SLGNN

SLGNN 实验代码，应用于求解双足机器人运动生成问题。

Experimental code for SLGNN, applied to motion generation for bipedal robots.

实验 / Experiments

文件夹 / Folder	内容 / Description
experiment_a	直线行走 / Straight walking
experiment_b	楼梯行走 / Stair climbing
experiment_c	速度可行域扫描 / Feasible-speed scanning
experiment_d	SLGNN、GNN 与 ZNN 对比 / Comparison of SLGNN, GNN and ZNN

环境与依赖 / Environment and Dependencies

• MATLAB R2026a，已在 Windows 上验证。

• 主实验仅需基础 MATLAB；可选的符号推导脚本 to_jacobianc.m 需要 Symbolic Math Toolbox。

• MATLAB R2026a, tested on Windows.

• The main experiments require only base MATLAB. The optional symbolic derivation script to_jacobianc.m requires Symbolic Math Toolbox.

运行 / Usage

在 MATLAB 中进入所选实验文件夹，运行 main。 必要输入数据已包含，各实验请分别运行。

Open the selected experiment folder in MATLAB and run main. Required input data are included. Run each experiment separately.

cd experiment_a
main

数据与可用性 / Source Data and Availability

论文配图对应的数据、单位、来源及完整性说明见 source_data。 提供 CSV、原生图形/结果及导出、查看脚本。Fig.4b、4g、4h 仅保留矢量图， 未找到原始数值数组；Fig.7 提供论文照片拼图，不包含硬件数值日志。

Data organized by manuscript figure, units, provenance and availability limits are documented in source_data, with CSV exports, native archives and export/visualization scripts. Fig.4b, 4g and 4h have vector figure files only; their original numerical arrays are unavailable. Fig.7 includes the published photographic montage, without numerical hardware logs.

数据与代码可用性声明 / Data and code availability statements.

开源协议 / License

本项目采用 BSD-3-Clause 协议，详见 LICENSE 文件。

This project is licensed under the BSD-3-Clause License. See the LICENSE file for details.

引用 / Citation

如果本代码对您的研究有帮助，请引用相关论文。 完整引用信息将在论文发表后补充。

If this code is useful for your research, please cite the associated paper. Full citation details will be added upon publication.