System and method for real-time model-based close-loop control function modification in technical processes of a vehicle control system or manufacturing plant

A vehicle control system for automated driving control, the vehicle control system comprising:

• a sensor system;
• and
• a processing system that includes one or more processors;
• wherein:
• the processing system is configured to:
• conduct automated driving control of the vehicle by executing a plurality of functions for guiding the vehicle along a path;
• during the automated driving control of the vehicle, obtain real-time sensor data from the sensor system, the sensor data comprising information characterizing an environment of the vehicle and/or one or more operating states of vehicle components or subsystems;
• and
• analyze the real-time sensor data to determine a current observation that includes input variables and output variables of a first function and of a second function, the first function comprises a perception function of the control system, and the second function comprises a control or actuation function of the control system;
• the analysis is a model-based analysis in which:
• a first model is provided for the first function, and a second model is provided for the second function;
• the first model models a functional relationship between an input variable and an output variable of the first function and an influence of a first disturbance variable as a function of a first parameterization and as a function of a predefined learned distribution over the first parameterization;
• and
• the second model models a functional relationship between an input variable and an output variable of the second function and an influence of a second disturbance variable as a function of a second parameterization and as a function of a predefined learned distribution over the second parameterization;
• the processing system is configured to:
• use the first and second models and the current observation to determine a prediction regarding an effect of:
• i) changing the first function under the influence of the first disturbance variable, and/or ii) changing the second function under the influence of the second disturbance variable;
• select, based on the prediction, whichever one of the first and second functions has been predicted, by the prediction, to have a higher likelihood of achieving a desired effect in operation of the vehicle, compared to the change of another one of the first function and second functions;
• and
• update the selected one of the first and second functions, during continued automated driving control of the vehicle, by modifying one or more operating parameters of the selected function based on the respective one of the parameterizations of the respective one of the models that is associated with the selected one of the first and second functions, thereby altering a behavior of the selected function in real-time operation;
• and
• the processing system is configured to perform the obtainment of the real-time sensor data, the analysis, the prediction determination, the selection, and the update as part of a closed-loop process that occurs during driving operation of the vehicle, such that the continued automated driving control of the vehicle is influenced by successive updates to the selected one of the first and second functions based on the real-time sensor data and model-based analysis.

A non-transitory computer-readable medium on which is stored a computer program including computer-readable instructions that are executable by a processing system of a vehicle for controlling automated driving operations of the vehicle, the instructions, when executed by the processing system, causing the processing system to perform a method, the method comprising:

• conducting automated driving control of the vehicle by executing a plurality of functions for guiding the vehicle along a path;
• during the automated driving control of the vehicle, obtaining real-time sensor data from a sensor system of the vehicle, the sensor data comprising information characterizing an environment of the vehicle and/or one or more operating states of vehicle components or subsystems;
• analyzing the real-time sensor data to determine a current observation that includes input variables and output variables of a first function and of a second function, wherein:
• the first function comprises a perception function of the control system, and the second function comprises a control or actuation function of the control system;
• and
• the analyzing is a model-based analysis in which:
• a first model is provided for the first function, and a second model is provided for the second function;
• the first model models a functional relationship between an input variable and an output variable of the first function and an influence of a first disturbance variable as a function of a first parameterization and as a function of a predefined learned distribution over the first parameterization;
• and
• the second model models a functional relationship between an input variable and an output variable of the second function and an influence of a second disturbance variable as a function of a second parameterization and as a function of a predefined learned distribution over the second parameterization;
• using the first and second models and the current observation to determine a prediction regarding an effect of:
• i) changing the first function under the influence of the first disturbance variable, and/or ii) changing the second function under the influence of the second disturbance variable;
• selecting, based on the prediction, whichever one of the first and second functions has been predicted, by the prediction, to have a higher likelihood of achieving a desired effect in operation of the vehicle, compared to the change of another one of the first function and second functions;
• and
• updating the selected one of the first and second functions, during continued automated driving control of the vehicle, by modifying one or more operating parameters of the selected function based on the respective one of the parameterizations of the respective one of the models that is associated with the selected one of the first and second functions, thereby altering a behavior of the selected function in real-time operation;
• wherein the method is performed as part of a closed-loop process that occurs during driving operation of the vehicle, such that the continued automated driving control of the vehicle is influenced by successive updates to the selected one of the first and second functions based on the real-time sensor data and model-based analysis.

A method for operating a manufacturing plant comprising a plurality of production cells configured to carry out respective manufacturing functions, the method comprising:

• conducting an automated operation of the manufacturing plant by a control system that executes a plurality of functions for managing the manufacturing functions performed by the production cells;
• during the automated operation of the manufacturing plant, obtaining, by a processing system that includes one or more processors, real-time sensor data from a sensor system of the manufacturing plant, the sensor data comprising information characterizing the plant;
• analyzing, by the control system, the real-time sensor data to determine a current observation that includes input variables and output variables of a first function and of a second function, wherein:
• the first function comprises a monitoring or quality control function applied to a first production cell, and the second function comprises a control or actuation function applied to a second production cell;
• and
• the analyzing is a model-based analysis in which:
• a first model is provided for the first function, and a second model is provided for the second function;
• the first model models a functional relationship between an input variable and an output variable of the first function and an influence of a first disturbance variable as a function of a first parameterization and as a function of a predefined learned distribution over the first parameterization;
• and
• the second model models a functional relationship between an input variable and an output variable of the second function and an influence of a second disturbance variable as a function of a second parameterization and as a function of a predefined learned distribution over the second parameterization;
• using, by the processing system, the first and second models and the current observation to determine a prediction regarding an effect of:
• i) changing the first function under the influence of the first disturbance variable, and/or ii) changing the second function under the influence of the second disturbance variable;
• selecting, by the processing system and based on the prediction, whichever one of the first and second functions has been predicted, by the prediction, to have a higher likelihood of achieving a desired effect in operation of the manufacturing plant, compared to the change of another one of the first function and second functions;
• and
• updating, by the processing system, the selected one of the first and second functions, during continued automated operation of the manufacturing plant, by modifying one or more operating parameters of the selected function based on the respective one of the parameterizations of the respective one of the models that is associated with the selected one of the first and second functions, thereby altering a behavior of the selected function in real-time operation;
• wherein the method is performed as part of a closed-loop process that occurs during operation of the manufacturing plant, such that the continued automated operation of the plant is influenced by successive updates to the selected one of the first and second functions based on the real-time sensor data and model-based analysis, thereby affecting characteristics of products produced by the manufacturing plant.