The development of expert systems for fuelwood energy crop selection

Miscellaneous · Fri Jan 01 00:00:00 EST 1988

OSTI ID:6807317

Afzal, M

A significant portion of the World's primary energy is derived from biomass. Firewood plays an important role in the supply of this renewable energy. Good yields of firewood depend on the selection of appropriate plant species. The information on this aspect is generally incomplete, uncertain, inconsistent and often unavailable where needed. An Artificial Intelligence technique was applied to this problem. The primary objective was to develop low-cost, PC-based Expert Systems (ES) for the selection of energy crops for the severely effected areas. As a result, two ES (ECE.1 and ECE.2) were designed, developed, and tested. A Linear Programming (LP) model was also formulated for the economic evaluation. Both ES were general purpose, with similar knowledge structures and utilized Backward Chaining inference. The ES were developed on a commercially available shell, and knowledge was represented in PRL (Production Rule Language). The ES were user friendly, transparent, capable of handling Certainty Factors and limited missing values. Each ES contained knowledge on 44 plant species (shrubs and trees). The ES knowledge could be modified and updated without changing the control structure. The LP model provided a quantitative economic analysis. It optimized the return of farm enterprise (selected from a developing country) before and after the introduction of an energy crop (Leucaena.leucocephala). The introduction of this energy crop suggested a switch over from the traditional wheat and maize cropping, due to a three fold increase in the farmer's return. It also increased that share of crop income as compared to the tractor income. These models performed efficiently under the limited testing. The ES, however, need additional testing and modifications for wider practical applications.

• Research Organization: Kansas State Univ., Manhattan, KS (United States)
• OSTI ID: 6807317
• Resource Relation: Other Information: Thesis (Ph.D.)
• Country of Publication: United States
• Language: English

Related Subjects

09 BIOMASS FUELS
BIOMASS PLANTATIONS
EXPERT SYSTEMS
PLANTS
SPECIES DIVERSITY
WOOD FUELS
PRODUCTION
AGRICULTURE
ARTIFICIAL INTELLIGENCE
ECONOMIC ANALYSIS
ECONOMICS
FORESTRY
KNOWLEDGE BASE
MAIZE
MATHEMATICAL MODELS
MODIFICATIONS
OPTIMIZATION
PERFORMANCE TESTING
SHRUBS
TREES
WHEAT
CEREALS
ENERGY SOURCES
FUELS
GRAMINEAE
LILIOPSIDA
MAGNOLIOPHYTA
TESTING
090800* - Biomass Fuels- Production- (1990-)