Prediction Error Of Model Research Articles

A Regularized Regression Thermal Error Modeling Method for CNC Machine Tools under Different Ambient Temperatures and Spindle Speeds

Establishing a mathematical model to predict and compensate for the thermal error of CNC machine tools is a commonly used approach. Most existing methods, especially those based on deep learning algorithms, have complicated models that need huge amounts of training data and lack interpretability. Therefore, this paper proposes a regularized regression algorithm for thermal error modeling, which has a simple structure that can be easily implemented in practice and has good interpretability. In addition, automatic temperature-sensitive variable selection is realized. Specifically, the least absolute regression method combined with two regularization techniques is used to establish the thermal error prediction model. The prediction effects are compared with state-of-the-art algorithms, including deep-learning-based algorithms. Comparison of the results shows that the proposed method has the best prediction accuracy and robustness. Finally, compensation experiments with the established model are conducted and prove the effectiveness of the proposed modeling method.

Synthesis, Characterization, and Evaluation of Co-MOF Based ZIF-67 for CO2 Corrosion Inhibition of X65 Steel: Insights from Electrochemical Studies and a Machine Learning Algorithm

Co-MOF based metal organic framework was synthesized by reacting a metal ion (cobalt nitrate hexahydrate) with an organic ligand (2-methylimidazole) via a wet chemical method. The resulting material was characterized using detailed analytical methods and further was used as a self-assembly corrosion inhibitor in sweet corrosive environment. The empirical data set via electrochemical studies was modeled using adaptive neuro fuzzy inference system (ANFIS). The observed results showed that Co-MOF could significantly impede the corrosion rate of X65 steel and protect it from CO2 corrosion. Increasing the concentration of Co-MOF in the test solution increased the inhibition efficiency up to 97% at 0.1 wt % Co-MOF with a mixed-type inhibition mechanism. In addition, the DFT/MD-simulation approach evidenced the adsorption disposition of Co-MOF in aqueous and gas phase which complement with the empirical findings. Also, the prognostic capability of the proposed algorithm based on the statistical parameters such as root-mean-square error (RMSE), chi square (χ2), model predictive error (MPE) and coefficient of determination (R2) were appraised. From the viewpoint of statistics, the explanatory model aligned credibly with the ANFIS algorithm. The overall findings confirmed a dense hybrid coating of the synthesized Co-MOF on X65 steel as responsible for the inhibition of the sweet corrosion.

Thermal Error Modeling of Numerical Control Machine Based on Beetle Antennae Search Back-propagation Neural Networks

Thermal errors are one key impact factor on the processing accuracy of numerical control machine. This study targeted at a certain vertical processing center presents a new algorithm for predictive modeling of thermal errors in numerical control machine. This algorithm is founded on back-propagation neural networks (BPNNs) and adopts beetle antennae search (BAS) to find the best weights and thresholds of BPNNs. It avoids the local minimization due to local extremums faced by traditional BPNNs. The intermingling rate and arithmetic computation efficiency of neural network algorithms are further improved. Then, a BAS-BP thermal error prediction model is built with the machine temperature changes and thermal errors as the input data. Compared with conventional BPNNs, the BPNN after particle swarm optimization suggests the convergence rate of BAS-BP is improved by 85%, the leftover mistakes between the genuine information and the anticipated information are under 1 um, and the overall prediction precision is above 90%. Thus, the new model has high precision, high anti-disturbance ability and strong robustness.

Thermal error prediction model of high-speed motorized spindle based on DELM network optimized by weighted mean of vectors algorithm

The thermal error compensation process of high-speed motorized spindle needs to establish an effective prediction model. In this study, firstly, the thermal characteristics of a motorized spindle are tested, and the data at different speeds are obtained. Considering the disadvantage that the number of clusters in traditional clustering methods needs to be specified artificially, particle clustering and grey correlation analysis are selected to optimize the temperature measuring points. Aiming at the influence of data noise and the slow training speed, the depth limit learning machine is used to establish the thermal error prediction model. Considering that the output of deep extreme learning machine (DELM) is easily influenced by random input weights and biases, weighted mean of vectors algorithm (INFO) proposed in 2022 is chosen to optimize the two influencing factors. By establishing the INFO-DELM thermal error prediction model, and using the screened data to train the INDO-DELM model, the final prediction result is obtained. The results show that the accuracy of the INFO-DELM model is improved by 3%–26% compared with the basic DELM model. In addition, the INFO-DELM model shows excellent stability in both forward and backward forecasting, which proves that the model has strong adaptability. The optimization method of measuring points and INFO-DELM model in this paper provide a new idea for the field of thermal error compensation of motorized spindle.

