Double-layered model predictive optimization control strategy for temperature uniformity of microwave heating process

Abstract

Based on the application requirements of the dynamic process analysis of heating materials using high-power microwave reactors with multiple microwave sources, a novel layered control framework is proposed for temperature uniformity during the process of heating heterogeneous materials using microwaves. The framework comprehensively considers the material temperature distribution, production process constraints and energy consumption elements. This double-layered model predictive optimization control framework is implemented through two layers. The upper layer is realized by a distributed data computation of the input electric power of microwave sources and temperature target tracking to find the real-time steady-state optimal input electric power based on the material temperature, process constraints, and energy consumption coefficients. In the lower layer, a dynamic predictive control method based on dynamical models of the heating process with multiple microwave sources is used to track the temperature targets and results calculated by the upper layer. A temperature target shifting strategy is used between the upper and lower layers to eliminate model prediction errors. The effectiveness of the proposed double-layered control framework was validated by heating alumina ceramic in a ducted microwave reactor with six microwave sources.