Method and apparatus for providing a sparse Gaussian process model for calculation in an engine control unit

摘要

The invention relates to a method for determining a sparse Gaussian process model for execution in a purely hardware-based model calculation unit (43), comprising the following steps: - providing support point data points, a parameter vector based thereon and corresponding hyperparameters; Determining or providing vertex virtual data points for the sparse Gaussian process model; and determining a parameter vector Q * y for the sparse Gaussian process model using a Cholesky decomposition of a covariance matrix KM between the virtual tributary data points and depending on the tributary data points, the parameter vector based thereon and the corresponding hyperparameters representing the sparse Gaussian process Define model.