Probabilistic Prediction Method Research Articles

Multi-market P2P trading of cooling–heating-power-hydrogen integrated energy systems: An equilibrium-heuristic online prediction optimization approach

In this paper, an equilibrium-heuristic online prediction optimization approach is proposed for multi-market peer-to-peer (P2P) electricity–hydrogen trading of integrated energy systems (IESs) with uncertainties. First, the IES, consisting of a hydrogen energy storage subsystem and a combined cooling, heating, and power subsystem, is constructed in the distribution network to improve energy utilization and market efficiency. Then, a bi-level optimization model for IESs, participating in the P2P electricity–hydrogen energy trading, the real-time electricity, and the ancillary service markets, is proposed, in which the top-level model can be formulated as a P2P electricity–hydrogen trading pricing model through the social welfare maximization problem, and the lower-level model is used to maximize the IES operating profit. To effectively solve the bi-level model, the game equilibrium-based ADMM distributed algorithm is used to obtain the P2P trading volume and prices of the top-level model, and a new hybrid heuristic algorithm, called hybrid sand cat swarm optimization and improved honey badger algorithm (SCIHB), is proposed to solve the lower level model. SCIHB utilizes a hybrid mechanism and an adaptive learning rate parameter to balance local exploitation and global exploration. Furthermore, a new online learning probabilistic prediction method based on the hidden Markov model and wavelet transform is introduced to describe the uncertainties of wind power, electricity prices, and frequency regulation prices. Finally, case studies are conducted on an IEEE 33-bus test system, and the numerical results verify the effectiveness of the proposed P2P trading model and solution approach.

Generative probabilistic prediction of precipitation induced landslide deformation with variational autoencoder and gated recurrent unit

Landslides, prevalent in mountainous areas, are typically triggered by tectonic movements, climatic changes, and human activities. They pose catastrophic risks, especially when occurring near settlements and infrastructure. Therefore, detecting, monitoring, and predicting landslide deformations is essential for geo-risk mitigation. The mainstream of the previous studies have often focused on deterministic models for immediate landslide prediction. However, in most of them, the aspect of prediction uncertainties are not sufficiently addressed. This paper introduces an innovative probabilistic prediction method using a Variational Autoencoder (VAE) combined with Gated Recurrent Unit (GRU) to forecast landslide deformations from a generative standpoint. Our approach consists of two main elements: firstly, training the VAE-GRU model to maximize the variational lower bound on the likelihood of historical precipitation data; secondly, using the learned approximated posterior distribution to predict imminent deformations from a generative angle. To assess the prediction quality, we use four widely-used metrics: Prediction Interval Coverage Probability (PICP), Prediction Interval Normalized Average Width (PINAW), Coverage Width-Based Criterion (CWC), and Prediction Interval Normalized Root Mean Square Width (PINRW). The results demonstrate that our proposed VAE-GRU framework surpasses traditional state-of-the-art (SOTA) probabilistic deformation prediction algorithms in terms of accuracy and reliability.

Probabilistic interval prediction method based on shape‐adaptive quantile regression

AbstractThis article introduces customized screening ensemble with shape‐adaptive quantile regression (CseAQR), a novel probabilistic interval forecasting method built upon the quantile regression model. CseAQR utilizes ensemble learning to perform adaptive quantile regression prediction, which can handle the heteroscedasticity feature in time series data by using a weighted adaptive allocation loss function to enhance the adaptability of the basic quantile regression model on the dataset. The model performance predictor is used to select the optimal ensemble learner combination, assign reasonable adaptive weights to it, and obtain a preliminary prediction interval through weighted aggregation. Combining ensemble learners not only improves the accuracy and robustness of prediction intervals but also ensures the commutativity required for conformal prediction. Finally, the conformal prediction method is applied to locally adjust the prediction interval, constructing a more consistently aligned prediction interval with the actual data on a narrower basis.

Probability prediction method for rockburst intensity based on rough set and multidimensional cloud model uncertainty reasoning

Probability prediction method for rockburst intensity based on rough set and multidimensional cloud model uncertainty reasoning

Optimizing prior distribution parameters for probabilistic prediction of remaining useful life using deep learning

In this study, a deep learning-based probabilistic remaining useful life (RUL) prediction model is proposed to improve the strong prior limitations of traditional probabilistic RUL prediction methods through a flexible prior distribution and strategy for sequential optimization of hyperparameters with regularization factor. It enables output richer probabilistic lifetime density distributions and confidence intervals with various parameters and overcome the problem of poor accuracy of short RUL predictions to some extent. Eventually, the model is effectively validated on a benchmark dataset, and the experimental results show that the probabilistic lifetime prediction model with optimized prior distribution parameters significantly improves prediction performance and demonstrates good learning performance and robustness of test results compared with traditional point estimation methods and parameter-free models. This study informs maintenance decisions and reliability assessments in engineering systems and guides the research and application of probabilistic-based prediction methods in deep learning framework.