Flight Delay Regression Prediction Model Based on Att-Conv-LSTM.

Accurate prediction results can provide an excellent reference value for the prevention of large-scale flight delays. Most of the currently available regression prediction algorithms use a single time series network to extract features, with less consideration of the spatial dimensional information contained in the data. Aiming at the above problem, a flight delay prediction method based on Att-Conv-LSTM is proposed. First, in order to fully extract both temporal and spatial information contained in the dataset, the long short-term memory network is used for getting time characteristics, and a convolutional neural network is adopted for obtaining spatial features. Then, the attention mechanism module is added to improve the iteration efficiency of the network. Experimental results show that the prediction error of the Conv-LSTM model is reduced by 11.41 percent compared with the single LSTM, and the prediction error of the Att-Conv-LSTM model is reduced by 10.83 percent compared with the Conv-LSTM. It is proven that considering spatio-temporal characteristics can obtain more accurate prediction results in the flight delay problem, and the attention mechanism module can also effectively improve the model performance.

Seasonal, Latitudinal, and Longitudinal Trends in Nighttime Ozone Vertical Structure on Mars From MAVEN/IUVS Stellar Occultations

AbstractStellar occultation measurements conducted by the Mars Atmosphere and Volatile and EvolutioN/Imaging UltraViolet Spectrograph instrument were able to make vertically resolved measurements of ozone density in the middle atmosphere of Mars that offered good coverage with respect to latitude, longitude, and local time. These measurements were used to identify systematic variations in the vertical structure of ozone with longitude that could be distinguished from general trends in the evolution of ozone with respect to season and latitude. A total of 583 individual nightside occultations between Martian years 32 and 36 were analyzed, of which 224 were confirmed to have ozone, all found between Ls = 15°–165°. Close to aphelion (Ls = 60°–90°), peak ozone densities between 30 and 40 km altitude were observed to be within error of model predictions at all measured latitudes, but diverged from model predictions before and after this time. At low latitudes, seasonal changes were seen to have the greatest effect on the observed vertical structure of ozone, with detached ozone layer densities at altitudes above 30 km usually varying within approximately a factor of two along a given latitudinal band at a given time of year. Nonetheless, evidence of a persistent regional enhancement of ozone abundance was observed over equatorial latitudes during the aphelion season, spanning a longitude range of approximately 50°–130°E longitude. Planetary waves were clearly observed at higher Southern latitudes during Southern winter, often resulting in order of magnitude variations in ozone density with longitude.

Hyperparameter optimization of tapping center machines model using robust whale optimization algorithm

To build a synchronization error prediction model for the machine tool efficiently, a robust whale optimization algorithm (RWOA) method proposed in this study is applied to the hyperparameter optimization of its model. The proposed RWOA method integrated a non-linear time-invariant inertia weighting (NTIW) method and a Taguchi-based adaptive parameter exploration (ATPE) to improve the performance of WOA and promote robustness. The NTIW method can improve the performance of other algorithms, so this study used the NTIW method in the WOA. In addition, the Taguchi method can get an excellent combination of variables with optimal values and stable performance, making the WOA robust. First, to verify the validity of the proposed RWOA method, 13 benchmark functions were used in this study. The results of the benchmark function tests include the mean, standard deviation, and p-value of the t-distribution test. The results show that 11 of the 13 functions differ significantly. In other non-significant difference functions, the means and standard deviations obtained by the proposed RWOA are considerably better than those obtained by WOA. Since the product cost of machine tools is higher, if a prediction model can be built effectively, it can reduce the cost. Therefore, in this study, the proposed RWOA was used to explore the best hyperparameter combination for the model. From the results, the model’s average MAPE (mean absolute percentage error) was 7.2604% for training data and 9.2603% for the testing data under 30 modeling runs. For the best one in 30 models, the MAPE was 6.8384% for the training data and 6.7372% for the testing data. This model was also introduced into the actual machine in this study, and the experimental results showed the MAPE 6.3447%. The proposed RWOA method effectively explores a suitable synchronization error model for the tool machine.

