Model-based Prediction Research Articles

Composite light ropes model-based dynamics force prediction model of high speed dry milling UD-CF/PEEK considering size effect

As an emerging material, Carbon fiber reinforced polyetheretherketone (CF/PEEK) is widely used in aviation, automotive, sports products, medical equipment and other high-end manufacturing industries due to its high shock resistance, thermostability, and recyclability. However, its dry-machining is an unclear dynamic material removal process which limits the machining efficiency and surface quality. To this end, we first presented composite light ropes model in machining composites considering size effect. Based on composite light ropes model, cutting region is affected by bending, stretching, pressing, bouncing and delamination, as the cutting mechanisms of carbon fiber and matrix are different. The carbon fibers distribution in chip is first introduced into the model to confirm the fiber deformation at different position. Based on composite light ropes model and elastic-plastic theory, the cutting force prediction model is obtained. To complete the model, the free-damped vibration force prediction model is established in empty cutting stage. The dynamics force prediction model is proofed with high prediction accuracy. Besides, the influences of cutting parameters on cutting and vibration force are analyzed by nonlinear regression analysis. It demonstrated that size effect is of great importance in establishing cutting force prediction model and revealing cutting mechanisms of CF/PEEK.

A model-based prediction and analysis of seasonal and tidal influence on pollutants distribution from city outfalls of river Ganges in West Bengal, India and its mapping using GIS tool

River Ganges (locally called as river Ganga) is one of the most scared rivers in India. The river is symbol of hope, faith and is worshipped for its wholesomeness due to its purity and sanctity. Pollution of river water due to anthropogenic activity is a very common issue worldwide. Similarly, river Ganga pollution in India throughout its entire courses, is a major concern due to city outfalls. This river, also named as river Hooghly in West Bengal, India, is exposed to outfalls carrying domestic wastewater of its both bank and their distribution in river Ganga is strongly influenced by season and tide. This study aimed to generate an idea of distance and direction wise changes of concentration of pollutants in wastewater in river Ganga. During 2014, the selection of five major outfalls was done by considering Dissolved Oxygen (DO), Biochemical Oxygen Demand (BOD), heavy metals, total fecal coliform level, and the study continued for next four consecutive years to find out the influence of tide and season. Geographical Information System (GIS) based maps provided a better reflection of these changes. Student’s t-test highlighted the significant changes in concentration of parameters season wise. A significant higher value of DO, BOD, nitrate nitrogen, and chloride were found in pre-monsoon season compared to monsoon season. Regression Equation generated for highly correlated parameters (coliform and heavy metals) helped to predict the level of one parameter with others. The zone of influence of BOD, DO, phosphorus and nitrate nitrogen from each of the five selected outfalls was very prominent. Acoustic Doppler current profiler at two of the five outfalls helped to estimate strip-wise depth average discharge which helped to estimate the value of water quality parameters by Plug Flow Model during high tide and low tide. A strong tidal variation was observed during low tide. This study helped to predict the influential zone from outfalls which will help to generate an alternative solution of river water use. This approach can be applied globally to prepare river water usage guidelines.

A novel machine learning model for safety risk analysis in flywheel-battery hybrid energy storage system

Flywheel energy storage system (FESS) has been regarded as the most promising hybrid storage technique to manage the battery charging process of electric vehicles. Thanks to properly regulating with the FESS, the battery life can be significantly prolonged. In order to ensure the safety of the hybrid storage system, it is imperative to monitor the mechanical operation condition of the FESS. Because the rolling bearing is a critical mechanical component in the FESS, the performance degradation monitoring and remaining useful life (RUL) prediction of the rolling bearing must be performed. This paper proposes a new machine learning method for the construction of a health indicator to quantitatively evaluate the bearing health status. In this new method, the original feature set is firstly selected through three feature evaluation indicators and the principle component analysis (PCA) is employed to fuse the original feature set as a new health indicator. Then the primary trend of the health indicator is extracted by the empirical mode decomposition (EMD); and the Kriging model-based prediction method is proposed to predict the bearing RUL. The feasibility and superiority of the proposed method is verified through experimental test and analysis result shows that the root mean square error (RMSE) of the prediction is as small as 0.0425.

