电讯技术
電訊技術
전신기술
TELECOMMUNICATIONS ENGINEERING
2014年
10期
1446-1450
,共5页
航空雷达%故障诊断%粗糙集%β上近似属性约简%规则融合
航空雷達%故障診斷%粗糙集%β上近似屬性約簡%規則融閤
항공뢰체%고장진단%조조집%β상근사속성약간%규칙융합
airborne radar%fault diagnosis%rough sets%β upper approximation reduction%rule fusion
航空雷达故障诊断的一个主要目标是求出某种程度上可能发生的故障类型,为此,首先引入β上近似属性约简,定义了β上近似决策辨识集和辨识矩阵,给出了求取β上近似约简集的方法;其次,提出了一种基于包含度的规则融合方法,该方法充分挖掘属性值信息,可以得到全部属性条件下的带可信度的决策规则,有效解决了在各类型故障样本数目相差较大的情形下的规则融合问题;最后,以某型航空雷达的整机测试为例,在基于β上近似约简对其测试参数进行约简的基础上,应用基于包含度的规则融合方法,得到了全部属性取值条件下的航空雷达故障诊断规则。结果表明:该方法直观有效、计算简便,为航空雷达故障诊断提供了新的思路和途径。
航空雷達故障診斷的一箇主要目標是求齣某種程度上可能髮生的故障類型,為此,首先引入β上近似屬性約簡,定義瞭β上近似決策辨識集和辨識矩陣,給齣瞭求取β上近似約簡集的方法;其次,提齣瞭一種基于包含度的規則融閤方法,該方法充分挖掘屬性值信息,可以得到全部屬性條件下的帶可信度的決策規則,有效解決瞭在各類型故障樣本數目相差較大的情形下的規則融閤問題;最後,以某型航空雷達的整機測試為例,在基于β上近似約簡對其測試參數進行約簡的基礎上,應用基于包含度的規則融閤方法,得到瞭全部屬性取值條件下的航空雷達故障診斷規則。結果錶明:該方法直觀有效、計算簡便,為航空雷達故障診斷提供瞭新的思路和途徑。
항공뢰체고장진단적일개주요목표시구출모충정도상가능발생적고장류형,위차,수선인입β상근사속성약간,정의료β상근사결책변식집화변식구진,급출료구취β상근사약간집적방법;기차,제출료일충기우포함도적규칙융합방법,해방법충분알굴속성치신식,가이득도전부속성조건하적대가신도적결책규칙,유효해결료재각류형고장양본수목상차교대적정형하적규칙융합문제;최후,이모형항공뢰체적정궤측시위례,재기우β상근사약간대기측시삼수진행약간적기출상,응용기우포함도적규칙융합방법,득도료전부속성취치조건하적항공뢰체고장진단규칙。결과표명:해방법직관유효、계산간편,위항공뢰체고장진단제공료신적사로화도경。
The acquisition of decisions with some extent reliability is one of important aids in airborne radar fault diagnosis. The β upper approximation reduction is firstly introduced, moreover β upper approximation discernibility sets and discernibility matrix are defined and the acquirement approach of βupper approxima-tion reduction is given. Secondly,the rule fusion method is offered to sufficiently mine the attribute-value in-formation and the decision-making rules with reliability cover all values of the attributes are obtained, which efficiently solve the rule fusion problem in the condition of the very different number among all fault kinds. Finally, taking a test of an airborne radar as an example, this paper cuts test parameter based onβupper ap-proximation reduction and applies the rule fusion method to obtain the airborne radar fault diagnosis rules which cover all values of the attributes. Results show that the methods in this paper are visually effective and easy calculation, which offer a new though and approach for airborne radar fault diagnosis.