Probabilistic Power Research Articles

A comparative study of different deep learning methods for time-series probabilistic residential load power forecasting

The widespread adoption of nonlinear power electronic devices in residential settings has significantly increased the stochasticity and uncertainty of power systems. The original load power data, characterized by numerous irregular, random, and probabilistic components, adversely impacts the predictive performance of deep learning techniques, particularly neural networks. To address this challenge, this paper proposes a time-series probabilistic load power prediction technique based on the mature neural network point prediction technique, i.e., decomposing the load power data into deterministic and stochastic components. The deterministic component is predicted using deep learning neural network technology, the stochastic component is fitted with Gaussian mixture distribution model and the parameters are fitted using great expectation algorithm, after which the stochastic component prediction data is obtained using the stochastic component generation method. Using a mature neural network point prediction technique, the study evaluates six different deep learning methods to forecast residential load power. By comparing the prediction errors of these methods, the optimal model is identified, leading to a substantial improvement in prediction accuracy.

State-of-the-Art Probabilistic Solar Power Forecasting: A Structured Review

In recent years, the installed capacity increment with regard to solar power generation has been highlighted as a crucial role played by Photovoltaic (PV) generation forecasting in integrating a growing number of distributed PV sites into power systems. Nevertheless, because of the PV generation’s unpredictable nature, deterministic point forecast methods struggle to accurately assess the uncertainties associated with PV generation. This paper presents a detailed structured review of the state-of-the-art concerning Probabilistic Solar Power Forecasting (PSPF), which covers forecasting methods, model comparison, forecasting horizon and quantification metrics. Our review methodology leverages the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) approach to systematically identify primary data sources, focusing on keywords such as probabilistic forecasting, Deep Learning (DL), and Machine learning (ML). Through an extensive and rigorous search of renowned databases such as SCOPUS and Web of Science (WoS), we identified 36 relevant studies (n=36). Consequently, expert scholars decided to develop three themes: (1) Conventional PSPF, (2) PSPF utilizing ML, and (3) PSPF using DL. Probabilistic forecasting is an invaluable tool concerning power systems, especially regarding the rising proportion of renewable energy sources in the energy mix. We tackle the inherent uncertainty of renewable generation, maintain grid stability, and promote efficient energy management and planning. In the end, this research contributes to the development of a power system that is more resilient, reliable, and sustainable.

A novel hybrid model based on evolving multi-quantile long and short-term memory neural network for ultra-short-term probabilistic forecasting of photovoltaic power

Probabilistic forecasting is extremely crucial in eliminating uncertainty in photovoltaic (PV) power generation. Quantile regression long and short-term memory neural network (QRLSTM) is widely recognized as promising methods for PV power probabilistic forecasting due to their strong generalization ability. However, these models train the model for each quantile individually, which lacks consideration of the correlation and monotonicity between quantiles, and multiple training leads to excessive computational complexity. Furthermore, the non-differentiable pinball loss function generated by QR places significant demands on the optimization algorithms. To address these issues, this paper proposes an evolutive distributed chaotic particle swarm optimization (EDCPSO)-optimized multi-quantile LSTM (MQLSTM) to achieve high-quality probabilistic PV power prediction. MQLSTM is a multi-output network structure that simultaneously outputs all quantile estimates and adopts a loss function with all quantile scores and non-crossing constraints to guide the training of the model. This approach not only improves the quality and reasonableness of quantile estimations, but also reduces computational difficulty. Then, from the perspective of evolutionary computation, considering the weight parameters of each connection layer in MQLSTM as decision variables, we convert the probabilistic forecasting into an optimization problem and propose a EDCPSO to solve the training difficulty. It implements a targeted distributed chaos strategy based on the evolutionary state to improve convergence speed and search capability. The proposed model is tested to be superior in real cases.

Probabilistic Power and Energy Balance Risk Scheduling Method Based on Distributed Robust Optimization

The volatility and uncertainty associated with the high proportion of wind and PV output in the new power system significantly impact the power and energy balance, making it challenging to accurately assess the risks related to renewable energy abandonment and supply guarantee. Therefore, a probabilistic power and energy balance risk analysis method based on distributed robust optimization is proposed. Firstly, the affine factor and the flexible ramp reserve capacity of thermal power are combined to establish a probabilistic index, which serves to characterize the risk associated with the power and energy balance. Drawing upon the principles of the conditional value at risk theory, the risk indexes of the load shedding power and curtailment power under a certain confidence probability are proposed. Secondly, the probability distribution fuzzy sets of uncertain variables are constructed using the distributionally robust method to measure the Wasserstein distance between different probability distributions. Finally, aiming at minimizing the operation cost of thermal power, the risk cost of power abandonment, and the risk cost of load shedding, a distributed robust optimal scheduling model based on a flexible ramp reserve of thermal power is established.

