A humanoid robot with a body joined to an omnidirectional mobile ground base, and equipped with: a body position sensor and a base position sensor to provide measures, actuators comprising at least 3 wheels located in the omnidirectional mobile base, extractors for converting the measures into useful data, a controller to calculate position, velocity and acceleration commands from the useful data using a robot model and pre-ordered position and velocity references, means for converting the commands into instructions for the actuators, wherein the robot model is a double point-mass model, and wherein the commands are based on a linear model predictive control law with a discretized time according to a sampling time period and a number of predicted samples, and expressed as a quadratic optimization formulation with: a weighted sum of objectives and a set of predefined linear constraints.