Power Load Forecasting Research Articles

Short term power load forecasting based on BES-VMD and CNN-Bi-LSTM method with error correction

Aiming at the strong non-linear and non-stationary characteristics of power load, a short-term power load forecasting method based on bald eagle search (BES) optimization variational mode decomposition (VMD), convolutional bi-directional long short-term memory (CNN-Bi-LSTM) network and considering error correction is studied to improve the accuracy of load forecasting. Firstly, a decomposition loss evaluation criterion is established, and the VMD optimal decomposition parameters under the evaluation criterion are determined based on BES to improve the decomposition quality of the signal. Then, the original load sequence is decomposed into different modal components, and the corresponding CNN-Bi-LSTM network prediction models are established for each modal component. In addition, considering the influence of various modal components, holiday and meteorological factors on the error, an error correction model considering short-term factors is established to mine the hidden information contained in the error to reduce the inherent error of the model. Finally, the proposed method is applied to a public dataset provided by a public utility in the United States. The results show that this method can better track the changes of load and effectively improve the accuracy of short-term power load forecasting.

Detection of Outliers in Time Series Power Data Based on Prediction Errors

The primary focus of smart grid power analysis is on power load forecasting and data anomaly detection. Efficient and accurate power load prediction and data anomaly detection enable energy companies to develop reasonable production and scheduling plans and reduce waste. Since traditional anomaly detection algorithms are typically for symmetrically distributed time series data, the distribution of energy consumption data features uncertainty. To this end, a time series outlier detection approach based on prediction errors is proposed in this paper, which starts by using an attention mechanism-based convolutional neural network (CNN)-gated recursive unit (GRU) method to obtain the residual between the measured value and its predicted value, and the residual data generally conform to a symmetric distribution. Subsequently, for these residual data, a random forest classification algorithm based on grid search optimization is used to identify outliers in the power consumption data. The model proposed in this paper is applied to both classical and real energy consumption datasets, and the performance is evaluated using different metrics. As shown in the results, the average accuracy of the model is improved by 25.2% and the average precision is improved by 17.2%, with an average recall improvement of 16.4% and an average F1 score improvement of 26.8% compared to the mainstream algorithms.

Communication-Efficient Federated Learning for Power Load Forecasting in Electric IoTs

With the construction of the modern power system, power load forecasting is significant to keep the electric Internet of Things in operation. However, it usually needs to collect massive power load data on the server and may face the problem of privacy leakage of raw data. Federated learning can enhance the privacy of the raw power load data of clients by frequently transmitting model updates. Concerning the increasing communication burden of resource-heterogeneous clients resulting from frequent communication with the server, a communication-efficient federated learning algorithm based on Compressed Model Updates and Lazy uploAd (CMULA-FL) was proposed to reduce the communication cost. CMULA-FL also integrates the error compensation strategy to improve the model utility. First, the compression operator is used to compress the transmitted model updates, of which large norms are uploaded to reduce the communication cost of each epoch and transmission frequency. Second, by measuring the error of compression and lazy upload, the error is accumulated to the next epoch to improve the model utility. Finally, based on simulation experiments on the benchmark power load data, the results show that the communication cost decreases at least 60% with controlled loss of model prediction compared with baseline.

Comprehensive Review of Machine Learning-Based Methods for Electricity Load Forecasting

With the improvement of data processing power and the continuous development of modern power grids, there is an increasing demand for accuracy in predicting power load. To study the field of power load forecasting, this article summarizes and categorizes different models into three types: traditional models, single machine learning models, and hybrid models, based on previous literature. Firstly, a general overview is provided of the application of different models in power load forecasting. Secondly, typical models from three categories are selected for a detailed introduction. In traditional models, the ARIMA model is chosen, while in single machine learning models, CNN, and LSTM are chosen. For the hybrid model, the ResNet-LSTM mixed neural network is selected for the introduction. Finally, four different datasets were used to test different models. The differences and patterns of the models were summarized, and suggestions were proposed for future research directions in the field of power load forecasting.

