Time Series Prediction Algorithm Research Articles

Energy-saving potential benchmarking method of office buildings based on probabilistic forecast

Building energy consumption represents a substantial portion of China's total energy usage profile. Conducting a quantitative assessment of the energy-saving potential in buildings is crucial, as it aids in estimating the investment risk and payback period for energy-efficient projects. This evaluation also informs the development and formulation of policies promoting building retrofits and energy conservation. However, this field's present state of research exhibits a significant gap. This study takes 8 typical office buildings as case objects and conducts relevant research on quantitative evaluation of energy-saving potential. On the basis of quantile regression, three corresponding probabilistic forecasting models for building energy consumption are established by combining the classic time series prediction algorithm Long Short-Term Memory (LSTM) in data-driven methods, Gradient Boosting Decision Tree (GBDT) in known building energy benchmarking, and a relatively simple linear model. The evaluation results demonstrate that the LSTM-QR model can serve effectively as a probabilistic forecasting model for estimating building energy-saving potential. Based on the LSTM-QR model results, a quantitative evaluation method for building energy-saving potential benchmarking is proposed, utilizing linearized Cumulative Distribution Function (CDF) curves. This method can quantitatively assess the energy-saving potential of the studied buildings. By analyzing and comparing these buildings' hourly energy consumption data over the past three years, it is evident that the quantitative scoring results from this method can provide directional guidance for building renovation projects to enhance energy efficiency.

Constrained Flooding Based on Time Series Prediction and Lightweight GBN in BLE Mesh.

Bluetooth Low Energy Mesh (BLE Mesh) enables Bluetooth flexibility and coverage by introducing Low-Power Nodes (LPNs) and enhanced networking protocol. It is also a commonly used communication method in sensor networks. In BLE Mesh, LPNs are periodically woken to exchange messages in a stop-and-wait way, where the tradeoff between energy and efficiency is a hard problem. Related works have reduced the energy consumption of LPNs mainly in the direction of changing the bearer layer, improving time synchronization and broadcast channel utilization. These algorithms improve communication efficiency; however, they cause energy loss, especially for the LPNs. In this paper, we propose a constrained flooding algorithm based on time series prediction and lightweight GBN (Go-Back-N). On the one hand, the wake-up cycle of the LPNs is determined by the time series prediction of the surrounding load. On the other, LPNs exchange messages through lightweight GBN, which improves the window and ACK mechanisms. Simulation results validate the effectiveness of the Time series Prediction and LlightWeight GBN (TP-LW) algorithm in energy consumption and throughput. Compared with the original algorithm of BLE Mesh, when fewer packets are transmitted, the throughput is increased by 214.71%, and the energy consumption is reduced by 65.14%.

Traffic flow prediction based on nonlinear weight decreasing PSO-SVR univariate time series prediction algorithm

The aim of this paper is to investigate the application of car flow prediction in the field of transport in order to solve the problem of urban traffic congestion. For this purpose, we adopt the nonlinear weight decreasing PSO-SVR univariate time series prediction algorithm to predict the car flow, and divide the data set into training set and test set according to the ratio of 7:3. By analysing the scatterplot and line graph between the predicted and actual values of the training and test sets, we find that the prediction effect is better, but there is a certain deviation. Specifically, the scatter plot shows a Y=X distribution, but the actual values have a wide range of variation relative to the predicted values. Meanwhile, the line graph shows that the actual traffic flow and the predicted traffic flow have the same trend, but the actual traffic flow value changes more, while the predicted traffic flow value changes less. This may be due to the long time span resulting in too many cycles, and shortening the time span can reduce the number of cycles and thus improve the prediction accuracy. Further analysis of the MAE values for the training and test sets shows that the MAE values for both the training and test sets are relatively small, 1.8437 and 2.6408, respectively, where the MAE for the test set is on the large side, but the overall prediction results of the model are better. Therefore, the nonlinear weight decreasing PSO-SVR univariate time-series prediction algorithm used in this paper can provide powerful decision support for traffic management departments and help them to better formulate traffic planning and management strategies.

