Noise in microchip electrophoresis signals significantly affects analytical accuracy. To address the limitations of existing filtering methods, a Multi-Objective Optimization Strategy-based Filtering Parameter Optimization Algorithm (MOSLSMA) is proposed. This method formulates the filter parameter optimization process as a dual-objective optimization problem, treating distortion and smoothness as competing objectives. By extending the SLSMA algorithm and incorporating non-dominated sorting techniques, the proposed approach approximates the Pareto optimal solution set. The performance of five commonly used filters is evaluated and compared experimentally. Results show that, compared to conventional multi-objective algorithms, MOSLSMA achieves a more uniformly distributed Pareto solution set. Furthermore, findings indicate that the Savitzky–Golay (SG) filter performs best in terms of signal-to-noise ratio, smoothness, and peak position preservation. This study provides an effective solution for smoothing microchip electrophoresis signals.