Efficient Algorithm for Solving Inverse Source Problems of a Nonlinear Diffusion Equation in Microwave Heating

Determination of the equivalent internal heat source from surface temperature measurements in microwave processing of materials is formulated as an inverse source problem of a nonlinear diffusion equation. The versatile generalized pulse-spectrum technique (GPST) inversion algorithm with the incorporation of multi-level grid method and hierarchical parallelism is developed for solving this type of inverse problems. Development of a simple 2D code has been completed. Numerical simulations are carried out to test the feasibility and capability of this improved GPST without the real measurement data. It is found that this new inversion algorithm not only does produce very good results but also is much more efficient and stable than its standard version.