Abstract: The scientific detection performance of a space-based gravitational wave detection constellation is jointly affected by instrumental noise, inter-satellite link operation, effective observation time, and the characteristics of target gravitational wave sources. A single metric, such as detector sensitivity or signal-to-noise ratio, is insufficient to fully characterize the scientific capability of such a mission under realistic system and operational constraints. To address this issue, this paper develops a performance evaluation system for space-based gravitational wave detection constellations. The system is organized around three categories of evaluation indicators: instrument performance, operational availability, and scientific response. Corresponding functional modules are established, including dynamic full-link simulation, effective detection time assessment, and signal-to-noise ratio calculation. Based on the MATLAB/Simulink platform, the heliocentric Taiji-like mission scheme is selected as a representative case for simulation verification. The results show that the system can obtain key performance quantities, including residual acceleration noise of the test masses, optical metrology system noise, and detector sensitivity. It can also simulate the influence of three-spacecraft link acquisition and link reconstruction after interruption on the effective scientific observation time. Furthermore, by combining the detector sensitivity, effective observation time, and characteristic strains of representative gravitational-wave sources, the system evaluates the signal-to-noise ratios of verification galactic binaries, an extreme mass-ratio inspiral system, and massive black hole binary systems. The obtained results demonstrate that the proposed system can characterize the visibility differences among different types of sources and provide a feasible framework for scientific performance assessment of space-based gravitational wave detection missions. The system can support mission scheme comparison, system-level requirement decomposition, and analysis of the degree to which scientific objectives can be satisfied.