電力中央研究所

報告書「電力中央研究所報告」は当研究所の研究成果を取りまとめた刊行物として、昭和28年より発行されております。 一部の報告書はPDF形式で全文をダウンロードすることができます。

※ PDFのファイルサイズが大きい場合には、ダウンロードに時間がかかる場合がございます。 ダウンロードは1回のクリックで開始しますので、ダウンロードが完了するまで、複数回のクリックはなさらないようご注意願います。

電力中央研究所 報告書(電力中央研究所報告)

報告書データベース 詳細情報


報告書番号

ER94001

タイトル(和文)

SOFTWARE FAULT REDUCTION USING CASE TOOLS

タイトル(英文)

SOFTWARE FAULT REDUCTION USING CASE TOOLS

概要 (図表や脚注は「報告書全文」に掲載しております)

概要 (英文)

ELECTRIC UTILITIES ARE INCREASINGLY DEVELOPING NEW TYPES OF SOFTWARE-BASED SYSTEMS SUCH AS KNOWLEDGE-BASED SYSTEMS OR NEURAL NETWORKS. HOWEVER,WOULD-BE USERS OF THESE SYSTEMS FACE THE ISSUE OF QUALITY ASSURANCE BECAUSE THE DEVELOPERS DO NOT PROVIDE QUALITY ASSURANCE AT THE SAME HIGH LEVELS AS CONVENTIONAL SOFTWARE DEVELOPERS. THE REASON IS MAINLY LACK OF A SYSTEMATIC PROCESS THAT PREVENTS AND DETECTS FAULTS AT EVERY STEP OF DEVELOPMENT. THIS REPORT DESCRIBES A METHODOLOGY NAMED FAULT REDUCTION METHODOLOGY THAT PROVIDES SUCH A PROCESSFOR REDUCING THE NUMBER OF FAULTS IN SOFTWARE BY USING CASE TOOLS. THEMETHODOLOGY IS BASED ON ENUMERATION OF FAULT TYPES AND ANALYSIS OF PREVENTIVE AND DETECTIVE METHODS FOR EACH TYPE OF FAULT. THE METHODOLOGY REDUCES THE POSSIBILITY OF FAULTS BY FOUR STEPS. (1)DESIGN A DEVELOPMENTPROCESS COMBINING SOFTWARE LIFECYCLE COMPONENTS CALLED KERNELS CONSIDERING FEATURES OF THE DEVELOPED SYSTEM. (2)SELECT REPRESENTATIONS OF THEARTIFACTS OF SOFTWARE KERNELS BASED ON FAULT CONTROL LEVEL FOR EACH REPRESENTATION TO REDUCE THE PROBABILITY OF FAULTS. (3)SELECT A SET OF V&V METHODS TO DETECT IDENTIFIED IMPORTANT TYPES OF FAULTS EFFECTIVELY BASED ON AN EFFECTIVENESS MEASURE. (4)USE CASE TOOLS THAT SUPPORT SELECTED REPRESENTATIONS OR V&V METHODS EFFECTIVELY.

報告書年度

1994

発行年月

1995/03/01

報告者

担当氏名所属

篠原 靖志

情報研究所情報科学部人工知能グループ

キーワード

和文英文
KNOWLEDGE-BASED SYSTEMS

RELIABILITY

VERIFICATION

DEVELOPMENT METHODOLOGY

CASE TOOLS
Copyright (C) Central Research Institute of Electric Power Industry