This paper proposes a digital twin framework for robust parallel control of the mobile gin pole in ultra-high voltage (UHV) transmission line construction, aiming to improve safety and operational efficiency under uncertain conditions. The new framework integrates kinetic analysis, machine learning models, and multi-objective optimization algorithms to address the challenges of heavy-lifting operations in complex terrains. The method conducts finite-element kinetic analysis based on the actual structure of the mobile gin pole. A Tyrannosaurus Rex Optimization Algorithm (TROA) is employed to enhance the performance of the Extra Randomized Trees (ET) model for predicting key parameters such as maximum axial stress and shear stress. The framework leverages the Non-Dominated Sorting Genetic Algorithm III (NSGA-III) to optimize safety and efficiency metrics by adjusting key control parameters. A digital twin system for the mobile gin pole was constructed to validate the proposed approach. Results indicate that: (1) The proposed prediction model achieved performance improvements with R2, RMSE, and MSE of 0.9642, 19.6, and 7.42, respectively. Compared with baseline machine learning models, the proposed model achieved significant improvements of 21.5%, 19.2%, and 5.1% in R2, RMSE, and MSE, respectively. (2) Experiments confirm that the proposed model maintains high prediction accuracy under noise interference and missing data scenarios, indicating strong robustness. (3) Under various operation conditions, the method reduces safety risks by up to 32.30% and improves operational efficiency by up to 42.73%. Case studies further verify the effectiveness of the proposed framework, demonstrating superior prediction accuracy, noise resistance, and computational efficiency compared to conventional control methods. The core methodological novelty of this study lies in integrating TROA, ET, NSGA-III, and digital twin technology into a unified framework for mobile gin poles. This framework adopts TROA-ET to convert finite-element-based kinetic analysis into a behavior–mechanics surrogate model. It further embeds the constructed surrogate model into an NSGA-III-driven digital twin parallel control architecture. In this way, the study contributes an integrated and computationally efficient solution for safety–efficiency co-optimization of mobile gin pole operations under uncertainty.
Chen et al. (Mon,) studied this question.