Non-intrusive Load Decomposition Research Articles

Non-intrusive Load Decomposition for Smart Buildings Based on Parallel Connectivity Networks and Attention Mechanism

Non-intrusive Load Decomposition for Smart Buildings Based on Parallel Connectivity Networks and Attention Mechanism

Research on non-intrusive load decomposition model based on parallel multi-scale attention mechanism and its application in smart grid

This paper presents the MUSENILM model, a non-intrusive load decomposition model incorporating a parallel multi-scale attention mechanism to enhance energy monitoring and management in smart grids. The core innovation of the proposed model is its ability to extract multi-scale features, enhancing the model's understanding of time series data and achieving significant performance improvements on the UK-DALE and REDD public datasets. Specifically, when MUSENILM identifies the fridge electricity consumption pattern on the UK-DALE public dataset, compared to previous models, the accuracy improves from 88% to 91%, and the F1 score increases from 87% to 90%; on the load decomposition tasks of the remaining four appliances, the F1 scores are all improved, while the mean absolute error (MAE) and cumulative absolute error (SAE) for the five appliances are also reduced. Additionally, it shows better results compared to previous models on the REDD dataset. Moreover, when the MUSENILM model is transplanted to embedded devices and applied in smart grids, it can effectively identify illegal lithium battery charging events of electric bicycles in different scenarios, which is crucial for ensuring grid security and optimizing energy distribution. This research not only provides an efficient method for the field of NILM but also offers practical solutions for violation monitoring and management in smart grids.

Non-Intrusive Load Monitoring Based on Multiscale Attention Mechanisms

With the development of smart grids and new power systems, the combination of non-intrusive load identification technology and smart home technology can provide users with the operating conditions of home appliances and equipment, thus reducing home energy loss and improving users’ ability to demand a response. This paper proposes a non-intrusive load decomposition model with a parallel multiscale attention mechanism (PMAM). The model can extract both local and global feature information and fuse it through a parallel multiscale network. This improves the attention mechanism’s ability to capture feature information over long time periods. To validate the model’s decomposition ability, we combined the PMAM model with four benchmark models: the Long Short-Term Memory (LSTM) recurrent neural network model, the Time Pooling-based Load Disaggregation Model (TPNILM), the Extreme Learning Machine (ELM), and the Load Disaggregation Model without Parallel Multi-scalar Attention Mechanisms (UNPMAM). The model was trained on the publicly available UK-DALE dataset and tested. The models’ test results were quantitatively evaluated using a confusion matrix. This involved calculating the F1 score of the load decomposition. A higher F1 score indicates better model decomposition performance. The results indicate that the PMAM model proposed in this paper maintains an F1 score above 0.9 for the decomposition of three types of electrical equipment under the same household user, which is 3% higher than that of the other benchmark models on average. In the cross-household test, the PMAM also demonstrated a better decomposition ability, with the F1 score maintained above 0.85, and the mean absolute error (MAE) decreased by 5.3% on average compared with that of the UNPMAM.

AttG-BDGNets: Attention-Guided Bidirectional Dynamic Graph IndRNN for Non-Intrusive Load Monitoring

Most current non-intrusive load monitoring methods focus on traditional load characteristic analysis and algorithm optimization, lack knowledge of users’ electricity consumption behavior habits, and have poor accuracy. We propose a novel attention-guided bidirectional dynamic graph IndRNN approach. The method first extends sequence or multidimensional data to a topological graph structure. It effectively utilizes the global context by following an adaptive graph topology derived from each set of data content. Then, the bidirectional Graph IndRNN network (Graph IndRNN) encodes the aggregated signals into different graph nodes, which use node information transfer and aggregation based on the entropy measure, power attribute characteristics, and the time-related structural characteristics of the corresponding device signals. The function dynamically incorporates local and global contextual interactions from positive and negative directions to learn the neighboring node information for non-intrusive load decomposition. In addition, using the sequential attention mechanism as a guide while eliminating redundant information facilitates flexible reasoning and establishes good vertex relationships. Finally, we conducted experimental evaluations on multiple open source data, proving that the method has good robustness and accuracy.