PROACTIVE HORIZONTAL SCALING METHOD FOR KUBERNETES

Context. The problem of minimizing redundant resource reservation while maintaining QoS at an agreed level is crucial for modern information systems. Modern information systems can include a large number of applications, each of which uses computing resources and has its own unique features, which require a high level of automation to increase the efficiency of computing resource management processes.
 Objective. The purpose of this paper is to ensure the quality of IT services at an agreed level in the face of significant dynamics of user requests by developing and using a method of proactive automatic application scaling in Kubernetes.
 Method. This paper proposes a proactive horizontal scaling method based on the Prophet time series prediction algorithm. Prometheus metrics storage is used as a data source for training and validating forecasting models. Based on the historical metrics, a model is trained to predict the future utilization of computation resources using Prophet. The obtained time series is validated and used to calculate the required number of application replicas, considering deployment delays.
 Results. The experiments have shown the effectiveness of the proposed proactive automated application scaling method in comparison with existing solutions based on the reactive approach in the selected scenarios. This method made it possible to reduce the reservation of computing resources by 47% without loss of service quality compared to the configuration without scaling.
 Conclusions. A method for automating the horizontal scaling of applications in Kubernetes is proposed. Although the experiments have shown the effectiveness of this solution, this method can be significantly improved. In particular, it is necessary to consider the possibility of integrating a reactive component for atypical load patterns.

Forecasting individual bids in real electricity markets through machine learning framework

With the increasing uncertainty caused by the complexity of the world’s energy environment and the increasing penetration rate of renewable energy, it is significant to estimate the future operation of power markets in advance. Forecasting individual bids in spot electricity markets is a promising new method for achieving so, but it has not been fully studied due to the difficulty of forecasting a bid function. The idealization of existing optimization-based models decreases their practical effects in real markets. Thus, we propose a scalable forecasting framework that incorporates several customized state-of-art machine learning methods according to the characteristics of the bidding data. First, several low-rank approximation algorithms are customized to encode the high-dimensional bidding curves into low-dimensional feature spaces and reconstruct them from the predicted feature space. Second, a transformer-based multidimensional time series prediction algorithm is proposed to predict the bidding feature based on both related factors and historical bidding records. To appropriately evaluate the performances of the forecasting methods, we introduce a dynamic criterion based on the economic implications of bids. The comprehensive framework is tested based on actual market data from the Australian national electricity market, and in the empirical example, the feasibility and effectiveness of the proposed framework are demonstrated.

Transformer based Kalman Filter with EM algorithm for time series prediction and anomaly detection of complex systems

Modeling time series signals is critical in modern intelligent systems, which are able to determine the health condition and detect potential anomalies of complex systems by comparing predictions and measured signals. The accurate prediction of sequential signals is still challenging due to the conditional independence of measurements. This study proposes a Transformer based Kalman Filter (TKF) for accurate time-varying signal prediction, which contributes to capturing temporal features among time-series data. Then, Expectation Maximization (EM) algorithm is adopted to estimate parameters for Kalman Filter with hidden states, which avoids parameters’ determination heavily rely on manual experience. We conduct experiments on real-world time-series signals collected from complex systems. The results show that the proposed TKF achieves higher prediction accuracy and has potential to be applied for anomaly detection of complex systems.

Comparative Analysis of Time Series Prediction Algorithms on Multiple Network Function Data of NWDAF

