dc.contributor.author Hosseini, Ebrahim. en_US dc.date.accessioned 2014-10-21T12:36:50Z dc.date.available 2014-10-21T12:36:50Z dc.date.issued 1998 en_US dc.identifier.other AAINQ31526 en_US dc.identifier.uri http://hdl.handle.net/10222/55542 dc.description The procedure presented builds a bridge between computational intelligence and hydrologic modeling in applying fuzzy and multivariable data to design in urban and rural watersheds. Watershed parameters such as soil moisture content, cover density and impermeability are viewed as continuous parameters using fuzzy set theory and are then analyzed in a fuzzy logic control expert system. This logic model, involving relations between these parameters uses methods that explicitly take vagueness into account. The theory of fuzzy sets, especially fuzzy modeling, is employed in a new way to represent watershed parameter relations as a set of fuzzy rules. This approach was implemented as an interactive modeling system, called the "Fuzzy Logic Expert Watershed Curve Number" (FLEW$\sp{\rm CN}).$ This model is based on logical relationships between parameters rather than experimental data. Fuzzy logic control methodology was used in development in three consecutive phases in which each phase produced a base for the next phase of development. These phases were firstly, a single-input-single-output (SISO) model for soil texture (% sand versus % clay), followed by double-input-single-output (DISO) models for % sand versus each individual watershed parameter. The final stage of fuzzy logic control was to develop a multi-input-single-output (MISO) model based on the parallel algorithm by forward reasoning strategy with regard to conditional and unconditional rule-based systems. The final model has an open-loop control structure (output has no effect on input). en_US dc.description The program is organized as a procedural process with elements of action (membership functions). Each membership function is defined in binary format as an object of an element. Each element has two different references as a pivot (only a specific element appears to act when the program is called) and as a global (all of the elements are affected when the program calls the elements). This process appears during execution time as a user interface. en_US dc.description The user interface links keyboard input of fuzzy data to the fuzzy control and graphically illustrates each input space. The user interface also indicates the inference action in a screen output space. The user interface provides an effective means to assess the effect of changes in combination of inputs on the output response. en_US dc.description The program is coded in the Turbo C environment on a DOS platform using the Borland graphic support system. This program is independent of any expert shell during the time of execution. en_US dc.description Verification was carried out by comparison with the Soil Conservation Service (SCS) method. The validation showed that the fuzzy logic model predicted a curve number in a similar range to that of the SCS method. However, predictably some differences were observed which can be attributed to fuzzy logic methodology. This is because fuzzy logic control produces a continuous model, whereas the SCS model is a discrete model. The advantage of the FLEW$\sp{\rm CN}$ model is that it is not limited to a specific range of inputs; any combination of inputs (% sand, % cover, % moisture, and % of impermeable area) can be translated to an output response (CN). en_US dc.description Thesis (Ph.D.)--DalTech - Dalhousie University (Canada), 1998. en_US dc.language eng en_US dc.publisher Dalhousie University en_US dc.publisher en_US dc.subject Hydrology. en_US dc.subject Engineering, Civil. en_US dc.subject Artificial Intelligence. en_US dc.title A multiple-input-single-output (MISO) fuzzy logic model for generation of watershed Soil Conservation Service (SCS) curve numbers. en_US dc.type text en_US dc.contributor.degree Ph.D. en_US
﻿ Find Full text