Training-free data-driven method for input-output modeling of complex process

A twin roll casting system, comprising:

• a pair of counter-rotating casting rolls having a nip between the casting rolls and capable of delivering cast strip downwardly from the nip;
• a tundish to deliver molten metal to a casting pool above the nip;
• a casting controller configured to receive a plant input comprising at least one process control setpoint;
• a sensor capable of measuring at least one parameter of the casting process;
• and
• a feedback controller coupled to a cast strip sensor to receive casting process measurement signals from the sensor and coupled to the casting controller, the feedback controller comprising a data-driven model comprising a database of state-input pairs;
• wherein the feedback controller is configured to execute the following steps at each time step:
• (a) measure a state-input similarity between a new state observation and samples in the database,
• (b) assign a weight to each consequent output of samples in the database based on the measured state-input similarity, and
• (c) sum the weighted outputs and predict an output of a new state observation;
• wherein the feedback controller is configured to provide the at least one process control setpoint to the casting controller based on the predicted output of the new state observation.