Novel Application of Survival Models for Predicting Microbial Community Transitions with Variable Selection for Environmental DNA.

Survival analysis is a prolific statistical tool in medicine for inferring risk and time to disease-related events. However, it is underutilized in microbiome research to predict microbial community-mediated events, partly due to the sparsity and high-dimensional nature of the data. We advance the application of Cox proportional hazards (Cox PH) survival models to environmental DNA (eDNA) data with feature selection suitable for filtering irrelevant and redundant taxonomic variables. Selection methods are compared in terms of false positives, sensitivity, and survival estimation accuracy in simulation and in a real data setting to forecast harmful cyanobacterial blooms. A novel extension of a method for selecting microbial biomarkers with survival data (SuRFCox) reliably outperforms other methods. We determine that Cox PH models with SuRFCox-selected predictors are more robust to varied signal, noise, and data correlation structure. SuRFCox also yields the most accurate and consistent prediction of blooms according to cross-validated testing by year over eight different bloom seasons. Identification of common biomarkers among validated survival forecasts over changing conditions has clear biological significance. Survival models with such biomarkers inform risk assessment and provide insight into the causes of critical community transitions. IMPORTANCE In this paper, we report on a novel approach of selecting microorganisms for model-based prediction of the time to critical microbially modulated events (e.g., harmful algal blooms, clinical outcomes, community shifts, etc.). Our novel method for identifying biomarkers from large, dynamic communities of microbes has broad utility to environmental and ecological impact risk assessment and public health. Results will also promote theoretical and practical advancements relevant to the biology of specific organisms. To address the unique challenge posed by diverse environmental conditions and sparse microbes, we developed a novel method of selecting predictors for modeling time-to-event data. Competing methods for selecting predictors are rigorously compared to determine which is the most accurate and generalizable. Model forecasts are applied to show suitable predictors can precisely quantify the risk over time of biological events like harmful cyanobacterial blooms.

Traffic Accident Severity Prediction Based on Random Forest

The prediction of traffic accident severity is essential for traffic safety management and control. To achieve high prediction accuracy and model interpretability, we propose a hybrid model that integrates random forest (RF) and Bayesian optimization (BO). In the proposed model, BO-RF, RF is adopted as a basic predictive model and BO is used to tune the parameters of RF. Experimental results show that BO-RF achieves higher accuracy than conventional algorithms. Moreover, BO-RF provides interpretable results by relative importance and a partial dependence plot. We can identify important influential factors for traffic accident severity by relative importance. Further, we can investigate how the influential factors affect traffic accident severity by the partial dependence plot. These results provide insights to mitigate the severity of traffic accident consequences and contribute to the sustainable development of transportation.

A digital twin-driven hybrid approach for the prediction of performance degradation in transmission unit of CNC machine tool

• DT model based meta-action, collected data from multi-sensors on transmission unit. • Monitoring and simulation of performance degradation on the wear of transmission. • Combining the data-driven and the model-based method by particle filter algorithm . • Integrating mechanism and real-time data, mutual modified to a High fidelity model . Precision performance prediction of transmission system is considered as a key technology to modern equipment health management. Given the importance of maintaining a transmission system's precision, this paper presents a hybrid approach framework driven by digital twin technology (DT), to predict performance degradation. Firstly, a DT model based on meta-action theory is established, and real-time monitoring and digital simulation, driven by DT data, is realized in order to analyze the precision of the transmission units in machine tools. Secondly, the wear of gear in transmission unit is studied through Achard wear theory, which considered the comprehensive influence of gear load and speed on surface wear of the gear pair tooth, based on the model driving method. The performance degradation of the transmission unit is obtained by using the RBF neural network algorithm based on the data-driven method to extrapolate the wear data to the field-measurable precision index value. In addition, the hybrid predictive approach of the performance degradation model through the particle filter algorithm is built, and the real-time data is used to update the current state estimation to improve the prediction accuracy. By combining the mechanism of the physical degradation processes with the real-time and historical data and turning them into a cooperative architecture, this prediction method uses the complementary advantages offered by the fusion of these methods to bridge the link between data-driven prediction and model-based prediction. Finally, the method has been successfully applied to the precision prediction of the transmission unit in CNCMT turntable, and it is compared with the single prediction method to verify the effectiveness and feasibility.