Wheel Odometry with Deep Learning-Based Error Prediction Model for Vehicle Localization

Wheel odometry is a simple and low-cost localization technique that can be used for localization in GNSS-deprived environments; however, its measurement accuracy is affected by many factors, such as wheel slip, wear, and tire pressure changes, resulting in unpredictable and variable errors, which in turn affect positioning performance. To improve the localization performance of wheel odometry, this study developed a wheel odometry error prediction model based on a transformer neural network to learn the measurement uncertainty of wheel odometry and accurately predict the odometry error. Driving condition characteristics including features describing road types, road conditions, and vehicle driving operations were considered, and models both with and without driving condition characteristics were compared and analyzed. Tests were performed on a public dataset and an experimental vehicle. The experimental results demonstrate that the proposed model can predict the odometry error with higher accuracy, stability, and reliability than the LSTM and WhONet models under multiple challenging and longer GNSS outage driving conditions. At the same time, the transformer model’s overall performance can be improved in longer GNSS outage driving conditions by considering the driving condition characteristics. Tests on the experimental vehicle demonstrate the model’s generalization capability and the improved positioning performance of dead reckoning when using the proposed model. This study explored the possibility of applying a transformer model to wheel odometry and provides a new solution for using deep learning in localization.

A variational Bayesian inference technique for model updating of structural systems with unknown noise statistics

Dynamic models of structural and mechanical systems can be updated to match the measured data through a Bayesian inference process. However, the performance of classical (non-adaptive) Bayesian model updating approaches decreases significantly when the pre-assumed statistical characteristics of the model prediction error are violated. To overcome this issue, this paper presents an adaptive recursive variational Bayesian approach to estimate the statistical characteristics of the prediction error jointly with the unknown model parameters. This approach improves the accuracy and robustness of model updating by including the estimation of model prediction error. The performance of this approach is demonstrated using numerically simulated data obtained from a structural frame with material non-linearity under earthquake excitation. Results show that in the presence of non-stationary noise/error, the non-adaptive approach fails to estimate unknown model parameters, whereas the proposed approach can accurately estimate them.

Remote Sensing Inversion of Typical Offshore Water Quality Parameter Concentration Based on Improved SVR Algorithm

Chlorophyll a concentration and suspended matter concentration, as typical water quality parameters related to spectral characteristics, are essential for characterizing the degree of eutrophication in water bodies. They have become crucial indicators for water quality assessment of inland water bodies. The support vector regression model (SVR) is suitable for small samples, has excellent generalization ability, and has high prediction accuracy. Still, it has the problem of difficult selection of model parameters and quickly falling into local extremes. To solve this problem, a hybrid Differential Evolution-Grey Wolf Optimizer (DE-GWO) algorithm is introduced into the parameter selection process of the support vector regression model, and an improved SVR algorithm (DE-GWO-SVR) is proposed for the remote sensing inversion of chlorophyll a concentration and suspended sediment concentration in water bodies. In this paper, the spectral reflectance of the water surface and the chlorophyll a and broken matter concentration values were obtained by field measurements in the Tangdao Bay waters of Qingdao, Shandong Province. The inverse model between the concentration values of the two water quality parameters and the corresponding sensitive factors was established by first determining the sensitive factors based on the response of the spectral reflectance to the two water quality parameters and introducing the DE-GWO optimization algorithm into the parameter selection process of the SVR model. Finally, the accuracy of the model was verified using Sentinel II satellite remote sensing spectral data, and then the inverse accuracy of the two water quality parameters was obtained. The mean relative error (MRE) of the chlorophyll a prediction model built by the DE-GWO algorithm optimizing the SVR is 25.1%, and the mean relative error (MRE) of the suspended matter prediction model is 32.5%. The inversion results were all better than the other models (linear regression, SVR, and GWO-SVR model). When the best model, built from the measured water surface spectral data, was applied to the Sentinel II satellite data, the improved SVR model outperformed the other models in terms of mean relative error. The experimental results confirm that the DE-GWO-SVR algorithm is an effective method for remote sensing inversion of chlorophyll a and suspended matter concentrations in water bodies, which can provide a reference for remote sensing inversion of chlorophyll a and suspended matter concentrations in Chinese offshore waters and subsequent scientific management of waters.