Flexible Load Multi-Step Forecasting Method Based on Non-Intrusive Load Decomposition

Accurate forecasting of flexible loads can capture the potential of their application and improve the adjustable space of the distribution network. Flexible load data, such as air conditioning (AC) and electric vehicles (EV), are generally included in the total load data, making it difficult to forecast them directly. To this end, this paper proposes a multi-step flexible load prediction model based on the non-intrusive load decomposition technique and Informer algorithm. The CNN-BiLSTM model is first used to decompose the flexible load from the total load via feature extraction and feature mapping of the flexible load to the overall load. The Informer model is then used to predict the flexible load and the residual load separately in multiple steps, and the prediction results are summed to obtain the overall prediction results. In this paper, the model is validated using two datasets, where in dataset 1, the prediction coefficients of determination for flexible load air conditioning and electric vehicles are 0.9329 and 0.9892. The predicted value of the total load is obtained by adding the flexible load to the residual load. At a prediction step of 1, the total load prediction coefficient of determination is 0.9813, which improves the prediction coefficient of determination by 0.0069 compared to the direct prediction of the total load, and prediction decision coefficient improves by 0.067 at 20 predicted steps. When applied to data set 2, the prediction coefficient of determination for flexible load air conditioning is 0.9646.

Non-intrusive residential electricity load decomposition via low-resource model transferring

Non-intrusive load decomposition (NILD) technology has a broad application prospect because it can deeply excavate the internal electricity consumption data of customers and obtain electricity consumption information. A non-intrusive load decomposition method based on model transfer is proposed to address the problem of insufficient accumulation of historical electricity consumption data of some household customers. Firstly, Time2vec, an embedding layer with time vectors, is introduced as an encoder to convert the time series into a continuous vector space, which improves the computational efficiency of the model for power time series information. Then, by introducing the Channel-wise attention mechanism, favorable residential electricity load characteristics can be effectively obtained. The model transfer method is used to fine-tune the pre-trained model based on complete data to overcome the problem of insufficient accumulation of historical electricity consumption data of some households. The experimental results show that the proposed method is a significant improvement over the existing state-of-the-art methods. The proposed method significantly improves the accuracy of load decomposition while reducing the model parameters. The model parameters are reduced by 62.4%, and the average absolute error is reduced by 13.8%–59.3%. Even with a small number of data training source domains, the proposed method is still able to accurately obtain the power consumption information of each appliance, proving that the proposed method overcomes the problem of insufficient accumulation of electricity consumption data of household users.

A Non-Intrusive Load Decomposition Method Based on Generative Adversarial Networks and Asynchronous Federated Learning

The non-intrusive load decomposition method is an essential approach of appreciating the water requirements of consumers. Traditional techniques of load breakdown have a number of drawbacks, including a poor level of water identification precision. In order to do this, a model for load decomposition based on generative adversarial networks has been developed. By applying a convolutional autoencoder to the total power noise, the generative network constructs the power breakdown sequence of the given home appliance. The discriminant network is then utilized to determine whether or not the created sequence is legitimate. In order to achieve a more natural and authentic sequence, the participants compete against one another. In light of the shortcomings of the centralized model training method, the network model implementation scheme based on asynchronous federated learning is utilized to replace the traditional convolution in order to realize the model’s lightweight nature. This allows the model to be applied to terminal devices such as smart water meters, and it is carried out in a cloud-edge collaborative manner. In addition, the model is carried out in a manner that is collaborative between the cloud and the edge. Training not only eases the strain of communication transmission but also protects the privacy of users and the integrity of their data.

