Abstract
Estimating the wear of the single electrical parts of a home appliance without resorting to a large number of sensors is desirable for ensuring a proper level of maintenance by the manufacturers. Deep learning techniques can be effective tools for such estimation from relatively poor measurements, but their computational demands must be carefully considered, for the actual deployment. In this work, we employ one-dimensional Convolutional Neural Networks and Long Short-Term Memory networks to infer the status of some electrical components of different models of washing machines, from the electrical signals measured at the plug. These tools are trained and tested on a large dataset (502 washing cycles approx 1000 h) collected from four different washing machines and are carefully designed in order to comply with the memory constraints imposed by available hardware selected for a real implementation. The approach is end-to-end; i.e., it does not require any feature extraction, except the harmonic decomposition of the electrical signals, and thus it can be easily generalized to other appliances.
Highlights
The reliability of a home appliance is an important component of its quality, tools for guaranteeing an appropriate level of maintenance are highly desirable for the manufactures
We employ one-dimensional Convolutional Neural Networks and Long Short-Term Memory networks to infer the status of some electrical components of different models of washing machines, from the electrical signals measured at the plug
Among the many tools that allow for making predictions from time series, we investigate the one-dimensional Convolutional Neural Networks (CNNs) and the Long Short-Term Memories (LSTMs)
Summary
The reliability of a home appliance is an important component of its quality, tools for guaranteeing an appropriate level of maintenance are highly desirable for the manufactures. We expand the mentioned study, focusing only on the classification problem (Fig. 1), and we take into account hardware and memory constraints, as a premise for the actual deployment on an appropriate micro-controller. Such constraints turn out to have significant consequences, that allow for complementing the analysis carried out in [3]. In [24], a comparison is carried out of several deep learning tools (such as recurrent neural networks, LSTMs and auto-encoders) aimed at predicting energy consumption patterns in a building, detecting anomalies of usage that generate an increase in energy consumption and the device that causes them. The three main trends are (1) scaled-down AI with pruning and weight quantization [2, 14, 17, 29, 30]; (2) the use of simplified or approximated activation functions [4, 20, 31]; (3) the use of recursive NN structures (like e.g., LSTMs and Gated recurrent Units, GRU) [28]
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
