Error Prediction Research Articles

Traveling control method adapted to different paddy ground conditions with feedforward compensation for crawler combine harvester based on online tracking error prediction

Enhancing the adaptability of agricultural machinery to different ground conditions is paramount for advancing path-tracking accuracy and control stability. Precise online perception of the steering motion state of a crawler combine harvester is essential for adapting to different paddy ground conditions (PGCs) for path-tracking control. In this study, a novel pose vector method (PVM) was proposed for the online determination of the steering radius of a crawler combine harvester. A support vector regression (SVR) model was used to forecast the error-correction factor for the steering trajectory arc radius, and a feedforward compensation control method was designed based on the steering control model. Field steering experiments revealed that the PVM outperformed the conventional least squares method (LSM) in calculating the trajectory arc radius, particularly when dealing with sparse pose data. The relative error of the PVM was several to over 10 times lower than that of the LSM fitting. The field experiments of straight-path tracking showed that the standard deviations of the lateral deviation were 0.078 m, 0.059 m, and 0.062 m, and the standard deviations of the heading deviation were 2.622°, 1.580°, and 1.593°, respectively, for the three different PGCs of soil compaction from low to high. Compared to the conventional pure tracking algorithm, the proposed PGCs adaptive tracking control algorithm enhances the adaptability of the crawler combine harvester to PGCs. These results offer a novel method for enhancing the online perception of the motion state of a crawler combine harvester and improving the adaptive control performance under different PGCs.

OntoFact: Unveiling Fantastic Fact-Skeleton of LLMs via Ontology-Driven Reinforcement Learning

Large language models (LLMs) have demonstrated impressive proficiency in information retrieval, while they are prone to generating incorrect responses that conflict with reality, a phenomenon known as intrinsic hallucination. The critical challenge lies in the unclear and unreliable fact distribution within LLMs trained on vast amounts of data. The prevalent approach frames the factual detection task as a question-answering paradigm, where the LLMs are asked about factual knowledge and examined for correctness. However, existing studies primarily focused on deriving test cases only from several specific domains, such as movies and sports, limiting the comprehensive observation of missing knowledge and the analysis of unexpected hallucinations. To address this issue, we propose OntoFact, an adaptive framework for detecting unknown facts of LLMs, devoted to mining the ontology-level skeleton of the missing knowledge. Specifically, we argue that LLMs could expose the ontology-based similarity among missing facts and introduce five representative knowledge graphs (KGs) as benchmarks. We further devise a sophisticated ontology-driven reinforcement learning (ORL) mechanism to produce error-prone test cases with specific entities and relations automatically. The ORL mechanism rewards the KGs for navigating toward a feasible direction for unveiling factual errors. Moreover, empirical efforts demonstrate that dominant LLMs are biased towards answering Yes rather than No, regardless of whether this knowledge is included. To mitigate the overconfidence of LLMs, we leverage a hallucination-free detection (HFD) strategy to tackle unfair comparisons between baselines, thereby boosting the result robustness. Experimental results on 5 datasets, using 32 representative LLMs, reveal a general lack of fact in current LLMs. Notably, ChatGPT exhibits fact error rates of 51.6% on DBpedia and 64.7% on YAGO, respectively. Additionally, the ORL mechanism demonstrates promising error prediction scores, with F1 scores ranging from 70% to 90% across most LLMs. Compared to the exhaustive testing, ORL achieves an average recall of 80% while reducing evaluation time by 35.29% to 63.12%.

A Hybrid-Model-Based CNC Machining Trajectory Error Prediction and Compensation Method

Intelligent manufacturing is the main direction of Industry 4.0, pointing towards the future development of manufacturing. The core component of intelligent manufacturing is the computer numerical control (CNC) system. Predicting and compensating for machining trajectory errors by controlling the CNC system’s accuracy is of great significance in enhancing the efficiency, quality, and flexibility of intelligent manufacturing. Traditional machining trajectory error prediction and compensation methods make it challenging to consider the uncertainties that occur during the machining process, and they cannot meet the requirements of intelligent manufacturing with respect to the complexity and accuracy of process parameter optimization. In this paper, we propose a hybrid-model-based machining trajectory error prediction and compensation method to address these issues. Firstly, a digital twin framework for the CNC system, based on a hybrid model, was constructed. The machining trajectory error prediction and compensation mechanisms were then analyzed, and an artificial intelligence (AI) algorithm was used to predict the machining trajectory error. This error was then compensated for via the adaptive compensation method. Finally, the feasibility and effectiveness of the method were verified through specific experiments, and a realization case for this digital-twin-driven machining trajectory error prediction and compensation method was provided.