Identification and Evaluation of Vulnerable Links in a Distribution Network with Renewable Energy Source Based on Minimum Discriminant Information

With the increase in the proportion of photovoltaic and wind power access, the scale and form of distribution networks are becoming more and more complex. The traditional single distribution network vulnerability assessment method is difficult to use to identify the vulnerable links in the distribution network. Therefore, this paper proposes a method for identifying and evaluating vulnerable links in distribution networks based on minimum discriminant information. First, considering the influence of distributed grid connection, an improved probabilistic power flow calculation method is proposed, which improves the calculation efficiency and accuracy. Second, considering the correlation degree, transmission capacity, and voltage stability of branches in the distribution network, the identification index of vulnerable lines is defined. Based on power quality and operating state, the identification index of vulnerable nodes in a distribution network is defined. Finally, based on the indicators of vulnerable nodes and vulnerable lines, the vulnerable links in the distribution network are comprehensively evaluated based on the principle of minimum discriminant information, and the vulnerable links of the entire distribution network are evaluated according to different degrees of vulnerability. The rationality and effectiveness of the proposed method are verified via an example analysis of actual power grid data.

Voltage over-limit risk assessment of wind power and photovoltaic access distribution system based on day-night segmentation and Gaussian mixture model

The access of distributed power sources such as wind power and photovoltaic (PV) with randomness and uncertainty makes the operation of distribution system more complicated. It is particularly necessary to comprehensively and efficiently evaluate the risk of voltage over-limit in distribution systems with wind power and photovoltaic. Firstly, aiming at the problem of centralized sampling of Monte Carlo method in risk assessment of distribution system, a day-night segmented Monte Carlo (MC) sampling method is proposed to generate a large number of day-night sampling scenarios. Secondly, the Gaussian mixture model is used to fit the probability density of the voltage obtained from probabilistic power flow calculations, concurrently, the utility preference exponential function characterizes the severity of the over-limit, and a comprehensive evaluation model of voltage over-limit risk is constructed. Finally, the effectiveness and accuracy of the proposed method are verified by taking the improved IEEE 33-bus distribution system as an example. These results offer actionable insights for future risk assessments of new energy grid connections.

Fractional Moments Based Adaptive Scaled Unscented Transformation for Probabilistic Power Flow of AC-DC Hybrid Grids

Fractional Moments Based Adaptive Scaled Unscented Transformation for Probabilistic Power Flow of AC-DC Hybrid Grids

Physics-informed reinforcement learning for probabilistic wind power forecasting under extreme events

With wind power penetration increases, accurate and reliable wind power forecasting is becoming gradually critical, and data driven model is a promising solution to implement this task. However, limited by deficient data samples under extreme conditions, scarcity of feature measurements fails to meet the number of training samples required, making the forecasting model exhibits low adaptability. This paper proposes a physics-informed reinforcement learning based method for probabilistic forecasting. Analytical physical expression of wind power output under extreme event is established to construct the error evaluation function for abrupt feature. Deep deterministic policy gradient-based quantile fitting model is then proposed, with abrupt feature embedded as the auxiliary input data for neural network. On this basis, parameter training technique under small data set is proposed to deal with the effect of extreme conditions on correction process of network. It extracts key historical transitions from experience replay pool to establish effective training samples, and model parameters are updated through the small-batch learning strategy to minimize long-term error feedback. Test result on the practical cold wave event of wind farms in China shows effectiveness of the proposed method.

