Power Load Data Research Articles

Recognition Method of Abnormal Power Consumption State of Power Users Based on Strategy Gradient Algorithm

Because the traditional method ignores the problem of power load data preprocessing, the accuracy of the recognition result of the power consumption status is not high, the recognition efficiency is not high, and the recognition effect is not good. For this reason, a method for identifying the abnormal power consumption status of power users based on the strategy gradient algorithm is proposed. The preprocessing of power load data mainly includes the completion of missing data and the feature extraction of power load data. Based on the results of the preprocessing, the abnormal increase in user power consumption is detected. Finally, the strategy gradient algorithm is used for initial training and training process testing to complete the identification of the abnormal state of power users. The experimental results show that the accuracy of the power status recognition result of the proposed method is higher, and the recognition time is always less than 2.0 s, indicating that the recognition effect of the method is better.

Interpolating high granularity solar generation and load consumption data using super resolution generative adversarial network

The vast majority of commonly accessible photovoltaics (PV) generation and load consumption datasets have low temporal resolutions, leading to inaccuracies in the modeling and optimisation of PV-integrated battery systems. This study addresses this problem by proposing an interpolation model based on a super resolution generative adversarial network (SRGAN) that generates 5-minute PV and load power data from 30-minute/hourly temporal resolutions. The proposed approach is validated by two different datasets including large amounts of residential data and compared to an alternative predictive model. The results indicate that the model can adequately capture the targeted data distributions and temporal characteristics with negligible statistical differences from the measured high resolution data. Moreover, it performs consistently across different types of PV/load profiles and on average it results in 0.32% and 0.28% normalised root mean squared errors (NRMSEs) in daily totals of 5-minute PV and load power values when using hourly data as inputs. Under a time-of-use (ToU) tariff, the interpolated 5-minute data leads to 44.7% and 41.7% error reductions compared to using hourly data for estimating electricity costs and battery saving potentials of a PV battery system. Hence, the proposed model can be potentially applied in a battery sizing tool to obtain more accurate sizing results when only low resolution data is available.

Least Squares Support Vector Machine Regression Based on Sparse Samples and Mixture Kernel Learning

Least squares support vector machine (LSSVM) is a machine learning algorithm based on statistical theory. Itsadvantages include robustness and calculation simplicity, and it has good performance in the data processingof small samples. The LSSVM model lacks sparsity and is unable to handle large-scale data problem, this articleproposes an LSSVM method based on mixture kernel learning and sparse samples. This algorithm reduces theinitial training set to a sub-dataset using a sparse selection strategy. It converts the single kernel function in theLSSVM model into a mixed kernel function and optimizes its parameters. The reduced sub-dataset is used fortraining LSSVM. Finally, a group of datasets in the UCI Machine Learning Repository were used to verify theeffectiveness of the proposed algorithm, which is applied to real-world power load data to achieve better fittingand improve the prediction accuracy.

A New Hyper-Parameter Optimization Method for Power Load Forecast Based on Recurrent Neural Networks

The selection of the hyper-parameters plays a critical role in the task of prediction based on the recurrent neural networks (RNN). Traditionally, the hyper-parameters of the machine learning models are selected by simulations as well as human experiences. In recent years, multiple algorithms based on Bayesian optimization (BO) are developed to determine the optimal values of the hyper-parameters. In most of these methods, gradients are required to be calculated. In this work, the particle swarm optimization (PSO) is used under the BO framework to develop a new method for hyper-parameter optimization. The proposed algorithm (BO-PSO) is free of gradient calculation and the particles can be optimized in parallel naturally. So the computational complexity can be effectively reduced which means better hyper-parameters can be obtained under the same amount of calculation. Experiments are done on real world power load data, where the proposed method outperforms the existing state-of-the-art algorithms, BO with limit-BFGS-bound (BO-L-BFGS-B) and BO with truncated-newton (BO-TNC), in terms of the prediction accuracy. The errors of the prediction result in different models show that BO-PSO is an effective hyper-parameter optimization method.

