Nonlinear Characteristics Research Articles

A Neuroadaptive Position-Sensorless Robust Control for Permanent Magnet Synchronous Motor Drive System with Uncertain Disturbance

The Permanent Magnet Synchronous Motor (PMSM) drive system is extensively utilized in high-precision positioning applications that demand superior dynamic performance across various operating conditions. Given the non-linear characteristics of the PMSM, a neuroadaptive sensorless controller based on B-spline neural networks is proposed to determine the control signals necessary for achieving the desired performance. The proposed control technique considers the system’s non-linearities and can be adapted to varying operating conditions, all while maintaining a low computational cost suitable for real-time operation. The introduced neuroadaptive controller is evaluated under conditions of uncertainty, and its performance is compared to that of a conventional PI controller optimized using the Whale Optimization Algorithm (WOA). The results demonstrate the viability of the proposed approach.

Revealing the Role of Electronic Effect to Modulate the Photophysics and Z-Scan Responses of o-Locked GFP Chromophores.

Three novel core green fluorescent protein (GFP) chromophore analogues, based on a doubly locked conformation and variable electronic effects by replacing one hydrogen with bromine, iodine, and methyl, respectively, have been synthesized to modulate the push-pull effect. These chromophores exhibited intramolecular H-bonding, as evidenced by single-crystal X-ray and 1H NMR studies. The fluorescence quantum yields (ϕf) of all of the chromophores were found to be more than an order of magnitude higher (∼0.2) than the unlocked chromophores (∼0.01). It was found that the electronic effect did affect the nonradiative rates, as the quantum yields were found to vary with respect to different analogues in the same solvents. The effect of the push-pull effect was also evident by a higher Stokes-shifted emission in the case of the methyl derivative with respect to the bromo- and iodo-analogues. Furthermore, the emission spectra of these GFP chromophores were found to show positive solvatochromism, which was supported by a quantum chemical calculation. A detailed study, correlating the observed spectral changes with various solvent functions and supported by computational results, established a facile proton transfer, followed by twisted intramolecular charge transfer (TICT) to be the major nonradiative channels of these chromophores. Besides, a completely novel usage of these chromophores was explored for the first time by studying their third-order nonlinear optical characteristics in DMSO using a single-beam Z-scan technique. All of the chromophores exhibited tunable nonlinear refraction (NLR) and nonlinear absorption (NLA) properties that depend upon different substituent groups present in the chromophores. Here, the NLR was due to the effect of self-defocusing, whereas the NLA was triggered by reverse saturable absorption, which is attributed to the two-photon absorption (TPA) process. Surprisingly, the efficiency of the TPA ability of the chromophores was found to be a function of the induced electronic effect. Hence, this work opens a new route for the utility of the ortho-locked GFP chromophores in the field of nonlinear optical applications.

Para-phenylenediamine Schiff base: highly fluorescent photostable solid-state organic dye

Schiff bases derived from the condensation of primary amines with aldehydes, such as para-phenylenediamine with salicylaldehyde, exhibit unique optical and structural properties ideal for photonic applications. This study synthesizes such a Schiff base, revealing its properties through detailed HNMR1, FTIR, and XRD characterizations. The compound forms a robust π-conjugated structure, showing a fluorescence emission peak at 560 nm and significant absorbance at 380 nm. A spin-coated laser cavity displayed a critical lasing threshold at 1 MW·cm−2, which could be optimized down to 0.6 kW·cm−2. Moreover, the compounds’ acceptor-donor-acceptor configuration raised outstanding nonlinear optical properties, including a substantial two-photon absorption cross-section of 2×10−11 cm ⋅ W−1·GM, enhancing its utility in high-resolution two-photon imaging and advanced photonic applications. Other nonlinear optical characteristics determined during these studies are the saturable absorption-induced nonlinear absorption coefficient (−2×10−11 cm ⋅ W−1), saturation intensity (2.5×1011 W·cm−2), and Kerr-induced nonlinear refractive index (5×10−16 cm2 ⋅ W−1). The combined linear and nonlinear optical properties, supported by sustained emission and minimal photobleaching under intense re-excitation, establish the para-phenylenediamine Schiff base and derivatives as promising materials for high-brightness, photostable organic light emitters, and solid-state lasers.

