Data-driven Modeling Approach Research Articles

Minimal model identification of drum brake squeal via SINDy

AbstractThe industrial standard in the design and development process of NVH(Noise Vibration Harshness) characteristic of brakes is the application of Finite Element(FE) models with a high number of degrees of freedom in the range of one or several millions. Nevertheless, parallel experimental investigations are still indispensable. On the other hand, minimal models with, due to the inclusion of the self-excitation process, at least two degrees of freedom are well known to be capable to explain qualitatively phenomena as instability of the desired non-vibrating solution or limit cycle oscillation but are in general very inaccurate in predicting the dynamics of a specific real brake. This is because the underlying physical assumptions are already too restrictive and model parameters (especially those referring to nonlinearities) are widely unknown. To overcome this problem, the data-driven modeling approach SINDy(Sparse Identification of Nonlinear Dynamics) is applied to identify appropriate nonlinear functions for a brake squeal minimal model. A problem thereby is the limited database. It turns out that the naive implementation of the method yielding the lowest possible residuum does not necessarily provide physically meaningful models and results, respectively. Instead, a constrained model that incorporates physical knowledge is used to robustly identify parameters and reproduce realistic dynamic behavior. Thereby, several appropriate models with coexisting limit cycles and stationary equilibrium are identified. In particular, it was found that the angular position of the brake drum has a significant influence on the model parameters and therefore must be taken into account in a model with long-term validity.

Data-driven modeling of multi-stable origami structures: Extracting the global governing equation and exploring the complex dynamics

In recent years, multi-stable origami structures have garnered increasing attention for their applications in dynamic scenarios such as robotic arm motions, impact energy absorption, and spectrum gap regulation. Understanding the intricate working mechanisms and exploring the rich dynamics of these structures necessitate the development of dynamic models. However, existing dynamic modeling methods for origami structures are often cumbersome, and the resulting dynamic models often lack interpretability. To overcome these limitations, we propose a novel data-driven dynamic modeling approach based on B-spline Galerkin method and subset selection strategy. This approach directly captures the dynamics of multi-stable origami structures using measured data, eliminating the need for reliance on empirical or prior knowledge. To validate the effectiveness of the proposed approach, we first evaluate it on the Duffing system, which has explicit expressions, successfully reconstructing the governing equation. Subsequently, we apply this method to the dynamic modeling of the origami ball structure with tri-stability and the multi-cell stacked Miura-origami (SMO) structure with high dimensionality through simulation, showing favorable results. Finally, using experimental data collected from a bi-stable SMO structure prototype, we employ the proposed method to obtain a global model that can accurately predict different dynamic behaviors over a broad range of excitation frequencies, including intra-well periodic vibrations, inter-well periodic vibrations, and inter-well chaotic vibrations. Overall, our method showcases outstanding efficacy in formulating interpretable, manageable, and comprehensive dynamic models. It plays a pivotal role in delving into the intricate dynamics of multi-stable origami structures.

Deciphering oxygen distribution and hypoxia profiles in the tumor microenvironment: a data-driven mechanistic modeling approach

Objective. The distribution of hypoxia within tissues plays a critical role in tumor diagnosis and prognosis. Recognizing the significance of tumor oxygenation and hypoxia gradients, we introduce mathematical frameworks grounded in mechanistic modeling approaches for their quantitative assessment within a tumor microenvironment. By utilizing known blood vasculature, we aim to predict hypoxia levels across different tumor types. Approach. Our approach offers a computational method to measure and predict hypoxia using known blood vasculature. By formulating a reaction-diffusion model for oxygen distribution, we derive the corresponding hypoxia profile. Main results. The framework successfully replicates observed inter- and intra-tumor heterogeneity in experimentally obtained hypoxia profiles across various tumor types (breast, ovarian, pancreatic). Additionally, we propose a data-driven method to deduce partial differential equation models with spatially dependent parameters, which allows us to comprehend the variability of hypoxia profiles within tissues. The versatility of our framework lies in capturing diverse and dynamic behaviors of tumor oxygenation, as well as categorizing states of vascularization based on the dynamics of oxygen molecules, as identified by the model parameters. Significance. The proposed data-informed mechanistic method quantitatively assesses hypoxia in the tumor microenvironment by integrating diverse histopathological data and making predictions across different types of data. The framework provides valuable insights from both modeling and biological perspectives, advancing our comprehension of spatio-temporal dynamics of tumor oxygenation.