Deep learning-based modeling method for probabilistic LCF life prediction of turbine blisk

Turbine blisk is one of the typical components of gas turbine engines. The fatigue life of turbine blisk directly affects the reliability and safety of both turbine blisk and aeroengine whole-body. To monitor the performance degradation of an aeroengine, an efficient deep learning-based modeling method called convolutional-deep neural network (C-DNN) method is proposed by absorbing the advantages of both convolutional neural network (CNN) and deep neural network (DNN), to perform the probabilistic low cycle fatigue (LCF) life prediction of turbine blisk regarding uncertain influencing parameters. In the C-DNN method, the CNN method is used to extract the useful features of LCF life data by adopting two convolutional layers, to ensure the precision of C-DNN modeling. The two close-connected layers in DNN are employed for the regression modeling of aeroengine turbine blisk LCF life, to keep the accuracy of LCF life prediction. Through the probabilistic analysis of turbine blisk and the comparison of methods (ANN, CNN, DNN and C-DNN), it is revealed that the proposed C-DNN method is an effective mean for turbine blisk LCF life prediction and major factors affecting the LCF life were gained, and the method holds high efficiency and accuracy in regression modeling and simulations. This study provides a promising LCF life prediction method for complex structures, which contribute to monitor health status for aeroengines operation.

A distribution-free method for probabilistic prediction

Machine learning techniques have been widely used and are mostly performed by predicting point estimation. Nevertheless, there are various scenarios that require more information beyond only point estimation. Probabilistic prediction is a typical research topic that provides probability distribution while predicting, which addressed with much attention recently. Most of the previous works make an assumption on the probabilistic distribution to a certain extent, which may lead to potential errors in the subsequent decision-making procedure. In this paper, we propose a distribution-free method for regression problems on real-value response under the probabilistic prediction framework and present an effective boosting-based method to perform the training process. Moreover, we introduce a further improved method to accomplish an unbiased mean estimation of the target distribution. Thorough experiments on multiple benchmark data are conducted to demonstrate the effectiveness of the proposed method with regard to different measures.

Failure Propagation Prediction of Complex Electromechanical Systems Based on Interdependence

Interdependence is an inherent feature of the cyber-physical system. Small damage to one component in the system may affect several other components, leading to a series of failures, thus collapsing the entire system. Therefore, the system failure is often caused by the failure of one or more components. In order to solve this problem, this paper focuses on a failure propagation probability prediction method for complex electromechanical systems, considering component states and dependencies between components. Firstly, the key component set in the system is determined based on the reliability measure. Considering the three coupling mechanisms of mechanical, electrical, and information, a topology network model of the system is constructed. Secondly, based on the topology network model and fault data, the calculation method of influence degree between components is proposed. Three state parameters are used to express the risk point state of each component in the system through mathematical representation, and the correlation coefficient between the risk point state parameters is calculated and measured based on the uncertainty evaluation. Then, the influence matrix between the system risk points is constructed, and the fault sequence is predicted by using the prediction function of an Artificial Neural Network (ANN) to obtain the fault propagation probability. Finally, the method is applied to the rail train braking system, which verifies that the proposed method is feasible and effective.

Prediction Method for Leakage Probability of Typical Fluid Machinery & Pipeline based on Cloud Model

Many dangerous chemicals are involved in the operation of fluid machinery, and the leakage of hazardous chemicals could trigger serious accidents, so the leakage probability prediction of fluid machinery and the effective maintenance before the equipment breakdowns are the critical factors in preventing accidents. A leakage probability prediction method based on a cloud model is proposed to aim at the standard fluid machinery with a frequent leakage probability in a chemical industry park. Due to the basic leakage probability of fluid machinery at home and abroad, the index system of leakage influencing factors is constructed and its influence degree is analysed. The modified typical fluid machinery leakage probability is then obtained using the cloud model to eliminate the subjectivity of quantitative results. The proposed method is utilized to estimate the leakage probability of Ethylene pipeline leakage of a pump in 1300# cracking gas compression area. The outcome 4.27 × 10−5 proves that this method can effectively solve the deviation problem in the leakage prediction process of domestic fluid machinery and guide the process of leakage probability prediction or accident prevention.