Predicting bifurcation and amplitude death characteristics of thermoacoustic instabilities from PINNs-derived van der Pol oscillators

Self-sustained thermoacoustic oscillations as observed in low-emission combustion- involved gas turbines and aero-engines involve complicated thermal fluid–acoustics interaction and rich nonlinear dynamics. Such pulsating oscillations are known as thermoacoustic instability. When it occurs, large-amplitude limit cycle oscillations (LCOs) of thermodynamic parameters are frequently observed. These LCOs could cause overheating, flame flashback, and even engine failures. Thus it is critical to understand and predict the generation mechanisms and nonlinear dynamics behaviours, and then develop corresponding control approaches to prevent or control the onset of such instabilities. In this work, we develop and extend the classical van der Pol oscillators by integrating a physics-informed neural networks (PINNs) algorithm with a modelled nonlinear Rijke-type thermoacoustic combustor. The theoretical Rijke tube system (with Galerkin expansion and modified King's law implemented) and a CFD simulation model are applied to provide ‘training/calibration data’ for the extended van der Pol (EVDP)-PINNs model. The optimized EVDP oscillators are confirmed to be capable of capturing the key nonlinear characteristics by comparing the transient growth behaviours of thermodynamic perturbations and LCO amplitude and frequency. Further investigations are conducted to obtain Hopf bifurcation and amplitude death (AD) characteristics. Comparison is then made to the coupled EVDP systems. Quite similar Hopf bifurcation features, but differences in regions of AD, are observed. In general, we demonstrate an applicable approach to intelligently ‘learn’ a nonlinear thermoacoustic system and to create reliable EVDP oscillator systems, which have great potential to contribute to the development and testing of control approaches, such as the coupling described in this work, which may replace costly experimental tests.

Large-amplitude nonlinear vibrations characteristics of bi-directional functionally graded plate having microstructure defects in hygro-thermal environment

The present article investigates the nonlinear hygro-thermal vibration characteristics of bi-directional functionally graded plates (BFGP) with microstructural defects. The distribution of material properties is anticipated to vary gradually along both thickness and length directions according to the modified power-law distribution. A novel expression has been developed to solve for the nonlinear temperature variation along the thickness of the BFGP. The micro-structural defects in terms of porosity have been incorporated in the formulation using a newly developed porosity model for a BFGP. A refined non-polynomial trigonometric higher-order shear deformation theory (HSDT) with a von Karman sense of geometric nonlinearity has been employed for the vibration response of BFGP. The finite element formulation has been carried out utilizing the C0 continuous Lagrangian quadrilateral element with nine nodes and eight degrees of freedom per node. The equations of motion have been derived using a variational approach. Various validation and comparative studies have been conducted to substantiate the current formulation's authenticity. The effect of the geometric nonlinearity, boundary constraints, volume fraction index along thickness (n) and volume fraction index along length (q), porosity index, temperature, and moisture concentration difference on the vibration frequency ratios (VFR) has been covered in depth. The VFR is more sensitive to the volume fraction indices (VFI) ' n′ than 'q'. Incorporating porosity decreases the VFR of BFGP, but it increases at higher amplitude ratios when BFGP becomes more porous. The VFR of BFGP exhibits nonlinear behavior with increasing moisture concentration and also varies non-linearly with the rise in amplitude ratio due to the hygro-thermal environment.

