Short-term Forecasting Research Articles

Short-Term forecasting of floating photovoltaic power generation using machine learning models

Floating photovoltaic (FPV) power generation requires accurate short-term forecasting to optimize operational efficiency and enhance grid integration. This study investigates the application of machine learning models for predicting FPV power generation using data from the Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA) solar installation, which has a capacity of 157.20 kWp. Data were collected at 15-minute intervals from January 15 to January 21, 2024, encompassing nine input features such as ambient temperature, transient horizontal irradiation, daily horizontal irradiation, AC voltages, and AC currents for phases A, B, and C, with the total active power in kW as the target variable. The dataset was divided into a training set (first five days) and a testing set (remaining two days), and five machine learning models—Neural Networks (NN), Random Forest (RF), Extreme Learning Machine (ELM), Support Vector Regression (SVR), and Long Short-Term Memory (LSTM)—were employed. The results indicate that the Neural Networks model consistently outperforms the other machine learning algorithms in terms of predictive accuracy. These findings underscore the efficacy of machine learning techniques in forecasting FPV power generation, which has significant implications for enhancing the operational efficiency and grid integration of floating solar installations.

Fusion of Hierarchical Optimization Models for Accurate Power Load Prediction

Accurate power load forecasting is critical to achieving the sustainability of energy management systems. However, conventional prediction methods suffer from low precision and stability because of crude modules for predicting short-term and medium-term loads. To solve such a problem, a Combined Modeling Power Load-Forecasting (CMPLF) method is proposed in this work. The CMPLF comprises two modules to deal with short-term and medium-term load forecasting, respectively. Each module consists of four essential parts including initial forecasting, decomposition and denoising, nonlinear optimization, and evaluation. Especially, to break through bottlenecks in hierarchical model optimization, we effectively fuse the Nonlinear Autoregressive model with Exogenous Inputs (NARX) and Long-Short Term Memory (LSTM) networks into the Autoregressive Integrated Moving Average (ARIMA) model. The experiment results based on real-world datasets from Queensland and China mainland show that our CMPLF has significant performance superiority compared with the state-of-the-art (SOTA) methods. CMPLF achieves a goodness-of-fit value of 97.174% in short-term load prediction and 97.162% in medium-term prediction. Our approach will be of great significance in promoting the sustainable development of smart cities.

Short-Term Wind Speed Prediction via Sample Entropy: A Hybridisation Approach against Gradient Disappearance and Explosion

High-variant wind speeds cause aberrations in wind power systems and compromise the effective operation of wind farms. A single model cannot capture the inherent wind speed randomness and complexity. In the proposed hybrid strategy, wavelet transform (WT) is used for data decomposition, sample entropy (SampEn) for subseries complexity evaluation, neural network autoregression (NNAR) for deterministic subseries prediction, long short-term memory network (LSTM) for complex subseries prediction, and gradient boosting machine (GBM) for prediction reconciliation. The proposed WT-NNAR-LSTM-GBM approach predicts minutely averaged wind speed data collected at Southern African Universities Radiometric Network (SAURAN) stations: Council for Scientific and Industrial Research (CSIR), Richtersveld (RVD), Venda, and the Namibian University of Science and Technology (NUST). For comparison purposes, in WT-NNAR-LSTM-GBM, LSTM and NNAR are respectively replaced with a k-nearest neighbour (KNN) to form the corresponding hybrids: WT-NNAR-KNN-GBM and WT-KNN-LSTM-GBM. We assessed WT-NNAR-LSTM-GBM’s efficacy against NNAR, LSTM, WT-NNAR-KNN-GBM, and WT-KNN-LSTM-GBM as well as the naïve model. The comparative study found that the WT-NNAR-LSTM-GBM model was the most accurate, sharpest, and robust based on mean absolute error, median absolute deviation, and residual analysis. The study results suggest using short-term forecasts to optimise wind power production, enhance grid operations in real-time, and open the door to further algorithmic enhancements.

