Alarm System Design Using Rank Order Filters

• Author / Creator

  Azad, Ishtiza I

• In the process industry, process variables are continuously monitored to ensure safety, reliability and efficiency of plant operations. Due to the advancement of the modern communication and computer technology, it is now possible to incorporate alarms to every process variable at little or no cost. As a result, operators are flooded with too many alarms beyond their capacity to respond accordingly. Many of these alarms are false or nuisance alarms. Therefore, an efficient and dependable alarm system is needed for greater safety and productivity. Motivated by this, this thesis focuses on the application of a class of nonlinear filters, namely rank order filters, on process data and develops quantitative relationship among filter parameters and alarm performance indices.

  In industries, filtering is a widely used alarm design technique. Moving average filters are most commonly applied filters in the industry because of their simplicity and ease of implementation. However, nonlinear filters have not been able to draw industrial attention due to nonlinearity and unknown relationship with alarm performance indices. Hence, we investigated the applicability of rank order filters and compared the performance with moving average filters under different input distributions. We established analytical relationships between filter order with different ranks and detection delay. Then we proposed a method to design filter order meeting the performance requirements.

  We obtained significant improvements in terms of reducing false and missed alarm rates and detection delay by applying rank order filters. The performance curve of a rank order filter lies between the performance curves of the moving average filter and the general optimal filter with the corresponding order. In the end, all the theoretical development and design techniques have been validated through numerical simulation and some industrial case study.

• Subjects / Keywords

  • Expected Detection Delay
  • Rank Order Filters
  • Alarm System Design
  • False and Missed Alarm Rates

• Graduation date

  Fall 2015

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3H70856C

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.