A Nonlinear Suspension Road Roughness Recognition Method Based on NARX-PASCKF

Road roughness significantly impacts vehicle safety and dynamic responses. For nonlinear suspension systems, the nonlinear characteristics often make it challenging for estimators to identify the actual road roughness accurately. This paper proposes a hybrid road roughness identification algorithm based on nonlinear auto-regressive with exogenous inputs (NARX) and a process noise adaptive square root cubature Kalman filter (PASCKF) to address this issue. Driven by vehicle acceleration data, an NARX-based road roughness identification system is constructed to mitigate the model uncertainties. Furthermore, a hybrid strategy is proposed. On the one hand, the accurate road roughness estimated by the NARX is converted into process noise covariance, enhancing the estimator’s accuracy and convergence rate. Another switching strategy is proposed to optimize the non-convergence issues of the PASCKF. Finally, simulation and actual vehicle experiment data demonstrate that this approach offers superior identification accuracy and adaptability compared to the standalone SCKF algorithm.

A Low-Computational-Complexity Digital Predistortion Model for Wideband Power Amplifier

This paper proposes a Composition Piecewise Memory Polynomial (CPMP) digital predistortion model based on a Vector Switched (VS) behavioral model to address the challenges of severe nonlinearity and strong memory effects in wideband power amplifiers (PAs). To tackle this issue, two thresholds are calculated and used to segment the envelope values of the input signal according to the nonlinear distortion characteristics of the PA. In this approach, a Generalized Memory Polynomial (GMP) model is employed for the lower segment, a Memory Polynomial (MP) model is employed for the middle segment, and a higher-order GMP model is employed for the upper segment. By sharing the fundamental MP among the proposed segmented models and leveraging a design methodology that configures different cross terms, memory depths, and polynomial orders for each segment, this model achieves superior linearization performance while simultaneously reducing the computational complexity associated with model extraction. The experimental results demonstrate that the adjacent channel power ratio (ACPR) of the predistorted PA output signal using the proposed model improves from −36 dBc to −54 dBc, matching the performance of the GMP model. Furthermore, this performance is 0.5 dBc better than the Piecewise Dynamic Deviation Reduction (PDDR) and Decomposed Vector Rotation (DVR) models. Notably, the complexity of the proposed parameter extraction process is 28.8% of the DVR model, 21.79% of the GMP model, and 12.83% of the PDDR model.

Nonlinear vibration analysis of multi-support oil-tank system in aero-engine based on asymptotic methods

This article is primarily focused on the complicated nonlinear vibration behavior of an aero-engine lubricating oil-tank system with multiple nonlinear pedestals. First, the rubber damping ring’s nonlinear factors are introduced into a cubic nonlinear dynamic model of lubricating oil-tank system. The nonlinear system parameters are identified by testing a simplified aero-engine lubricating oil-tank structure. Then, the nonlinear dynamic model is solved via asymptotic approach, and results are confirmed using Runge-Kutta numerical analysis approach. The system’s vibration responses and amplitude-frequency curves are also acquired, and influence of system parameters on nonlinearity is analyzed. Finally, an experimental investigation on the simplified oil-tank is performed. The nonlinear vibration characteristics of the rubber damping ring and its vibration reduction mechanism are verified using vibration transmissibility. The experimental results are consistent with numerical results. This paper’s recommended analysis approach has engineering significance for vibration and noise reduction design of aero-engine components.

Two-Dimensional MXene Flakes with Large Second Harmonic Generation and Unique Surface Responses.