Short-term load forecasting: cascade intuitionistic fuzzy time series—univariate and bivariate models

AbstractShort-term load forecasting (STLF) is essential for developing reliable and sustainable economic and operational strategies for power systems. This study presents a forecasting model combining cascade forward neural network (CFNN) and intuitionistic fuzzy time series (IFTS) models for STLF. The proposed cascading intuitionistic fuzzy time series forecasting model (C-IFTS-FM) offers the advantage of CFNN using the links of both linear and nonlinear to model fuzzy relations between inputs and outputs. Moreover, it offers a more reliable and realistic approach to uncertainty, taking notice of also the degree of hesitation. C-IFTS-FM works in univariate structure when it uses only hourly load data, and in bivariate structure when it uses hourly load data and hourly temperature time series together. The conversion of time series into IFTS is realized with intuitionistic fuzzy c-means (IFCM). Thus, the membership and non-membership values for each data point are produced. In modelling process, membership and non-membership values, in addition to actual lagged observations, are used as input of the CFNNs. The effectiveness of C-IFTS-FM on test sets for both structures was discussed comparatively via different error criteria, in addition, the convergence time was examined, and also the fit of forecasts and observations was presented with different illustrations. Among different combinations of hyperparameters, in the best case, approximately 86% better accuracy is achieved than the best of the others, while even in the case of the worst of hyperparameters combination, the accuracy was improved by over 20% for the PSJM data sets. For HEXING, CHENGNAN, and EUNITE data sets, these progress rates reached approximately 90% in the best case.

Improved informer PV power short-term prediction model based on weather typing and AHA-VMD-MPE

Precise prediction of PV power in the short term is crucial for maintaining power system stability and balance. However, the performance of conventional time series prediction models on short-term long series prediction is scarcely sufficient because of the stochastic and turbulent character of PV power data. This work suggests a PV short-term power forecast model based on weather type, AHA-VMD-MPE decomposition reconstruction, and improved Informer combination to tackle this issue. Firstly, a SUM-ApEn-K-mean++ multidimensional clustering method to group the dataset by weather conditions. Then an AHA-VMD-MPE decomposition model is proposed to decompose the historical power data Finally the Informer model is improved and the improved model is utilized to predict the PV power under various weather conditions. The model exhibits great accuracy and stability in short-term PV power prediction, as demonstrated by the experimental results, which were validated using measured data from many PV power plants.

Spatiotemporal variations of precipitation patterns in the middle and lower reaches of Yangtze River Basin

Context With escalating global climate change, regional flood disasters have become increasingly prevalent. Precipitation, as a primary influencing factor, has garnered significant attention. Aims This study is based on precipitation data to investigate the spatiotemporal characteristics of precipitation in the middle and lower reaches of Yangtze River Basin (MLYB), trying to explore more concise methods for precipitation forecasting. Methods Statistical methods were employed to analyse historical precipitation patterns, followed by forecasting future trends using statistical time series models. Key results Precipitation in the MLYB exhibited a decreasing trend during 1961–2010, which shifted to an increasing trend after 2011, becoming more pronounced since 2017. Precipitation patterns in the MLYB were clearly increasing in the east and decreasing in the west, with the Taihu Basin showing the greatest rise. The ARIMA model predicted a significant increase in precipitation after 2022. Conclusions In recent years, precipitation in the MLYB has significantly increased, especially in downstream areas. Although the ARIMA model offers an effective and reasonably simple method for short-term forecast, it struggles with complex terrain influences. Implications These findings provide a theoretical basis for flood prevention in the MLYB, as well as a reference for precipitation prediction simulations in data-limited regions.

