Abstract: In response to the problem that existing assistant decision-making modes are unable to fully support the complex decisionmaking thinking process of commanders, an assistant decision-making mode and its implementation method that fits the thought exploration process of commanders are proposed. Firstly, an assistant decision-making mode based on decision-making assumptions and scene matching is proposed. This mode supports the commander´s exploration and trial in complex decision-making problems, by verifying the commander´s decision-making assumptions through multiple factor analysis. Then, a precise design method for assistant decision-making based on "scene-parameter" matching is proposed to reduce the complexity of parameter setting in the process of repeated analysis of multiple factors. By classifying the setting modes of assistant decision-making parameters, the dependency relationship between default parameter types and decision scene elements is proposed. The low-level parameter setting job that is less related to decision-making problems is reduced. The timeliness and pertinence of assistant decision-making are improved. Finally, the typical cases have been used to demonstrate the strong practicality and feasibility of the above assistant decision-making mode and method.