MXenes are a family of two-dimensional (2D) materials with broad and varied applications in biology, materials science, photonics, and environmental remediation owing to their layered structure and high surface area-to-volume ratio. MXenes have exhibited significant nonlinear optical characteristics, which have been primarily explored in the context of photonics applications, yet the second-harmonic generation (SHG) behavior of MXenes remains an unexplored aspect of their optical properties. Herein, we demonstrate and quantify large second-order responses of 2D Ti3C2Tx MXenes both in aqueous solutions and on a silicon substrate for the first time. MXene flakes showed strong second-harmonic scattering (SHS) in a dilute suspension with a sensitivity of less than 0.1 μg/mL. Angle-dependent SHS experiments further found that the second-order responses originate from coherent 2D dipole radiation. Through confocal and atomic force microscopies, we found that the intense SHG signal from free-standing MXene flakes increases exponentially with decreasing thickness, while two-photon fluorescence increases linearly with thickness. The second-order susceptibility of the MXenes was determined to be 3.6 pm V-1 with a thickness of 10 nm, almost twice of that for an often-used SHG crystal, beta barium borate. We further explored surface properties of the MXene sheets by investigating the SHS responses upon addition of organic dye molecules to the system. It was found that the adsorption of crystal violet (CV) obeys a Langmuir adsorption model while the addition of malachite green (MG) resulted in almost no change in SHG intensity, even though the adsorption capacities for both CV (61.3 ± 1.7 mg/g) and MG (54.8 ± 2.8 mg/g) are similar. Such a stark difference in adsorption characteristics between cationic organic CV and MG dyes is likely due to their distinct orientational orderings on the MXene surfaces. This work opens many possibilities for the further employment of the family of 2D materials in photonics, optics, and surface catalysis applications.

A neural network approach for unstructured mesh quality evaluation

PurposeThe quality of the unstructured mesh has a considerable impact on the stability and accuracy of aerodynamic simulation in computational fluid dynamics (CFD). Typically, engineers spend a significant portion of their time on mesh quality evaluation to ensure a valid, high-quality mesh. The extensive manual interaction and a priori knowledge required to undertake an accurate and timely evaluation process have become a bottleneck in the idealized efficient CFD workflow. This paper aims to introduce a neural network-based quality evaluation approach for unstructured meshes to enable higher efficiency and the level of automation.Design/methodology/approachThe paper investigates the capability of deep neural networks for the quality evaluation of unstructured meshes. For training the network, we build a training dataset for mesh quality learning algorithms. The dataset contains a rich variety of unstructured aircraft meshes with different mesh sizes, densities, cell distribution, growth ratios and cell numbers to ensure its diversity and availability. We also design a neural network, AircraftNet, to learn the effect of mesh quality on the convergent properties of the numerical solutions. The proposed network directly manipulates raw point data in mesh source files rather than passing it to an intermediate data representation. During training, AircraftNet extracts non-linear quality features from high-dimensional data spaces and then automatically predicts the overall quality of the input unstructured mesh.FindingsThe paper provides a series of experimental results on GPUs. It shows that AircraftNet is able to effectively analyze the quality-related features like mesh density and distribution from the extracted features and achieve high prediction accuracy on the proposed dataset with even a small number of training runs.Research limitations/implicationsBecause of the limited training dataset, the research results may lack generalizability. Therefore, researchers are encouraged to test the proposed propositions further.Originality/valueThe paper publishes a benchmarking dataset for mesh quality learning algorithms and designs a novel neural network approach for unstructured mesh quality evaluation.

Physics-informed neutral network with physically consistent and residual learning for excavator precision operation control

The data-driven methodologies can establish accurate Inverse Dynamics Model (IDM) of the excavator thus improving control precisions. However, the inherent black-box nature of these models often results in overfitting to the dataset, leading to predictions that deviate from the constraints of physical system. Consequently, this can lead to controller failures, introducing unpredictable behavior that threatens operation precision. In addition, the uncertainty of the external disturbance poses great challenge to the precision of controller. This study presents a physics-informed neural network to build accurate IDM with physical consistency. The Rigid Body Dynamics (RBD) of the excavator are coupled within a Deep Lagrangian Network (DeLaN), while a Convolutional Neural Network (CNN) and a Long Short-Term Memory Network (LSTM) are employed to assimilate the residual nonlinear characteristics, such as hydraulic flexibilities and stick–slip friction. To the uncertainty of the external disturbance, the Prescribed Performance Inverse Dynamics Controller combination with the DeLaN-CNN-LSTM model (PPIDC-DCL) is constructed for precise control by constraining the control error within a finite region. The experimental results demonstrate that the model captures the underlying structure of the dynamic and builds the IDM with high accuracy and robustness. Moreover, the PPIDC-DCL controller effectively constrains the control error and realizes precision control. The proposed method has potential applications and provides novel insights for achieving precise operation control of excavators.