Development and Validation of an Interpretable Conformal Predictor to Predict Sepsis Mortality Risk: Retrospective Cohort Study.

Early and reliable identification of patients with sepsis who are at high risk of mortality is important to improve clinical outcomes. However, 3 major barriers to artificial intelligence (AI) models, including the lack of interpretability, the difficulty in generalizability, and the risk of automation bias, hinder the widespread adoption of AI models for use in clinical practice. This study aimed to develop and validate (internally and externally) a conformal predictor of sepsis mortality risk in patients who are critically ill, leveraging AI-assisted prediction modeling. The proposed approach enables explaining the model output and assessing its confidence level. We retrospectively extracted data on adult patients with sepsis from a database collected in a teaching hospital at Beth Israel Deaconess Medical Center for model training and internal validation. A large multicenter critical care database from the Philips eICU Research Institute was used for external validation. A total of 103 clinical features were extracted from the first day after admission. We developed an AI model using gradient-boosting machines to predict the mortality risk of sepsis and used Mondrian conformal prediction to estimate the prediction uncertainty. The Shapley additive explanation method was used to explain the model. A total of 16,746 (80%) patients from Beth Israel Deaconess Medical Center were used to train the model. When tested on the internal validation population of 4187 (20%) patients, the model achieved an area under the receiver operating characteristic curve of 0.858 (95% CI 0.845-0.871), which was reduced to 0.800 (95% CI 0.789-0.811) when externally validated on 10,362 patients from the Philips eICU database. At a specified confidence level of 90% for the internal validation cohort the percentage of error predictions (n=438) out of all predictions (n=4187) was 10.5%, with 1229 (29.4%) predictions requiring clinician review. In contrast, the AI model without conformal prediction made 1449 (34.6%) errors. When externally validated, more predictions (n=4004, 38.6%) were flagged for clinician review due to interdatabase heterogeneity. Nevertheless, the model still produced significantly lower error rates compared to the point predictions by AI (n=1221, 11.8% vs n=4540, 43.8%). The most important predictors identified in this predictive model were Acute Physiology Score III, age, urine output, vasopressors, and pulmonary infection. Clinically relevant risk factors contributing to a single patient were also examined to show how the risk arose. By combining model explanation and conformal prediction, AI-based systems can be better translated into medical practice for clinical decision-making.

A comprehensive error evaluation method for non-rigid body gantry machine tools considering quasi-static crosstalk based on optimized multilateration measurement

For large-scale machine tools, the impact of non-rigid body behavior on volumetric error accuracy cannot be ignored, which hasn't been considered in general volumetric error model. This paper presents a comprehensive error evaluation method for large-scale gantry machine tools considering non-rigid body effect, aiming to provide guidance for volumetric error calibration process. An extended kinematic model is established based on the analysis of crosstalk influence on geometric error motions. An optimized multilateration measurement with improved accuracy and robustness is carried out for volumetric error identification, verified by Monte Carlo simulations. The main novelty of the optimization is to introduce a space uniformity index to generate specialized calibration point sets for any measurement points. A new volumetric error prediction method based on classified descriptions of geometric error motions is proposed, according to their relationship with crosstalk. Experimental results show that the proposed method achieves higher prediction accuracy in crosstalk-sensitive directions.

Disparities in Medical School Clerkship Grades Associated With Sex, Race, and Ethnicity: A Person-Centered Approach.

To determine whether students' self-reported race/ethnicity and sex were associated with grades earned in 7 core clerkships. A person-centered approach was used to group students based on observed clerkship grade patterns. Predictors of group membership and predictive bias by race/ethnicity and sex were investigated. Using data from 6 medical student cohorts at Johns Hopkins University School of Medicine (JHUSOM), latent class analysis was used to classify students based on clerkship grades. Multinomial logistic regression was employed to investigate if preclerkship measures and student demographic characteristics predicted clerkship performance-level groups. Marginal effects for United States Medical Licensing Exam (USMLE) Step 1 scores were obtained to assess the predictive validity of the test on group membership by race/ethnicity and sex. Predictive bias was examined by comparing multinomial logistic regression prediction errors across racial/ethnic groups. Three clerkship performance-level groups emerged from the data: low, middle, and high. Significant predictors of group membership were race/ethnicity, sex, and USMLE Step 1 scores. Black or African American students were more likely (odds ratio [OR] = 4.26) to be low performers than White students. Black or African American (OR = 0.08) and Asian students (OR = 0.41) were less likely to be high performers than White students. Female students (OR = 2.51) were more likely to be high performers than male students. Patterns of prediction errors observed across racial/ethnic groups showed predictive bias when using USMLE Step 1 scores to predict clerkship performance-level groups. Disparities in clerkship grades associated with race/ethnicity were found among JHUSOM students, which persisted after controlling for USMLE Step 1 scores, sex, and other preclerkship performance measures. Differential predictive validity of USMLE Step 1 exam scores and systematic error predictions by race/ethnicity show predictive bias when using USMLE Step 1 scores to predict clerkship performance across racial/ethnic groups.

