Page 73 - 《软件学报》2020年第11期

康燕妮 等:软件需求变更管理的系统动力学仿真建模 3389

Fig.10 Requirements change management subsystem
图 10 需求变更管理子系统

在该子系统中,需求变更请求创建率(issue created rate)表示项目中每周需求变更请求创建的数量,单位为
Dmnl,计算公式如(16)所示.
issue created rate = ⎫
⎬ (16)
-
IF THEN ELSE "( Base line issue created rate "≥ 0:AND:Number of users≥ 0,"Base line issue created ",0) ⎭
-
处于“to do”状态的需求变更请求(number of to do issue)是需求变更等待实现的数量,单位为 issue,计算公式
如(17)所示.
⎛ ⎛ accepted issue rate Return to do rate − ⎞ ⎞⎛ ⎞
+
⎜ ⎜ ⎟ ≥ 0, ⎟⎜ ⎟
⎜
do
Number of to issue = INTEG IF THEN ELSE ⎜ ⎝ issue change to sprint rate ⎠ ⎜ ⎟ ⎟ ,95 ⎟ ⎟ (17)
⎜
+
⎜ ⎜ ⎛ ⎜ accepted issue rate Return to do rate − ⎞ ⎟ ,0 ⎟ ⎟
⎜ ⎜ issue change to sprint rate ⎟ ⎟
⎝ ⎝ ⎠ ⎠⎝ ⎠
已实现需求变更总数(resolved & closed issue)的单位为 issue,计算公式如(18)所示.
“resolved & closed issue”=INTEG(“resolved & closed rate”,0) (18)
无效需求变更请求百分比(percent of invalid issue)表示无效需求变更请求在需求变更总数所占百分比,单
位为 Dmnl,计算公式如(19)所示.
percent of invalid issue = ⎫
⎬ (19)
IF THEN ELSE(TOTAL issue number ≠ 0,Number of invalid issue /TOTAL issue number ,0) ⎭
接受需求变更请求百分比(percent of accept issue)表示接受需求变更请求在需求变更总数所占百分比,单
位为 Dmnl,计算公式如(20)所示.
percent of accept issue = IF THEN ELSE (TOTAL issue number ≠ 0, ⎫
(Number of sprint issue + Number of to do issue + Number of waiting to rework issue + ⎪ ⎬ (20)
Number of waiting to test issue + "resolved & closed issue ) "/TOTAL issue number , 0) ⎪ ⎭

1.4 开源软件需求变更管理系统动力学模型检测
建模工作完成后,对模型进行检测是为了保证模型结构和行为的有效性,确定模型与建模目的相符,发现模
型的缺陷并提高模型的有用性.从建立模型真实性的角度来说,在系统动力学领域,仿真模型完全的有效性是不
可能的,我们只能从仿真模型的目的出发,选择最为恰当的模型.因此在建模的过程中,我们从两方面来尽量保