Performance evaluation of multivariate deep-time convolution neural architectures for short-term electricity forecasting: Findings and failures

Deep learning (DL) models hold great promise in enhancing the decision-making abilities of electricity market participants and system operators in the short term, as they excel at capturing the intricate interdependencies and uncertainties in electricity consumption. Nevertheless, currently, the adoption of DL models for electricity prediction is not keeping up with the advancements in DL research. Within this framework, the aims of this study are to examine the probabilistic, multivariate, and multihorizon time series prediction of deep time convolution neural (DTCN) models. The effectiveness of DTCN in forecasting electricity consumption over daily, weekly and monthly time horizons is emperically assessed. We analyze both probabilistic and deterministic predictions. The evaluation data comprises real residential electricity consumption, average temperature and humidity with hourly granularity, collected over one year from a residential building. For evaluation, we considered accuracy and uncertainties, sensitivity to hyperparameters, the impact of exogeneous variables, as well as computational efficiency. Results show that the probabilistic models offer greater advantages when it comes to forecasting on multihorizons, especially for diverse datasets with non-uniform time series models. Unlike traditional models, the DTCN models demonstrated superior performance in capturing complex patterns and reducing forecast errors in short-term scenarios. However, their performance can be influenced by the sensitivity of hyperparameters. These findings can act as a benchmark for quantitatively assessing the performance of novel multivariate DL models in predicting electricity demand.

Earthquake Precursors: The Physics, Identification, and Application

The paper presents the author’s vision of the problem of earthquake hazards from the physical point of view. The first part is concerned with the processes of precursor’s generation. These processes are a part of the complex system of the lithosphere–atmosphere–ionosphere–magnetosphere coupling, which is characteristic of many other natural phenomena, where air ionization, atmospheric thermodynamic instability, and the Global Electric Circuit are involved in the processes of the geosphere’s interaction. The second part of the paper is concentrated on the reliable precursor’s identification. The specific features helping to identify precursors are separated into two groups: the absolute signatures such as the precursor’s locality or equatorial anomaly crests generation in conditions of absence of natural east-directed electric field and the conditional signatures due to the physical uniqueness mechanism of their generation, or necessity of the presence of additional precursors as multiple consequences of air ionization demonstrating the precursor’s synergy. The last part of the paper is devoted to the possible practical applications of the described precursors for purposes of the short-term earthquake forecast. A change in the paradigm of the earthquake forecast is proposed. The problem should be placed into the same category as weather forecasting or space weather forecasting.

Which daily equity returns improve output forecasts?

We document the improvements in short term forecasts of real output growth for the United States and the euro area from incorporating daily financial data and using mixed data sampling (MIDAS) regressions. Furthermore, we show that a significant share of forecast improvements are driven by information embedded in stock returns of large, capital-intensive firms. In comparison, labor-intensive firms contribute less to improvements in output forecasts within a MIDAS framework.

Leveraging ecological indicators to improve short term forecasts of fish recruitment

AbstractForecasting the recruitment of fish populations with skill has been a challenge in fisheries for over a century. Previous large‐scale meta‐analyses have suggested linkages between environmental or ecosystem drivers and recruitment; however, applying this information in a management setting remains underutilized. Here, we use a well‐studied database of groundfish assessments from the West Coast of the USA to ask whether environmental variables or ecosystem indicators derived from long‐term monitoring datasets offer an improvement in our ability to skilfully forecast fish recruitment. A secondary question is which types of modelling approaches (ranging from linear models to non‐parametric methods) yield the best forecast skill. Third, we examine whether simultaneous forecasting of multiple species offers an advantage over generating species‐specific forecasts. We find that for approximately one third of the 29 assessed stocks, ecosystem indicators from juvenile surveys yields the highest out of sample predictive skill compared to other covariates (including environmental variables from Regional Ocean Modeling System output) or null models. Across modelling approaches, our results suggest that simpler linear modelling approaches do as well or better than more complicated approaches (reducing out of sample Root Mean Square Error by ~40% compared to null models), and that there appears to be little benefit to performing multispecies forecasts instead of single‐species forecasts. Our results provide a general framework for generating recruitment forecasts in other species and ecosystems, as well as a benchmark for future analyses to evaluate skill. The most promising applications are likely for species that are short lived, have relatively high recruitment variability, and moderate amounts of age or length data. Forecasts using our approach may be useful in identifying covariates or mechanisms to include in operational assessments but also provide qualitative advice to managers implementing ecosystem based fisheries management.