Non-intrusive load decomposition algorithm based on temporal convolutional network and long short-term memory model combined with strongly correlated power sequences

The deep learning neural network method is used to complete the non-intrusive load decomposition task, which often has the problems of poor model performance and low decomposition accuracy. In order to solve the problem that the decomposition accuracy is reduced because the model cannot fully learn all the power characteristics of a specific load in deep learning, a non-intrusive method based on the combination of TCN-LSTM two-layer neural network and strongly correlated power sequence is proposed. Firstly, the load decomposition algorithm learns the time-attribute power information mapping relationship through TCN, obtains the load state pre-identification result and coupling it with the prior data, obtains the independent strongly correlated power sequence of each electrical device to replace the original aggregate power sequence to improve the proportion of the target signal in the mixed signal, and inputs it into the LSTM to capture advanced time-domain information. The measured power sequence is used as a label to implement the neural network model. After training, the decomposed power sequences of five experimental loads were obtained and compared with the actual power sequences. In the part of case analysis, the UK-DALE public data set is used to verify the algorithm in the paper, and by comparing the algorithm in the paper with two typical algorithms, the effectiveness of the model in the paper is verified, and the high-precision load decomposition task is completed.

Non-Intrusive Load Decomposition Based on Instance-Batch Normalization Networks

At present, the non-intrusive load decomposition method for low-frequency sampling data is as yet insufficient within the context of generalization performance, failing to meet the decomposition accuracy requirements when applied to novel scenarios. To address this issue, a non-intrusive load decomposition method based on instance-batch normalization network is proposed. This method uses an encoder-decoder structure with attention mechanism, in which skip connections are introduced at the corresponding layers of the encoder and decoder. In this way, the decoder can reconstruct a more accurate power sequence of the target. The proposed model was tested on two public datasets, REDD and UKDALE, and the performance was compared with mainstream algorithms. The results show that the F1 score was higher by an average of 18.4 when compared with mainstream algorithms. Additionally, the mean absolute error reduced by an average of 25%, and the root mean square error was reduced by an average of 22%.

A non-intrusive load decomposition method of resident by multi-scale attention mechanism

Non-intrusive load monitoring (NILM) is one of the important technologies in home energy management and power demand response scenario. However, the presence of multi-mode appliances and appliances with close power values have affected in diminishing the accuracy of identification based NILM algorithms. To tackle these challenges, the work proposes a resident load decomposition method combining multi-scale attention mechanism and convolutional neural network. At the first stage, the attention scores of the normal load data at the previous few moments of the attention model are smoothed dynamically against the abnormal scores at the current moment. The load identification attention model is optimized by constraint factors. Then, on this basis, convolution filters of different sizes are used to model the mixed load data of different electrical equipment, to mine more abundant characteristic information. Finally, to illustrate the proposed processes and validate its effectiveness, taking the PLAID data set as an example, the method proposed in the article is compared with respect to the existing NILM techniques. The experimental results show that the method based on the multi-scale attention mechanism in this paper can greatly improve the effect of load decomposition. Moreover, it reduces the confusion problem of electrical appliance identification with similar load characteristics.

DRA-net: A new deep learning framwork for non-intrusive load disaggregation

The non-intrusive load decomposition method helps users understand the current situation of electricity consumption and reduce energy consumption. Traditional methods based on deep learning are difficult to identify low usage appliances, and are prone to model degradation leading to insufficient classification capacity. To solve this problem, this paper proposes a dilated residual aggregation network to achieve non-intrusive load decomposition. First, the original power data is processed by difference to enhance the data expression ability. Secondly, the residual structure and dilated convolution are combined to realize the cross layer transmission of load characteristic information, and capture more long sequence content. Then, the feature enhancement module is proposed to recalibrate the local feature mapping, so as to enhance the learning ability of its own network for subtle features. Compared to traditional network models, the null-residual aggregated convolutional network model has the advantages of strong learning capability for fine load features and good generalisation performance, improving the accuracy of load decomposition. The experimental results on several datasets show that the network model has good generalization performance and improves the recognition accuracy of low usage appliances.