Application of the Southwell Method to Determine the Critical Load of Compression Rods Made of Nonlinear Materials

Experimental determination of the critical force of a compression bar made of linear material does not pose major problems. The widely used Southwell method is also applicable to bars with cross-sections varying in length. The purpose of the research undertaken was to demonstrate that this method can also be successfully applied to bars made of elastic materials exhibiting nonlinear σ(ε) characteristics. Using analytical relationships for a rod satisfying the assumptions of Euler-Bernoulli theory, F(δ) relationships were derived for a rod with an initial arc imperfection, and Southwell diagrams were constructed on this basis. Nonlinear equilibrium paths F(δ) were also determined numerically using the COSMOS/M program for this purpose. For the pre-assumed different nonlinear characteristics, full confirmation of the validity of the application of the Southwell method for determining the critical force was obtained.

The impact of proton ion irradiation on Se70Te20In10 thin films: Linear and non-linear optical properties for photonic applications

The examination of the linear and non-linear optical features was conducted on amorphous ternary Se70Te20 In10 thin films, which were exposed to proton irradiation energy of 3 MeV at varying fluences of 5×1013, 5×1014, 5×1015 and 5×1016 ions/cm2. The bulk sample was produced using the melt quenching method, and subsequently, thin films were obtained through electron beam evaporation. Calculations of linear and non-linear optical parameters were carried out based on transmittance and reflectance data acquired from UV–Vis–NIR spectroscopy spanning 300–2500 nm. Field emission scanning electron microscopy (FESEM) and X-ray diffraction (XRD) confirmed an amorphous phase in the films. The optical parameters showed strong dependency on the proton radiation fluence. In addition, the existence of minima or maxima for a number of parameters at a fluence of 5×1014 ions/cm2 showed that this was the optimum fluence. The linear and non-linear properties showed that the sample have potential applications in photonic devices.

A comparative study of three types of strength criteria for rocks

Currently, the unified strength criterion (USC), the three-dimensional Hoek-Brown (3D H-B) criterion and the generalized unified strength theory (GUST) are the three types of typical rock strength criteria. In this paper, a comparative study of the three types of strength criteria is performed for rocks. Based on the nonlinear characteristics of rock strength on meridian and deviatoric planes, the USC can predict rock strength under triaxial stress state. The USC is composed of two failure functions on meridian and deviatoric planes. The failure surface of the USC in principal stress space satisfies smoothness and convexity. The predicted strength for five types of rock under the true triaxial tests were compared among the USC, the GUST, and the 3D H-B criterion. The results indicate that the USC can effectively reflect the influence of the intermediate principal stress on rock strength and accurately predict rock strength under both triaxial tension (σ1=σ2>σ3) and triaxial compression (σ1>σ2=σ3). Additionally, the conventional triaxial tests were conducted on other eight types of rock to measure the strength on meridian plane. The predicted strengths for the eight types of rock on meridian plane were compared between the USC and the original H-B, which suggests that the USC is suitable for various types of rock and provides higher accuracy.

Based on Deep Learning Model and Flink Streaming Computing Short Term Photovoltaic Power Generation Prediction for Suburban Distribution Network

With the advancement of global energy internet construction, accurate prediction of new energy generation power such as photovoltaic is an important foundation for ensuring the safety and economic working of new power systems. A short-term photovoltaic power generation prediction method for suburban distribution networks based on deep learning model fusion and Flink flow calculation is proposed to address the challenges of complex power grids, diversified disturbance factors, and isolated monitoring points. This method uses Bi directional Long Short Term Memory(BiLSTM) to extract cross sequential nonlinear characteristic of photovoltaic power generation time series data. Compared with standard LSTM, BiLSTM can consider both historical and future information simultaneously, thus extracting richer extracted features from power generation time series data. This method also integrates attention mechanism to capture the importance distribution of historical temporal features for power generation prediction, effectively solving the problem of long-term temporal dependence in standard LSTM models. The Flink streaming computing framework embeds a trained BiLSTM-Attention photovoltaic power generation prediction model, enabling real-time prediction and monitoring analysis of photovoltaic power generation at various monitoring points in the suburban distribution network. This article uses a dataset of a suburban photovoltaic power station for validation, and trains the model with historical power generation data, meteorological factors, weather types, seasons, and other information as inputs. The BiLSTM-Attention fusion model studys the temporal characteristics of power generation, and has high accuracy in predicting short-term photovoltaic power generation in different scenarios. The Flink streaming computing platform can not only process high throughput predicted power data, but also has low time delay.

