MATH30006 Quality Control and Forecasting SEMESTER: First
CONTACT: Professor T Subba Rao (Ferranti/C5) CREDIT RATING: 10
Aims: To present various standard acceptance sampling plans, quality control methods for monitoring the quality of manufactured items, time series methods for the modelling of data and forecasting using these models.
Intended Learning Outcomes: On completion of the course students will be able to:
  • Design single and double sampling plans with optimum properties;
  • Construct quality control limits and assess their merits;
  • Fit simple time series models to real data and calculate forecasts and their standard errors.
Pre-requisites: 114, 257 (Statistical Theory and Methods).
Dependent Courses: None.
Course Description: Modern quality control has evolved from a purely statistically based sampling procedure to a toolkit of techniques widely used under quality assurance (QA). Statistical Process Control (SPC) and Total Quality Management (TQM) are largely concerned with trouble shooting and early detection of abnormal process operation. The modern quality specialist requires a broad knowledge of the operation of such procedures. Forecasting is an important part of general management; and time series models are often used for forecasting. A brief description of the methods and techniques will be given.
Teaching Mode: 2 Lectures per week
1 Tutorial per week
Private Study: 5 hours per week
Recommended Texts: Jerry Banks, Principles of Quality Control, 1989, Wiley.
C Chatfield, The Analysis of Time Series: Theory and Practice, 1996, Chapman-Hall.
D C Montgomery, Statistical Quality Control, (2nd edition), 1991, Wiley.
G B Wetherill, Sampling Inspection and Quality Control, 1969, Methuen.
G B Wetherill, D W Brown, Statistical Process Control - Theory and Practice, 1991, Chapman and Hall.
Assessment Methods: Coursework: 20%
Coursework Mode: Tests in Weeks 8 and 11
Examination: 80%
Examination is of 2 hours duration at the end of the First Semester.
No of lectures: Syllabus
2 Basic concepts of acceptance sampling; operating characteristic function (OC), producer's risk and consumer's risk. Average run length.
4 Sampling plans (single and double) based on attributes. Rectifying sampling plans. Average outgoing limit (AOQL). Design of a single sampling plan.
4 Sequential sampling plans. Sequential probability ratio tests. Average sampling number (ASN).
4 Principles and objects of control charts. Assignable and random causes of variation. Shewhart's control charts for defectives, defects, means and variances. Average run length (ARL).
4 Cumulative sum control chart. Johnson's approach. Efficiency of Cusum charts.
6 Trend and seasonality in time series. Second order stationarity. Conditions for stationarity and inventibility of time series models. Discussion of AR, MA, ARMA and ARIMA models. Mean square error forecasts. Calculation of forecasts from the time series models.

   Last revised August, 2006