Non-intrusive Load Identification Method based on SAGOA-GMM Algorithm

Non-intrusive Load Identification play an important role in daily life. It can monitor and predict grid load while statistics and analysis of user electricity information. Aiming at the problems of low non-intrusive load decomposition ability and low precision when two electrical appliances are started and stopped at the same time, a new type of clustering and decomposition algorithm is proposed. The algorithm first analyses the measured power and use DBSCAN to filter out the noise of the collected data. Secondly, the remaining power points are clustered using the Adaptive Gaussian Mixture Model (AGMM) to obtain the cluster centres of the electrical appliances, and finally correlate the corresponding current waveform to establish a load characteristic database. In terms of load decomposition, a mathematical model was established for the magnitude of the changing power and current. The Grasshopper optimization algorithm (GOA) is optimized by introducing simulated annealing (SA) to identify and decompose electrical appliances that start and stop at the same time. The result of the decomposition is checked by the current similarity test to determine whether the result of the decomposition is correct, thereby improving the recognition accuracy. Experimental data shows that the combination of DBSCAN and GMM can can identify similar power characteristics. The introduction of SA makes up for the weakness of GOA and gives full play to the advantages of GOA's high identification efficiency. Finally, the test is carried out through the load detection data of the simultaneous start and stop of the two equipment. The test results show that the proposed method can effectively identify the simultaneous start and stop of two loads and can solve the problem of low recognition rate caused by the similar load power, which lays the foundation for the development of non-intrusive load identification in the future.

Commercial building load characteristics modeling considering equipment innate laws and various staff behaviors under demand response mechanism

Abstract At present, the traditional load model relies on a large number of historical data, ignoring the relationship between load demand and user behavior and the temporal and spatial distribution of load. To solve this problem, this paper proposes a modeling method of energy consumption characteristics of commercial office building considering the physical characteristics of energy consuming equipment and staff behavior factors. Firstly, the load of commercial office building is classified based on user behavior. Secondly, according to the proposed load classification method, a time distribution model associated with user behavior is established for each type of load energy consumption equipment, and the total load model of commercial office building is obtained. On this basis, the proposed model is expanded in both time and space, so that the total load model of commercial office building can be used to analyze energy consumption in different time scales and different regional areas. Through the combination of non-intrusive load decomposition and Markov chain, the energy consumption behavior of users is analyzed and simulated, and a refined load forecasting method of commercial office building considering equipment and user behavior under demand response mechanism is proposed. Finally, the example analysis shows that the proposed method can no longer depend on huge amounts of similar data for driving, can effectively reduce the impact of the original data on the load feature extraction, and has the ability to achieve load forecasting independently.

Non-intrusive load decomposition based on CNN–LSTM hybrid deep learning model

With the rapid development of science and technology, the problem of energy load monitoring and decomposition of electrical equipment has been receiving widespread attention from academia and industry. For the purpose of improving the performance of non-intrusive load decomposition, a non-intrusive load decomposition method based on a hybrid deep learning model is proposed. In this method, first of all, the data set is normalized and preprocessed. Secondly, a hybrid deep learning model integrating convolutional neural network (CNN) with long short-term memory network (LSTM) is constructed to fully excavate the spatial and temporal characteristics of load data. Finally, different evaluation indicators are used to analyze the mixture. The model is fully evaluated, and contrasted with the traditional single deep learning model. Experimental results on the open dataset UK-DALE show that the proposed algorithm improves the performance of the whole network system. In this paper, the proposed decomposition method is compared with the existing traditional deep learning load decomposition method. At the same time, compared with the obtained methods: spectral decomposition, EMS, LSTM–RNN, and other algorithms, the accuracy of load decomposition is significantly improved, and the test accuracy reaches 98%.

Sequence to Point Learning Based on an Attention Neural Network for Nonintrusive Load Decomposition

Nonintrusive load monitoring (NILM) analyzes only the main circuit load information with an algorithm to decompose the load, which is an important way to help reduce energy usage. Recent research shows that deep learning has become popular for this problem. However, the ability of a neural network to extract load features depends on its structure. Therefore, more research is required to determine the best network architecture. This study proposed two deep neural networks based on the attention mechanism to improve the current sequence to point (s2p) learning model. The first model employs Bahdanau style attention and RNN layers, and the second model replaces the RNN layer with a self-attention layer. The two models are both based on a time embedding layer. Therefore, they can be better applied in NILM. To verify the effectiveness of the algorithms, we selected two open datasets and compared them with the original s2p model. The results show that attention mechanisms can effectively improve the model’s performance.