Assessment of blend uniformity in a stream sampler device using Raman spectroscopy

This study describes the first implementation of Raman spectrometer in a stream sampler for the in-line monitoring of low drug concentration in poor flowability powder blends. Raman spectra were continuously acquired as the powder blends flowed through the stream sampler operating with a paddle wheel speed of 10 RPM and used to develop the calibration models. A calibration model was developed to quantify caffeine concentration from 1.50 to 4.50% w/w using Partial Least Squares Regression (PLS-R). Three test set blends were used to assess the prediction errors of the calibration model. Caffeine concentration was predicted for the test set blends with a root mean square error of prediction of 0.21% w/w and a low bias of −0.03% w/w. The calibration model showed good prediction performance with an estimated sample mass of 83 mg. Variographic analysis demonstrated the low process variance of the real-time spectral acquisition through minimum practical error and sill values. The results showed the ability of the Raman spectrometer coupled with the stream sampler to monitor low drug concentration for poor flowability blends.

A Method for Increasing the Robustness of Stable Feature Selection for Biomarker Discovery in Molecular Medicine Developed Using Serum Small Extracellular Vesicle Associated miRNAs and the Barrett's Oesophagus Disease Spectrum.

The biomarker development field within molecular medicine remains limited by the methods that are available for building predictive models. We developed an efficient method for conservatively estimating confidence intervals for the cross validation-derived prediction errors of biomarker models. This new method was investigated for its ability to improve the capacity of our previously developed method, StaVarSel, for selecting stable biomarkers. Compared with the standard cross validation method, StaVarSel markedly improved the estimated generalisable predictive capacity of serum miRNA biomarkers for the detection of disease states that are at increased risk of progressing to oesophageal adenocarcinoma. The incorporation of our new method for conservatively estimating confidence intervals into StaVarSel resulted in the selection of less complex models with increased stability and improved or similar predictive capacities. The methods developed in this study have the potential to improve progress from biomarker discovery to biomarker driven translational research.

Explainability-based Trust Algorithm for electricity price forecasting models

Advanced machine learning (ML) algorithms have outperformed traditional approaches in various forecasting applications, especially electricity price forecasting (EPF). However, the prediction accuracy of ML reduces substantially if the input data is not similar to the ones seen by the model during training. This is often observed in EPF problems when market dynamics change owing to a rise in fuel prices, an increase in renewable penetration, a change in operational policies, etc. While the dip in model accuracy for unseen data is a cause for concern, what is more, challenging is not knowing when the ML model would respond in such a manner. Such uncertainty makes the power market participants, like bidding agents and retailers, vulnerable to substantial financial loss caused by the prediction errors of EPF models. Therefore, it becomes essential to identify whether or not the model prediction at a given instance is trustworthy. In this light, this paper proposes a trust algorithm for EPF users based on explainable artificial intelligence techniques. The suggested algorithm generates trust scores that reflect the model’s prediction quality for each new input. These scores are formulated in two stages: in the first stage, the coarse version of the score is formed using correlations of local and global explanations, and in the second stage, the score is fine-tuned further by the Shapley additive explanations values of different features. Such score-based explanations are more straightforward than feature-based visual explanations for EPF users like asset managers and traders. A dataset from Italy’s and ERCOT’s electricity market validates the efficacy of the proposed algorithm. Results show that the algorithm has more than 85% accuracy in identifying good predictions when the data distribution is similar to the training dataset. In the case of distribution shift, the algorithm shows the same accuracy level in identifying bad predictions.

Data Modeling of Sewage Treatment Plant Based on Long Short-Term Memory with Multilayer Perceptron Network

As wastewater treatment usually involves complicated biochemical reactions, leading to strong coupling correlation and nonlinearity in water quality parameters, it is difficult to analyze and optimize the control of the wastewater treatment plant (WWTP) with traditional mathematical models. This research focuses on how deep learning techniques can be used to model the data from a specific WWTP so as to optimize the required energy consumption. In the operation of a wastewater treatment plant, various sensors are used to record the treatment process data; these data are used to train deep neural networks (DNNs). A long short-term memory with multilayer perceptron network (LMPNet) model is proposed to model the water quality parameters and site control parameters, such as COD, pH, NH3-N, et al., and the LMPNet model prediction error is then measured by criteria such as the MSE, MAE, and R2. The experimental results show that the LMPNet model demonstrates great accuracy in the modeling of the control of WWTPs. A life-long learning strategy is also developed for the LMPNet in order to adapt to the environment that may change over time. By developing performance evaluation metrics, the purification performance can be analyzed, and the prediction reference can be provided for the subsequent control optimization and energy saving plan.

