Synergistic optimization of thawing efficiency and water-holding capacity in chicken breast via magneto-electric...

Traditional chicken breast thawing suffers from severe phase-transition instability and quality degradation. To overcome limitations of standalone fields, a novel magneto-electric coupling thawing (MECT) system integrating pulsed magnetic fields (PMF) and high-voltage electrostatic fields (HVEF) was developed. A decoupled multi-input single-output back-propagation neural network (MISO-BPNN) framework precisely mapped the nonlinear kinetics. Featuring a "Slope-Forced Alignment" penalty mechanism, this MISO architecture eliminated systematic bias and task interference, demonstrating exceptional predictive fidelity (R > 0.97) for thawing time and moisture loss. To resolve the inherent efficiency-quality trade-off, a multi-objective Desirability function was globally optimized via parallel heuristic algorithms (GA, PSO, SA). The industrial-adapted optimum (4.0 mT, 0.2 Hz, 2.0 kV/cm) synergistically yielded a reduced thawing time (8.37 h), thawing loss (3.93%), and cooking loss (16.82%). With experimental relative errors strictly below 5%, MECT proved highly effective in enhancing both efficiency and water-holding capacity. This MISO-based predictive modeling framework provides a robust paradigm for intelligent and adaptive control of complex multi-field coupled food processing systems.