The digital twin, a concept that aims to establish a real-time mapping between physical space and virtual space, can be used for real-time analysis, reliability assessment, predictive maintenance, and design optimization of products. This article presents an enabling technology named shape–performance integrated digital twin (SPI-DT) and takes a boom crane as an example to illustrate how to design the SPI-DT step by step for the structural analysis of complex heavy equipment. The SPI-DT contains different types of models, such as an analytical model, a numerical model, and an artificial intelligence (AI) model. In addition, it leverages multisource dynamic data obtained by placing different sensors at multiple measurement positions. In the SPI-DT, the AI model plays a central role, invoking the numerical model and sensor data as the input to predict the structural performance of key components of heavy equipment, while the analytical model analyzes the structure of noncritical components with sensor data as input. This significantly improves the computational efficiency of the digital twin used for the structural analysis of complex heavy equipment, making the digital twin computationally affordable, and thus can be used for the safety assessment and damage protection of the equipment in the operation, as well as the design optimization of next-generation products. Moreover, to visually demonstrate the models and data in the SPI-DT, a three-dimensional application used to display and record the shape and performance information in real time during the operation of the boom crane is developed.

One main challenge in tunnel constructions is to predict the tunnel geological conditions without excavation to ensure safety during the construction process. This paper proposes a data-driven framework for real-time interpreting the operating data of tunnel boring machines (TBMs) without interrupting tunneling operations, and eventually automate the tunneling operation. In this framework, we first convert the indexes of the original data from discontinuous operating time to continuous operating displacement. After screening outliers, to more exhaustively explore the inherent characteristics of TBM operating data, we then augment features by using the first-order and the second-order difference information. There are two main concerns for developing a desired geological-type predictor: 1) since multiple geological types could coexist in one tunnel section, the predictor should have multiple outputs; and 2) since the geological types are specified by the values of 7 kinds of physical-mechanical indexes of geological types, this geological characteristic should also be encoded into the predictor’s structure. Therefore, we design a feedforward multiple-output artificial neural network (ANN) with two hidden layers as the predictor, where the second hidden layer has 7 nodes that correspond to 7 kinds of physical-mechanical indexes. The experimental results show that: 1) the feature augmentation (FA) method indeed improves the prediction performance; 2) the ANN predictor has the best performance on the test set when the second hidden layer has 7 nodes; 3) the proposed ANN predictor outperforms many widely-used learning models, e.g., XGboost, random forest (RF) and support vector regression (SVR); and 4) the predictor is capable of accurately predicting the geological types of stratum.

Hybrid or ensemble surrogate models developed in recent years have shown a better accuracy compared to individual surrogate models. However, it is still challenging for hybrid surrogate models to always meet the accuracy, robustness, and efficiency requirements for many specific problems. In this paper, an advanced hybrid surrogate model, namely, extended adaptive hybrid functions (E-AHF), is developed, which consists of two major components. The first part automatically filters out the poorly performing individual models and remains the appropriate ones based on the leave-one-out (LOO) cross-validation (CV) error. The second part calculates the adaptive weight factors for each individual surrogate model based on the baseline model and the estimated mean square error in a Gaussian process prediction. A large set of numerical experiments consisting of up to 40 test problems from one dimension to 16 dimensions are used to verify the accuracy and robustness of the proposed model. The results show that both the accuracy and the robustness of E-AHF have been remarkably improved compared with the individual surrogate models and multiple benchmark hybrid surrogate models. The computational time of E-AHF has also been considerately reduced compared with other hybrid models.

Digital twin has the potential for increasing production, achieving real-time monitor, and realizing predictive maintenance by establishing a real-time high-fidelity mapping between the physical entity and its digital model. However, the high accuracy and instantaneousness requirements of digital twins have hindered their applications in practical engineering. This article presents a universal framework to fulfill the requirements and to build an accurate and trustworthy digital twin by integrating numerical simulations, sensor data, multifidelity surrogate (MFS) models, and visualization techniques. In practical engineering, the number of sensors available to measure quantities of interest is often limited, and complementary simulations are necessary to compute these quantities. The simulation results are generally more comprehensive but not as accurate as the sensor data. Therefore, the proposed framework combines the benefits of both simulation results and sensor data by using an MFS model based on moving least squares (MLS), named MLS-based multifidelity surrogate (MFS-MLS). The MFS-MLS was developed as an essential part to calibrate the continuous field of the simulation by limited sensor data to obtain accurate results for the digital twin. Then, single-fidelity surrogate models are built on the whole domain using the calibrated results of the MFS-MLS as training samples and sensor data as inputs to predict and visualize the quantities of interest in real time. In addition, the framework was validated by a truss test case, and the results demonstrate that the proposed framework has the potential to be an effective tool to build accurate and trustworthy digital twins.