Deep Convolutional Transformer Network for Stock Movement Prediction

The prediction and modeling of stock price movements have been shown to possess considerable economic significance within the finance sector. Recently, a range of artificial intelligence methodologies, encompassing both traditional machine learning and deep learning approaches, have been introduced for the purpose of forecasting stock price fluctuations, yielding numerous successful outcomes. Nonetheless, the identification of effective features for predicting stock movements is considered a complex challenge, primarily due to the non-linear characteristics, volatility, and inherent noise present in financial data. This study introduces an innovative Deep Convolutional Transformer (DCT) model that amalgamates convolutional neural networks, Transformers, and a multi-head attention mechanism. It features an inception convolutional token embedding architecture alongside separable fully connected layers. Experiments conducted on the NASDAQ, Hang Seng Index (HSI), and Shanghai Stock Exchange Composite (SSEC) employ Mean Absolute Error (MAE), Mean Square Error (MSE), Mean Absolute Percentage Error (MAPE), accuracy, and Matthews Correlation Coefficient (MCC) as evaluation metrics. The findings reveal that the DCT model achieves the highest accuracy of 58.85% on the NASDAQ dataset with a sliding window width of 30 days. In terms of error metrics, it surpasses other models, demonstrating the lowest average prediction error across all datasets for MAE, MSE, and MAPE. Furthermore, the DCT model attains the highest MCC values across all three datasets. These results suggest a promising capability for classifying stock price trends and affirming the DCT model’s superiority in predicting closing prices.

Two-Step Iterative Medical Microwave Tomography

In the field of medical imaging, microwave tomography (MWT) is based on the scattering and absorption characteristics of different tissues to microwaves and can reconstruct the electromagnetic property distribution of biological tissues non-invasively and without ionizing radiation. However, due to the inherently nonlinear and ill-posed characteristics of MWT calculations, actual imaging is prone to overfitting or artifacts. To address this, this paper proposes a two-step iterative imaging approach for rapid medical microwave tomography. This method establishes corresponding objective functions for microwave imaging across multiple frequencies and conducts iterative calculations on images at varying resolutions. This effectively enhances image clarity and accuracy while alleviating the issue of prolonged computational time associated with imaging complex structures at high resolution due to insufficient prior information during iterative processes. In the electromagnetic simulation section, we simulated a three-layer brain model and conducted imaging experiments. The results demonstrate that the algorithm significantly enhances imaging resolution, accurately pinpointing cerebral hemorrhages at different locations using an eight-antenna array and successfully reconstructs tomography images with a hemorrhage area radius of 1 cm. Lastly, experiments were conducted using a medical microwave tomography platform and four simplified human brain models, achieving millimeter-level accuracy in MWT.

Entire loosening stage monitoring of bolted joints via nonlinear electro-mechanical impedance spectroscopy