Anomaly Recognition and Correction Algorithm for Big Data of Distribution Network Load

With the continuous improvement of the intelligent degree of distribution network, load information presents the growth of big data level, large-scale, mixed, inaccurate monitoring or collection of load data often appear, which brings some difficulties to the scheduling and forecasting work. Therefore, it is necessary to identify and correct these abnormal data. Based on the analysis of the causes and distribution characteristics of different types of abnormal load data, this paper identifies and modifies the abnormal data based on the improved density estimation algorithm, so as to realize the rapid processing and repair of the distribution network load big data. Finally, the power load data of a province in 2020 is selected to clean the load data. The results show that the algorithm proposed in this paper can quickly and accurately realize the identification and correction of abnormal load data.

Deep-Learning Forecasting Method for Electric Power Load via Attention-Based Encoder-Decoder with Bayesian Optimization

Short-term electrical load forecasting plays an important role in the safety, stability, and sustainability of the power production and scheduling process. An accurate prediction of power load can provide a reliable decision for power system management. To solve the limitation of the existing load forecasting methods in dealing with time-series data, causing the poor stability and non-ideal forecasting accuracy, this paper proposed an attention-based encoder-decoder network with Bayesian optimization to do the accurate short-term power load forecasting. Proposed model is based on an encoder-decoder architecture with a gated recurrent units (GRU) recurrent neural network with high robustness on time-series data modeling. The temporal attention layer focuses on the key features of input data that play a vital role in promoting the prediction accuracy for load forecasting. Finally, the Bayesian optimization method is used to confirm the model’s hyperparameters to achieve optimal predictions. The verification experiments of 24 h load forecasting with real power load data from American Electric Power (AEP) show that the proposed model outperforms other models in terms of prediction accuracy and algorithm stability, providing an effective approach for migrating time-serial power load prediction by deep-learning technology.

Research on load forecasting method of large Power Grid based on Deep confidence Network

In view of the existing electric power load forecasting models exist in the measuring precision is not high and the disadvantage of low stability, this paper puts forward a deep belief network - based extreme learning machine (Deep Belief Network – Extreme Learning Machine, DBN - ELM) of power load forecasting model. On the basis of traditional DBN, this model introduces ELM as the regression layer, and combines DBN’s advantages in feature extraction with ELM’s strong generalization ability to improve the accuracy of power load prediction. The experiment shows that compared with the traditional DBN neural network, the model can better fit the variation trend of power load data time series, and has higher prediction accuracy.

A multi-energy scheduling strategy for orderly charging and discharging of electric vehicles based on multi-objective particle swarm optimization

In this paper, a multi-energy scheduling model based on the ordered charging and discharging of EVs in typical urban residential areas is established, and the photovoltaic system is introduced into the model. To improve models with single objective in previous research, the minimum variance of total power load and the minimum scheduling cost are taken as optimization objectives of the model, and the multi-objective particle swarm optimization algorithm is used to solve the problem. To reduce the gap between theoretical results and practical application, the actual power load data of a distribution area in Shanghai in spring and summer is taken, and in the case of different number of EVs and weather conditions, the multi-energy scheduling system is simulated. The results show that: 1) The total power load of the area adopting the scheduling algorithm is significantly less than that of the area with disordered charging, where the growth rate of peak electricity consumption can be reduced by up to 127.2%; 2) With the increase of the number of EVs participating in the scheduling, the profitability of the system increases significantly. When 70 EVs participate in dispatching, the profit is up to 150% higher than that of 10 EVs.