With the emergence and vigorous development of 5G technology, there is a significant surge in network usage and traffic, resulting in heightened complexity within network and IT environments. This exponential increase in activity produces a plethora of events, making conventional systems inadequate for the efficient management of 5G networks. In comparison to 4G technology, 5G technology brings forth a host of new features, one of which is the network data analytics function (NWDAF). This function grants network operators the flexibility to either employ their own data analytics methodologies based on machine learning (ML) and deep learning (DL) into their networks. In this paper, we present a dataset named “NWDAF-NFPP” for network function performance time series prediction, collected from a laboratory at China Telecom. The dataset is carefully anonymized to ensure maximum realism and comprehensiveness, while safeguarding sensitive information. It encompasses eight categories of network functions, with data collected at five-minute intervals. The availability of this dataset provides valuable resources for researchers to conduct time series prediction research on network element performance. Following data collection, a total of six models were employed for network element performance prediction, encompassing both machine learning and deep learning approaches. This diverse set of models was carefully chosen to ensure comprehensive coverage of different techniques and algorithms. Through the comparison and analysis of these models, we aim to evaluate their predictive capabilities and identify the most effective approach for network element performance prediction. This comparative analysis will provide valuable insights into the strengths and limitations of each model, aiding in informed decision-making for network optimization and management strategies in the future.

Ultra-short-term forecasting model of power load based on fusion of power spectral density and Morlet wavelet.

An accurate ultra-short-term time series prediction of a power load is an important guarantee for power dispatching and the safe operation of power systems. Problems of the current ultra-short-term time series prediction algorithms include low prediction accuracy, difficulty capturing the local mutation features, poor stability, and others. From the perspective of series decomposition, a multi-scale sequence decomposition model (TFDNet) based on power spectral density and the Morlet wavelet transform is proposed that combines the multidimensional correlation feature fusion strategy in the time and frequency domains. By introducing the time-frequency energy selection module, the "prior knowledge" guidance module, and the sequence denoising decomposition module, the model not only effectively delineates the global trend and local seasonal features, completes the in-depth information mining of the smooth trend and fluctuating seasonal features, but more importantly, realizes the accurate capture of the local mutation seasonal features. Finally, on the premise of improving the forecasting accuracy, single-point load forecasting and quantile probabilistic load forecasting for ultra-short-term load forecasting are realized. Through the experiments conducted on three public datasets and one private dataset, the TFDNet model reduces the mean square error (MSE) and mean absolute error (MAE) by 19.80 and 11.20% on average, respectively, as compared with the benchmark method. These results indicate the potential applications of the TFDNet model.

Adaptive action-prediction cortical learning algorithm under uncertain environments

The cortical learning algorithm (CLA) is a time series prediction algorithm. Memory elements called columns and cells discretely represent data with their state combinations, whereas linking elements called synapses change their state combinations. For tasks requiring to take actions, the action-prediction CLA (ACLA) has an advantage to complement missing state values with their predictions. However, an increase in the number of missing state values (i) generates excess synapses negatively affect the action predictions and (ii) decreases the stability of data representation and makes the output of action values difficult. This paper proposes an adaptive ACLA using (i) adaptive synapse adjustment and (ii) adaptive action-separated decoding in an uncertain environment, missing multiple input state values probabilistically. (i) The proposed adaptive synapse adjustment suppresses unnecessary synapses. (ii) The proposed adaptive action-separated decoding adaptively outputs an action prediction separately for each action value. Experimental results using uncertain two- and three-dimensional mountain car tasks show that the proposed adaptive ACLA achieves a more robust action prediction performance than the conventional ACLA, DDPG, and the three LSTM-assisted reinforcement learning algorithms of DDPG, TD3, and SAC, even though the number of missing state values and their frequencies increase. These results implicate that the proposed adaptive ACLA is a way to making decisions for the future, even in cases where information surrounding the situation partially lacked.

Prediction of ship trajectory based on deep learning

The rapid development of computer technology strongly promotes the study of maritime traffic safety. The application of artificial intelligence technology makes the ship’s trajectory prediction not limited to complex physical models, and improves the generality of ship’s track prediction. To address the problems of relatively poor prediction accuracy in existing ship trajectory prediction research and high coupling of algorithms during ship navigation while the prediction model needs to be improved, this paper combines grey forecast prediction model and long and short-term memory (LSTM) neural network to establish a neural network model for ship trajectory prediction. This model can effectively increase the number of messages in LSTM model, and reduce the loss value of the model and improve the accuracy of ship track prediction. The prediction method in this paper is verified by AIS data of a ship. The results show that compared with other time series prediction algorithms, this algorithm has higher prediction accuracy.