Model-Based Prediction and Optimal Control of Pandemics by Non-Pharmaceutical Interventions

A model-based signal processing framework is proposed for pandemic trend forecasting and control, by using non-pharmaceutical interventions (NPI) at regional and country levels worldwide. The control objective is to prescribe quantifiable NPI strategies at different levels of stringency, which balance between human factors (such as new cases and death rates) and cost of intervention per region/country. Due to infrastructural disparities and differences in priorities of regions and countries, strategists are given the flexibility to weight between different NPIs and to select the desired balance between the human factor and overall NPI cost. The proposed framework is based on a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">finite-horizon optimal control</i> (FHOC) formulation of the bi-objective problem and the FHOC is numerically solved by using an ad hoc <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">extended Kalman filtering/smoothing</i> framework for optimal NPI estimation and pandemic trend forecasting. The algorithm enables strategists to select the desired balance between the human factor and NPI cost with a set of weights and parameters. The parameters of the model are partially selected by epidemiological facts from COVID-19 studies, and partially trained by using machine learning techniques. The developed algorithm is applied on ground truth data from the Oxford COVID-19 Government Response Tracker project, which has categorized and quantified the regional responses to the pandemic for more than 300 countries and regions worldwide, since January 2020. The dataset was used for NPI-based prediction and prescription during the XPRIZE Pandemic Response Challenge.

Linear Model Identification for Personalized Prediction and Control in Diabetes.

Type-1 diabetes (T1D) is a disease characterized by impaired blood glucose (BG) regulation, forcing patients to multiple daily therapeutic actions, including insulin administration. T1D management could considerably benefit of accurate BG predictions and automated insulin delivery. For both tasks, the large inter- and intra-individual variability in glucose response represents a major challenge. This work investigates different techniques to learn individualized linear models of glucose response to insulin and meal, suitable for model-based prediction and control. We focus on data-driven techniques for linear model-learning and compare the state-of-art parametric pipeline with a novel non-parametric approach based on Gaussian regression and Stable-Spline kernel. On data collected by 11 T1D individuals, the effectiveness of different models was evaluated by measuring root mean squared error (RMSE), coefficient of determination (COD), and time gain associated with BG predictors. Among the tested approaches, the non-parametric technique results in the best prediction performance: median RMSE = 29.8 mg/dL, and median COD = 57.4%, for a prediction horizon (PH) of 60 min. With respect to the state-of-the-art parametric techniques, the non-parametric approach grants a COD improvement of about 2%, 7%, 21%, and 41% for PH = 30, 60, 90, and 120 min (paired-sample t-test p ≤ 0.001, p = 0.003, p = 0.03, and p = 0.07 respectively). Non-parametric linear model-learning grants statistically significant improvement with respect to the state-of-art parametric approach. The higher PH, the more pronounced the improvement. The use of a linear non-parametric model-learning approach for model-based prediction and control could bring to a more prompt, safe and effective T1D management.