Machine learning-based short-term solar power forecasting: a comparison between regression and classification approaches using extensive Australian dataset

Solar energy production is an intermittent process that is affected by weather and climate conditions. This can lead to unstable and fluctuating electricity generation, which can cause financial losses and damage to the power grid. To better control power production, it is important to predict solar energy production. Big data and machine learning algorithms have yielded excellent results in this regard. This study compares the performance of two different machine learning approaches to solar energy production prediction: regression and classification. The regression approach predicts the actual power output, while the classification approach predicts whether the power output will be above or below a certain threshold. The study found that the random forest regressor algorithm performed the best in terms of accuracy, with mean absolute errors and root mean square errors of 0.046 and 0.11, respectively. However, it did not predict peak power values effectively, which can lead to higher errors. The long short-term memory algorithm performed better in classifying peak power values. The study concluded that classification models may be better at generalizing than regression models. This proposed approach is valuable for interpreting model performance and improving prediction accuracy.

An adaptive composite time series forecasting model for short-term traffic flow

Short-term traffic flow forecasting is a hot issue in the field of intelligent transportation. The research field of traffic forecasting has evolved greatly in past decades. With the rapid development of deep learning and neural networks, a series of effective methods have been proposed to address the short-term traffic flow forecasting problem, which makes it possible to examine and forecast traffic situations more accurately than ever. Different from linear based methods, deep learning based methods achieve traffic flow forecasting by exploring the complex nonlinear relationships in traffic flow. Most existing methods always use a single framework for feature extraction and forecasting only. These approaches treat all traffic flow equally and consider them contain same attribute. However, the traffic flow from different time spots or roads may contain distinct attributes information (such as congested and uncongested). A simple single framework usually ignore the different attributes embedded in different distributions of data. This would decrease the accuracy of traffic forecasting. To tackle these issues, we propose an adaptive composite framework, named Long-Short-Combination (LSC). In the proposed method, two data forecasting modules(L and S) are designed for short-term traffic flow with different attributes respectively. Furthermore, we also integrate an attribute forecasting module (C) to forecast the traffic attributes for each time point in future time series. The proposed framework has been assessed on real-world datasets. The experimental results demonstrate that the proposed model has excellent forecasting performance.

Two-stage meta-ensembling machine learning model for enhanced water quality forecasting

Accurate short-term forecasting of water quality variables (WQVs) such as dissolved oxygen (DO) and chlorophyll-a (Chl-a) is crucial for the effective management of aquatic resources. This study introduces a robust two-stage optimization-ensembling framework that integrates the Grey Wolf Optimizer (GWO) and the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to enhance the forecasting capabilities of machine learning (ML) models. Focusing on Small Prespa Lake, Greece, we implemented an array of diverse ML techniques, including eXtreme Gradient Boosting (XGB), Gradient Boosting Regressor (GBR), Light Gradient-Boosting Machine (LightGBM), and Multilayer Perceptron (MLP). These models were fine-tuned using GWO to optimize their performance over critical WQVs predicted six hours in advance. Our methodology employed rigorous data preprocessing techniques, including lag time feature engineering and principal component analysis (PCA), to handle the high dimensionality of the dataset. Optimal lag times ranging from 6 to 24 hour were evaluated, with the 24-hour lag proving to be the most effective in utilizing historical data to enhance forecasting accuracy. The GWO not only facilitated hyperparameter tuning but also demonstrated a notable improvement (7.6%) in the Kling-Gupta Efficiency (KGE) over conventional randomized search methods. Subsequently, the NSGA-II was utilized for multi-objective optimization, constructing powerful model ensembles that outperformed the individual GWO-optimized models by up to a 7% in KGE. In comparision to a standard genetic algorithm-based ensemble, the NSGA-II ensemble demonstrated superior effectiveness in balancing solution quality. This innovative approach not only establishes a new benchmark in water quality forecasting but also contributes substantially to proactive environmental monitoring and management strategies.