Deviatoric stress-strain curve construction with strain-softening account via the damage-modified Duncan-Chang, arctangent and informer models: a comparative analysis

Engineering construction in cold regions cannot be separated from permafrost research. This study aimed to determine the mechanical properties and changing laws of artificially frozen clay through triaxial tests. Two models have been established: a physical model based on the tradi-tional phenomenological constitutive theory and a deep learning model based on the data-driven constitutive theory, taking into account the softening phenomenon. The accuracy and applica-bility of the models were verified, followed by a comparative analysis. The results of the analysis are as follows. The Duncan-Chang model can describe the characteristics of the hardening-type deviatoric stress-strain curve, but it cannot describe the characteristics of the softening-type de-viatoric stress-strain curve. The Modified Duncan-Chang (MDC) model fails to accurately de-scribe the characteristics of a smooth deviatoric stress-strain curve. The Strain-Damage Modified Duncan-Chang (SD-MDC) model exhibits a good fit in both the ascending and descending seg-ments of the curve, but it lacks effectiveness in the convergence segment of the S-shaped sof-tening curve. For this reason, this paper has chosen the arctangent function to establish a Strain-Damage Modified arctangent constitutive model (SD-MAM). This model accurately re-flects the stress evolution process of different types of frozen soils. Additionally, the Informer time series prediction algorithm was utilized to develop the Informer permafrost deviatoric stress prediction model which achieved an R2 value above 99%. In comparison to the SD-MAM model, the Informer model demonstrates higher precision, does not rely on assumptions, is cost-effective, and has a wide range of applications. However, it lacks physical meaning, and interpretability, and requires further discussion regarding the reliability of the results. This study offers valuable insights into the development and application of constitutive models for frozen soils.

Short-term forecasting model for residential indoor temperature in DHS based on sequence generative adversarial network

With the rapid development of the economy and the continuous improvement of people's living conditions, building thermal comfort has become one of the essential objectives of the development of the smart district heating system (SDHS). The accurate prediction approach of indoor temperature is the primary prerequisite and basis for achieving optimal thermal comfort regulation. However, the buildings' indoor temperature has significant thermal inertia and nonlinear characteristics due to the influence of multiple factors. The traditional time-series prediction algorithm can hardly accurately extract the indoor temperature variation pattern and cannot fully meet the satisfactory regulation requirements of SDHS. Therefore, an indoor temperature prediction model based on a sequence generative adversarial network (SGAN) is proposed in this paper. The new SGAN algorithm is trained by iterative adversarial training of the generator and discriminator, and the LSTM model built into the generator can effectively extract the high-level nonlinear abstract features of indoor temperature to achieve its accurate prediction. The detailed comparative experimental results show that the proposed indoor forecasting algorithm based on SGAN has obvious performance advantages compared to state-of-the-art algorithms, such as random forest regression (RFR), gradient boosting regression (GBR), support vector regression (SVR), adaptive boost (AdaBoost), multilayer perception (MLP), and long-short term memory(LSTM). The SGAN's mean absolute percentage error (MAPE) index reaches 2.3%.

Analysis of the Abnormality of Traction Energy Consumption in Urban Rail Transit System

Traction energy consumption (TEC) is a critical part of the total energy consumption in urban rail transit (URT) systems. Energy consumption patterns and abnormal analysis of TEC guarantee the energy-saving URT operation. With the rapid development of urbanization, the current energy consumption is becoming more and more prominent with some inherent drawbacks, such as complex original data, complicated statistical analysis, and abnormal energy consumption. This paper proposes a method for time accrual abnormal analysis of TEC. The system architecture of TEC typical values is presented, composed of three elements: research object, evaluation index, and time scale. The time series prediction algorithm calculates the typical values of the cumulative energy consumption index in each energy consumption mode. For the abnormality in TEC mode, the distance of the string vector is used as the similarity measure. Then, the similarity-based anomaly analysis method is used to judge the pattern abnormality. By comparing the advantages and disadvantages of engineering practice and theoretical research methods, we analyze the applicability of traditional anomaly detection algorithms to perform anomaly analysis of TEC in URT systems. The adopted time accrual abnormal analysis achieves a high fault detection rate, outperforming other models.

