Mechanical Engineering:

Title:           Development of computer model to predict optimum composition with increased sampling rate of multi-fuel biomass for direct combustion based electricity generation
Authors:       Ranipet Hafeez BASHA, Shuichi TORII
Abstract:     User interface based computer model to simulate multi-fuel biomass heat of combustion was developed by us at Thermal Engineering Laboratory of Kumamoto University, Japan by using Java 2 Secondary Edition (J2SE) version 5.0 programming platform with Apache Tomcat servlet application. In order to automate the mixing of auxiliary fuel like agricultural wastes, forest residues, etc. with main fuel i.e. Refuse Derived Fuel (RDF) this computer model was enhanced using predictive modelling which uses data mining techniques. The software will be capable to automatically alter the ratios of different biomass types and check all the possible combinations to deduce the most optimum mixture based on fuel availability and cost parameters. In the previous paper predictor for 5% sampling was developed, however in this paper the sampling rate was reduced to 1% for greater accuracy. Knapsack problem based on combinatory optimization was applied in developing this predictor model, also probability method that motivated this study was disclosed. The graph generated for 1% sampling had to be divided into sub-graphs namely primary and secondary series so that the characteristics can be described in simple, however due to the complexity of equations and huge computational time the results for tertiary and quaternary series were not considered.

Keywords:   Biomass, Computer Simulator, Computer Predictor, Combustion, Sampling rate
Pages:         125-128

Download Full Paper