A Pyramid-CNN Based Deep Learning Model for Power Load Forecasting of Similar-Profile Energy Customers Based on Clustering

With rapid advancements in renewable energy sources, billing mechanism (AMI), and latest communication technologies, the traditional control networks are evolving towards wise, versatile and collaborative Smart Grids (SG). The short term power load forecasting of individual as well as group of similar energy customers is critical for effective operation and management of SG. Forecasting power load of individual as well as group of similar energy customers is challenging compared to aggregate load forecasting of a residential community. The main reason is the high volatility and uncertainty involved for the case of individual and group of similar energy customers. Several machine/deep learning models have been developed in the recent past for forecasting load of individual energy customers, but such explorations are ineffective due to the requirement of one trained model for every energy customer, which is practically not feasible. We plan to build a deep learning model using convolutional neural network (CNN) layers in pyramidal architecture for effective load forecasting for a group of similar energy-profile customers. Initially, we grouped a subset of energy customers from database of Smart Grid Smart City (SGSC) into clusters using DBSCAN approach. The CNN layers are used for extracting feature from historical load of each cluster. The extracted feature of similar energy-profile customers (grouped based on clustering) is combined to make training-databases for each cluster. We have used the power load data from SGSC project, which contain thousands of individual household energy customers data. The developed Pyramid-CNN model is trained based on these sets of databases. The trained model is evaluated on randomly selected customers from few clusters. We obtained significantly improved forecasting results for randomly selected user from different clusters. Our adapted strategy of clustering based model training resulted in upto 10 percent MAPE improvement for the energy customers. The essence of our work is that energy customers can be grouped into clusters and then representative model could be developed/trained, which can accurately forecast power load for individual energy-customer. This approach is highly feasible, as we do not need to train a model per energy customer and still achieve competitive forecasting results.

Short-term power load forecasting based on I-GWO-KELM algorithm

In this paper, I-GWO-KELM algorithm is used for short-term power load forecasting. Normalize the power data and meteorological data of the short-term power load, and use GWO to optimize the regularization coefficient of KELM and the RBF kernel parameters. To apply the model to short-term power load forecasting to obtain simulations for the next 24 hours and 168 hours curve. Experiments show that the improved model I3-GWO-KELM proposed in this paper has the best effect. The improvement of GWO in this paper is effective and feasible. In the application of short-term power load forecasting, the IGWO-KELM model is more accurate than the ELM and KELM models.

Long-term load combination forecasting method considering the periodicity and trend of data

In order to solve the problems of insufficient accuracy of long-term power load forecasting and poor applicability of the model, this paper considers the coupling of a number of macro indicators, such as regional economic development and social development indicators, with the time series data of regional power load. BP neural network and Autoregressive integrated moving average model (ARIMA) are used to integrate and improve the forecasting model, so as to improve the trend forecasting ability of annual load forecasting model. The non parametric function method is used to forecast the periodic load data in the monthly load data, the annual load forecast is combined with the monthly load forecast to improve the overall forecasting accuracy of the model. Finally, through the comparison of grey prediction and other models and the verification of MAPE error analysis method, the prediction accuracy of the model method considering the combination of data periodicity and trend is significantly improved, which is suitable for the long-term prediction of regional power load.

Short Term Load Forecasting Based on SBiGRU and CEEMDAN-SBiGRU Combined Model

With the continuous development of global science and technology industry, the demand for power is increasing, so short-term power load forecasting is particularly important. At present, a large number of load forecasting models have been applied to short-term load forecasting, but most of them ignore the error accumulation in the iterative training process. To solve this problem, this article proposes a combined measurement model which combines stacked bidirectional gated recurrent unit (SBiGRU), complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and error correction. In the first stage, SBiGRU model is established to study the time series characteristics of load series under the influence of temperature and holiday types. The error series generated in the prediction process of SBiGRU model reflects the error characteristics of load series; In the second stage, the error sequence is decomposed into several intrinsic mode functions (IMF) components and trend components by CEEMDAN algorithm. The SBiGRU model is established again for each component to learn and predict, and the predicted values of all components are reconstructed to get the error prediction results; Finally, sum the two-stage prediction results to correct the error. The accuracy of SBiGRU-CEEMDAN-SBiGRU combination model is evaluated by two public power load data. The experimental results show that the SBiGRU-CEEMDAN-SBiGRU combination model has better accuracy and stability than the traditional model.