An active learning-based incremental deep-broad learning algorithm for unbalanced time series prediction

Time series are a kind of streaming data, which are chaotic and sequential. As real-world time series data are often not available at once and drift with time growth, Incremental Learning (IL) is well suited for Time Series Prediction (TSP). Most previous incremental TSP algorithms are limited by the assumption of data balance. However, real-world time series data are often unbalanced, with long-tailed distribution and other characteristics resulting in the failure of IL algorithms. In this paper, a balanced-driven Active Learning (AL) strategy is proposed to deal with data imbalance problems in IL processes. What’s more, by integrating the advantages of Deep Learning (DL) and the Broad Learning System (BLS), a novel Deep-Broad Learning (DeepBL) network with its incremental learning algorithm is proposed. The proposed Active Learning-based Incremental Deep-Broad Learning (AI_DeepBL) algorithm is applied to real-world univariate and multivariate time series datasets and achieves superior performance compared with classical and state-of-the-art TSP algorithms.

A novel conformable fractional nonlinear grey multivariable prediction model with marine predator algorithm for time series prediction

To promote the development of multivariate prediction modelling in small sample environments, this paper constructs a new multivariate prediction model named CFDNGBM(r,N) by integrating the marine predator algorithm (MPA), the modified conformable fractional-order accumulation operation (MCFAO) and the grey prediction model. In CFDNGBM(r,N), the MCFAO and time-delay polynomial are used to enhance the model prediction performance, the backward difference operation is used to activate the unbiasedness, and the unbiased regularization algorithm is used to estimate the model parameters. In addition, the MPA is used to optimize the hyperparameters in the model. The experimental results show that the CFDNGBM(r,N) outperforms the 12 benchmark algorithms and all the optimization measures are effective, both of which confirm the effectiveness of the proposed methods.

Time Series Prediction Method of Industrial Process With Limited Data Based on Transfer Learning

Industrial time series, as a kind of data that responds to production process information, can be analyzed and predicted for effective monitoring of industrial production processes. There are problems of data shortage and algorithm cold start in industrial modeling process caused by complex working conditions, change of data acquisition environment and short running time of equipment. As a result, the accuracy of the existing data-driven industrial time series prediction algorithm is greatly limited. To address the above problems, we propose a new time series prediction method for industrial processes under limited data based on dynamic transfer learning in this work. This method aims to effectively use historical data of similar equipment or working conditions rather than discard them to help establish an industrial time series prediction model with limited target data. In this method, first, historical data are divided into multiple batches, and then a new multi-source transfer learning framework with dynamic maximum mean difference (MMD) loss is established according to the distribution distance between each batch of historical data and the limited target data at the current moment. The framework also combines multi-task learning methods to establish multi-step prediction model for online learning in industrial processes. Compared with other commonly used methods, experiments on two real-world datasets of solar power generation prediction and heating furnace temperature prediction demonstrate the effectiveness of the proposed method.

STP-Net: A Signal Prediction Method of Laminar Decompression Robot Based on Swin Transformer

