Control of cyber-physical systems under uncertainty

A computing system, comprising:

• at least one processor;
• and
• memory that comprises computer-executable instructions that, when executed by the at least one processor, cause the at least one processor to perform acts including:
• providing a probabilistic framework, wherein the probabilistic framework enables constraints to be defined for synthesis of control inputs of a cyber-physical system;
• forming the constraints for the cyber-physical system based on the probabilistic framework, wherein the constraints are based on predictions that comprise probability distributions over sets of possible values of phenomena, wherein the phenomena pertain to at least one of the cyber-physical system or an environment in which the cyber-physical system operates;
• and
• synthesizing the control inputs for the cyber-physical system that satisfy the constraints.

A method of controlling a cyber-physical system, comprising:

• combining a first prediction generated utilizing a first classifier and a second prediction generated utilizing a second classifier, wherein the first prediction comprises a first probability distribution and the second prediction comprises a second probability distribution, and wherein constraints that maintain safe operation of the cyber-physical system in an environment are based on a combination of the first prediction and the second prediction;
• synthesizing control inputs for the cyber-physical system that satisfy the constraints;
• and
• causing the cyber-physical system to operate in the environment based on the control inputs.