Sensor Management Method of Giving Priority to Confirmed Identified Targets.

The optimization objective function of sensor management for target identification is commonly established based on information theory indicators such as information gain, discrimination, discrimination gain, and quadratic entropy, which can control the sensors to reduce the overall uncertainty of all targets to be identified but ignores the speed of target being confirmed as identified. Therefore, inspired by the maximum posterior criterion of target identification and the target identification confirmation mechanism, we study a sensor management method that preferentially allocates resources to identifiable targets. Firstly, in the distributed target identification framework based on Bayesian theory, an improved identification probability prediction method that provides feedback the global identification results to local classifiers is proposed, which can improve the accuracy of identification probability prediction. Secondly, an effective sensor management function based on information entropy and expected confidence level is proposed to optimize the identification uncertainty itself rather than its variation, which can increase the priority of targets that satisfy the desired confidence level. In the end, the sensor management for target identification is modeled as a sensor allocation problem, and the optimization objective function based on the effective function is constructed, which can improve the target identification speed. The experimental results show that the correct identification rate of the proposed method is comparable to the methods based on information gain, discrimination, discrimination gain, and quadratic entropy in different scenarios, but the average time to confirm the identification is the shortest.

Probabilistic Forecasting of Residential Energy Consumption Based on SWT-QRTCN-ADSC-NLSTM Model

Residential electricity consumption forecasting plays a crucial role in the rational allocation of resources reducing energy waste and enhancing the grid-connected operation of power systems. Probabilistic forecasting can provide more comprehensive information for the decision-making and dispatching process by quantifying the uncertainty of electricity load. In this study, we propose a method based on stationary wavelet transform (SWT), quantile regression (QR), Bidirectional nested long short-term memory (BiNLSTM), and Depthwise separable convolution (DSC) combined with attention mechanism for electricity consumption probability prediction methods. First, the data sequence is decomposed using SWT to reduce the complexity of the sequence; then, the combined neural network model with attention is used to obtain the prediction values under different quantile conditions. Finally, the probability density curve of electricity consumption is obtained by combining kernel density estimation (KDE). The model was tested using historical demand-side data from five UK households to achieve energy consumption predictions 5 min in advance. It is demonstrated that the model can achieve both reliable probabilistic prediction and accurate deterministic prediction.

A parallel GRU with dual-stage attention mechanism model integrating uncertainty quantification for probabilistic RUL prediction of wind turbine bearings

The accurate probabilistic prediction of remaining useful life (RUL) of bearings plays an important role in ensuring the safe operation of wind turbine maintenance decision making. However, it is still a challenge to improve prediction accuracy and quantify prediction uncertainty, leading to the inability of precision and reliable prediction results. Accordingly, this paper proposed a novel probabilistic RUL prediction method combined the parallel GRUs with dual-stage attention mechanism (PDAGRU) prediction model with the non-parametric uncertainty quantification approach to overcome the limitations. In the PDAGRU model, the dual-stage attention mechanism is developed to improve the capability of degradation information extraction. Meanwhile, the parallel structure can enhance prediction accuracy and help quantify model uncertainty. The proposed uncertainty quantification approach with less prior knowledge can provide probabilistic RUL prediction results based on the kernel density estimation (KDE) and Monte Carlo (MC) dropout. Moreover, a first prediction time (FPT) determination method based on the isotonic regression is developed to more accurately reflect the degradation trajectory of bearings’ RUL. Two cases including simulated data and real-world data are deployed to verify the effectiveness of the proposed method. Compared with other methods, the superiority of the proposed method is verified. The RUL prediction accuracy and interval coverage probability for wind turbine bearings are high to 90.4% and 99.7% respectively.

Wind Power Probability Density Prediction Based on Quantile Regression Model of Dilated Causal Convolutional Neural Network

Aiming at the wind power prediction problem, a wind power probability prediction method based on the quantile regression of a dilated causal convolutional neural network is proposed. With the developed model, the Adam stochastic gradient descent technique is utilized to solve the cavity parameters of the causal convolutional neural network under different quantile conditions and obtain the probability density distribution of wind power at various times within the following 200 hours. The presented method can obtain more useful information than conventional point and interval predictions. Moreover, a prediction of the future complete probability distribution of wind power can be realized. According to the actual data forecast of wind power in the PJM network in the United States, the proposed probability density prediction approach can not only obtain more accurate point prediction results, it also obtains the complete probability density curve prediction results for wind power. Compared with two other quantile regression methods, the developed technique can achieve a higher accuracy and smaller prediction interval range under the same confidence level.

