A Hybrid Method of Cascade-Filtering and Committee Decision Mechanism for Non-Intrusive Load Monitoring

There are, in general, two methods of load monitoring which are intrusive load monitoring and non-intrusive load monitoring (NILM). The NILM method has attracted considerable research interests, since it only requires a set of voltage and current sensors to be installed at the electrical service entry (ESE) for load disaggregation. However, the main challenge of NILM is to accurately analyze data from the ESE and determine the electrical consumption of each appliance. A single load signature (LS) has some limitations and may not be used for the load disaggregation under all circumstances. Thus, the methods such as committee decision mechanism (CDM), which employ multiple LSs for load disaggregation, have been proposed. However, the CDM requires extensive calculations. This paper presents a new NILM method that combines a cascade-filtering approach with the CDM. The proposed method enhances the load identification capability of the NILM. A platform is built to validate the proposed method. The experimental results indicate that the identification accuracy of the proposed method is higher compared with the existing NILM methods. Moreover, the computation burden is much less when compared with the CDM method, demonstrating the high computational efficiency of the proposed method.