Polarized light orientation is a new navigation method based on atmospheric polarized light field, which has the advantages of anti-interference and no accumulation of errors, can provide a technical way for the autonomous orientation of micro-unmanned platforms under the condition of satellite rejection. Under complex weather conditions, the orientation accuracy of polarized light decreases, and the existing orientation error analysis and research are not systematic enough. In this paper, the orientation error model of polarized light is constructed based on Rayleigh atmosphere model. Secondly, the model is modeled for solar position error, horizontal attitude angles error, incident light polarization angle error, light intensity error, time error and position error, and the simulation quantitative analysis is carried out. The results show that: when the solar height angle is close to 90°, the solar height angle error is the main error source affecting the orientation. When solar altitude is small, the horizontal attitude angles error, the incident light polarization angle error and the light intensity error are the main influencing factors of orientation accuracy, of which horizontal attitude angles error is the most important. In position error, longitude error is the main error source relative to latitude error. Under the positioning accuracy of 0.2° and UTC timing, orientation error caused by position error and time error is less than 0.1°. In addition, a feasible way to reduce the orientation error caused by the light intensity error is found through error analysis. When the noise amplitude reaches half of the light intensity amplitude, the simulation results show that the orientation error is reduced from 0.85° to 0.23°. This discovery lays a foundation for further improving the environmental applicability and availability of the microarray polarized light compass.