Vehicle Trajectory Prediction Considering Driver Uncertainty and Vehicle Dynamics Based on Dynamic Bayesian Network

Vehicle trajectory prediction is a crucial but intricate problem for lateral driving assistance systems because of driver uncertainty. This article presents a probabilistic vehicle-trajectory prediction method based on a dynamic Bayesian network (DBN) model integrating the driver’s intention, maneuvering behavior, and vehicle dynamics. By selecting a most-relevant-feature vector using joint mutual information, we design a Gaussian mixture model- hidden Markov model and employ the model as a node in the DBN to identify the driver’s intention. Then, a reference path is generated using the road information. The uncertainties of drivers are captured in steering- and longitudinal-control using a stochastic driver model and a Markov chain, respectively. A vehicle dynamic model ensures that the predicted vehicle trajectory adheres to the vehicle dynamics, which improves the prediction accuracy. A particle filter is used to recursively estimate the vehicle trajectory, including position coordinates and the lateral distance from the vehicle center of gravity to the road edge. We evaluate the proposed DBN trajectory prediction method in both lane-keeping and lane-changing scenarios based on a dataset collected from a real-time dynamic driving simulator. Results show that the proposed method can achieve accurate long-term trajectory prediction.

Probabilistic Prediction Method of Erosion Volume and Deposition Area from Rainfall Observation Data

We propose a methodology to estimate the spatial distribution of the probability of sediment deposition due to debris flow from rainfall data by combining the probability prediction of erosion volume based on an ordinal logistic regression and a sediment transport simulation. By using the Receiver Operating Characteristic (ROC) curve and Area Under Curve (AUC) we have selected the best combination of shortand long-term rainfall indices used as explanatory variables in the ordinal logistic model. The results showed that the regression model using 60-minute and 48-hour rainfall indices performed well and that the regression model using three events improved the predictability of local disasters in 2014. Furthermore, we performed Monte Carlo debris-flow simulations using rainfall data from 2014 using the model. We confirmed that the spatial distribution of disaster probability is consistent with the actual damage.

Probability forecasting method of Three Gorges Reservoir inflow flood based on error distribution estimation

准确、及时的入库洪水预报,对三峡水库综合效益的发挥和长江流域水旱灾害防御、水资源利用、流域综合管理等具有重要作用。基于预报误差的最优分布估计和分布函数动态参数假定,提出了一种三峡水库入库洪水概率预报方法,并进行了洪水概率预报业务试验。结果表明:本文所提方法科学可行,计算快捷,使用方便,便于在实时作业预报中应用推广;概率预报结果较确定性预报结果,在水量预报、预警效果等方面均有所改善,1~5 d预见期预报的确定性系数提高0.1%~3.4%,水量误差减少0.1%~4.8%,可为三峡水库实时调度提供更可靠的预报信息;所提出的三峡水库入库洪水概率预报业务化产品,可提供更多风险信息,为三峡水库的科学调度,尤其是洪水资源化利用提供更好的优化决策支撑。;Accurate and timely inflow flood forecasting plays an important role in the comprehensive benefit of the Three Gorges Reservoir, and the prevention of floods and droughts, the utilization of water resources, the comprehensive management for the Yangtze River basin. Based on the optimal distribution estimation of the prediction error and the assumption of the dynamic parameters for the distribution function, a probability prediction method for the inflow flood of the Three Gorges Reservoir is proposed, and the operational test of the probability prediction of the flood is carried out. The results show that the physical rational of the presented method is clear, and it is easy to be applied and popularized in real-time operation forecast. Compared with the deterministic prediction results, the probability forecast has improved in water volume prediction and early warning effect. For example, the certainty coefficient of forecast in 1-5 d is increased by 0.1%-3.4%, and the water volume error is reduced by 0.1%-4.8%, which can provide more reliable forecast information for the real-time operation of the Three Gorges Reservoir. The proposed operational product of probability forecast of inflow flood process can provide more risk information, and provide better support for scientific and fine operation of the Three Gorges Reservoir, especially for optimal decision-making of flood resource utilization.