A multimodal multitask deep learning framework for vibrotactile feedback and sound rendering

Data-driven approaches are often utilized to model and generate vibrotactile feedback and sounds for rigid stylus-based interaction. Nevertheless, in prior research, these two modalities were typically addressed separately due to challenges related to synchronization and design complexity. To this end, we introduce a novel multimodal multitask deep learning framework. In this paper, we developed a comprehensive end-to-end data-driven system that encompasses the capture of contact acceleration signals and sound data from various texture surfaces. This framework introduces novel encoder-decoder networks for modeling and rendering vibrotactile feedback through an actuator while routing sound to headphones. The proposed encoder-decoder networks incorporate stacked transformers with convolutional layers to capture both local variability and overall trends within the data. To the best of our knowledge, this is the first attempt to apply transformer-based data-driven approach for modeling and rendering of vibrotactile signals as well as sounds during tool-surface interactions. In numerical evaluations, the proposed framework demonstrates a lower RMS error compared to state-of-the-art models for both vibrotactile signals and sound data. Additionally, subjective similarity evaluation also confirm the superiority of proposed method over state-of-the-art.

Offset-Free Koopman Model Predictive Control of Thermal Comfort Regulation for a Variable Refrigerant Flow-Dedicated Outdoor Air System-Combined System

Abstract Variable refrigerant flow (VRF) system has been an appealing solution of air conditioning for residential and commercial buildings, due to its flexibility and cost effectiveness, while lack of ventilation capability is a major drawback. Incorporation of dedicated outdoor air system (DOAS) is a typical practice. However, good coordination between DOAS and VRF is critical for achieving desired thermal comfort is challenging due to the possible complexity of mixed sensible and latent heat exchanges. In this paper, to handle the nonlinear dynamic characteristics of VRF-DOAS system, we propose an offset-free Koopman model predictive control (MPC) strategy for thermal comfort regulation, in which the MPC design is computationally more efficient due to the convex problem formulation and the use of reduced-order Koopman models, and the offset-free MPC structure enhances the robustness to model uncertainties and unmeasured disturbances. A control-oriented model is obtained by hybridizing the first-principle and data-driven modeling approach. The proposed controls strategy is evaluated with a Modelica simulation model of a VRF-DOAS system. A Dymola-Python cosimulation platform is developed via the functional mockup interface (FMI), for which the MPC algorithms are implemented in Python. Simulation results show significantly better performance of the offset-free Koopman MPC in thermal comfort regulation.

Performance prediction and operating parameters optimization for proton exchange membrane fuel cell based on data-driven surrogate model and particle swarm optimization

The operating parameters of proton exchange membrane fuel cell (PEMFC) are critical to its performance and working life. This study presents a data-driven modeling approach combining a surrogate model with the particle swarm optimization algorithm to optimize operating parameters for PEMFC and obtain the maximum power density. The results show that the operating parameters significantly influence power density under high current densities, with inlet temperature having the most significant effect. Lower inlet temperature, relative humidity in the cathode and anode, along with higher operating pressure yield improved output performance. The Genetic Algorithm-Backpropagation Neural Network based surrogate model exhibits excellent predictive performance with correlation coefficients of 0.99896 and 0.99815 for the training and test sets, respectively. Optimized conditions achieve a 3.3% increase in power density compared to initial settings, with only a 0.15% error in simulation calculations. This data-driven approach provides valuable insights for maximizing PEMFC efficiency and performance.