Prediction of knock propensity using stochastic modeling in a spark-ignition engine

To comply with stringent CO2 regulations, enhanced thermal efficiency has been prioritized in internal combustion engine development; however, this has strongly driven the development of engines with operating conditions more prone to knock. Current knock sensors have its limitations to decompose knock signal by degradation so that it required a cross-referencing signal. In addition, knock control intervention is currently preceded by the occurrence of the knock, leading to decrease in thermal efficiency by retarding spark timing. In the present work, a novel prediction model for knock propensity (incidence) is presented, aiming to enable active control of knock or autoignition, and to support conventional knock sensor for cross-referencing by facilitating virtual knock sensor. A zero-dimensional model-based prediction of the in-cylinder pressure is demonstrated to prevent using in-cylinder pressure transducer, along with other incorporated predictive sub-models for the residual gas fraction, heat loss, burn duration, and heat release rate. Ignition delay correlation and Livengood-Wu relation are used to predict the onset of knock, and a burn point-based criterion is newly proposed for application in stochastic modeling for determining the knock propensity. The predicted knock propensity from the combined holistic model shows a remarkable agreement with experimental results.

Prediction of wind turbine blade icing fault based on selective deep ensemble model

In advance accurate prediction of wind turbine blade icing fault is of fundamental importance. Deep learning is the mainstream prediction technique while existing research about the prediction of wind turbine blade icing fault has primarily derived a single deep learning model. This study introduces the group method of data handling (GMDH) technique and proposes a GMDH-based selective deep ensemble (GSDE) model. First, the model combines the convolution neural network (CNN) with recurrent neural network (RNN), long short-term memory (LSTM), and gated recurrent unit (GRU) to construct CNN–RNN, CNN–LSTM, and CNN–GRU, respectively. Together with CNN, four cost-sensitive deep neural networks based on focal loss are formed and used as the basic prediction models. Second, a series of training sets are constructed by the Chi-square test. Four different basic prediction models are trained on each training set, and the prediction results of all base predictors are obtained. Third, the GMDH technique is applied to the cost-sensitive selective deep ensemble for final prediction results. Experiments are conducted to deeply verify the prediction performance of the GSDE model on two wind turbine datasets collected by the supervisory control and data acquisition (SCADA) system. Results show that the proposed model outperforms five existing ensemble models and five single deep learning models.

Serum amyloid A predicts poor functional outcome in patients with ischemic stroke receiving endovascular thrombectomy: a case control study

BackgroundPost-stroke inflammation contributes to poor outcomes, but its impact on patients with stroke receiving endovascular thrombectomy (EVT) remains unknown.MethodsWe enrolled adult patients with stroke who received EVT, with blood sampling...

Intelligent Vehicle Path Tracking Control Based on Improved MPC and Hybrid PID

In this study, to improve the accuracy of path tracking in intelligent vehicles, we propose an intelligent vehicle path-tracking control method based on improved model predictive control (MPC) combined with hybrid proportional-integral-derivative (PID) control theory. In the lateral control, a constraint on the side deflection of the front wheel is added based on traditional MPC and a relaxation factor is introduced to improve the stability of vehicle control for the driving stability. In longitudinal control, a hybrid PID controller is designed for different road conditions to improve the accuracy of control of vehicle speed. We present the results of a co-simulation using CarSim and MATLAB/Simulink and a test with a sample vehicle, which show that the proposed path tracking controller can greatly improve the path tracking accuracy and stability of an intelligent vehicle. The model-based prediction, rolling optimization solution, feedback control, and the addition of a constraint on the side deflection of the front wheel as well as a relaxation factor can ensure the lateral driving stability of an intelligent vehicle. The proposed approach achieved a lateral error of less than 1%, and the yaw angle was controlled within 4°. The longitudinal speed control based on hybrid PID controller can improve the response speed of the system and meet the real-time requirements of vehicle driving.

On the Uniqueness of the Quasi-Moment-Method Solution to the Pathloss Model Calibration Problem

