Intelligent Approach To Slope Stability Evaluation Based On Case-based Reasoning

Slope is a non-linear and uncertain kinetic system affected by many complicated factors. A large amount of subjective and objective uncertainty lies in the slope stability evaluation, which weaken seriously the reliability, effectiveness, and usefulness of the existing methods of the slope stability evaluation. In view of the incompleteness and uncertainty of the various factors affecting the slope stability, a series of new methods of slope stability evaluation, such as fuzzy comprehension evaluation method, grey cluster evaluation method, reliability evaluation method, and neural network evaluation method, etc. have been proposed. The proposal of these methods has reflected from various aspects the properties affecting the uncertainty of the slope stability. In consideration of special complexity of the slope stability evaluation, the researchers have fully realized the importance of the engineering geological work and the first-hand geological data, and put stress on the unification of theories, methods and techniques. They have noticed the comparison and application in slope stability evaluation of the large number of well-studied slope cases with regard to the stability condition. Based upon this, intelligent approach to slope stability evaluation using case-based reasoning is presented for the first time in this dissertation.The main achievement of this dissertation are briefly mentioned as follows:1. The studying work purpose and signification of the dissertation are briefly introduced. The uncertainty faced in the slope stability evaluation are systematically analyzed, and the developing tendency studying the uncertain problems of the slope stability are presented.2. The main uncertain methods of the slope stability evaluation, such as fuzzy comprehension evaluation method, grey cluster evaluation method, reliability evaluation method, and neural network evaluation method, etc. are fully discussed, and their developing history and research situation are introduced in detail, and some problems which need to be studied are presented, and the studying contentsIVand work achievement are introduced.3. Based on a mathematical theory of evidence produced by G.Shafer, the D-S information fusion method for the determination of rockmass mechanical parameters is presented for the first time, by which the discernment frame of rockmass mechanical parameters is established. The D-S rule of the information fusion of rockmass mechanical parameters is presented. Meanwhile, the basic probability assignments, belief functions, plausibility functions, and similar probability functions are given. This method is applied to an example of analysis and selection for rockmass mechanical parameters of the rock engineering, and the satisfactory results are obtained.4. Based on the extension theory produced by W. Cai, the matter-element method of the slope stability evaluation is presented for the first time, and the classical field, the limited field, and the matter-element for appraising are determined. The dependent degrees of the matter-element for appraising of slope stability are calculated. The matter-element model established can comprehensively reflect the influences of the multiple factors on slope stability. These factors are mainly the compressive strength of rocks, the quality index of rocks RQD, the velocity of elastic wave of rockmass, the slope height, the annual rainfall, and the structural character grade of rockmass. The conclusions obtained are consistent with the engineering practice.5. In view of the incompleteness and uncertainty of the information of slope stability evaluation, the intelligent approach to slope stability evaluation using case-based reasoning is presented for the first time. Considering the sensitivity of attribute weights to the environment, the algorithm of attribute weights is set up on the basis of the concept of changeable weights, and the slope base case indexing model is set up. In this model, fuzzy analogy pre...