This article proposes a nonlinear electro-mechanical impedance spectroscopy (NEMIS) methodology for the entire loosening stage monitoring of bolted joints. Unlike the conventional electro-mechanical impedance spectroscopy (EMIS) implemented in the frequency domain, NEMIS utilizes a temporal interrogating signal to procure the developed nonlinear impedance spectrum. It can harness the structural resonance information akin to the conventional EMIS, while concurrently capturing the contact acoustic nonlinearity (CAN) features with only one solitary Piezoelectric Wafer Active Sensor (PWAS). To elucidate the underlying principles of NEMIS, a comparative analysis juxtaposing the conventional EMIS, nonlinear ultrasonic techniques, and the proposed NEMIS was conducted. Subsequently, a reduced-order analytical model was established to simulate the rough contact interfaces of the bolted joints under various looseness states. Parametric studies were carried out to demonstrate the resonance deviation and nonlinear features resulting from the bolt looseness. Ultimately, the experimental implementation of NEMIS for the entire loosening stage monitoring of the bolted joint was performed compared with the conventional EMIS method. It was proved that NEMIS could detect both the incipient bolt loosening at an embryo stage and severe loosening of the bolted joint at the terminal period, integrating the continuous monitoring capability of conventional EMIS and the sensitivity of nonlinear ultrasonic methodology. The quantification on the severity of bolt loosening was accomplished via both linear and nonlinear damage indices, exhibiting a consistent trend with the gradations of bolt looseness. The article culminates in a summary, concluding remarks, and suggestions for future work.

Data driven cost-sensitive boosted tree for interpretable banking systemic risk prediction

Systemic risk (SR) in the banking sector poses a significant threat to both the financial system and the real economy. Its inherent characteristics of nonlinearity, non-equilibrium, and interconnectedness make it challenging to analyze using conventional statistical methods. In this paper, a cost-sensitive gradient boosting tree algorithm, FLXGBoost, is proposed for predicting SR. FLXGBoost considers the boosted tree, XGBoost as the base framework, boosting trees as the fundamental framework, guaranteeing the robustness of SR prediction. Additionally, to tackle the challenge of extreme data imbalance prevalent in SR prediction tasks, a cost-aware loss function, focal loss, is embedded into the boosted tree to enable FLXGBoost a risk-aware fashion. Moreover, a tree-derived interpretable algorithm SHAP is incorporated into this cost-sensitive solution, making FLXGBoost an accurate and interpretable risk-aware model. Experimental results on a financial risk prediction dataset pertaining to banking SR evince the capacity of FLXGBoost to significantly reduce the misclassification rate of risk banks, thereby mitigating substantial losses attributed to erroneous predictions of risky scenarios. Moreover, compared with classical imbalanced machine learning-based SR prediction approaches, the diverse evaluation metrics of FLXGBoost show that it is a competitive solution for accurate SR prediction. Besides, the explanatory analysis further demonstrates that FLXGBoost is a promising solution to address the issue of biased predictions in imbalanced banking SR in the interpretation perspective.

Real Case Study of 600 m3 Bubble Column Fermentations: Spatially Resolved Simulations Unveil Optimization Potentials for l-Phenylalanine Production With Escherichia coli.

Large-scale fermentations (»100 m³) often encounter concentration gradients which may significantly affect microbial activities and production performance. Reliably investigating such scenarios in silico would allow to optimize bioproduction. But related simulations are very rare in particular for large bubble columns. Here, we pioneer the spatially resolved investigation of a 600 m³ bubble column operating for Escherichia coli based l-phenylalanine fed-batch production. Microbial kinetics are derived from experimental data. Advanced Euler-Lagrange (EL) computational fluid dynamics (CFD) simulations are applied to track individual bubble dynamics that result from a recently developed bubble breakage model. Thereon, the complex nonlinear characteristics of hydrodynamics, mass transfer, and microbial activities are simulated for large scale and compared with real data. As a key characteristic, zones for upriser, downcomer, and circulation cells were identified that dominate mixing and mass transfer. This results in complex gradients of glucose, dissolved oxygen, and microbial rates dividing the bioreactor into sections. Consequently, alternate feed designs are evaluated splitting real feed rates in two feeds at different locations. The opposite reversed installation of feed spots and spargers improved the product synthesis by 6.24% while alternate scenarios increased the growth rate by 11.05%. The results demonstrate how sophisticated, spatially resolved simulations of hydrodynamics, mass transfer, and microbial kinetics help to optimize bioreactors in silico.