Load event detection is fundamental to event-based nonintrusive load monitoring (NILM) solution. It is directly related to whether the transient and steady-state signatures of appliances can be accurately extracted. In the real world, the load composition is random and complicated, the power consumption patterns of appliances are diverse, and multiple events may occur simultaneously or near each other, all these make it difficult for any one single event detection method to achieve satisfactory robustness. In coping with this, an adaptive two-stage event detection method is proposed in this article. First, an adaptive detection threshold whose value can be adjusted adaptively according to the load fluctuations is adopted, thus improving the ability to detect events of different amplitudes. Then, considering the different event geometric features, an improved edge detection method and a window-based detection method combining moving average with moving t-test are proposed for step-like events and long-transient events, respectively, and they are executed consecutively to achieve the effective capture of the complete transient process of appliances. Furthermore, an event separation step is taken to locate and separate the near-simultaneous events which often appear in low-frequency data. Specific event detection performance evaluation metrics are also designed. Comparison test results on private and public datasets show that the proposed method achieves high detection accuracy for different events of various appliances and maintains strong robustness in different operation scenarios.

Load event matching is the key for event-based non-intrusive load monitoring (NILM). It aims to find the load event sequence corresponding to the appliance’s operation cycle from all detected load events. Firstly, a novel combinatorial optimization model for load event matching is established. To reduce the dimension of the estimation matrix in the proposed optimization model, the concept of balanced window (BW) is introduced to indicate a period of load operation within which events can be fully matched. Then an iterative load event optimal matching strategy based on nested BWs is proposed, where most inner BWs are processed firstly. For the event matching in one specific BW, a two-stage optimization algorithm is designed. In the first stage, a set of candidate load event sequences is generated by using the depth first search algorithm with pruning, which reduces the solution space of the combinatorial optimization problem. In the second stage, the optimal load event sequences that result in the optimal fit of the aggregated load power is selected. Comparative experiments on the publicly-accessible REDD datasets have shown that the proposed method achieves better performance against the existing unsupervised NILM methods, and is promising for practical implementation.

Non-intrusive load monitoring (NILM) can obtain fine-grained electricity consumption information of each appliance by analyzing the voltage and current data measured at a single point on the bus, which is of great significance for promoting and improving the efficiency and sustainability of the power grid and enhancing the energy efficiency of users. NILM mainly includes data collection and preprocessing, event detection, feature extraction, and appliance identification. One of the most critical steps in NILM is signature extraction, which is the basis for all algorithms to achieve good state detection and energy disaggregation. With the generalization of machine learning algorithms, different algorithms have also been used to extract unique signatures of appliances. Recently, the development and deployment of the voltage–current (V-I) trajectory signatures applied for appliance identification motivated us to present a comprehensive review in this domain. The V-I trajectory signatures have the potential to be an intermediate domain between computer vision and NILM. By identifying the V-I trajectory, we can detect the operating state of the appliance. We also summarize existing papers based on V-I trajectories and look forward to future research directions that help to promote the field’s development.

Load event detection is the fundamental step for the event-based nonintrusive load monitoring (NILM). However, existing event detection methods with fixed parameters may fail in coping with the inherent multitimescale characteristics of events and their detection accuracy is easily affected by the load fluctuations. In this regard, this article extends our previous two-stage event detection framework, and proposes a novel multitimescale event detection method based on the minimum description length (MDL) principle. Following the completion of step-like event detection, a long-transient event detection scheme with variable length sliding window is designed, which is intended to observe and characterize the same event at different timescales. In that, the context information in the aggregated load data is mined by motif discovery, then based on the MDL principle, the proper observation scales are selected for different events and the corresponding detection results are determined. For the postprocessing, a load fluctuation segment location method based on voice activity detection (VAD) algorithm is proposed to identify and remove the unreasonable events caused by load fluctuations. Based on new evaluation metrics, the comparison tests on public and private datasets demonstrate that our method achieves higher detection accuracy and integrity for events of various appliances across different scenarios.