Constructing nonlinear data-driven models from pitching wing experiments using multisine excitation signals

Accurate modelling of unsteady nonlinear aerodynamic loads is crucial for the effective design and control of aerodynamic systems. Data-driven modelling approaches have emerged as powerful tools for capturing intricate system dynamics without prior knowledge of the underlying physical principles. This paper focuses on the construction of nonlinear data-driven models for pitching wing systems using multisine excitation signals and their relative merits compared to swept sine signals. The polynomial nonlinear state-space (PNLSS) modelling framework is used, which has been proven in the system identification community to offer flexibility and richness in representing diverse nonlinear phenomena. Data are gathered with a dedicated experimental wind tunnel setup. A NACA 0018 wing is made to oscillate in pitch according to a swept sine and a multisine with a frequency up to 2 Hz. The experiments are repeated to cover the attached flow, light stall, and deep stall regimes. PNLSS models are trained on multisines and benchmarked against models trained on swept sines to assess their comparative performance. Then, all models are successfully cross-validated using single sine data at different angles of attack. We conclude that one single data-driven model can accurately represent the lift force across different linear (attached flow) and nonlinear (light and deep stall) operating regimes.

Neural-Integrated Meshfree (NIM) Method: A differentiable programming-based hybrid solver for computational mechanics

While deep learning and data-driven modeling approaches based on deep neural networks (DNNs) have recently attracted increasing attention for solving partial differential equations, their practical applications to real-world scientific and engineering problems remain limited due to the relatively low accuracy and high computational costs. In this study, we present a differentiable programming-based hybrid meshfree approach within the field of computational mechanics. This approach seamlessly integrates neural network-based deep learning architectures with physics-based meshfree discretization techniques, thus referred to as the neural-integrated meshfree (NIM) method. To effectively approximate the solution, NIM employs a hybrid scheme called NeuroPU approximation, which combines continuous DNN representations with partition of unity (PU) basis functions associated with the underlying spatial discretization. This neural-numerical hybridization not only enhances the solution representation through functional space decomposition but also reduces both the size of DNN model and the need of automatic differentiation for spatial gradient computations, leading to a significant improvement in training efficiency. Under the NIM framework, we propose two truly meshfree solvers: the strong form-based NIM (S-NIM) and the local variational form-based NIM (V-NIM). In the S-NIM solver, the strong-form governing equation is directly considered in the loss function, while the V-NIM solver employs a local Petrov–Galerkin approach that allows the construction of variational residuals based on arbitrary overlapping subdomains. This ensures both the satisfaction of underlying physics and the preservation of the meshfree property. We perform extensive numerical experiments on both stationary and transient benchmark problems to assess the effectiveness of the proposed NIM methods in terms of accuracy, scalability, generalizability, and convergence properties. Moreover, comparative analysis with other physics-informed machine learning methods demonstrates that NIM, especially V-NIM, significantly enhances both accuracy and efficiency in end-to-end predictive capabilities.

Comprehensive Composite Mould Filling Pattern Dataset for Process Modelling and Prediction

The Resin Transfer Moulding process receives great attention from both academia and industry, owing to its superior manufacturing rate and product quality. Particularly, the progression of its mould filling stage is crucial to ensure a complete reinforcement saturation. Contemporary process simulation methods focus primarily on physics-based approaches to model the complex resin permeation phenomenon, which are computationally expensive to solve. Thus, the application of machine learning and data-driven modelling approaches is of great interest to minimise the cost of process simulation. In this study, a comprehensive dataset consisting of mould filling patterns of the Resin Transfer Moulding process at different injection locations for a composite dashboard panel case study is presented. The problem description and significance of the dataset are outlined. The distribution of this comprehensive dataset aims to lower the barriers to entry for researching machine learning approaches in composite moulding applications, while concurrently providing a standardised baseline for evaluating newly developed algorithms and models in future research works.

