多传感器融合与融合控制试验数据集
收藏资源简介：
本数据集提供了智能线控底盘核心算法模块——“融合控制软件”的第三方性能验证测试完整数据包。数据集依据国家标准GB/T 25000.51-2016及软件需求规格，于2025年12月在中国汽车工程研究院专业测试场，通过Speedgoat实时仿真机与实车构成的硬件在环（HIL）测试环境采集生成。数据集核心内容为覆盖软件全部四项关键性能需求的7个测试用例的完整执行记录与量化结果，具体包括：1）模型纵向参数精度测试数据（半/全油门加速至60 km/h，前进车速精度>96.5%，纵向加速度精度>98.6%）；2）模型侧向/垂向参数精度测试数据（60/70 km/h单移线工况，横摆角速度、侧向加速度、侧倾角速度精度>96.3%，车身侧偏角精度>88.1%，车身侧倾角精度>91.6%）；3）路面附着估计精度测试数据（40/60 km/h对接路面，估计精度>90.9%）；4）纵横垂向协同控制能力及模式切换时间测试数据（切换时间120 ms）。数据集以结构化测试报告（.pdf）为核心，并附有原始测试记录图表、精度计算过程及符合性对照表。所有测试用例执行通过率为100%，数据经测试方（中国汽研）与委托方（清华大学）校验。本数据集可作为智能底盘域控制软件性能评估的基准，为相关算法的验收、对标、可靠性研究及行业标准制定提供可直接引用的数据。
This dataset provides the complete set of third-party performance verification test data for the core algorithm module of the intelligent steer-by-wire chassis—the "Integrated Control Software". Collected and generated in December 2025 at the professional proving ground of China Automotive Engineering Research Institute (CAERI) using a hardware-in-the-loop (HIL) test environment composed of a Speedgoat real-time simulator and a real vehicle, this dataset complies with the national standard GB/T 25000.51-2016 and the software requirements specification. The core content of the dataset consists of complete execution records and quantitative results of 7 test cases covering all four key performance requirements of the software, specifically as follows: 1) Test data for longitudinal parameter accuracy of the model: half/full-throttle acceleration to 60 km/h, with forward vehicle speed accuracy exceeding 96.5% and longitudinal acceleration accuracy exceeding 98.6%; 2) Test data for lateral/vertical parameter accuracy of the model: single-lane-change maneuvers at 60/70 km/h, with yaw rate, lateral acceleration and roll rate accuracy exceeding 96.3%, vehicle sideslip angle accuracy exceeding 88.1%, and vehicle roll angle accuracy exceeding 91.6%; 3) Test data for road surface friction estimation accuracy: split-mu road surfaces at 40/60 km/h, with estimation accuracy exceeding 90.9%; 4) Test data for longitudinal-lateral-vertical cooperative control capability and mode switching time: mode switching time of 120 ms. The dataset takes structured test reports (.pdf) as its core, and is accompanied by original test record charts, accuracy calculation procedures, and compliance comparison tables. The pass rate of all test cases is 100%, and the data has been verified by the testing party (China Automotive Engineering Research Institute, CAERI) and the client (Tsinghua University). This dataset can serve as a benchmark for performance evaluation of intelligent chassis domain control software, providing directly citable data for acceptance testing, benchmarking, reliability research and industry standard formulation of relevant algorithms.
