O.3.1.1 Development of Steady-State Reference Model for Heat Pump System Based on Real-time Machine Learning Algorithm

Artificial intelligence (AI) is getting more and more important in all fields of engineering technologies. Typically, thermal engineering systems are comprised of heat exchangers and fluid machinery and normally it is complicated and time consuming to analyze the systems mathematically. For heat pumps, a number of studies have been carried out to identify their operating status mathematically, however, accurate models are very difficult to develop due to numerous cases of different installation and operating conditions. Even the same heat pumps would have inevitable uncertainties by where and how they are installed. As an alternative way to estimate the performance of heat pumps, a methodology to develop a reference model using machine learning algorithm is introduced on in this study. For steady-state identification, a steady-state detector with a simple low pass filter is applied to measured signals of key features. Once steady state of the system is identified, the real-time measurements are collected to train the system model. From the study, the semi-expert based learning algorithm is effective to develop reference models of heat pump systems.

Download

Publication type Conf Proceedings Paper

Publication date 18 May 2017

Authors Minsung Kim, Yong Soo Lee

Keywords Heat pump; Steady-state; Reference model; Machine learning; Fault detection and diagnosis

Order nr -

Download