ES-dRNN: A Hybrid Exponential Smoothing and Dilated Recurrent Neural Network Model for Short-Term Load Forecasting.

Short-term load forecasting (STLF) is challenging due to complex time series (TS) which express three seasonal patterns and a nonlinear trend. This article proposes a novel hybrid hierarchical deep-learning (DL) model that deals with multiple seasonality and produces both point forecasts and predictive intervals (PIs). It combines exponential smoothing (ES) and a recurrent neural network (RNN). ES extracts dynamically the main components of each individual TS and enables on-the-fly deseasonalization, which is particularly useful when operating on a relatively small dataset. A multilayer RNN is equipped with a new type of dilated recurrent cell designed to efficiently model both short and long-term dependencies in TS. To improve the internal TS representation and thus the model's performance, RNN learns simultaneously both the ES parameters and the main mapping function transforming inputs into forecasts. We compare our approach against several baseline methods, including classical statistical methods and machine learning (ML) approaches, on STLF problems for 35 European countries. The empirical study clearly shows that the proposed model has high expressive power to solve nonlinear stochastic forecasting problems with TS including multiple seasonality and significant random fluctuations. In fact, it outperforms both statistical and state-of-the-art ML models in terms of accuracy.

WITHDRAWN: Short-term Load Forecasting Method Based on LSTM Optimized by the Hybrid Dung Bettle Optimization Algorithm

Forecasting electricity load is a crucial responsibility for the power system. Precise short-term power load forecasting (STLF) is a valuable tool for regional production planning, energy conservation, emission reduction, and grid coordination and dispatch. This study suggests a model for STLF that makes use of an optimized Long and Short-Term Memory (LSTM) neural network through the Hybrid Dung Beetle Optimization algorithm (HDBO-LSTM). Firstly, the Hybrid Dung Beetle Optimization (HDBO) algorithm is used to optimize the LSTM network's hyperparameters to address the problem that the LSTM network's random hyperparameter selection will materially influence the accuracy of load forecasting. Secondly, to address the problems that the original Dung Beetle Optimization (DBO) algorithm has poor global exploration ability and is easy to falls into local optimization, multiple improvement strategies are proposed to enhance the DBO algorithm, and the HDBO algorithm, which has a stronger searching ability and faster convergence speed, is proposed. By comparing with several algorithms on 10 benchmark functions, HDBO shows better search performance. Finally, the HDBO-LSTM load forecasting model is developed and evaluated using the Electrotechnical Mathematical Modeling Competition's power load dataset. The outcomes show that HDBO-LSTM outperforms the other comparison models in terms of accuracy and predicting efficacy.

Multi-Derivational Parsing of Vague Languages- The New Paradigm of Syntactic Pattern Recognition.

The new paradigm of syntactic pattern recognition, SPR, which uses multi-derivational parsing of vague languages is introduced in the paper. The methodology proposed addresses the issue of the recognition of vague/distorted patterns which is one of the important open problems in the area. The concept of the vague language of patterns and the efficient parsing method based on the class of dynamically programmed grammars are introduced. A vague language is defined with vague primitives which are vectors of "neighboring" primitives associated with measures of distance, probability, fuzziness, etc. The use of vague primitives allows us to identify b best structural templates during multi-derivational parsing that can be used for getting more adequate final result. The generic architecture of SPR system based on the approach proposed together with the system's applications for short-term electrical load forecasting and for analysis of ultrasound images in order to diagnose congenital defects of fetal palates are presented. The results of the experimental studies are discussed.

