软件安全性3M评价中评价因子修正的研究

安全性苛求系统需要防范意外事件造成对人体，环境或大宗财产的损害，为此，对软件需要进行严格的安全性测试，并对软件防范意外事件的能力进行评价，然而，完全依靠试验统计或事故累积数据进行安全性评价。会由于测试开销过大或等待累积时间过长影响评价的准确性和及时性，需要利用软件开发过程中涉及的诸多要素和在各个阶段来自多个方面所提供的信息参与安全性评价，但是，在这些信息中包含了某些主观的成分有待验证，在多元，多模型，多阶段（3M）安全性软件评价方法的框架中，有可能利用后续测试得到的信息不断对评价因子进行修正，提高评价的准确性，讨论了后续测试对3M模型中的复杂性，技术支持和开发水平因子的修正方法，着重论述对开 水平因子的修正，提出了后续测试信息和先期测试信息进行比较和处理的过程和一些规则。

Modification for Assessment Parameters Made by Subsequent Testing of Software Safety 3M Assessment

Safety-critical systems need to prevent damages to people lives,environments or mass assets caused by unexpected events.Therefore,software has to undergo safety testing strictly,and to be assessed for the capacity of avoiding unexpected events.However,safety assessment depending completely on experimentally statistical or accumulative accident data will be hampered in its accuracy and timeliness by high testing expenditure or long accumulative waiting time.Many related factors in the process of software development and information from various aspects in every phase are included for safety assessment.But there are some subjective elements to be verified in the information.In the multi-source,multi-model and multi-stage(3M) framework of the safety-critical software assessment method,it's possible to modify assessment parameters continuously by utilizing information obtained from subsequent testing to improve assessment accuracy.The modification method of complexity,technological support and development level parameter in the 3M is discussed with the modification of development level parameter emphasized on.Finally,some procedures and rules in processing and comparison between previous and subsequent testing information are presented in this paper.