In order to quantify the design process and construct the decision model of general arrangement for fighters， a boundary design method and two clustering analysis methods for general arrangement are proposed. Firstly， a boundary fluctuation risk indicator is introduced， which can be utilized to update an initial cabin definition to a refined cabin definition by non-negativity adjustment. Simultaneously， based on the ranking of boundary fluctuation risk indicators from high to low， p-cut strong boundaries can be selected and a method for constructing p-cut strong regions using these boundaries is defined. The p-cut strong regions can serve as classes in clustering analysis for general arrangement. Then， two methods are proposed for implementing clustering analysis： Naive Bayes-based clustering （NBC） method and fuzzy analytic hierarchy process-based clustering （FAHPC） method. By feature recognition of equipment， NBC method applies naive Bayes classifier for classification after feature encoding. FAHPC method first constructs an indicator system for general arrangement and uses expert scoring method to determine the membership degree of each indicator relative to the evaluation comment set， and then uses fuzzy analytic hierarchy process to determine weights between the criteria layer and the indicator layer. Finally， normalized comprehensive scores are used as membership degrees to determine classification results. An engineering example validates the feasibility and practicality of the proposed methods.