A Forecasting Approach for IoT-Based Energy and Power Quality Monitoring in Buildings

This paper proposes a novel energy consumption and power quality forecasting technique based on the combination of poly-exponential (PE) and random forest (RF). This technique paves away seamlessly integrating the forecast to the Internet-of-things (IoT)-based monitoring system due to its capability of improving computational efficiency. For the experimental setup, the monitoring system in a building supplied the data for forecasting consisting of daily energy consumption and power quality parameters. The developed algorithm for efficient computational forecasting has already considered the space limitation in online real-time monitoring with a minor effect on the model prediction error (less than 0.3% in our experimental setup). The algorithm can alert the user once an indication of energy waste (more than the determined boundary) or the actual voltage violated the tolerance might occur. An experimental dataset taken from the Bappeda Pontianak building in West Kalimantan, Indonesia, shows that the proposed algorithm required fewer samples than the RF yet could predict the daily energy consumption and voltage inside a building accurately up to 96%. Note to Practitioners—As the energy and power quality monitoring using IoT is applied for buildings nowadays, how to handle the big data and forecast the energy and power for evaluation and strategy in buildings is significant. A standard solution is to apply sophisticated technology with a high computational resource. However, this will increase the cost either in the investment or maintenance. The algorithm in this paper is a new approach for IoT-based monitoring and forecasting of energy consumption and power quality by combining PE and RT models, which is relatively cheap in computational complexity yet does not require much space to store in an online setup. The proposed algorithm first creates the model with the RT algorithm and then tracks the residual generated by this algorithm with the PE model and extended Kalman filter. We then show that this model can predict the daily energy consumption accurately. The algorithm based on this model can be integrated into the monitoring system and advise the user when the current energy consumption has passed the “normal” energy consumption concerning the information from previous days. Experiments on a building located in Pontianak show that this algorithm is applicable and can precisely predict daily energy consumption and power quality with a limited amount of data. In future research, we plan to expand this technique further, such that it is not only for monitoring purposes but also for control.

Health Status Prediction of Lithium Battery Based on LSTM Model with Optimization Algorithms

Consider the current status of health (SOH) of lithium batteries, which presents challenging existing issues of accurately predicting and calculating. In this paper, an LSTM model and multi-optimization algorithm were used to estimate the battery health state. Taking advantage of the fast convergence speed and wide global optimization range of the optimization algorithm, optimized the number of layers and neurons in the LSTM model so the LSTM model was established, used to predict the health status of lithium batteries, and compared with the LSTM model prediction method without optimization. The results showed that the error of the battery health prediction model based on the proposed prediction model was less than 3%, the prediction accuracy was higher than the LSTM model without optimization, and the model had better accuracy and stability.

A Modeling Method for Thermal Error Prediction of CNC Machine Equipment Based on Sparrow Search Algorithm and Long Short-Term Memory Neural Network.

To better solve the problem of thermal error of computerized numerical control machining equipment (CNCME), a thermal error prediction model based on the sparrow search algorithm and long short-term memory neural network (SSA-LSTMNN) is proposed. Firstly, the Fuzzy C-means clustering algorithm (FCMCA) is used to screen the key temperature-sensitive points of the CNCME. Secondly, by taking the temperature rise data of key temperature-sensitive points as input and the corresponding time thermal error data as output, we established the SSA-LSTMNN thermal error prediction model. The SSA is used to optimize the parameters of LSTMNN and make its performance play the best. Taking the VMC1060 vertical machining center as the research object, we carried out the experiment. Finally, the prediction effect of the proposed model is compared with the article swarm optimized algorithm and LSTM neural network (PSOA-LSTMNN), the LSTMNN, and the traditional recurrent neural network (TRNN) model. The results show that the average values of the predicted residual fluctuations of the SSA-LSTMNN model are all more than 44% lower than those of the other three models under different operating conditions, which has a strong practicality.

