Method for estimating model uncertainties with the aid of a neural network and an architecture of the neural network

A system, comprising:

• a non-transitory memory;
• and
• at least one processor coupled to the non-transitory memory, the non-transitory memory storing a program of a neural network including a neural process, the neural network configured to estimate uncertainties in a model, the model modeling a technical system and/or a system behavior of the technical system, wherein the at least one processor is configured to execute the program to perform operations including to:
• determine a model uncertainty as a variance (σz2) of a Gaussian distribution and as a mean value (μz) of the Gaussian distribution via latent variables (z) from a set of contexts (Dc);
• determine a mean value (μy) of an output of the model as a function of an input location (x) using a neural decoder network, wherein the latent variables (z) are used as weights of the neural decoder network, and wherein, for determining the mean value (μy) of the output of the model, the neural decoder network receives, as input, the input location (x) and the latent variables are not forwarded to the neural decoder network;
• ascertain an inadmissible deviation of the system behavior of the technical system from a standard value range based on the mean value (μy) of the output of the model;
• and
• output, in response to ascertaining the inadmissible deviation of the system behavior of the technical system from the standard value range, a warning signal.

A training method for parameterizing a neural network, the neural network being configured to estimate uncertainties in a model, the neural network configured to:

• determine a model uncertainty as a variance (σz2) of a Gaussian distribution and as a mean value (μz) of the Gaussian distribution via latent variables (z) from a set of contexts (Dc), and
• determine a mean value (μy) of an output of the model as a function of an input location (x) using a neural decoder network based on the Gaussian distribution, wherein the latent variables (z) are used as weights of the neural decoder network,
• wherein the model models a technical system and/or a system behavior of the technical system, and the neural network includes at least one neural decoder network, wherein the latent variables (z) are used as the weights of the neural decoder network, and wherein the neural network includes at least one neural encoder network and/or at least one aggregator module, the at least one neural encoder network and/or the at least one aggregator module being configured to determine the model uncertainty as the variance (σz2) of the Gaussian distribution and the mean value (μz) of the Gaussian distribution using the latent variables (z) from the set of contexts (Dc), and wherein the training method comprises the following:
• training of weights for the at least one neural encoder network and/or the at least one aggregator module, wherein the latent variables (z) are the weights of the neural decoder network.