Data-driven prediction of aerodynamic noise of transonic buffeting over an airfoil

Accurately predicting buffet frequency and aerodynamic noise level is crucial in transonic buffet noise reduction studies. In this study, the Random Forest (RF) algorithm is employed to predict the Power Spectral Density (PSD) and Overall Sound Pressure Level (OASPL) distribution over the supercritical airfoil RAE2822. The study indicates that the RF algorithm exhibits greater advantages over the Multi-Layer Perceptron (MLP). This algorithm does not suffer from the problems of high-frequency divergence or inflection distortion, maintaining a reduction in prediction errors for OASPL and PSD by approximately three to four orders of magnitude. Additionally, this paper proposes a priori criterion for evaluating the accuracy of PSD prediction. When there is a strong correlation between PSD at adjacent points, a data-driven modeling approach can achieve higher prediction accuracy. However, when the root mean square error of the cross-correlation functions, auto-correlation functions, and the PSD between adjacent points are both high, the generalization ability of pure data-driven modeling is insufficient, necessitating the additional monitoring points to ensure prediction accuracy.

Autonomous optimization of process parameters and in-situ anomaly detection in aerosol jet printing by an integrated machine learning approach

Aerosol jet printing (AJP) is a newly-developed technology for printed electronics in additive manufacturing (AM) industry. Compared to conventional printed electronics, AJP technology has the capability of fabricating high-resolution patterns (∼ 10 µm) with multiple materials and high degrees of freedom. However, insufficiently optimizing the printing process, including reliance on traditional trial-and-error methods and theoretical models, may lead to a decline in production quality, consequently limiting the widespread adoption of AJP technology. In this paper, an integrated machine learning approach is developed for autonomous optimization of process parameters and in-situ anomaly detection in AJP. In the developed approach, a coaxial camera is incorporated with a stepwise data-driven modeling approach for optimal operating window identification and process model development of AJP, thereby systematically optimizing the main process parameters prior to printing. Subsequently, a convolutional neural network (CNN) model is developed and integrated with an alignment camera for in-situ process monitoring and printing status evaluation, enabling the anomaly detection of the AJP process. Afterwards, leveraging process similarity through transfer learning allows rapid calibration of detected system drifts or variations in process stability, thereby enhancing the product consistency of AJP. Generally, by utilizing state-of-the-art machine learning techniques, the autonomous optimization of process parameters and anomaly detection in AJP achieve accuracies of 95.3% and 92.7%, respectively. This indicates the effectiveness and systematic nature of the developed approach compared to traditional trial-and-error methods and theoretical models. Furthermore, due to its data-driven characteristics, this approach has the potential to extend its applicability to other noncontact ink writing techniques, facilitating further research for process optimization in AM.

Self-Supervised Learning and Multisource Heterogeneous Information Fusion Based Quality Anomaly Detection for Heavy-Plate Shape