Prediction of the Waviness Error in Ultra-Precision Fly Cutting Using the Direct Integration Method

Abstract Fly cutting is widely used in manufacturing of large-scale, high-precision optical components. However, the discontinuity of fly cutting machining leads to significant relative vibrations between the tool and the workpiece. The cutting process generates periodic waves along the cutting direction, which will deteriorate the wavefront characteristics of optical components. Based on the machining dynamics, this paper proposes a direct integration method to predict the waviness error of the machined surface. The cutting force model of fly cutting is established. The multi-mode characteristics of the spindle-tool system are measured by the experimental method. Then, the influence of uncertainties on the calculation results is analyzed by the variance-based sensitivity analysis method. Finally, the plane cutting experiment verifies that the direct integration method effectively predicts the waviness error and its variation trend, and the waviness prediction research is important for optimization of the machining parameters.

Automatic hip osteoarthritis grading with uncertainty estimation from computed tomography using digitally-reconstructed radiographs.

Progression of hip osteoarthritis (hip OA) leads to pain and disability, likely leading to surgical treatment such as hip arthroplasty at the terminal stage. The severity of hip OA is often classified using the Crowe and Kellgren-Lawrence (KL) classifications. However, as the classification is subjective, we aimed to develop an automated approach to classify the disease severity based on the two grades using digitally-reconstructed radiographs from CT images. Automatic grading of the hip OA severity was performed using deep learning-based models. The models were trained to predict the disease grade using two grading schemes, i.e., predicting the Crowe and KL grades separately, and predicting a new ordinal label combining both grades and representing the disease progression of hip OA. The models were trained in classification and regression settings. In addition, the model uncertainty was estimated and validated as a predictor of classification accuracy. The models were trained and validated on a database of 197 hip OA patients, and externally validated on 52 patients. The model accuracy was evaluated using exact class accuracy (ECA), one-neighbor class accuracy (ONCA), and balanced accuracy. The deep learning models produced a comparable accuracy of approximately 0.65 (ECA) and 0.95 (ONCA) in the classification and regression settings. The model uncertainty was significantly larger in cases with large classification errors ( ). In this study, an automatic approach for grading hip OA severity from CT images was developed. The models have shown comparable performance with high ONCA, which facilitates automated grading in large-scale CT databases and indicates the potential for further disease progression analysis. Classification accuracy was correlated with the model uncertainty, which would allow for the prediction of classification errors. The code will be made publicly available at https://github.com/NAIST-ICB/HipOA-Grading .

Adaptive thermal error prediction for CNC machine tool spindle using online measurement and an improved recursive least square algorithm

Establishing models for predicting and compensating for spindle thermal errors is cost-effective and necessary to improve the accuracy of machine tools for smart manufacturing. However, the prediction performance of existing methods deteriorates significantly with dynamic working conditions of machine tools because training from static conditions leads to the inability to adapt to dynamic conditions. Therefore, an adaptive thermal error modeling method using online measurement and an improved recursive least square algorithm is proposed to fill this research gap, which updates the thermal error model adaptively to ensure that dynamic working conditions are learned in real time. Particularly, Spearman's rank correlation coefficient method is first adopted for temperature-sensitive point selection to capture the nonlinear relationship between temperature and thermal error variables. Furthermore, a variable-forgetting factor-based recursive least square (VFF-RLS) algorithm is proposed to improve the prediction performance, in which the proposed variable forgetting factor is adaptively updated according to real-time thermal error data collected by online measurement. The experimental results showed that the proposed VFF-RLS method can maintain a high prediction accuracy of 1.75 μm and robustness of 0.16 μm on both constant and dynamic working conditions. The effectiveness of the VFF-RLS method is validated by verification experiments.