Spinal surgery robots have a great application value in laminar decompression surgery. For a safe surgery, the robot needs to accurately identify the cutting state of the lamina. Therefore, it is very important to deal with various sensing signals in the form of a time series. However, various state recognition algorithms proposed so far cannot completely avoid cutting through the lamina, leaving hidden dangers for nerve thermal damage caused by high-temperature liquid splashing. We propose a long time series prediction algorithm called STP-Net, which combined with the existing algorithms for recognizing the lamina cutting state can stop the lamina cutting in advance, thus preserving a thin inner layer of cortical bone and blocking high-temperature liquid splashing. STP-Net has the following advantages: (1) Dimension reduction is first performed on Swin (shifted window) Transformer to obtain 1D Swin Transformer, which was used to develop STP-Net for long time series prediction, which not only has high accuracy but also shows a linear relationship between computational complexity and input sequence length L, namely, O L . (2) The token merging layer is proposed and applied to the encoder of STP-Net, which reduces the computation cost and improves the global information extraction capability of the algorithm. (3) STP-Net uses a generative decoder to output the prediction sequence directly through a single operation, which not only has high efficiency but also avoids the accumulation of errors. Taking force signals as an example, when 400 numbers are used to predict 100 numbers, the average mean square error (MSE) and mean absolute error (MAE) of standard STP-Net are 1.22 × 10−3 and 2.42 × 10−2, respectively. STP-Net, combined with the existing method for recognizing the lamina cutting status, can stop the robot from cutting in advance in 87.14% of the cases. In addition, the results of practical lamina cutting experiments confirmed the effectiveness of STP-Net.

AE-DIL: A double incremental learning algorithm for non-stationary time series prediction via adaptive ensemble

Many dynamic processes in the real world can be modeled as time series, so time series prediction is significant for social and economic development. The inherent non-stationarity of time series obtained from actual projects may make it difficult to predict accurately. To alleviate this problem, in this paper, a Double Incremental Learning algorithm via Adaptive Ensemble, termed as AE-DIL for short, is proposed for non-stationary time series prediction. AE-DIL provides a general online prediction framework consisting of two modules. The first detects changes based on the statistical hypothesis test and self-adaptive sliding window technology. The second updates the prediction model based on double incremental learning and adaptive ensemble learning. The effectiveness of the proposed algorithm is empirically underpinned by the experiments conducted on seven benchmark time series datasets, compared with several baselines and state-of-the-art models.

Research on Mathematical Modeling Competition and Ideological and Political Education Based on Analysis and Prediction Algorithm

This paper selects the research of mathematical modeling competition and correlation analysis of ideological and political education based on analysis and prediction algorithm, including corresponding analysis algorithm, time series prediction algorithm, bp neural network prediction algorithm, and grey association analysis algorithm. First of all, when the corresponding analysis method is used, based on the extraction of a large number of ideological and political documents, the indicators of the past three years were obtained, and the four indicators of "teacher system, reform and innovation, cultural integration, and student classroom" were clustered to obtain the status quo and development trend of the four indicators. For the prediction of the future ideological and political education, we used the time-series algorithm and the BP neural network algorithm to derive the number of published articles and the development trends in the ideological and political field in the next two years.Using grey correlation analysis, with "mathematical modeling, politics" as the theme, to all relevant literature retrieval, extract representative 100 articles, to all the factors involved in the article clustering, red elements, traditional culture, classroom atmosphere, personal value, national situation five related factors, analyze the influence of grey correlation, get the results from high to low order: red elements 0.611,0.596, traditional culture, classroom atmosphere 0.594, personal value of 0.593, national situation 0.581, are strong influence factors, so as to prove the rationality of the model.

Tax-Scheduler: An interactive visualization system for staff shifting and scheduling at tax authorities

Given a large number of applications and complex processing procedures, how to efficiently shift and schedule tax officers to provide good services to taxpayers is now receiving more attention from tax authorities. The availability of historical application data makes it possible for tax managers to shift and schedule staff with data support, but it is unclear how to properly leverage the historical data. To investigate the problem, this study adopts a user-centered design approach. We first collect user requirements by conducting interviews with tax managers and characterize their requirements of shifting and scheduling into time series prediction and resource scheduling problems. Then, we propose Tax-Scheduler, an interactive visualization system with a time-series prediction algorithm and genetic algorithm to support staff shifting and scheduling in the tax scenarios. To evaluate the effectiveness of the system and understand how non-technical tax managers react to the system with advanced algorithms and visualizations, we conduct user interviews with tax managers and distill several implications for future system design.