National freight transport planning: towards a Strategic Planning Extranet Decision Support System (SPEDSS)

This thesis provides a `proof-of-concept' prototype and a design architecture for a Object Oriented (00) database towards the development of a Decision Support System (DSS) for the national freight transport planning problem. Both governments and industry require a Strategic Planning Extranet Decision Support System (SPEDSS) for their effective management of the national Freight Transport Networks (FTN). This thesis addresses the three key problems for the development of a SPEDSS to facilitate national strategic freight planning: 1) scope and scale of data available and required; 2) scope and scale of existing models; and 3) construction of the software. The research approach taken embodies systems thinking and includes the use of: Object Oriented Analysis and Design (OOA/D) for problem encapsulation and database design; artificial neural network (and proposed rule extraction) for knowledge acquisition of the United States FTN data set; and an iterative Object Oriented (00) software design for the development of a `proof-of-concept' prototype. The research findings demonstrate that an 00 approach along with the use of 00 methodologies and technologies coupled with artificial neural networks (ANNs) offers a robust and flexible methodology for the analysis of the FTN problem domain and the design architecture of an Extranet based SPEDSS. The objectives of this research were to: 1) identify and analyse current problems and proposed solutions facing industry and governments in strategic transportation planning; 2) determine the functional requirements of an FTN SPEDSS; 3) perform a feasibility analysis for building a FTN SPEDSS `proof-of-concept' prototype and (00) database design; 4) develop a methodology for a national `internet-enabled' SPEDSS model and database; 5) construct a `proof-of-concept' prototype for a SPEDSS encapsulating identified user requirements; 6) develop a methodology to resolve the issue of the scale of data and data knowledge acquisition which would act as the `intelligence' within a SPDSS; 7) implement the data methodology using Artificial Neural Networks (ANNs) towards the validation of it; and 8) make recommendations for national freight transportation strategic planning and further research required to fulfil the needs of governments and industry. This thesis includes: an 00 database design for encapsulation of the FTN; an `internet-enabled' Dynamic Modelling Methodology (DMM) for the virtual modelling of the FTNs; a Unified Modelling Language (UML) `proof-of-concept' prototype; and conclusions and recommendations for further collaborative research are identified.