To　explore　the　difference　of　the　factors　influencing　stress　distribution　for　large-scale　component,　an　automatic　mechanism　of　acquiring　stress　distribution　characteristics　and　influence　knowledge　was　established.　For　the　large-scale　component　with　continuous　changing　geometry,　a　stress　survey　method　was　proposed　to　extract　maximum　stress　of　each　sub-region　for　every　sample.　After　evaluating　the　danger　situation,　some　characteristic　regions　were　determined　and　the　characteristic　stress　sets　were　acquired.　Furthermore,　the　influences　of　structural　parameters　for　characteristic　stresses　and　lightweight　index　were　analyzed.　By　constructing　knowledge　reasoning　model　based　on　multi-states　adjusting　strategy,　the　main　influence　factors　as　well　as　their　saliency　under　different　expectations　were　acquired　to　reflect　the　priority　of　structural　parameters　for　adjusting.　Finally,　the　gooseneck-type　boom　was　taken　as　an　example,　which　demonstrates　that　the　process　of　modeling,　analysis,　feature　extraction　and　knowledge　acquisition　can　be　realized　automatically　and　the　useful　knowledge　can　be　acquired　efficiently　and　flexibly　for　the　intelligent　optimization　of　large-scale　component.　©　2019,　Editorial　Department　of　JOURNAL　OF　MECHANICAL　STRENGTH.　All　right　reserved.