Interpretable feature selection and deep learning for short-term probabilistic PV power forecasting in buildings using local monitoring data

Accurate probabilistic forecasting of photovoltaic (PV) power is crucial for optimizing energy scheduling in smart buildings and ensuring the low-carbon, efficient operation of building energy management systems (BEMS). However, existing feature selection techniques fail to guarantee that the selected features genuinely impact the output of forecasting models. Additionally, traditional black-box deep learning models lack clarity on whether their output truly relies on those selected features. These challenges limit the accuracy of forecasting models. To address these challenges, a novel methodology named temporal importance model explanation and temporal fusion transformers (TIME-TFT) model is proposed. Firstly, the TIME method is employed for feature selection, and interpretable outputs are used to identify important global features. Secondly, the TFT model is then utilized for forecasting tasks, providing interpretable outputs to track back to the features that TFT model pays attention to. Finally, the consistency between the interpretable outputs of TIME method and TFT model is examined to confirm predictions are based on genuinely learned selected features. Empirical studies demonstrate the superiority of the proposed TIME-TFT system, outperforming comparable models with an R2 of 0.9546. In summary, the interpretable outputs not only improve the accuracy of predictions but also provides visual evidence for predictions, thereby bolstering effectiveness and credibility in engineering practices.

Impact of electric vehicles and wave energy systems on OPF of power networks using hybrid Osprey-PSO approach

In this article, a novel optimization algorithm named hybrid Osprey Optimization and Particle Swarm Optimization (OOPSO) is proposed to solve the Probabilistic Optimal Power Flow (POPF) problem in power systems with high penetration of renewable energy resources (RERs), such as photovoltaic, wind, and wave energy and integration of electric vehicles (EVs). The algorithm's results are compared with established optimization algorithms like Osprey Optimization Algorithm (OOA), Particle Swarm Optimization (PSO), and Heap Optimization. Applied to the 30 and 118-bus IEEE test systems, the OOPSO algorithm efficiently deals with the uncertainties produced by RERs generation. This research study includes EV charging profiles for weekdays and weekends. It considers home and work-level charging scenarios. The RERs and EVs penetration ensure system reliability while minimizing generation costs and satisfying power flow constraints. Compared to existing algorithms, the OOPSO algorithm achieves superior performance, faster convergence, and higher solution accuracy. The simulation results show the effectiveness of OOPSO. It appears to be a robust tool for POPF, integrating RERs in power systems with EVs. This work highlights the potential of OOPSO as a feasible approach for power system optimization in the context of RERs and EV integration, paving the way for further research into unconventional grid optimization techniques.

Probabilistic power flow based on physics-guided graph neural networks

Under the high-dimensional and nonlinear stochastic power system environment, artificial intelligence (AI) is becoming a promising alternative to the urgent demand for probabilistic power flow (PPF). However, traditional AI only learns from experiences in data and is unable to satisfy physical constraints. This may result in an undesired infeasible inference for PPF. To solve this issue, a physics-guided graph neural network (PG-GNN)-based PPF analysis method is proposed. At the outset, alternating current power flow equation is analyzed to produce gradients for connecting physical component and AI. The proposed PG-GNN then incorporates the sensitivities into the training tensor graph such that it enables in-depth physical learning. At the same time, the GNN serves to enhance generalizability in varying topology. The well-trained PG-GNN is finally leveraged to surrogate traditional power flow solver in Monte-Carlo method, such that real-time PPF is realized. Numerical tests on benchmarks demonstrate that, in unseen scenarios, the proposed method enables only little gap versus the exact baseline, but beats other rivals with over tenfold computing efficiency. Moreover, comparisons with traditional AIs manifest that our method holds better generalizability.

Collaborative monitoring of wind turbine performance based on probabilistic power curve comparison

A wind farm is usually equipped with multiple wind turbines of the same type. These wind turbines often work under same complex conditions. Accurate performance degradation monitoring is crucial for ensuring the reliable operation of wind farms and reducing maintenance costs. Motivated by this, this article develops a new wind turbine performance degradation monitoring scheme, which is based on pairwise comparison of the probability power curves of different wind turbines in a wind farm. Firstly, covariate matching is used to eliminate the inherent differences in meteorological variables of different turbines within the same data segment. Next, two probabilistic wind power curves, the quantile power curve and density power curve, model the functional relationship between the meteorological variables and wind power output. Then, deviation vectors are generated by calculating the deviation of probabilistic power curves between each pair of wind turbines. Finally, a directional Hotelling T2 control chart is proposed to monitor the deviation vectors. We apply the new method on the real data of a wind farm in East Britain. Results show that the proposed monitoring technique can monitor wind turbine performance degradation more precisely and comprehensively than the existing approaches.

