Non-intrusive Load Monitoring Method Based on SAGA-FCM Algorithm
Abstract
Existing non-intrusive algorithms have some deficiencies such as low recognition accuracy and the accuracy decreases significantly when the amount of monitoring data is too large. An improved fuzzy C-means algorithm (FCM) is proposed in this paper Which will make up the shortcomings of existing algorithms. In order to optimize the clustering process ,the simulated annealing algorithm (SA) and genetic algorithm (GA) are introduced to fasten the process. The experimental and comparison results fully demonstrate the superiority of the new FCM in terms of recognition, robustness, stability and reliability. Furthermore, the new FCM is applied to practical household appliances. All the comparison results consistently indicate that the new FCM is highly competitive and can be used in big data monitoring environment.
Keywords
Non-intrusive load monitoring, Delta feature extraction, Genetic simulated annealing algorithm, The fuzzy C-means algorithm
DOI
10.12783/dtetr/acaai2020/34212
10.12783/dtetr/acaai2020/34212
Refbacks
- There are currently no refbacks.