The incorporation of biochar (BC) as a partial substitute for cement in concrete formulations provides a promising pathway for mitigating the environmental impacts associated with carbon dioxide emissions during cement production. To achieve optimal composition of BC-enhanced concrete, multiple objectives including mechanical strength, economic factors, and embodied CO must be balanced while considering a multitude of variables constrained by complex non-linear relationships. Here, we introduced an intelligent hybrid optimization algorithm that combines Particle Swarm Optimization (PSO), Least Squares Support Vector Machine (LSSVM), and Non-dominated Sorting Genetic Algorithm II (NSGAII) to predict the performance of BC-enhanced concrete and optimize its mix proportions for multiple objectives. Our results demonstrate that the PSO optimization algorithm for searching LSSVM hyper parameters outperforms other optimization algorithms, exhibiting higher generalization performance and improved overall accuracy (R = 0.95). Furthermore, the proposed model framework effectively presents a complete Pareto front for the BC-enhanced concrete mix ratio. This triangular model system for concrete mix ratio can determine the optimal solution, tailored to the preferences of the owner's unit. Ultimately, PSO-LSSVM-NSGAII intelligent hybrid optimization algorithm enhances the efficiency of mix proportion design for BC-enhanced concrete.

中文翻译：

---

生物炭增强混凝土混合物：开创性的多目标优化

在混凝土配方中加入生物炭（BC）作为水泥的部分替代品，为减轻水泥生产过程中二氧化碳排放对环境的影响提供了一条有前途的途径。为了实现BC增强混凝土的最佳组成，必须平衡包括机械强度、经济因素和隐含二氧化碳在内的多个目标，同时考虑受复杂非线性关系约束的多个变量。在这里，我们介绍了一种智能混合优化算法，该算法结合了粒子群优化（PSO）、最小二乘支持向量机（LSSVM）和非支配排序遗传算法II（NSGAII）来预测BC增强混凝土的性能并优化其多个目标的混合比例。我们的结果表明，用于搜索 LSSVM 超参数的 PSO 优化算法优于其他优化算法，表现出更高的泛化性能和更高的整体精度（R = 0.95）。此外，所提出的模型框架有效地呈现了 BC 增强混凝土配合比的完整帕累托前沿。这个混凝土配合比三角模型系统可以根据业主单位的喜好确定最佳解决方案。最终，PSO-LSSVM-NSGAII智能混合优化算法提高了BC增强混凝土配合比设计的效率。