Though the importance of the engineering rock mass properties such as deformation modulus, cohesion and friction angle is emphasized, the researches on the reliability of these values are rarely performed. By using modified Takagi-Sugeno (TS) fuzzy inference system, a fuzzy model was developed based on the data obtained from domestic tunnel construction sites, and the ability of prediction was compared with the empirical equations conveniently used for estimating these values, which finds that fuzzy model is better than empirical equation. Sufficiently large database, however, is required for training the fuzzy model for more accurate prediction.
Obwohl die Wichtigkeit der technischen Eigentuemer der Felsenmassen wie zum Beispiel Verformungsmodulus, Bindekraft und Reibungswinkel betont werden, wird die Forschung auf der Zuverlaßigkeit dieser Werte selten durchgefuehrt. Mit modifiziertem Takagi Sugenos (TS) Flaumigenschlußsystem wird die Flaumigenmodelle von den Daten, die von inlandischen Tunnelbaustellen erhalten werden, entwickelt. Zum Schatzen dieser Werte ist es viel guenstig, Flaumigenmodell zu benutzen, als zu adoptieren die empirische Gleichung fuer die Fahigkeit der Voraussage. Genuegende große Datenbank wird aber zum Ausbilden flaumigen Modells fuer genauere Voraussage erfordert.
Quoique l'importance sur la caracteristique d'ingenierie de la masse de roche comme le module de deformation, la cohesion et l'angle de friction, soit soulignee, la recherche sur la confiance de ces valeurs est rarement effectuee. Avec le système modifie de Takagi-Sugeno(TS) fuzzy inference, le nouveau modèle fuzzy s'est developpe. Après avoir estime sa valeur, nous pouvons savoir que ce modèle fuzzy, base sur les donnees venant du chantier de construction de tunnel national, est mieux que l'equaton empirique. Pourtant plus de donnees de base sont indispensables pour produire le modèle fuzzy plus precis.
Introduction The design and stability analysis of a structure in rock require engineering rock mass properties such as deformation modulus, cohesion and friction angle. It is very important to put accurate values of engineering rock mass properties in the numerical analysis for the stability analysis of the underground structures. It is, however, hard to estimate these values because of difficult procedures and variation of the values. Indirect estimation methods include uncertainties related to the limited data availability, variability of rock type and the heterogeneous nature of the rock masses. Some attempts have been made to increase the reliability of these values using statistical method such as Monte Carlo simulation, which still has limitation in estimating probability density functions of the variables. Therefore, it is imperative to develop a reliable evaluation method for engineering rock mass properties.
In the present study, a prediction model by using fuzzy set theory was proposed for the evaluation of engineering rock mass properties. Also comparisons between prediction results from fuzzy model and those of empirical equation were performed. In the following sections, the background procedure, results and conclusions of this study are introduced.
Modified Takagi-Sugeno (TS) Fuzzy Model Modified TS fuzzy model by Chen et al (1998) was used for prediction model. Its consequent parts of fuzzy rules consisted of linear combinations of input variables.