Short-term load forecasting of mixed industrial users in distribution networks using DARTS-LSTM method

Abstract Load forecasting has perennially stood as a fundamental pillar within the power industry, with precision in prognostication constituting a paramount necessity for critical functions including grid management, optimization of energy storage, and response to demand fluctuations. In the present study, we introduce a comprehensive global modeling prediction framework aimed at forecasting large-scale power demand within time series data, alongside the development of a novel short-term power load forecasting model. By integrating all time series data pertaining to the same regression task and employing a unified univariate prediction function across the entire time series dataset, we efficaciously forecast large-scale heterogeneous loads aggregated at various levels within distribution networks. The proposed framework leverages the cutting-edge DARTS (Differentiable Architecture Search) deep learning framework for implementation. Through comprehensive experimentation on real-world data, our model framework demonstrates superior performance in terms of overall prediction accuracy, scalability, and robustness in handling missing and non-stationary time series data.

Research on a short-term forecast of industry load based on PSO-LSTM

Abstract Load prediction is crucial to maintaining a stable power grid and encouraging the use of innovative energy types like scenescape energy. The industrial load of a region accounts for a significant proportion of the total load, but the industrial load is difficult to predict due to its large fluctuations and instability. This study proposes a short-term power load prediction model, merging the Particle Swarm Optimization (PSO) algorithm and Long Short-Term Memory (LSTM) neural network. The PSO algorithm fine-tunes the LSTM network’s hyperparameters, which are then implemented on the LSTM layer. The trained model, using real-world datasets, is then tested for accuracy. Our model’s effectiveness is backed by practical examples.

O'Hare Airport roadway traffic prediction via data fusion and Gaussian process regression

This study proposes an approach of leveraging information gathered from multiple traffic data sources at different resolutions to obtain approximate inference on the traffic distribution of Chicago's O'Hare Airport area. Specifically, it proposes the ingestion of traffic datasets at different resolutions to build spatiotemporal models for predicting the distribution of traffic volume on the road network. Due to its good adaptability and flexibility for spatiotemporal data, the Gaussian process (GP) regression was employed to provide short-term forecasts using data collected by loop detectors (sensors) and supplemented by telematics data. The GP regression is used to make predictions of the distribution of the proportion of sensor data traffic volume represented by the telematics data for each location of the sensors. Consequently, the fitted GP model can be used to determine the approximate traffic distribution for a testing location outside of the training points. Policymakers in the transportation sector can find the results of this work helpful for making informed decisions relating to current and future transportation conditions in the area.

An enhanced model based on deep operator network for very short-term forecasting of ship motion

Very short-term forecasting of ship motion involves forecasting future ship movements based on learned characteristics from historical motion data. However, ship motion exhibits not only temporal features but also features in the frequency domain, which are often overlooked. This paper introduces a novel method called Fourier Transform and Multilayer Perceptron-net enhancement based on Deep Operator Network (DeepONet), abbreviated as FMD. This approach effectively captures and learns ship motion patterns in both the temporal and frequency domains. Specifically, the branch net of DeepONet learns temporal features, while the trunk net performs Fourier filtering to capture the underlying ship motion patterns. In addition, the learning effectiveness of Fourier filtering is complemented by using MPL-net to enhance the extraction of detailed features in motion data. To evaluate the prediction performance of FMD, this study explores the optimal filtering frequency of the FMD model using experimental ship model motion data. Comparative testing with the DeepONet model includes multi-step prediction, coupled data forecasting, and generalization studies. Testing results demonstrate that the proposed FMD model improves prediction accuracy from 11.78% to 33.10% under Mean Squared Error (MSE) compared to the DeepONet model. Even under sea conditions ranging from mild to intense, the FMD model maintains high accuracy, with an improvement of over 30% in accuracy compared to DeepONet at longer step lengths under MSE conditions. Testing results indicate the superiority and advancement of FMD in prediction accuracy, generalization, and versatility, showcasing significant advantages in very short-term forecasting of ship motion.