Non-intrusive load monitoring algorithm based on household electricity use habits

The construction of smart grid is an important part of improving the utilization rate of electric energy. As an important way for the construction of smart grid, non-intrusive load decomposition methods have been extensively studied. In this type of method, limited by transmission cost and network bandwidth, low-frequency data has been widely used in practical applications. However, the accuracy of device identification in this case faces challenges. Due to the relatively single characteristics of low-frequency data, it is difficult to express the operating status of complex electrical appliances, resulting in low decomposition performance. In this paper, a non-intrusive load is proposed based on household electrical habits by studying the relationship between household electricity consumption habits and load status decomposition method. The Gaussian mixture model and time information are used to model the probability distribution of the electrical appliance state. This probability distribution is then used as the observation probability distribution of the factor hidden Markov model. In such a way, the BH-FHMM model is proposed. Finally, load decomposition is carried out through the load decomposition process of the FHMM model. In order to verify the performance of the proposed method, an experimental comparison is conducted based on the REDD data set. According to the results, a significant improvement in equipment recognition accuracy is obtained.

Sequence-to-Point Learning Based on Temporal Convolutional Networks for Nonintrusive Load Monitoring

Nonintrusive load monitoring (NILM) is performed by monitoring the total electricity consumption data from the customer’s meter and decomposing it with a load decomposition algorithm to get real-time electricity consumption data for individual appliances. Nonintrusive load decomposition is a single-channel blind source separation problem, which is difficult to implement because of the inherently unidentifiable problem. Recent studies have shown that deep learning is gradually becoming the mainstream approach to solve the NILM problem owing to the availability of data, computing power, and deep network training algorithm models. Among them, the sequence-to-point (seq2point) (Zhang) load decomposition model achieves state-of-the-art prediction because it concentrates on the representation ability of the network at one point. In this article, we optimize the structure of the network model based on the seq2point model, use a temporal convolutional network (TCN) with a more flexible perceptual field of view to train the load decomposition model, and replace the traditional activation function Relu with the activation function GELU. Finally, we test the algorithm in the public dataset UK-DALE and select the mean absolute error and F-score as the main evaluation indexes of the algorithm performance. Comparing the decomposition results of the two algorithms for the same period dataset, the results show that the temporal convolutional network model has substantially improved the performance index of load decomposition while reducing the model parameters.

Non-intrusive load decomposition model based on Group Bayesian optimization and post-processing

Non-intrusive load decomposition can decompose the power consumption of a single appliance from the household bus data, which is of great significance for users to adjust their own power consumption strategy. In order to solve the problem of large amount of computation in hyperparameter optimization of load decomposition model based on deep residual network, a Group Bayesian optimization method is proposed. This method can obtain better hyperparameter combination with less computational cost. In addition, in order to solve the problem of irrelevant activation of the model decomposition results, an improved post-processing method is proposed to improve the comprehensive performance of the model. Finally, the public data set REFIT is used to verify the proposed method, and the results show that the proposed method has a low decomposition error.

Research on Nonintrusive Load Decomposition of Enterprises Based on Bidirectional LSTM

To detect the operating condition of equipment and understand the environmental management situation of enterprises in real-time, this paper studies the non-intrusive load decomposition of enterprises based on bidirectional LSTM. In this paper, we first obtain the load characteristic parameters of different equipment in different states, and then obtain the electrical power measured from the main power meter of decontamination equipment in TOP-5 through the softmax layer of bidirectional LSTM, and then change the softmax layer to decompose the load data from the main power meter.

Research on Non-intrusive Load Decomposition Based on FHMM

Non-intrusive load monitoring and decomposition, as one of the important parts of intelligent power utilization system, can deeply analyze users’ internal load components and obtain user’s electricity consumption information from different scales, which is of great significance to users and power companies. In this paper, a non-intrusive load decomposition method based on factorial hidden Markov model using low frequency data is proposed. Kmeans-II algorithm is used to cluster the working state of a single load, the results of which are used to calculate the parameters of the HMM for individual load model. The total load model is represented by a factorial hidden Markov model, which transforms the load decomposition into an optimization problem with maximum probability. The improved Viterbi algorithm based on event detection is proposed to solve this optimization problem, so as to obtain the working state sequence and realize load decomposition. Finally, the correctness and practicability of the method are verified by an example.