ABSTRACT

The estimation of soil and rock parameters based on field instrumentation data is a traditional and a common procedure in Geomechanics. However, the use of optimization techniques to perform backanalysis in a more systematic manner is a relatively recent development (Gioda & Sakurai, 1987). From a mathematical point of view, the identification of parameters results in a minimization problem. Nowadays, the use of system identification and optimization techniques allows the performance of this type of analyses in a more rational and objective manner. Examples of backanalysis in the context of tunnel excavation problems are described in Ledesma et al (1996, 1997) and Gens et al (1988, 1996) among others. In those works, a maximum likelihood approach is presented to estimate the required parameters. The main difficulty when estimating parameters is that several combinations of them may give similar results in terms of measurement variables and therefore the solution is not unique. Because of that, part of

1 INTRODUCTION

The work presented in this paper has been developed in the context of the European Science Foundation Project “Micro-Measurement and Monitoring System for Ageing Underground Infrastructures”. The research project focus on the monitoring of the deterioration of underground structures by means of new technologies based on wireless sensor networks, including computer vision and miniature electromechanical sensors. Ageing of tunnels is observed as a deterioration of the lining in most cases. Opening of cracks, strains or relative displacements or rotations between two points constitute typically the main variables that can be measured on the lining surface. Other basic measurements like pore water pressures in the surrounding soil or absolute displacements could be considered as well in further developments.