We present AutoSiMP, an autonomous pipeline that transforms a natural-language structural problem description into a validated, binary topology without manual configuration. The pipeline comprises five modules: (1) an LLM-based configurator that parses a plain-English prompt into a validated specification of geometry, supports, loads, passive regions, and mesh parameters; (2) a boundary-condition generator producing solver-ready DOF arrays, force vectors, and passive-element masks; (3) a three-field SIMP solver with Heaviside projection and pluggable continuation control; (4) an eight-check structural evaluator (connectivity, compliance, grayness, volume fraction, convergence, plus three informational quality metrics); and (5) a closed-loop retry mechanism. We evaluate on three axes. Configuration accuracy: across 10 diverse problems the configurator produces valid specifications on all cases with a median compliance penalty of $+0.3\%$ versus expert ground truth. Controller comparison: on 17 benchmarks with six controllers sharing an identical sharpening tail, the LLM controller achieves the lowest median compliance but $76.5\%$ pass rate, while the deterministic schedule achieves $100\%$ pass rate at only $+1.5\%$ higher compliance. End-to-end reliability: with the schedule controller, all LLM-configured problems pass every quality check on the first attempt $-$ no retries needed. Among the systems surveyed in this work (Table 1), AutoSiMP is the first to close the full loop from natural-language problem description to validated structural topology. The complete codebase, all specifications, and an interactive web demo will be released upon journal acceptance.

翻译：我们提出AutoSiMP，一种能够将自然语言描述的结构问题自动转化为经过验证的二元拓扑构型的自主化流程，无需人工配置。该流程包含五个模块：（1）基于大语言模型的配置器，将纯英语提示解析为包含几何、支撑、载荷、被动区域和网格参数的经过验证的规范说明；（2）边界条件生成器，生成求解器就绪的自由度数组、力向量和被动单元掩码；（3）采用Heaviside投影与可插拔延拓控制的三场SIMP求解器；（4）八项结构评估器（连通性、柔度、灰度、体积分数、收敛性及三项信息质量指标）；（5）闭环重试机制。我们从三个维度进行评价。配置准确性：在10个不同问题上，配置器在所有案例中均生成有效规范，中位柔度惩罚相对专家基准为+0.3%。控制器对比：在17个基准测试中（采用共享相同锐化尾部的六种控制器），大语言模型控制器实现最低中位柔度但通过率仅76.5%，而确定性调度方案实现100%通过率且柔度仅增加+1.5%。端到端可靠性：采用调度控制器时，所有大语言模型配置的问题首次尝试即通过全部质量检查，无需重试。在本工作调研的系统中（表1），AutoSiMP是首个实现从自然语言问题描述到验证后结构拓扑完整闭环的系统。完整代码库、所有规范说明及交互式网页演示将在期刊接收后发布。