Much of the planet's energy consumption, pollutions, waste generation etc. happens in cities., which makes it important to consider urban areas in efforts aiming at sustainable development. Transportation and land use planning has become essential as a decision aid tool. Development of LUTI models (Land Use and Transportation Integrated models) has increased during the last 20 years. Calibration of large-scale LUTI models is a challenging task. It is usually partitioned into a set of smaller, partial parameter estimation problems of individual components of a model, and an integrated calibration of the composite model, taking into account the mutual interactions between these components, is most often lacking. This work presents a reformulation of the calibration of the Tranus model as an optimisation problem. This methodology is applied to the estimation of the endogenous variable called " shadow prices " , this variable acts as an error term of the localisation utility function. We also present a test methodology for validating the calibration against synthetic data that perfectly fit observations. We present this methodology on a small example that permits us to get a visual assessment of the solution.