Measurement error prediction-based reliability assessment framework for electric metering devices under harsh natural environments

Reliability assessment for electric metering devices (EMDs), which includes environmental stress analysis, measurement error (ME) prediction, and reliability estimation, can be utilized for predictive instrument maintenance, especially under harsh environmental stresses. Nevertheless, the actual evaluation process is limited by data noise and the unavailability of future environmental inputs for the degradation model. To this end, we first extract the main environmental factors affecting ME and then fuse them using the weighted principal component analysis (WPCA) method to provide future environmental input values for the prediction model. Next, a bi-directional (BD) outlier detection method based on MCD robust analysis and the Thompson tau method is proposed to detect outliers from horizontal and longitudinal perspectives. Then, an ME prediction method called multi-stress fusion nonlinear degradation (MFND) is put forward, which considers the cyclic variation of ME over time and the effect of environmental stresses on ME. Real-world datasets from a high-dry-hot area manifest that our proposed BD-based MFND reliability assessment framework enjoys pleasurable prediction performance. Compared to some well- known methods, our framework excels in outlier detection and ME prediction.

Construction of a sustainable model to predict the moisture content of porang powder (Amorphophallus oncophyllus) based on pointed-scan visible near-infrared spectroscopy

Abstract The moisture content of porang powder (PP) is an inherent quality parameter. Therefore, several analytical methods, such as oven drying and Karl–Fischer titration, were applied to determine the content. However, these techniques are noted to have various disadvantages, such as being time-consuming, requiring sample preparation, being labor-intensive, and producing chemical waste. This study aims to investigate the potential of visible near-infrared (Vis-NIR) spectroscopy as a nondestructive and sustainable analytical technology to predict moisture content in PP. In this study, we developed a traditional machine learning algorithm, a partial least squares regression (PLSR), in tandem with two spectral bands, which are Vis-NIR (400–1,000 nm) and NIR (954–1,700 nm). To upgrade the performance of PLSR, we applied seven preprocessing techniques: mean normalization, maximum normalization, range normalization, multiplicative scatter correction, standard normal variate (SNV), and Savitzky–Golay first and second derivatives. We found that PLSR using NIR spectral bands was more effective; the preprocessed mean normalization exhibited the best results with a coefficient of determination ( R p 2 ) \left({R}_{p}^{2}) of 0.96 and a standard error prediction (SEP) of 0.56 using five latent variables. Furthermore, we also extracted 39 optimum wavelengths using variable importance in projection and achieved better performance ( R p 2 {R}_{p}^{2} = 0.95, SEP = 0.56%wb, and 5 LVs) via SNV preprocessed NIR spectra.

Temporally organized representations of reward and risk in the human brain

The value and uncertainty associated with choice alternatives constitute critical features relevant for decisions. However, the manner in which reward and risk representations are temporally organized in the brain remains elusive. Here we leverage the spatiotemporal precision of intracranial electroencephalography, along with a simple card game designed to elicit the unfolding computation of a set of reward and risk variables, to uncover this temporal organization. Reward outcome representations across wide-spread regions follow a sequential order along the anteroposterior axis of the brain. In contrast, expected value can be decoded from multiple regions at the same time, and error signals in both reward and risk domains reflect a mixture of sequential and parallel encoding. We further highlight the role of the anterior insula in generalizing between reward prediction error and risk prediction error codes. Together our results emphasize the importance of neural dynamics for understanding value-based decisions under uncertainty.

Robot joint space grid error compensation based on three-dimensional discrete point space circular fitting

The poor absolute positioning accuracy of industrial robots has limited their application in fields such as aerospace manufacturing. To address this issue, the spatial grid compensation method has been proposed as an effective solution. In this paper, we propose a sampling method based on three-dimensional discrete point space circular fitting for grid points to significantly reduce the sampling workload and improve compensation accuracy compared to traditional joint space grid compensation methods. Additionally, we use the Kriging interpolation algorithm instead of the inverse distance weighting (IDW) algorithm for spatial interpolation prediction of pose error. Based on this, the proposed sampling and interpolation prediction method in this paper was verified on a Comau NJ500–2.7 manipulator equipped with a fiber-laying end effector. The experimental results demonstrate that using our proposed sampling method yields pose data of grid points that have only a small deviation from directly sampled results and are only slightly higher than the robot's repeat positioning accuracy. Moreover, our proposed method can significantly reduce the sampling workload by 60% under the experimental conditions of this study (sampling 10 groups of data within 90 degrees range), with increasing sampling range leading to more obvious efficiency improvements. Finally, we show that compared to the IDW algorithm, the Kriging interpolation algorithm yields better results and improves the mean absolute positioning accuracy of robots after compensation by 30%.

The measurement error prediction method of voltage transformer based on TCN-RBF

Accurate prediction of voltage transformer error is an important guarantee for the stable and economic operation of the power grid. This paper presents a measurement error prediction method for voltage transformers based on time convolution and radial basis neural networks (TCN-RBF). Firstly, the ensemble empirical mode decomposition is used to decompose the original error sequence into multiple components. Secondly, a TCN prediction model was established based on each component and historical error sequence to achieve error prediction. To further improve the prediction progress, the RBF model was used to further fit the predicted results with the environmental parameter data and confirm the final predicted value. The operation error of the transformer in a substation is tested. The test results show that compared with LSTM, SVM, BP, and other prediction models, this method has better performance in error prediction.