An optimised LSTM algorithm for short-term load forecasting

Load forecasting is a basic work of power dispatching, planning and other departments, and has always attracted attention from inside and outside the industry. The purpose of this is to make the extracted features nonlinear and to be able to learn more complex knowledge. The improved LSTM network is introduced to make a short-term forecast of wind power load. The dataset used in this experiment is the foreign GEFcom2014-Load wind power dataset. Because the dimension units of the 24-dimensional impact indicators in the dataset are different, in order to eliminate the dimensional impact between the indicators, the data is normalised in the experiment by adopting the min-max standardisation method. The experimental results show that the training error, validation error, and test error of the improved method are all reduced by 1%, 18% and 16%, compared with the second, third, and fourth groups.

Firefly Algorithm-Based Semi-Supervised Learning With Transformer Method for Shore Power Load Forecasting

Firefly Algorithm-Based Semi-Supervised Learning With Transformer Method for Shore Power Load Forecasting

Short-Term Power Load Forecasting with Hybrid TPA-BiLSTM Prediction Model Based on CSSA

Since the existing prediction methods have encountered difficulties in processing the multiple influencing factors in short-term power load forecasting, we propose a bidirectional long short-term memory (BiLSTM) neural network model based on the temporal pattern attention (TPA) mechanism. Firstly, based on the grey relational analysis, datasets similar to forecast day are obtained. Secondly, the bidirectional LSTM layer models the data of the historical load, temperature, humidity, and date-type and extracts complex relationships between data from the hidden row vectors obtained by the BiLSTM network, so that the influencing factors (with different characteristics) can select relevant information from different time steps to reduce the prediction error of the model. Simultaneously, the complex and nonlinear dependencies between time steps and sequences are extracted by the TPA mechanism, so the attention weight vector is constructed for the hidden layer output of BiLSTM and the relevant variables at different time steps are weighted to influence the input. Finally, the chaotic sparrow search algorithm (CSSA) is used to optimize the hyperparameter selection of the model. The short-term power load forecasting on different data sets shows that the average absolute errors of short-term power load forecasting based on our method are 0.876 and 4.238, respectively, which is lower than other forecasting methods, demonstrating the accuracy and stability of our model.

Short‐term load forecasting of multi‐scale recurrent neural networks based on residual structure

SummaryAccurate short‐term load forecasting plays an important role in reducing power generation costs, maintaining supply and demand balance, and stabling the power grids operation. In recent years, deep learning models based on recurrent neural networks (RNN) have been widely used in short‐term load forecasting. Nevertheless, RNN cannot extract multi‐scale features of load data, resulting in low forecasting accuracy. A model for short‐term power load forecasting of residual multiscale‐RNN (RM‐RNN) was proposed in this study. RM‐RNN uses the multilayer RNN network structure. Specifically, each layer sets the dilated convolution with different dilated coefficients to extract the multi‐scale features of the load data. Adjacent networks transfer feature information for feature fusion through the residual structures. The experiment used random sampling data training model, and compared RM‐RNN with multiple deep learning models. The experimental results demonstrated that the mean error of RM‐RNN prediction is the lowest, indicating that dilated convolution can effectively extract multi‐scale features of load data. This result verified the effectiveness of residual structure fusion features, and improved the accuracy of short‐term load forecasting.

Terminal Voltage Prediction of Distribution and Substation Based on Spatial-Temporal Distribution Characteristics of the Station Load

A terminal voltage prediction method based on the spatial-temporal distribution characteristics of the load in the substation area is proposed. Considering that the power flow and load belong to the influence factors of the terminal voltage, the power flow analysis method based on the parameters and power distribution of the low-voltage line is used. After the general model of the low-voltage line is built according to the spatial distribution characteristics of the load in the substation area. Combined with the injected power of the parent node of the terminal node in the constructed line model, the terminal node voltage is obtained; After the load information of different time periods in the station area is predicted by the distribution and substation load prediction method based on wavelet-clustering, the voltage analysis results under the change of power flow and load moment are taken as the prediction samples of the terminal voltage prediction method in the station area based on the fusion of power flow calculation and load forecasting. Based on the low-voltage limit, the deep residual network is used to realize the state prediction of whether the terminal voltage is normal or not. The test shows that the prediction result of this method for the terminal voltage of distribution and substation in a certain area is in line with the actual situation and has the ability to predict the terminal voltage of distribution and substation.

