In a Magnetically Coupled Resonant Wireless Power Transfer (MCR-WPT) system, the magnetic coupling coil is one of the key factors that determines the system’s output power, transmission efficiency, anti-offset capability, and so on. This article proposes a coil parameter optimization method for a wireless power transfer system based on crowding distance division and adaptive genetic operators. Through optimizing the design of decision variables, such as the numbers of transmitting and receiving coil turns, the spacings between transmitting and receiving coil turns, the inner radii of the transmitting and receiving coils, and the vertical distance of the coil, the best transmission performance can be achieved. This study improves the NSGA-II algorithm through proposing a genetic operator algorithm for average crowding and high crowding populations based on adaptive operators, as well as a genetic operator algorithm for low crowding populations based on information entropy. These improved algorithms avoid problems inherent to traditional genetic operators such as fixed genetic proportions, do not easily cause the algorithm to fall into a local optimal solution, and show better convergence in the ZDT1–ZDT3 test functions. The optimization design method in this article is not only independent of commercial software such as ANSYS Maxwell 2021 R1, but can also significantly improve the calculation speed compared with traditional simulation software.

Concrete mix design increasingly requires rapid screening of mixture proportions against mechanical and resource-efficiency targets. This study develops an explainable, cement-reduction-oriented computational screening framework based on the UCI Concrete Compressive Strength dataset. The dataset records Portland cement, fly ash, blast furnace slag, aggregates, water, superplasticizer, curing age, and compressive strength, but does not report clinker factor, material-specific emission factors, or durability performance. Cement dosage is therefore minimized only as a surrogate objective; the study does not claim quantified embodied-carbon optimization. A LightGBM surrogate was trained using raw mixture variables and domain-informed ratios. On the held-out test set, the model achieved an $R^2$ of 0.946 and a mean absolute error of 2.639 MPa. Shapley Additive Explanations were used to examine the statistical influence of mixture variables, with the water-to-binder ratio emerging as the dominant predictor. Constraint-filtered Monte Carlo sampling and non-dominated sorting were then used to screen Pareto-efficient binder allocations. For a 28-day target of 45 MPa, repeated searches identified candidates with a mean Portland cement dosage of $164.4 {\mathrm{kg/m}}^3$, 44.2% below the mean of empirical mixtures in the same strength band. The evaluated numerical modules were embedded in a Streamlit prototype in which a DeepSeek large language model performs only intent parsing and report generation. The main contribution is this tool-augmented separation of language interaction from deterministic engineering computation. The resulting mixtures remain computational candidates and should next be validated experimentally and assessed using material-specific life-cycle carbon and durability data.