Statistical analysis of the impacts of COVID-19 pandemic on the small and large-scale tourism sectors in developing countries.

The worldwide COVID-19 pandemic has affected the tourism sector by closing borders, reducing both the transportation of tourists and tourist demand. Due to the country-wide lockdown, most activities in the hotel, motel, restaurant, and transportation sectors have been postponed. Consequently, the article investigates four research issues by examining the consequences of global tourism in the private sector before and after COVID-19. As an analytical method, the article suggested qualitative research methodologies to collect information from tourism employees. The opinions of the respondents were gathered through online emails in the questionnaire survey. Further, the article considers people's future desire for specific tourism destinations based on visitor arrivals. Forecasting tourist demand is an essential component of good and efficient tourism management. Consequently, the article proposes an attention-based long short-term memory model for exact demand forecasting. The experimental findings reveal that the model's minimal prediction error accuracy is 0.45%, which indicates that it has a more robust prediction effect, a faster convergence rate, and a greater prediction accuracy. Seasonality has emerged as one of the most distinguishing and defining characteristics of the global tourist business. Accordingly, the article mandated to compare the seasonal and non-seasonal effects of the tourist sector throughout the years 2020-2021. Moreover, Governments must analyse the crises' long-term consequences and, as a result, define the components that constitute government advantages supplied to the tourist sector during the pandemic era. As a result, many governmental policies, especially those about social welfare, may perceive a fresh start during the post-pandemic period, respectively.

An improved model for building energy consumption prediction based on time-series analysis

The development and popularisation of renewable energy is necessary. The application of renewable energy technology in buildings is an important research direction. Moreover, the prediction of renewable energy consumption in this direction is an essential research content. In view of this, a building energy consumption prediction model of renewable energy based on time-series analysis and a support vector machine (SVM) is proposed. The performance test of this model showed that its loss value was as low as 1.5% in the training set, and the loss value was 4.1% in the test set. In addition, it showed the highest accuracy rate of 95.5% in the neural network accuracy test, which is significantly higher than that of traditional algorithms. About the overall energy consumption prediction ability of the model, the experimental results showed that the lowest error of the energy consumption prediction model was 2.3%, the average relative error of the traditional SVM model in the same data set was 6.8% and that of the chaotic time-series model was 4.1%. Compared with the prediction ability of the traditional models currently used, the prediction ability of the energy consumption prediction model has been greatly improved, and it has the potential to be put into practical application.

A multi-task and multi-channel convolutional neural network for semi-supervised neonatal artefact detection

Objective. Automated artefact detection in the neonatal electroencephalogram (EEG) is crucial for reliable automated EEG analysis, but limited availability of expert artefact annotations challenges the development of deep learning models for artefact detection. This paper proposes a semi-supervised deep learning approach for artefact detection in neonatal EEG that requires few labelled data by training a multi-task convolutional neural network (CNN). Approach. An unsupervised and a supervised objective were jointly optimised by combining an autoencoder and an artefact classifier in one multi-output model that processes multi-channel EEG inputs. The proposed semi-supervised multi-task training strategy was compared to a classical supervised strategy and other existing state-of-the-art models. The models were trained and tested separately on two different datasets, which contained partially annotated multi-channel neonatal EEG. Models were evaluated using the F1-statistic and the relevance of the method was investigated in the context of a functional brain age (FBA) prediction model. Main results. The proposed multi-task and multi-channel CNN methods outperformed state-of-the-art methods, reaching F1 scores of 86.2% and 95.7% on two separate datasets. The proposed semi-supervised multi-task training strategy was shown to be superior to a classical supervised training strategy when the amount of labels in the dataset was artificially reduced. Finally, we found that the error of a brain age prediction model correlated with the amount of automatically detected artefacts in the EEG segment. Significance. Our results show that the proposed semi-supervised multi-task training strategy can train CNNs successfully even when the amount of labels in the dataset is limited. Therefore, this method is a promising semi-supervised technique for developing deep learning models with scarcely labelled data. Moreover, a correlation between the error of FBA estimates and the amount of detected artefacts in the corresponding EEG segments indicates the relevance of artefact detection for robust automated EEG analysis.