Hybrid model for intra-day probabilistic PV power forecast

Renewable generation forecasting has become an essential element of power system management because of the ever-increasing share of renewable power fed into the grid. Solar energy is one of the most common and well-known sources of energy in existing networks. Due to its intermittent and non-linear nature, accurate prediction of solar irradiance is essential for ensuring the reliable integration of photovoltaic (PV) plants with the grid system and for effectively managing supply and demand. The dispatching problem is typically addressed by generating a forecast for a few sample plants and scaling the result of these forecasts by the PV power generation capacity connected to the electric system. This research presents a comprehensive comparison of various methods for predicting PV power generation in fixed and single axis tracking PV systems at two different locations. These methods include time-series statistical regression, parametric physical approaches, machine learning (ML), and ensemble techniques, aimed at intra-day forecasts ranging from 1 to 6 h ahead. The study introduces a novel method for developing a parametric linear model for fixed PV systems and uses model output statistics (MOS) to improve forecast accuracy. Additionally, this study details various optimization algorithms designed to refine model parameters for creating a hybrid model. It also examines key factors influencing PV power forecasts and applies post-processing Kalman filter techniques. The study provides a detailed analysis of deterministic and probabilistic forecasting models, offering not just the most probable power production estimate, but also conveying the associated uncertainty. The results demonstrate a highly skilled ML based hybrid model to forecast PV power for different technologies at two different locations.

An Adaptive Approach for Probabilistic Wind Power Forecasting Based on Meta-Learning

An Adaptive Approach for Probabilistic Wind Power Forecasting Based on Meta-Learning

A Stochastic Recurrent Encoder Decoder Network for Multistep Probabilistic Wind Power Predictions.

In this article, a stochastic recurrent encoder decoder neural network (SREDNN), which considers latent random variables in its recurrent structures, is developed for the first time for the generative multistep probabilistic wind power predictions (MPWPPs). The SREDNN enables the stochastic recurrent model under the encoder-decoder framework to engage exogenous covariates to produce better MPWPP. The SREDNN consists of five components, the prior network, the inference network, the generative network, the encoder recurrent network, and the decoder recurrent network. The SREDNN is equipped with two critical advantages compared with conventional RNN-based methods. First, the integration over the latent random variable builds an infinite Gaussian mixture model (IGMM) as the observation model, which drastically increases the expressiveness of the wind power distribution. Secondly, hidden states of the SREDNN are updated in a stochastic way, which builds an infinite mixture of the IGMM for describing the ultimate wind power distribution and enables the SREDNN to model complex patterns across wind speed and wind power sequences. Computational experiments are conducted on a dataset of a commercial wind farm having 25 wind turbines (WTs) and two publicly assessable WT datasets to verify the advantages and effectiveness of the SREDNN for MPWPP. Experimental results show that the SREDNN achieves a lower negative form of the continuously ranked probability score (CRPS*) as well as a superior sharpness and comparable reliability of prediction intervals by comparing against considered benchmarking models. Results also show the clear benefit gained from considering latent random variables in SREDNN.

Probabilistic optimal power flow computation for power grid including correlated wind sources

AbstractThis paper sets out to develop an efficient probabilistic optimal power flow (POPF) algorithm to assess the influence of wind power on power grid. Given a set of wind data at multiple sites, their marginal distributions are fitted by a newly developed generalized Johnson system, whose parameters are specified by a percentile matching method. The correlation of wind speeds is characterized by a flexible Liouville copula, which allows to model the asymmetric dependence structure. In order to improve the efficiency for solving POPF problem, a lattice sampling method is developed to generate wind samples at multiple sites, and a logistic mixture model is proposed to fit distributions of POPF outputs. Finally, case studies are performed, the generalized Johnson system is compared with Weibull distribution and the original Johnson system for fitting wind samples, Liouville copula is compared against Archimedean copula for modelling correlated wind samples, and lattice sampling method is compared with Sobol sequence and Latin hypercube sampling for solving POPF problem on IEEE 118‐bus system, the results indicate the higher accuracy of the proposed methods for recovering the joint cumulative distribution function of correlated wind samples, as well as the higher efficiency for calculating statistical information of POPF outputs.

