A knowledge-based system has been developed to provide a tool for storing and using correlations and "typical" values for the estimation of ground properties. The system was implemented in the ProKappa software, running under X windows on a Sun Spark 2 workstation. The system developed is intended to provide geotechnical engineers with a decision-support tool and to demonstrate the applicability of knowledge-based system technology to the ground property evaluation problem.
The development of this system involved the identification and collection of the domain knowledge (knowledge associated with the ground, ground properties, "typical" values and correlations, which was elicited from the relevant technical literature). Generic forms for the representation of correlations and "typical" values in the system were developed (which provide a consistent form for the representation).
The system also incorporates an inference engine, which includes the process that makes use of this knowledge to produce estimations for ground properties; a user interface to facilitate the use of the system; and finally knowledge acquisition modules for updating of existing knowledge, to ensure that the system will maintain its functionality in the future.
CHAPTER 1 INTRODUCTION 1
1.1 General. 1
1.2 Overview of the thesis 3
CHAPTER 2 THE APPLICATION OF KNOWLEDGE-BASED SYSTEMS TO THE ESTIMATION OF GROUND PROPERTIES. 6
2.1 Introduction. 6
2.2 KBS components. 8
2.3 Development stages of KBS. 9
2.4 KBS in Geotechnical Engineering. 15
2.5 Discussion. 20
2.6 Conclusions. 22
CHAPTER 3 DEVELOPMENT TOOLS. 23
3.1 Introduction. 23
3.2 The ProKappa software. 23
3.2.1 Objects, slots, facets and applications. 24
3.2.2 Programming languages. 27
3.2.3 User interface development tools. 31
3.2.4 Monitors and Active Relations. 34
3.3.5 Summary. 38
CHAPTER 4 REPRESENTING THE GROUND AND ITS PROPERTIES. 39
4.1 Introduction. 39
4.2 A model for representing the ground. 41
4.2.1 Rocks. 42
4.2.2 Soils. 46
4.3 A model for representing ground parameters. 49
4.4 Implementation in the system. 54
4.5 The representation of "typical" values. 56
4.6 User interface facilities. 62
4.7 A knowledge acquisition module for typical values. 69
4.8 Summary. 78
CHAPTER 5 REPRESENTING CORRELATIONS IN A STRUCTURED FORM. 81
5.1 Introduction 81
5.2 Representing correlations in a structured form 85
5.2.1. Variables 85
5.2.2. The estimation procedure 88
5.2.3. Applicability 88
5.2.4. Reliability 90
5.2.5. Comments 92
5.3 Implementation in the system 92
5.3.1 The Correlation application 93
Introduction 93
Variable slots 95
Parameter slots 97
Data_Check! slot 100
The Estimation Procedure 104
Applicability Slots 105
Reliability 106
5.3.2 The Correction module 106
5.3.3 The CorrUI module 107
5.4 Summary 119
CHAPTER 6 A KNOWLEDGE ACQUISITION MODULE FOR THE IMPLEMENTATION OF CORRELATIONS. 123
6.1 Introduction 123
6.2 Establishment of the basic parameter 124
6.3 Variables and parameters 129
6.4 Implementation of estimation procedures. 140
6.5 Applicability definition. 145
6.6 Reliability definition. 151
6.7 Comments. 152
6.8 Updating correlations. 152
6.9 Overview of the Update module. 154
CHAPTER 7 DISCUSSION - FUTURE DEVELOPMENT 156
7.1 Discussion. 156
7.1 Future development. 162
CHAPTER 8 CONCLUSIONS 165
REFERENCES 167
Appendix A Correlations and Corrections A-1
Part 1: Correlations. A-1
Part 2: Corrections A-38
References A-43
Appendix B Typical values for ground properties B-1
Appendix C The ProTalk Code C-1