Research on short-term forecasting of power load based on big data BP neural network

The power network system is an indispensable part of the economy development, which directly affects the stable operation of various industries and people’s daily life. During the stable operation of the power system, the prediction of the power load plays an important role in the load scheduling of the power system. Aiming at the problem of short-term load forecasting of power system, this paper established a short-term forecasting model of power load based on the BP neural network forecasting model through the collection of big data and modified the network weights and thresholds through model training. Finally, a short-term prediction of the power load of a certain community was carried out. The results show that the prediction model based on BP neural network can accurately predict the short-term power load with small prediction errors and good prediction performance. It can meet the precision requirements of power system operation scheduling.

Short-term Power Load Forecasting Based on Improved BP Neural Network from Genetic Algorithm and Simulated Annealing Algorithm

As one of the common methods of load forecasting, the traditional BP neural network algorithm has been favored by scholars for many years due to its strong learning ability and adaptive ability, and is often used by scholars to forecast power load. However, BP neural network also has its own shortcomings, that is, low algorithm efficiency, local minimization and other problems. In this paper, GA algorithm and SA algorithm are used to modify the weights of BP neural network, and a short-term load forecasting model, namely BP-GSA model, is established. The least mean square error prediction model is obtained through training. Then using BP-GSA model and traditional BP neural network model, using MATLAB software to predict the daily power load of a city from July 1 to July 10. The forecast results show that the load forecast curve obtained by BP-GSA model is closer to the actual load curve. That is to say, BP-GSA load forecasting error is less than the traditional BP neural network load forecasting error, which proves that BP-GSA model has better forecasting effect.

Ultra-short term power load forecasting based on CEEMDAN-SE and LSTM neural network

Ultra-short-term power load forecasting refers to the use of load and weather information from the prior few hours to forecast the load for the next hour, which is very important for power dispatch and the power spot market establishment. Based on time series decomposition-reconstruction modeling and neural network forecasting, this study constructed a CEEMDAN-11May 13, 2014, was chosen as the starting time because the meteorological data (e.g., temperature, rainfall, wind speed, etc.) obtained by the authors for Changsha City started from that time; and to highlight the cyclical nature of the electricity load demand, given the lack of meteorological data afterApril 2018, May 13, 2017, was chosen as the cut-off date in this paper. However, we believe that the starting and ending time does not affect the idea and method of model construction in this paper.SE-LSTM model and used it to forecast the ultra-short-term electricity load in Changsha, China, considering meteorological and holiday factors. The article first decomposed the power load data from May 13, 2014, to May 13, 2017, at 24 time points per day for three years to obtain six component series, and then reconstructed them into a two-component series based on the sample entropy analysis to reflect the fluctuation and trend characteristics of the power load. Then, the LSTM neural network model was used to predict and superimpose the reconstructed component series to obtain the final prediction results. It was found that the RMSE, MAE, and MAPE of the CEEMDAN-SE-LSTM model were 62.102, 47.490, and 1.649 %, respectively, which were significantly better than those of the ARMA, LSTM single-prediction, EEMD-LSTM, and CEEMDAN-LSTM models. This study greatly improves the accuracy of ultra-short-term power-load forecasting, provides support for ultra-short-term power dispatching in Changsha, and provides a reference for other cities to develop short-term and ultra-short-term power load forecasting models.

