(2) REGRESSION
CONTROL CHART
Group: process
control
There
is no old definition
NEW DEFINITION
These charts are also known as
regression control charts. In many processes the process average may
continuously run either upward or downward after production of every unit of
product. This is a natural phenomenon and therefore, it is an acceptable trend.
Examples are effects of wearing of the punch, die, cutting tools or drill bits.
However, such a trend in the process mean is acceptable only within some upper
and lower limits (in most cases the specification limits). The trend charts are
developed to monitor and control these types of processes. The centerline of the trend chart will have an
upward or downward trend, and the upper and lower control limits will be
parallel to the centerline. The intersection of centerline a and the slope b
can be evaluated from the observations collected from the process [1]. the
equations for the control limits are
These charts are useful for
detecting changes in the process and also to decide whether or not a tool
change is required. These charts are very helpful in controlling processes in
machine shops and other production machines.
Mandel (1969) presented a
regression control chart and provided a number of administrative applications.
The general concept differs only slightly from a prediction interval for Y.
In fact, the assumptions
for a regression control chart are exactly the same as those for simple linear
regression. The general idea is to monitor the dependent variable using a
control chart approach. The centerline on the chart is at
with
k = 2 or 3, which would seem to be the natural choice for obtaining the control
limits. The argument given is that the use of control limits obtained from Y ±
2s will provide tighter control for extreme values of Xo, which is where tight
control is generally desired. Certainly that could be debated. An
out-of-control state, relative to the fitted regression model, is suspected of
existing when a value of Y plots outside the control limit lines defined by Eq.
(12.6), or outside the lines defined by Eq. (12.5), if preferred. The control
lines using Eq. (12.6) will be curved since the value of the expression under
the radical depends on the value of Xo.
Thomas
P. Ryan, Statistical Methods for Quality Improvement, pg.417
Krishna
B. Misra, Handbook of Performability Engineering, pg. 193
(3) SYSTEMS
DEVELOPMENT LIFE-CYCLE (SDLC)
Group:
database design
There
is no old definition
NEW DEFINITION
System Development Life Cycle
SDLC is also sometimes known as SLC. And it essentially refers to the
development of a computer system for the company. Through the use of various
systems developed and bought from various vendors an information system is
developed. The independent systems would have to be integrated to produce a
single coordinated system that would answer the needs of the company.
SDLC undergoes through a series
of phases before it is completed. But because it is a cycle, some steps have to
be iterated for the system to be efficient. SDLC begins when the company
management decides what they actually need from the system. They would usually
decide what capabilities the system should have and they will also define the
general objective of the project. After that a feasibility study is done to
check whether the project is quite doable in terms of the cost of developing
the system versus the benefits it will bring. Depending on the feasibility
study, the company may decide to upgrade their current system instead of
developing a new one. But if they decide to push through with it, fact finding
is done to see as to the how the current system is working and how the users
are at them. Al this stage, the detailed requirements of the end users are
defined. Several models would then be designed and then system specifications
will be done. Then the company would be ready to implement the system and
review its performance. If the new system seems fit and efficient, the old
system will then be replaced with it. The new system would then be used in the
operations of the company.
(Jeremy
Lewis, SDLC 100 Success Secrets, pg. 111)
Peter
Rob,Carlos Coronel, Database Systems: Design, Implementation, and Management,
Pg.375)
No comments:
Post a Comment