Physical field parameter dynamic matching bone soup cooking control method based on reinforcement learning

Abstract

This invention relates to the field of food processing control technology, and discloses a method for controlling the cooking of bone broth by dynamically matching physical field parameters based on reinforcement learning. The method includes the following steps: A pulse power-off is executed by an electromagnetic heater to create a thermal shock environment within the bone broth fluid medium and induce natural physical rebound. Temperature decay data is collected under zero power conditions. The viscoelastic relaxation time is calculated through nonlinear fitting and inversion. A deep reinforcement learning model performs inference decisions on the viscoelastic relaxation time based on a locking mechanism formula to determine the optimal angular frequency. The electromagnetic heater performs variable frequency heating based on the optimal angular frequency, maintaining the bone broth in a viscoelastic resonance state. The detection cycle is automatically restarted when resonance deviation occurs. This invention achieves indirect measurement of fluid rheological properties and dynamic matching of heating frequency through active thermal excitation and real-time feedback. It utilizes the physical resonance effect to improve energy-mass coupling efficiency and promote emulsification and nutrient extraction.

Images