Plate shape quality anomaly detection suffers from the problems of less labeled and multisource heterogeneous (MSH) data due to the complex production process with multi-process and multi-equipment. To address these problems, this paper proposes a self-supervised learning framework based on MSH contrastive learning (MSH-CL). In the proposed framework, an encoder is designed to extract the MSH information, which contains an MSH feature extract network (MSH-FEN) and an MSH feature fusion network (MSH-FFN). In MSH-FEN, a two-pathway CNN network is employed to extract the MSH feature. Then, a cross-MSH (CMSH) data fusion strategy based on the attention mechanism is designed to characterize the relationship between the MSH data in the MSH-FFN. The learning process consists of two phases, namely the self-supervised feature learning phase based on the contrastive learning (CL) and the supervised fine-tuning phase. The former phase is responsible for training the encoder with massive unlabeled data. Then, a classifier followed by the encoder is built with less labeled data in the supervised fine-tuning phase, which is applied to plate shape quality anomaly detection. Finally, the experiments are carried out on the real-world data set collected from a heavy-plate production process. The experimental results demonstrate the effectiveness and superiority of the proposed approach. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This paper focuses on the problem of shape quality anomaly detection for the heavy-plate production process, which is crucial to steel-making. However, the most quality-relevant data-driven modeling approaches suffer from the problem of requiring massive labeled data. In addition, the data generated in the practice heavy-plate production process have the characteristics of multisource heterogeneous (MSH). Motivated by the challenge of the less labeled and MSH data modeling, this paper proposed a self-supervised feature learning approach to learn a feature encoder to model the MSH data. Then, fine-tuning is used to obtain the model to detect the plate shape quality. In the feature encoder, a cross-MSH (CMSH) data fusion strategy is proposed. Finally, the effectiveness of the proposed approach is demonstrated by comparing it with other advanced algorithms. The proposed approach can be extended to other heavy-plate quality anomaly detection problems with less labeled and MSH data.

Production planning optimization framework for integrated refinery, ethylene and aromatics industrial chains considering environmental performance

Nonlinear modeling for the processing units and enterprise-wide optimization are essential to improve the overall profitability of production planning in the petrochemical enterprise. This study proposed a novel data-driven modeling approach for the ethylene plant and enterprise-wide production planning optimization framework for integrated the refinery, the ethylene plant and the aromatics complex considering environmental impact. The data-driven model was developed for different feedstocks in the ethylene plant to better predict product yield. The integrated model increased potential profitability of the petrochemical enterprise through considering the material balances of the intermediate products between the refinery, the ethylene plant and the aromatics complex. The integrated model planned utilization of dry gas resources, which reduced the waste of C2 and C3 resources and increased the production of olefin products. The enterprise-wide multi-period production planning model was constructed to optimize the planning operation of the refinery, the ethylene plant and the aromatics complex simultaneously, which was solved with the commercial optimization solver SBB. An industrial scenario was investigated to demonstrate the performance of the proposed model. Computational results for several cases indicated that integrated approach achieves a significant improvement in overall profit and provide valuable guidance for emission reduction.

Thermal resistance optimization of ultra-thin vapor chamber based on data-driven model and metaheuristic algorithm

The ultra-thin vapor chamber(UTVC) is extensively utilized across various fields due to its excellent heat dissipation performance and good temperature uniformity. The data-driven modeling approach is well suited to predict the thermal resistance of the UTVC due to its great flexibility and accuracy. In this paper, a novel approach is proposed to optimize the UTVC thermal resistance, which combines the radial basis function neural network(RBFNN) model with an improved adaptive differential fish swarm evolution algorithm(ADFEA). The mean square error of the RBFNN model was 0.00016 on the training set and 0.00027 on the test sets, which indicates that the model is able to accurately predict the thermal resistance of the UTVC. The data are obtained from experiments on a mesh wick UTVC with dimensions of 124 × 14 × 1 mm. A novel optimization algorithm, ADFEA, has been designed to enhance optimization capability and convergence accuracy. This algorithm combines differential algorithm and artificial fish swarm algorithm, incorporating a parameter adaptation mechanism. The optimal operating parameters of the UTVC are obtained by ADFEA optimization and the accuracy of the optimized results is verified by experiment. The proposed optimization method provides new insights for the design and optimization of UTVC.

