In high-heat-flux environments, traditional cooling channels often fail to satisfy concurrent requirements for high heat transfer efficiency, temperature uniformity, and minimal pumping power. This study proposes an engineering-oriented topology optimization method for fluid-solid conjugate heat transfer to address the conflict between thermal performance and flow resistance under non-uniform heat sources. We introduce a pseudo-three-dimensional conjugate heat transfer model governed by Darcy’s law. This formulation retains three-dimensional effects, such as sidewall conduction and non-uniform surface heat flux. Moreover, the governing equations are reduced to two dimensions, thereby significantly enhancing computational efficiency. To resolve the discrepancy between Darcy flow and high-Reynolds-number turbulence, the permeability parameter is calibrated against high-fidelity turbulence simulations, ensuring macroscopic consistency with realistic flow behavior. Using this calibrated model, we perform multi-condition topology optimization for various inlet-outlet configurations under non-uniform heat sources. The optimized designs are reconstructed into three-dimensional geometries and validated via numerical simulations. Compared to conventional straight channel designs, the optimized configurations exhibit better performance, demonstrating reduced peak temperatures, enhanced temperature uniformity, and controlled pressure drops. These findings validate the efficacy of the proposed method for advanced thermal management applications.