A liquid crystal variable retarder (LCVR) is the core device to realize fast and high-precision broadband polarization imaging, and its ability to suppress the noise will have an impact on the polarization measurement results. In order to obtain better imaging quality and measurement accuracy, it is crucial to solve the optimization problem of the LCVR. In this paper, the optimal objective function for solving the optimization problem of the LCVR is analytically derived and verified based on the genetic algorithm in the band range of 350-700 nm. Meanwhile, considering that the minimum number of four measurements at this time cannot achieve the optimal state, the relationship between the number of measurements and the overall performance relative to the error propagation (optimized conditions number) is discussed. The results show that a better optimal set of angles can be obtained by using the optimal objective function. In this paper, a set of the most favorable angles is obtained, and the optimized average of the CN is 2.0000, which is reduced by 0.32% compared with previous optimization results and is closer to the ideal value of the CN. In addition, in this paper, the noise immunity of the set of most favorable angles is simulated and analyzed, and the optimized system can effectively improve the measured performance of the wide-band liquid crystal variable retarder polarimeter.