Aiming at the problem of poor error-correction performance of the Successive Cancellation (SC) decoding algorithm for polar codes, a parameter adaptive noise hybrid perturbation SC (PANHPSC) decoding algorithm is proposed. The algorithm innovatively introduces a dual dynamic adjustment mechanism. On one hand, a scaling factor α enables the perturbation power to automatically adapt to the channel noise. On the other hand, based on the attempt count and the received signal amplitude, a biased correction amount can dynamically be calculated by utilizing an increment factor γ and an error-evidence factor β. Thus the perturbation amplitude is linearly increased during the decoding process and scaled according to the confidence level. Simulations demonstrate that the proposed algorithm can achieve the performance superior to traditional perturbation schemes without requiring the complex parameter tuning and approach the performance level of the SCL decoding algorithm in the relatively low complexity.