The patterns detected are called anomalies [ 3 ]. The CMSs store a large amount of operation data in power plants [ 4, 5, 6 ], and the use of this data represents an important field for future smart power plants (SPPs) [ 7, 8 ]. Therefore, anomaly detection using data-driven approaches is a popular research topic for SPPs.
The scores of all batteries are lower than a predefined threshold, i.e., 50% in this work, implying that all abnormal batteries are accurately predicted to be “abnormal”. In our test, the first abnormal battery has the highest score (44.6%), and its aging trajectory is given in Figure 4c.
Diverse pieces of equipment are utilized in a power plant. In this paper, a general anomaly detection framework for power plants is proposed. The framework utilizes a large amount of operating data relating to the equipment and adopts the LSTM-AE network to establish the NBM.
With these issues in mind, the early-stage identification of the battery lifetime abnormality remains an unsolved problem in the field of battery manufacturing and management. In this work, we make the first attempt to identify the lifetime abnormality of lithium-ion batteries using only the first-cycle aging data.
These seven batteries are, therefore, defined as “abnormal”. From the data monitoring point of view, these abnormal samples are also defined as “positive samples”, while the normal batteries are termed as “negative samples” in the following discussions. Illustration of our battery aging data. a) Initial resistance versus capacity of 215 batteries.
As a faulty battery tends to exhibit a notable deviation in measurements and estimations compared to the normal cluster, this disparity can serve as a fault indicator. For example, Lai et al. proposed a SOC correlation-based early-stage ISC detection method for the online detection of ISCs.