Chiller equipment monitoring system and method

Abstract

The present disclosure relates to a chiller equipment monitoring system and method. The system comprises an information collection unit for acquiring and recording operational data patterns of a chiller over defined time segments; a machine learning module for learning the operational data and establishing a normal operation pattern and a corresponding tolerance parameter for each time segment; an anomaly detection unit for performing pattern recognition between the real-time operational data and the learned normal pattern, and generating an anomaly signal when similarity falls below the tolerance threshold; and an alert unit for sending alert messages. Unlike traditional fixed-threshold methods, this system utilizes time-based pattern recognition to achieve higher anomaly detection accuracy and supports personalized model training for different devices and time segments, thereby enhancing prediction sensitivity and system scalability.

Images