The invention discloses a parallel-based self-adaptive decision-making efficiency optimization method which mainly solves the problems of low efficiency, insufficient real-time performance and large search overhead in the existing decision-making process based on search. The method is characterized in that in a self-adaptive decision-making process, dynamic allocation scheduling is conducted on self-adaptive search tasks by adopting a parallel task allocation scheduling method, parallel processing is conducted on the self-adaptive search tasks by utilizing the multi-threading technology, parallel on the two levels of in-between self-adaptive search tasks and within self-adaptive search tasks is achieved, and finally optimized self-adaptive decision-making is obtained. The parallel-based self-adaptive decision-making efficiency optimization method increases the search rate of self-adaptive adjustment scheme, improves the search efficiency, meets the requirements that self-adaptive systems need to adjust rapidly to changes, and can be applied to self-adaptive systems.