Linearized Matrix Inversion and Eigenvalue Methods for Characterization of Thermal System Performance for Control Applications Using TMG Thermal Simulation and MATLAB

Successful implementation of temperature control is required in a variety of engineering applications. A method using the I-deas TMG finite volume thermal solver and MATLAB with the Control System Toolbox to characterize the control system performance of a thermal model was described by Saeger [1][2]; this paper describes an alternate approach to this method.

The alternate method presented herein uses the TMG thermal solver to create linearized capacitance and conductance matrices of the thermal system. Then, MATLAB uses these matrices to create the frequency response curve using either matrix inversion or eigenvalue solution methods. Once the frequency response is generated, the MATLAB Control System Toolbox can be used to develop the control system.