Data-driven integrated study on operational performance degradation detection and recovery control of VFDR

The Variable Flow Ducted Rocket (VFDR) serves as the preferred propulsion system for hypersonic aircraft. However, it is prone to health degradation issues due to component failures, aging, and changes in operational conditions, which significantly challenge flight safety. This study introduces a data-driven integrated control approach for modeling, diagnosing, and remedying performance deterioration. Initially, a method for open-set modeling of performance degradation is developed using the Linear Parameter-Varying (LPV) modeling technique and gap metric theory, tackling the hurdles of data collection, overlapping fault classes, and modeling of unidentified faults in the VFDR system; thus, providing vital data support for detecting performance degradation. Moreover, employing the LPV model for performance degradation, a hybrid deep neural network combining multi-scale CNN and LSTM is trained to diagnose system performance issues online using multi-dimensional time-domain sequential data. Subsequently, a smooth transition control method utilizing sliding mode compensation is applied to regain control of system performance based on diagnostic outcomes and a pre-established VFDR performance recuperation controller. Simulations of three types of performance deterioration scenarios caused by changes in actuators, sensors, and the system's operational environment are conducted to demonstrate the efficiency of the proposed approach. The simulation outcomes affirm that the proposed approach is capable of effectively modeling system performance deterioration, diagnosing faults, and controlling recovery, offering promising technical support for the secure operation of intricate closed-loop control systems with stringent precision and reliability requirements.

Evolutionary modeling approach based on multiobjective genetic programming for strip quality prediction

In the iron and steel industry, hardness is one of the key indicators of strip quality in the continuous annealing production line (CAPL). However, the complex production process and the strong coupled nonlinearity between process parameters make it difficult to develop accurate mechanism models and pose a challenge for data-driven modeling approaches. More importantly, most of the data-driven learning methods lack interpretability and cannot characterize the mathematical relationship between process parameters and product quality, which in turn makes it extremely hard to understand the process mechanism. Therefore, this paper proposes an interpretable modeling approach (IMA) based on feature decomposition and ensemble to construct interpretable analytical models between process parameters and strip quality. In the IMA, a data-mechanism fusion-based feature decomposition (DM_FD) method is first applied to cope with high-dimensional input feature problems. Then, an improved multiobjective genetic programming algorithm (iMOGP) is developed to construct interpretability sub-models. Finally, a sparse optimization ensemble method (SOE) is used to integrate the sub-models to achieve interpretability and good generalization. Experimental results based on practical strip data demonstrate that the proposed IMA can cope well with high-dimensional input features and achieve model interpretability compared with commonly used machine learning methods and genetic programming (GP)-based modeling methods while ensuring better accuracy and generalization.

Gaussian process regression approach for predicting wave attenuation through rigid vegetation

Numerical modelling in the coastal environment often requires highly skilled users and can be hindered by high computation costs and time requirements. Machine Learning (ML) techniques have the potential to overcome these limitations and complement existing methods. This is an exploratory investigation utilising a Gaussian Process (GP) data-driven modelling approach that can reproduce, for the given range of conditions in this study, the results of a widely used process-based model, XBeachX, when applied to the challenging problem of wave attenuation through vegetation. This study utilises efficient sampling strategies for data exploration, providing a valuable framework for future studies. The GP model was trained on a synthetic dataset generated using the numerical model XBeachX, which was calibrated using laboratory measurements. Our findings indicate that well-trained ML models can strongly complement traditional modelling approaches, especially in an environment where data sources are increasingly available. We have also explored the underlying interactions of the GP model's input features and their relationship to the model's output through a sensitivity analysis.

