Abstract: As a new promising sound absorption material, micro-perforated panel absorbers (MPPAs) have been successfully applied in various noise control occasions, even capable of working at harsh environments, including high temperature circumstances. According to Maa's model, the sound absorption performance of a MPPA can be theoretically predicted as four parameters were chosen, i.e., the perforation diameter d, the panel thickness t, the distance between centers of adjacent perforations b, and the depth of the air gap D. However, the Maa's model turns to be inaccurate at high temperatures. In this work, the temperature T is included into Maa's model as the fifth parameter besides the four parameters. Simulation results show that better prediction accuracy could be obtained at high temperatures with the revised model. Furthermore, an improved particle swarm optimization algorithm is proposed to search for optimal structure parameters of mono-layer and double-layer MPPAs at given temperatures, which will facilitate the design work for MPPAs working at high temperatures.