A probabilistic inverse prediction method for predicting plutonium processing conditions

In the past decade, nuclear chemists and physicists have been conducting studies to investigate the signatures associated with the production of special nuclear material (SNM). In particular, these studies aim to determine how various processing parameters impact the physical, chemical, and morphological properties of the resulting special nuclear material. By better understanding how these properties relate to the processing parameters, scientists can better contribute to nuclear forensics investigations by quantifying their results and ultimately shortening the forensic timeline. This paper aims to statistically analyze and quantify the relationships that exist between the processing conditions used in these experiments and the various properties of the nuclear end-product by invoking inverse methods. In particular, these methods make use of Bayesian Adaptive Spline Surface models in conjunction with Bayesian model calibration techniques to probabilistically determine processing conditions as an inverse function of morphological characteristics. Not only does the model presented in this paper allow for providing point estimates of a sample of special nuclear material, but it also incorporates uncertainty into these predictions. This model proves sufficient for predicting processing conditions within a standard deviation of the observed processing conditions, on average, provides a solid foundation for future work in predicting processing conditions of particles of special nuclear material using only their observed morphological characteristics, and is generalizable to the field of chemometrics for applicability across different materials.

Probabilistic prediction of ground-motion intensity for regions lacking strong ground-motion records

Probabilistic prediction of ground-motion intensity in regions lacking strong ground-motion records is a vital issue for seismic structural design. Several approaches have been suggested for this purpose, and nearly all of them directly import and adjust the ground-motion prediction equation (GMPE) of ground-motion measures such as the peak ground acceleration (PGA) or spectral acceleration (SA) from data-rich regions. However, as the transmissions of PGA and SA from the source to the site correspond to nonlinear processes, this import and adjustment may lead to an unrealistic evaluation of ground motion. In this study, a novel probabilistic prediction method of ground-motion intensity for such regions is proposed. In contrast to the current approach wherein the GMPE of ground-motion measures such as PGA or SA is directly used, a Fourier amplitude spectral (FAS) model is suggested to express the seismic transmission process from the source to the site. The ground-motion intensities of PGA or SA are obtained from the FAS model using the random vibration theory. The exceedance probability of ground-motion intensity is calculated based on Monte Carlo simulations. As the FAS conforms to the linear system theory and the determination of FAS model does not require too many ground-motion records, the proposed method should be convenient for the probabilistic prediction of ground-motion intensity in regions lacking strong ground-motion records.

A probability prediction method for the classification of surrounding rock quality of tunnels with incomplete data using Bayesian networks

The classification of surrounding rock quality is critical for the dynamic construction and design of tunnels. However, obtaining complete parameters for predicting the surrounding rock grades is always challenging in complex tunnel geological environment. In this study, a new method based on Bayesian networks is proposed to predict the probability for the classification of surrounding rock quality of tunnel with incomplete data. A database is collected with 286 cases in 10 tunnels, involving nine parameters: rock hardness, weathering degree, rock mass integrity, rock mass structure, structural plane integrity, in-situ stress, groundwater, rock basic quality, and surrounding rock level. Moreover, the Bayesian network structure is built using the collected database and quantitatively verified by strength analysis. Then, the accuracy, precision, recall, F-measure and receiver operating characteristic (ROC) curves are utilized for model evaluation. The average values of accuracy, precision, recall, F-measure, and area under the curve (AUC) are approximately 89.2%, 91%, 92%, 91%, and 0.98, respectively. These results indicate that the established classification model has high accuracy, even with small sample size and imbalanced samples. Ten additional sets of tunnel cases (incomplete data) are also used for verification. The results reveal that compared with the traditional Q-system (Q) and rock mass rating (RMR) classification methods, the proposed classification model has the lowest error rate and is capable of using incomplete data to predict sample results. Finally, sensitivity analysis suggests that the rock hardness and rock mass integrity have the strongest impact on the quality of tunnel surrounding rock. Overall, the findings of this study can serve as a useful reference for future rock mass quality evaluation in tunnels, underground powerhouses, slopes, etc.

Dealing with multi-criteria decision analysis in time-evolving approach using a probabilistic prediction method

Multi-criteria Decision Analysis (MCDA) is a methodology that has been classically used to rank alternatives according to a set of decision criteria. The MCDA techniques have been shown to be an efficient tool in a number of real-life engineering problems. Nevertheless, most of the proposed approaches in the field do not consider the temporal characteristic of the criteria values, which can be an interesting information to be explored in order to predict future rankings. The present work proposes a novel MCDA methodology in which the past data of the criteria are considered to predict their future values. Our approach is based on a tensorial formulation, together with the use of the recursive least mean squares method in the prediction step. In addition, we consider a probabilistic prediction and use the Stochastic Multi-criteria Acceptability Analysis, which allows the decision maker to observe the degree of uncertainty in the ranking. Numerical experiments with synthetic and actual data attest to the proposal’s relevance in scenarios in which the criteria values change over time.