Investigations in this paper focus on establishing the uniqueness properties of the Quasi-Moment-Method (QMM) solution to the problem of calibrating nominal radiowave propagation pathloss prediction models. Nominal (basic) prediction models utilized for the investigations, were first subjected to QMM calibrations with measurements from three different propagation scenarios. Then, the nominal models were recast in forms suitable for Singular Value Decomposition (SVD) calibration before being calibrated with both the SVD and QMM algorithms. The prediction performances of the calibrated models as evaluated in terms of Root Mean Square Prediction Error (RMSE), Mean Prediction Error (MPE), and Grey Relational Grade-Mean Absolute Percentage Error (GRG-MAPE) very clearly indicate that the uniqueness of&nbsp; QMM-calibrations of basic pathloss models is more readily observable, when the basic models are recast in forms specific to SVD calibration. In the representative case of calibration with indoor-to-outdoor measurements, RMSE values were recorded for QMM-calibrated nominal models as 5.2639dB for the ECC33 model, and 5.3218dB for the other nominal models. Corresponding metrics for the alternative (rearranged) nominal models emerged as 5.2663dB for the ECC33 model and 5.2591dB for the other models. A similar general trend featured in the GRG-MAPE metrics, which for both SVD and QMM calibrations of all the alternative models, was recorded as 0.9131, but differed slightly (between 0.9138 and 0.9196) for the QMM calibration of the nominal models. The slight differences between these metrics (due to computational round-off approximations) confirm that when the &nbsp;components of basic models are linearly independent, the QMM solution is unique. Planning for wireless communications network deployment may consequently select any basic model of choice for QMM-calibration, and hence, identify relative contributions to pathloss by the model’s component parts.

Prediction of stock market price movements based on sentiment analysis on various news headlines

Due to the recent advances in social media, stock-related news is available on every social media platform, imposing a value to improve stock price prediction performance. We used linear discriminant analysis (LDA) to screen chip indicators, and we build a basic stock prediction model. We have gathered stock-related news headlines and then quantify the unstructured data into sentiment scores using text mining technology which can be correlated with the stock price changes. We have proposed an augmented prediction model, and our results suggest that prediction accuracy for stock prices can be improved, and it will be advantageous for stock investors regarding their investment strategies. In this research, we have analysed more than 49,000 news headlines to determine the relationship between the news headlines and stock price on a given date. Using novel data visualisation and natural language processing (NLP) techniques, we have implemented data visualisations showing how the company share prices are affected based on the sentiment scores we are getting from analysing the news headlines.

Application of Training Load Prediction Model based on Improved BP Neural Network in Sports Training of Athletes

With the enhancement of data mining technology, competitive sports informatization has become an inevitable development trend. It has become a common phenomenon to use data mining technology to help athletes train scientifically, assist coaches in rational decision-making, and improve team competitiveness. In competitive sports, cyclists' adaptation to training has a complex relationship with their physical performance. In order to explore the correlation between data and provide better training data for athletes, this study proposes a load prediction model based on BP neural network (Back propagation, BP). Considering the local convergence and random assignment of traditional BP model, an adaptive genetic algorithm with improved selection operator is used to determine the initial weights and thresholds of BP neural network to improve the accuracy of the prediction model. The experimental results show that the improved adaptive genetic algorithm improves the overall optimization ability of the BP neural network, the improved BP neural network model has good stability in the convergence process, and the algorithm can search for better weight thresholds. Compared with the basic BP neural network prediction model, the accuracy of the optimized prediction model is increased by 11.86%, and the average error value is reduced by 26.21%, which is a guide to improve the training effect of the cycling team's competitive sports.

Review on the Change Trend, Attribution Analysis, Retrieval, Simulation, and Prediction of Lake Surface Water Temperature

Lake surface water temperature (LSWT) is indicative of changes in climate and geomorphology. It has been found that LSWT has shown faster warming than air temperature in recent decades, which has greatly affected the aquatic ecosystem of lakes and even led to ecological collapse. It was mainly caused by climate warming and geomorphological changes caused by human activities. However, lakes with different properties respond differently to these changes, resulting in temporal and spatial heterogeneity in the changes of LSWT. After an extensive literature review, it was found that the research on LSWT is mainly focused on the following aspects: (1) qualitative or quantitative analysis of natural and anthropogenic factors affecting LSWT changes; (2) impacts of LSWT changes on lake ecology; (3) acquisition of LSWT data, such as retrieval based on remote sensing (RS) imagery, model-based simulation prediction; (4) study of temporal and spatial variation characteristics of LSWT long time series. Dynamically monitoring LSWT changes, predicting and quantifying the response of lakes with different properties to abrupt or gradual climate changes and changes in human activity intensity, can inform modeling of ecosystem response processes and long-term management planning of large freshwater ecosystems. Therefore, this paper provides a systematic review of LSWT from the above four aspects to provide empirical insight and methodological reference for conducting LSWT research.