Analysis of user electricity consumption behavior based on density peak clustering with shared neighbors and attractiveness

SummaryUser behavior analysis is the research foundation of power load forecasting and power abnormal detection, and it is the theoretical support for smart grid planning and the construction of energy internet. Aiming at the complex characteristics of high‐dimensional, noisy, and multi‐redundant of power load data, this article used the principal component analysis (PCA) to reduce the dimensionality of power data. The density peaks clustering algorithm with shared neighbor and attractiveness (DPC‐SNA) was then used to cluster the data with reduced dimensionality to extract the user's electricity consumption characteristics. The DPC‐SNA algorithm first constructs a sample similarity measurement criterion that shares the similarity of neighbors, on which the local density is defined accordingly. The new local density can effectively distinguish the contributions of local samples and global samples. Incorporating the idea of universal gravitation, a new calculation method of sample attractiveness was defined, and the remaining samples were allocated by the attractiveness matrix. Experiment was performed using the actual load data of special transformer users in a certain area. The results show that there were five typical electricity consumption behavior characteristics in this area, namely, the evening‐peak production type, all‐day production type, multi‐peak production type, daytime production type, and night production type. The corresponding load peak time periods were also obtained.

Consensus-based dispatch optimization of a microgrid considering meta-heuristic-based demand response scheduling and network packet loss characterization

The uncertainty inherent in power load forecasts represents a major factor in the mismatches between supply and demand in renewables-rich electricity networks, which consequently increases the energy bills and curtailed generation. As the transition to a power grid founded on the so-called grid-of-grids becomes more evident, the need for distributed control algorithms capable of handling computationally challenging problems in the energy sector does so as well. In this light, the consensus-based distributed algorithm has recently been shown to provide an effective platform for solving the complex energy management problem in microgrids. More specifically, in a microgrid context, the consensus-based distributed algorithm requires reliable information exchange with customers to achieve convergence. However, packet losses remain an important issue, which can potentially result in the failure of the overall system. In this setting, this paper introduces a novel method to effectively characterize such packet losses during information exchange between the customers and the microgrid operator, whilst solving the microgrid scheduling optimization problem for a multi-agent-based microgrid. More specifically, the proposed framework leverages the virulence optimization algorithm and the earth-worm optimization algorithm to optimally shift the energy consumption during peak periods to lower-priced off-peak hours. The effectiveness of the proposed method in minimizing the overall active power mismatches in the presence of packet losses has also been demonstrated based on benchmarking the results against the business-as-usual iterative scheduling algorithm. Also, the robustness of the overall meta-heuristic- and multi-agent-based method in producing optimal results is confirmed based on comparing the results obtained by several well-established meta-heuristic optimization algorithms, including the binary particle swarm optimization, the genetic algorithm, and the cuckoo search optimization.

An adaptive ensemble framework with representative subset based weight correction for short-term forecast of peak power load

In the ensemble learning-based load prediction systems, recent research studies have achieved remarkable results in terms of extracting the complementary merits of multiple individual models and reducing the forecasting uncertainty, but they lack appropriate guidance for the sample-specific ensemble weighting of actual load data showing dynamics and variability. To this end, this study proposes an adaptive ensemble framework to address the above challenging and still tricky problem, and an adaptive weight-based ensemble learning model (AW-ELM) is framed to forecast short-term electric load by introducing the representative subset based weight correction with respect to the data characteristics. In the model, a weight library is formed as representatives extracting from the given training data. Based on the representatives, probability distribution modeling is built via a distance method. Then, the adaptive weights of the test data are determined by using the best matching concept. Experiments of two electric power companies in Jiangxi Province, China are implemented to evaluate the performance of the ensemble model. The results show that the ensemble model achieves better forecasting accuracy than the average combined forecasting model and the comparison models.

Individualized Short-Term Electric Load Forecasting Using Data-Driven Meta-Heuristic Method Based on LSTM Network.