Hybrid model based on similar power extraction and improved temporal convolutional network for probabilistic wind power forecasting

Accurate wind power generation forecasting is of great significance to improve the operation of power system. Probabilistic forecasting has a higher application value in power grid because it can provide more abundant forecasting information than deterministic forecasting. In addition, multi-step forecasting can provide forecasting results in a longer time range, so that decision makers can make longer-term planning and strategic arrangements. In this paper, we propose a novel multi-step improved temporal convolutional network based on quadratic spline quantile function (MITCN-QSQF) for probabilistic wind power forecasting. First, we combine maximum information coefficient, Gaussian similarity and adaptive resample to propose an effective similar power generation feature extraction method (MGR) for power generation. Then the temporal convolutional network is improved to construct the multi-step time series forecasting model MITCN. By combining the proposed model and the powerful probabilistic forecasting method quadratic spline quantile function (QSQF), high-quality probabilistic forecasting of wind power is achieved. Through comprehensive simulations on an open-source dataset, the superiority and efficiency of the proposed method are verified. Compared with some advanced benchmarks, the proposed model can obtain more accurate deterministic and probabilistic forecasting results.

A transfer learning-based hybrid model with LightGBM for smart grid short-term energy load prediction

Efficient energy management is crucial given 2024's 4.1% worldwide electricity demand increase. This urgency emphasizes the necessity for various, sustainable energy sources in distribution grids. Short-term load prediction approaches using probabilistic power generation and energy storage are crucial for energy usage prediction. Urban energy planners use simulation, probability optimization and modelling to create sustainable energy systems. This study offers a novel hybrid model for smart grids: short-term energy load prediction using transfer learning (TL) and optimized lightGBM (OLGBM). Our two-phase solution tackles Short-term Load Forecasting complexities. First, aberrant supplements and quick deviation selection eliminate missing values and identify key features during data pre-processing. Second, TL-OLGBM learns dynamic time scales and complex data patterns with Bayesian optimization of hyperparameters to improve forecasting accuracy. Additionally, our architecture easily combines the newest Smart and Green Technology, enabling energy system innovation. Comparative performance research shows that our technique outperforms similar models in mean absolute percentage error, accuracy and root mean square error. This hybrid model is a reliable short-term energy load forecast solution that fits the dynamic terrain of smart and green technology integration in modern energy systems.

Enhancing Probabilistic Solar PV Forecasting: Integrating the NB-DST Method with Deterministic Models

Accurate quantification of uncertainty in solar photovoltaic (PV) generation forecasts is imperative for the efficient and reliable operation of the power grid. In this paper, a data-driven non-parametric probabilistic method based on the Naïve Bayes (NB) classification algorithm and Dempster–Shafer theory (DST) of evidence is proposed for day-ahead probabilistic PV power forecasting. This NB-DST method extends traditional deterministic solar PV forecasting methods by quantifying the uncertainty of their forecasts by estimating the cumulative distribution functions (CDFs) of their forecast errors and forecast variables. The statistical performance of this method is compared with the analog ensemble method and the persistence ensemble method under three different weather conditions using real-world data. The study results reveal that the proposed NB-DST method coupled with an artificial neural network model outperforms the other methods in that its estimated CDFs have lower spread, higher reliability, and sharper probabilistic forecasts with better accuracy.

Probabilistic wind power forecasting resilient to missing values: An adaptive quantile regression approach

Probabilistic wind power forecasting approaches have significantly advanced in recent decades. However, forecasters often assume data completeness and overlook the challenge of missing values resulting from sensor failures, network congestion, etc. Traditionally, this issue is addressed during the data preprocessing procedure using methods such as deletion and imputation. Nevertheless, these ad-hoc methods pose challenges to probabilistic wind power forecasting at both parameter estimation and operational forecasting stages. In this paper, we propose a resilient probabilistic forecasting approach that smoothly adapts to missingness patterns without requiring preprocessing or retraining. Specifically, we design an adaptive quantile regression model with parameters capable of adapting to missing patterns, comprising two modules. The first is a feature extraction module where weights are kept static and biases are designed as a function of missingness patterns. The second is a non-crossing quantile neural network module, ensuring monotonicity of quantiles, with higher quantiles derived by adding non-negative amounts to lower quantiles. The proposed approach is applicable to cases under all missingness mechanisms including missing-not-at-random cases. Case studies demonstrate that our proposed approach achieves state-of-the-art results in terms of the continuous ranked probability score, with acceptable computational cost.