Using active non-destructive state excitation of a physical system to model and control operations of the physical system

A computer-implemented method comprising:

• generating, by one or more first computing systems, a behavioral model that is for a physical system with a fuel cell and that has non-linear response behavior to linear changes in one or more inputs to the physical system, wherein the behavioral model represents behavior of the physical system at multiple levels for each of at least a first type of factor and a second type of factor that include charge level of the fuel cell and resistive load applied to the fuel cell and that affect the behavior of the physical system, by:
• supplying a plurality of input changes to the physical system by, for each of a plurality of input groups that each includes a combination having one of the multiple first levels for the first type of factor and one of multiple second levels for the second type of factor, supplying one or more of the input changes into the physical system using the one first level of the first type of factor for that input group, wherein the plurality of input changes include electrical signals supplied to at least one of a cathode of the fuel cell or an anode of the fuel cell;
• measuring changes in responses of the physical system to, for each of the plurality of input groups, the supplied one or more input changes for that input group;
• and
• combining the measured changes in the responses to produce the generated behavioral model for the physical system, to cause the generating of the behavioral model to include generating one or more incremental parametric non-linear behavioral models for the fuel cell;
• and
• controlling, by one or more second computing systems and after the generating of the behavioral model, further operations of the physical system by using the generated behavioral model and state information for the physical system to determine one or more control actions to perform, and performing the one or more control actions to modify the further operations of the physical system.

A non-transitory computer-readable medium having stored contents that cause one or more computing systems to perform automated operations, the automated operations comprising:

• generating, by at least one first computing system of the one or more computing systems, a behavioral model for a fuel cell that represents behavior of the fuel cell at multiple charge levels and with multiple resistive loads, by:
• supplying a plurality of electrical signals into the fuel cell by, for each of a plurality of combinations each including one of the multiple charge levels and one of the multiple resistive loads, supplying one or more of the electrical signals into the fuel cell while the fuel cell has the one charge level for that combination and the one resistive load for that combination;
• measuring changes in responses of the fuel cell to, for each of the plurality of combinations, the supplied one or more electrical signals for that combination;
• and
• combining the measured changes in the responses to produce the generated behavioral model for the fuel cell;
• and
• controlling, by at least one second computing system of the one or more computing systems and after the generating of the behavioral model, further operations of the fuel cell by using the generated behavioral model and current state information for the fuel cell to determine one or more control actions to perform, and performing the one or more control actions to modify the further operations of the fuel cell.

A system comprising:

• one or more hardware processors of one or more computing systems;
• and
• one or more memories with stored instructions that, when executed by at least one of the one or more hardware processors, cause the system to perform automated operations including at least:
• generating a behavioral model for a fuel cell that represents behavior of the fuel cell at multiple charge levels and with multiple resistive loads, by:
• supplying a plurality of electrical signals into the fuel cell by, for each of a plurality of combinations each including one of the multiple charge levels and one of the multiple resistive loads, supplying one or more of the electrical signals into the fuel cell while the fuel cell has the one charge level for that combination and the one resistive load for that combination;
• measuring changes in responses of the fuel cell to, for each of the plurality of combinations, the supplied one or more electrical signals for that combination;
• and
• combining the measured changes in the responses to produce the generated behavioral model for the fuel cell;
• and
• controlling, after the generating of the behavioral model, operations of an indicated fuel cell by using the generated behavioral model and state information for the indicated fuel cell to determine one or more control actions to perform, and performing the one or more control actions to modify the operations of the indicated fuel cell.