Extended Dynamic Mode Decomposition with Invertible Dictionary Learning

The Koopman operator has received attention for providing a potentially global linearization representation of the nonlinear dynamical system. To estimate or control the original system, the invertibility problem is introduced into the data-driven modeling, i.e., the observables are required to be reconstructed the original system’s states. Existing methods cannot solve this problem perfectly. Only linear or nonlinear but lossy reconstruction can be achieved. This paper proposed a novel data-driven modeling approach, denoted as the Extended Dynamic Mode Decomposition with Invertible Dictionary Learning (EDMD-IDL) to address this issue, which can be interpreted as a further extension of the classical Extended Dynamic Mode Decomposition (EDMD). The Invertible Neural Network (INN) is introduced in the proposed method, where its inverse process provides the explicit inverse on the dictionary functions, thus allowing the nonlinear and lossless reconstruction. An iterative algorithm is designed to solve the extended optimization problem defined by the Koopman operator and INN by combining the optimization algorithm based on the gradient descent and the classical EDMD method, making the method successfully obtain the finite-dimensional approximation of the Koopman operator. The method is tested on various canonical nonlinear dynamical systems and is shown that the predictions obtained in a linear fashion and the ground truth match well over the long-term, where only the initial status is provided. Comparison experiments highlight the superiority of the proposed method over the other EDMD-based methods. Notably, a typical example in fluid dynamics, cylinder wake, illustrates the potential of the method to be further extended to the high-dimensional system with tens of thousands of states. By combining the Proper Orthogonal Decomposition technique, nontrivial Kármán vortex sheet phenomenon is perfectly reconstructed. Our proposed method provides a new paradigm for solving the finite-dimensional approximation of the Koopman operator and applying it to data-driven modeling.

Estimating weakening on hillslopes caused by strong earthquakes

The weakening of hillslopes during strong earthquakes increases landsliding rates in post-seismic periods. However, very few studies have addressed the amount of coseismic reduction in shear strength of hillslope materials. This makes estimation of post-seismic landslide susceptibility challenging. Here we propose a method to quantify the maximum shear-strength reduction expected on seismically disturbed hillslopes. We focus on a subset of the area affected by the 2008 Mw 7.9 Wenchuan, China earthquake. We combine physical and data-driven modeling approaches. First, we back-analyze shear-strength reduction at locations where post-seismic landslides occurred. Second, we regress the estimated shear-strength reduction against peak ground acceleration, local relief, and topographic position index to extrapolate the shear-strength reduction over the entire study area. Our results show a maximum of 60%–75% reduction in near-surface shear strength over a peak ground acceleration range of 0.5–0.9 g. Reduction percentages can be generalized using a data-driven model.

Physics-infused deep neural network for solution of non-associative Drucker–Prager elastoplastic constitutive model

In the present work, a physics-informed deep learning-based constitutive modeling approach has been introduced, for the first time, to solve non-associative Drucker–Prager elastoplastic solid governed by a linear isotropic hardening rule. A purely data-driven surrogate modeling approach for representing complex and highly non-linear elastoplastic constitutive response prevents accurate predictions due to the absence of prior physical information. To mitigate this, we design an efficient physics-constrained training approach leveraging prior physics-driven optimization procedures. It has been achieved by formulating a highly physics-augmented multi-objective loss function that includes elastoplastic constitutive relations, Drucker–Prager yield criterion, non-associative flow rule, Kuhn–Tucker consistency conditions, and various boundary conditions. Utilizing multiple densely connected independent feed-forward deep neural networks fed with high-fidelity numerical solutions in a data-driven loss function, the model obtains the accurate elastoplastic solution by minimizing the proposed loss function. The strength and robustness of the approach have been demonstrated by accurately solving the benchmark problem where a plastically deformed isotropic shallow stratum has been subjected to compressive pressure under plane strain Drucker–Prager yield condition. To optimize the performance and trainability of the model, extensive experiments on network architecture and various degrees of data-driven estimate shed light on significant improvement in terms of the accuracy of the elastoplastic solution, particularly, that exhibits sharp, or very localized features. Moreover, we propose a transfer learning-based PINNs modeling approach that elucidates the possibility of predicting solutions for different sets of applied stress and material parameters. Requiring significantly less training data, the framework can simultaneously enhance the accuracy of the solution and adaptability of training by demonstrating rapid convergence in critical loss components. The current study highlights a systematic development of a novel physics-informed deep learning approach which is quite generic in nature, yet robust and highly physics-augmented for transferability of known knowledge for vastly accelerated convergence with improved accuracy of predicting an accurate description of non-associative elastoplastic solution in the regime of continuum mechanics.