CVT error prediction method based on improved attention mechanism transformer model

In order to enhance the performance of capacitive voltage transformer (CVT) error prediction, we propose a transformer model (AM-transformer) error prediction method based on an improved attention mechanism. The method starts from both input data and model improvement, adopting Pearson’s correlation coefficient to calculate the environmental factors that have a greater impact on the CVT measurement error, and enhancing the ability of local learning of the attention mechanism by adding causal convolution and data segmentation operations. The experimental results show that the mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) of the method reach 0.016, 4.21% and 0.021, respectively, realizing a high prediction accuracy.

Temperature error compensation method of fiber optic gyroscope based on deep learning

This work provides an error compensation strategy based on deep learning to address the temperature error of a fiber optic gyroscope (FOG). The Attention structure was used to improve the Long Short-Term Memory Network (LSTM), and the improved network was utilized to establish the prediction and compensation model of FOG error. From the two perspectives of the learning ability of the neural network and the improvement of FOG performance, this paper selects three indicators: prediction accuracy, mean square error, and FOG bias stability to comprehensively evaluate the performance of the compensation model. Through comparison, the compensation effect is significantly enhanced.

Error state prediction of capacitor voltage transformer based on TimesNet and gate control unit

The stability and reliability of the error state of capacitor voltage transformer (CVT) are poor, which affects the fairness of electricity trade settlement and the safe operation level of power grid. However, the CVT error data is superimposed by different periodic information such as daily period, monthly period and quarterly period, which makes the prediction of transformer error state challenging. In this paper, a CVT error state prediction method based on TimesNet and gate control unit (GRU) is proposed. The TimesNet network is employed to capture the intraperiod-variation and interperiod-variation characteristics of the ratio difference data of a-phase, b-phase and c-phase, and the feature data and GRU model are employed to predict the output. The simulation results demonstrate that the mean square error (MSE) and mean absolute error (MAE) of the proposed model are 0.0002 and 0.0101, respectively, indicating that this model has lower prediction error and higher prediction accuracy.

Preceding word information for predicting speech errors in English as foreign language speech

Speech errors, including disfluency errors (e.g., filled pauses (“uh”, “um”), repetition (“I me-mean right now.”)), and mispronunciation of speech segments (e.g., “think” as /sɪŋk/) are natural occurrences in speech production and they can affect speech fluency and proficiency. Detecting these errors is important, especially in assessing second language (L2) learners. Non-native speakers often produce speech errors, even in read speech, due to increased cognitive load when simultaneously producing the current word and processing the upcoming word. By analyzing two L2 speech corpora having different types of errors (disfluency and mispronunciation), the study aims to assess the effectiveness of incorporating preceding word information in speech error prediction. The results indicate that longer syllable duration in the word preceding the mispronunciation error correlates with error production. The study extends its investigation to identify acoustic and lexical features influenced by the presence of speech errors. The study found that speech errors increase duration of speech. The study also found that words with more syllables may cause L2 speakers to produce speech errors.

Analysis of Rotor Anisotropy and Position Tracking Error Prediction for Surface-Mounted PMSM Under High Frequency Signal Injection

Analysis of Rotor Anisotropy and Position Tracking Error Prediction for Surface-Mounted PMSM Under High Frequency Signal Injection

Indirect Prediction of Spindle Rotation Error Through Vibration Signal Based on Supervised Local Mean Decomposition Filter Fusion and Bi-LSTM Classification Network

Abstract Spindle rotation error directly correlates with the quality of mechanical processing. Currently, the error was mainly converted through measuring the distance information of standard component installed at the tool position, and it can't complete the normal machining because the tool is occupied. Therefore, a novel self-adaptive supervised learning method through easy-collected vibration signal that don't affect the machining to indirect predict the error. This method includes three steps: First, the original vibration signal is decomposed by local mean decompression (LMD) method to obtain two critical components; subsequently, the two components are fused as a signal by a weighted-average approach; finally, the fused signal and corresponding error are self-adaptive supervised trained by the setting termination condition to modify fusion coefficient and network parameters. The method is used to analyze the data-set of spindle platform, which has collected the experimental data at speeds 1000, 2000, 3000, and 4000 more than 170 groups, and the indirect prediction accuracy reached 94.12%, 92.35%, 97.68%, and 90.59%, respectively. Additionally, the experimental results were compared and demonstrated by three aspects with current different algorithms.