Application of depth learning in prediction of glutenite reservoir

Sedimentary facies of gravel reservoir has the characteristics of multiple sources and short flow. Sand body thickness changes rapidly in plane and space. Model-based multi-attribute regression seismic reservoir prediction method often results in unreasonable inter-well prediction results because of insufficient generalization ability of the model. For this reason, a deep learning reservoir prediction method based on fault blocks is proposed. Within a fault block range (regardless of wells outside the fault block), the seismic attributes extracted from the sliding time window of the isochronous sedimentary surface are used as input, and the sandstone thickness data of the target layer of the well point are used as expected output to train the depth neural network model of the multi-layer perceptor. Then, the trained model is used to predict the sandstone thickness distribution in the well distribution area. The application of actual data in A oilfield shows that the predictive coincidence rate of sandstone thickness in known wells is more than 95%, and the drilling occurrence rate of sandstone in horizontal section after completion of a posterior horizontal well is more than 80%. This method has high prediction accuracy and good generalization ability.

Model-based wear prediction of milling machine blades

Technological capabilities are enabling the implementation of various prognostic methodologies for industrial assets. Whereas companies are driving for finding new ways to improve asset operation and service offerings to their customers, model-based simulations can be deployed to model machine behavior for instance in usage cases unconventional for the asset. Modern embedded technologies in peripheral milling machines offer asset operation-related data collection and sharing online. Due to this capability, understanding of the asset operational behavior can be assessed remotely, and the data can be used to estimate the wear progress of the milling blades. However, creating a model-based simulation model is not heavily dependent on online data, yet the model construction requires knowledge about asset operation and physical behavior to enable purpose-fit and simple enough simulation construction. In this research, a model-based simulation model is created to predict peripheral milling machine blade wear in terms of average vibration and torque parameters. A blade wear variability in different usage profiles is being tested with a simulation test case. The results are proving that the model-based simulation model can be accurately used to emulate asset physical behavior by the means of torque and vibration parameters. Based on the results, changes in the asset vibration average trend can be distinctly utilized to estimate wear progress on the spindle cutting blades. Further, predicted vibration levels and blade lifetime estimations can be considered in PPX type of business model profitability or lifecycle related calculations where ownership of machines is retained by the manufacturing company.

Fuzzy logic model-based punch force prediction for deep drawing of high strength steel

The process of sheet metal forming is referred to as Deep Drawing. It is applied to make hollow–curved metal surfaces using a die and punch which creates the hollow burrow. This process is widely used in various industries to form sheet metals into automatic parts of certain requirements. The punch force applied in the die depends upon various parameters. The punch force is also responsible for the effective outcome of the deep drawing of the metal sheet. Thus, a set of process parameters that majorly affects the punch force is decided and those experiments were done to obtain the punch force at different values for different input parameters. Producing punch force requires huge energy expenditure. Hence, for the process to be both economical and efficient – lower punch forces are preferred. But when multiple parameters are considered, it becomes difficult to obtain the lowest punch force value. Hence, with the L27 orthogonal array, the number of combinations of the experiments that need to be done is reduced phenomenally. These experimental data were used to create a fuzzy logic model using MATLAB software. Therefore, a Fuzzy Logic Model was generated with the considered parameters being the input and the punch force being the output. This model will be used for punch force prediction as well as finding the optimized input parameters to achieve the lowest punch force.

Model-based Software Defect Prediction from Software Quality Characterized Code Features by using Stacking Ensemble Learning

Model-based Software Defect Prediction from Software Quality Characterized Code Features by using Stacking Ensemble Learning