A dimensionality reduction method of power load data based on the combination of VMD-OMP-Kmeans

With the power data becoming more and more complex, it is of great significance to reduce the dimension and reconstruct the power data for the subsequent processing and application of power big data. This paper presents an algorithm based on variational mode decomposition (VMD) decomposition and orthogonal matching pursuit (OMP) reconstruction to reduce the dimension of power load data. Firstly, VMD method is used to decompose and filter the power load data; Secondly, K-means clustering algorithm is used to synthesize the decomposed data mode into a cluster center data set. Then, the data set is used as a dictionary; Thirdly, an OMP algorithm is proposed to select the elements to be reconstructed in the dictionary. And the corresponding data weights are given to reconstruct the load data with high-precision. As a result, the dimensionality of massive load data is reduced automatically. Finally, this paper builds a flexible load data dimensionality reduction model based on VMD method to verify the proposed method.

Research on Short-Term Forecast of Power Load Based on Adam-Gru

To forecast the short-term power load accurately and reliably, a short-term power load model based on gate recurrent unit was proposed. In order to improve the accuracy of the model, the Adam algorithm is used to calculate the GRU hyperparameter Optimization, so as to construct an Adam-GRU short-term power load forecasting model. The performance of the forecasting model was evaluated by the Root Mean Square Error and running time, then the forecast results were compared with LSTM method. The experimental results show that Adam-GRU has a higher accuracy than the LSTM model, an increase of 15.54%, and a time complexity reduction of 15.75%. Compared with the RNN model, the accuracy has improved 10.84%. The results show that the short-term power load forecasting model and parameter optimization method based on Adam-GRU can effectively forecast the power load data.

Research on information security of users’ electricity data including electric vehicle based on elliptic curve encryption

In the smart grid and big data environment, accurate and large amount of power load data for users can be obtained with the wide application of non-intrusive load monitoring technology. In the research process of customers’ information, information security protection of users’ electricity data has become a research hotspot urgently. This article proposes a new type of load decomposition method for electric vehicle load information and compares it with hidden Markov model algorithm to verify its accuracy. On this basis, the elliptic curve encryption algorithm is used to encrypt the users’ electricity data, and the function and effectiveness of the encryption algorithm are verified by comparing the load decomposition of the electric vehicle with the unencrypted data.

Short-term load forecasting model based on EEMD-LSTS-ARIMA

To improve the accuracy of the short-term power load forecasting, this paper designed a model, namely, the first to use based on empirical mode decomposition (EMD) method based on the collection of ensemble empirical mode decomposition (EEMD) in order to improve signal decomposition in the process of modal aliasing phenomenon and false component, and using the decomposition method of power load data decomposition, modal components with different characteristics, and according to its characteristics can be divided into high frequency components, intermediate frequency and low frequency components, Then the Long Short-Term Memory network (LSTM) is used for load forecasting. Using the autoregressive Integrated moving average model (Arima) to correct the predicted residual error, the load forecasting model based on EEMD-LSTM-ARIMA is established. Finally, the data of Queensland, Australia is used as the research sample. On the above prediction model, the data of this region is input into the model for power load prediction. At the end of the prediction, the final prediction results are compared with the EEMD-LSTM model through root mean square error (RMSE) and mean absolute percentage error (MAPE). Through comparison, it is found that the method used in this paper has a high prediction accuracy.

