The compact linear collider (CLIC) is the particles accelerator project proposed by the european organization for nuclear research (CERN) for high energy physics after the large hadron collider (LHC). Because of the nanometric scale of the CLIC leptons beams, the emittance growth budgetis very tight. It induces alignment tolerances on the positions of the CLICcomponents that have never been achieved. The last step of the CLIC alignment will be done according to the beam itself. It falls within the competence of the physicists. However, in order to implement the beam-based feedback, a challenging pre-alignment is required: 10 μm at 3σ along a 200 m sliding window. For such a precision, the proposed solution must be compatible with a feedback between the measurement and repositioning systems. The CLIC pre-alignment will have to be active. This thesis does not demonstrate the feasibility of the CLIC active prealignment but shows the way to the last developments that have to be done for that purpose. A method is proposed. Based on the management of the Helmert transformations between Euclidian coordinate systems, from the geodetic networks to the metrological measurements, this method is likely to solve the CLIC pre-alignment problem. Large scale facilities have been built and Monte-Carlo simulations have been made in order to validate the mathematical modelings of the measurement systems and of the alignment references. When this is done, it will be possible to extrapolate the modelings to the entire CLIC length. It will be the last step of the demonstration of the CLIC pre-alignment feasibility