Short-term load forecasting is viewed as one promising technology for demand prediction under the most critical inputs for the promising arrangement of power plant units. Thus, it is imperative to present new incentive methods to motivate such power system operations for electricity management. This paper proposes an approach for short-term electric load forecasting using long short-term memory networks and an improved sine cosine algorithm called MetaREC. First, using long short-term memory networks for a special kind of recurrent neural network, the dispatching commands have the characteristics of storing and transmitting both long-term and short-term memories. Next, four important parameters are determined using the sine cosine algorithm base on a logistic chaos operator and multilevel modulation factor to overcome the inaccuracy of long short-term memory networks prediction, in terms of the manual selection of parameter values. Moreover, the performance of the MetaREC method outperforms others with regard to convergence accuracy and convergence speed on a variety of test functions. Finally, our analysis is extended to the scenario of the MetaREC_long short-term memory with back propagation neural network, long short-term memory networks with default parameters, long short-term memory networks with the conventional sine-cosine algorithm, and long short-term memory networks with whale optimization for power load forecasting on a real electric load dataset. Simulation results demonstrate that the multiple forecasts with MetaREC_long short-term memory can effectively incentivize the high accuracy and stability for short-term power load forecasting.

Prophet-Based Medium and Long-Term Electricity Load Forecasting Research

The Prophet model decomposes the power load series of the region into trend terms, seasonal terms and holiday terms to visualize the data in the power load The model is then combined separately to achieve electricity load forecasting. The prediction results of the Prophet model were compared with those of the traditional ARIMA model and LSTM model, and the average absolute percentage error was 7.4342%, the root mean square error was 60376.8026MW, all of which were better than those of the ARIMA model and RNN model, which verified the effectiveness and feasibility of the model in power load forecasting.

Short-term power load forecasting based on gray relational analysis and support vector machine optimized by artificial bee colony algorithm.

Short-term power load forecasting is essential in ensuring the safe operation of power systems and a prerequisite in building automated power systems. Short-term power load demonstrates substantial volatility because of the effect of various factors, such as temperature and weather conditions. However, the traditional short-term power load forecasting method ignores the influence of various factors on the load and presents problems of limited nonlinear mapping ability and weak generalization ability to unknown data. Therefore, a short-term power load forecasting method based on GRA and ABC-SVM is proposed in this study. First, the Pearson correlation coefficient method is used to select critical influencing factors. Second, the gray relational analysis (GRA) method is utilized to screen similar days in the history, construct a rough set of similar days, perform K-means clustering on the rough sets of similar days, and further construct the set of similar days. The artificial bee colony (ABC) algorithm is then utilized to optimize penalty coefficient and kernel function parameters of the support vector machine (SVM). Finally, the above method is applied on the basis of actual load data in Nanjing for simulation verification, and the results show the effectiveness of the proposed method.

Research on Adversarial Domain Adaptation Method and Its Application in Power Load Forecasting

Domain adaptation has been used to transfer the knowledge from the source domain to the target domain where training data is insufficient in the target domain; thus, it can overcome the data shortage problem of power load forecasting effectively. Inspired by Generative Adversarial Networks (GANs), adversarial domain adaptation transfers knowledge in adversarial learning. Existing adversarial domain adaptation faces the problems of adversarial disequilibrium and a lack of transferability quantification, which will eventually decrease the prediction accuracy. To address this issue, a novel adversarial domain adaptation method is proposed. Firstly, by analyzing the causes of the adversarial disequilibrium, an initial state fusion strategy is proposed to improve the reliability of the domain discriminator, thus maintaining the adversarial equilibrium. Secondly, domain similarity is calculated to quantify the transferability of source domain samples based on information entropy; through weighting in the process of domain alignment, the knowledge is transferred selectively and the negative transfer is suppressed. Finally, the Building Data Genome Project 2 (BDGP2) dataset is used to validate the proposed method. The experimental results demonstrate that the proposed method can alleviate the problem of adversarial disequilibrium and reasonably quantify the transferability to improve the accuracy of power load forecasting.

A novel approach for power load forecast based on GAN data augmentation

At present, the concept of Smart Grid is intimately related to huge volumes of data generation, known as Big Data. These play a key role as far as the improvement of the network operation is concerned and, above all, when it comes to energy consumption. Hence the load forecast based on Big Data processing represents a reliable alternative in carrying out an accurate load prediction. As such, a new approach for the power load forecast, based on GAN data augmentation, for a hospital, within a Smart Grid, will be presented in this work.