This thesis deals with an overview of statistical process control methods and defines a concept of a central management system for economical management of energy systems as a platform for efficient energy management. Cloud infrastructure was chosen in which users are accessing applications via a web interface, thus requiring devices with low computing power, such as tablet PCs or mobile phones.
Review of the data, which is used in the analysis of the energy efficiency, was made by identifying different types and sources of data. Data model was made in which the data is organised in several dimensions which allows On-Line Analytical Processing (OLAP).
Special statistical methods should be part of the system functionality for evaluating the dependencies between process parameters and external influences, and for detecting special causes of variability. Suitability of selected methods was verified on the real case scenario and the theory of use of control charts and the method of the cumulative sum of deviations (CUSUM) was confirmed.
Special attention was given to the module, alarms and information which specifies types of alarms and their origin. Different types of alarms were assigned to different communication channels. Strengths and weaknesses of communication channels were exposed. A Responsibilities management system was defined with rules under which alarms are assigned to actors. Methods based on action plans were defined for monitoring the effectiveness of the actions undertaken for eliminating alarm causes.
A prototype of the user interface was designed as a responsive web interface, accessible through a standard web browser without any additional software or hardware requirements.
|