Research on Short-term Power Load Forecasting Based on Bi-GRU

The precise, accurate and efficient short-term load forecasting can guide the power supply companies to rationally arrange power dispatch plans, help improve the stability of grid operation, and significantly, improving power utilization, thereby optimizing corporate marketing strategies and increasing corporate economic returns. Short-term load forecasting methods based on deep learning have gained greater attention from academia and power companies. Among them, the load forecasting model based on recurrent neural network has gained excellent forecasting results compared with traditional machine learning models. The advantage of the cyclic neural network is that it can extract the degree of relevance of the data in the time dimension, but the unidirectional network only considers the impact of historical data on the current forecast. This paper proposes a load forecasting model based on a bidirectional gated recurrent unit, which further improves the relevance of data. However, it introduces meteorological factors and the influence of holidays to improve the accuracy of forecast results. Taking the power load data of a district of a city in China as an example, the prediction results of this model meet the actual requirements and show better prediction performance over Bi-LSTM, LSTM, GRU, and other models.

Deep-Learning Prediction Model with Serial Two-Level Decomposition Based on Bayesian Optimization

The power load prediction is significant in a sustainable power system, which is the key to the energy system’s economic operation. An accurate prediction of the power load can provide a reliable decision for power system planning. However, it is challenging to predict the power load with a single model, especially for multistep prediction, because the time series load data have multiple periods. This paper presents a deep hybrid model with a serial two‐level decomposition structure. First, the power load data are decomposed into components; then, the gated recurrent unit (GRU) network, with the Bayesian optimization parameters, is used as the subpredictor for each component. Last, the predictions of different components are fused to achieve the final predictions. The power load data of American Electric Power (AEP) were used to verify the proposed predictor. The results showed that the proposed prediction method could effectively improve the accuracy of power load prediction.

A combined model based on SSA, neural networks, and LSSVM for short-term electric load and price forecasting

Electricity, a kind of clean energy, has been widely used in people’s production and daily life. However, it is very difficult to estimate the electricity energy production in advance and store the rest of the electric energy due to the climate, environment, population and other factors. Based on data preprocessing and artificial intelligence optimization algorithm, this paper introduces a combined forecasting method. The proposed method contains six individual methods, and each individual method has its own usage. Singular spectrum analysis (SSA) is adopted to reduce noise from the original data; three individual forecasting methods, Jordan neural network, the echo state network, least squares support vector machine, are applied to obtain the intermediate forecasting results; two optimization algorithms, particle swarm optimization and simulated annealing, are used to optimize the parameters of the combined model. This paper not only validates the superiority of the combined model compared to the single predictive model through the simulation experiments of power load data and electricity price data. The mean absolute percent error (MAPE) of the combined power load and electricity price forecast results are 1.14% and 7.58%, respectively, which are higher than the MAPE error of the corresponding single models prediction results. It has also been verified that the process of eliminating noise by the SSA plays a positive role in the accuracy of the combined forecasting model. In addition, two series of experiments on the power load data lead to two very interesting conclusions. One of the conclusions is that as the size of the test data increases, the prediction accuracy of the model decreases; the other is that the predicted result calculated through the optimized combined weight is better than the combined result calculated using the average weight, and the average weight is used. Weighted combination does not improve the prediction accuracy of a single model.

Edge sensing data-imaging conversion scheme of load forecasting in smart grid

Edge sensing data in smart grid provides vast valuable information, which promotes further innovated smart power applications in Internet of things (IoT) oriented smart cities and society. While in power load prediction, the potential relationships between the time series of power load data and the characteristics of temperature, weather and date, have not been explored comprehensively, which degrades the accuracy of load prediction in smart grid. In order to extract the generalized features and latent relationships in power load related edge sensing data, a power load prediction scheme based on edge sensing data-imaging conversion (DIC) is proposed to improve the forecasting accuracy in smart cites and society. DIC employs empirical mode decomposition (EMD) for power load time series data and combines it with characteristic time series including temperature, weather and date to form an image-like structure. And a DIC-based convolutional neural network (DI-CNN) is presented to implement convolution. Experimental results show that, compared with long short-term memory (LSTM), support vector machines (SVM), and CNN, the proposed DIC scheme improves the training speed by 61.7 %, reduces root mean square error (RMSE) by 32.9 % at least, and enhances the prediction accuracy by 1.4 %.