was implemented on a SUN Work Station using COMMON LISP. A hybrid representation technique involving rules and frames were adopted. Frames were used for the knowledge base which comprises of the properties of mixable materials and tanks. Rules were used to activate the frames, give decisions and to update the Knowledge Base. Breakdown considerations were also considered. The complexity of the search involved in obtaining decisions were analysed. A system validation was done. |
| Year | 1989 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Sadananda, Ramakoti; |
| Examination Committee(s) | Kanchana Kanchanasut;Tabucanon, Mario T.; |
| Scholarship Donor(s) | Government of Japan; |
| Degree | Thesis (M.Eng.) - Asian Institute of Technology, 1989 |