This dataset corresponds to Fig. 2 of the paper and records the CPU time for computing distributed controlled invariant sets (DCISs) in the connected vehicle platooning example as the number of vehicles varies. The data are generated from a heterogeneous longitudinal platooning model, where each vehicle state consists of the incremental headway Δh and incremental velocity Δv, and the control input is the incremental driving force ΔF. The example samples vehicle parameters, linearizes the continuous-time dynamics, discretizes the model with a 0.05 s sampling period, and solves a linear-programming-based DCIS synthesis problem under prescribed safety sets, input constraints, and information structures. The dataset supports scalability and computational-complexity evaluation of distributed safety-control algorithms for sparse interconnected systems.
The data are derived from the connected vehicle platooning numerical example in the paper "Computing Distributed Controlled Invariant Sets for Interconnected Linear Systems With Information Structures." Vehicle masses and drag-related coefficients are randomly generated within the intervals specified in the paper. The equilibrium headway is 3 m and the equilibrium velocity is 20 m/s. The continuous-time model is linearized and discretized for DCIS computation. PlatoonData reports the CPU time of DCIS computation for different numbers of vehicles, with all experiments conducted on a 1.80-GHz laptop with 16 GB RAM.
The dataset is produced through simulation and optimization. Heterogeneous vehicle parameters are generated according to the intervals specified in the paper, and a linearized platooning model is constructed. The model is then discretized with a sampling period of 0.05 s, with constraints imposed on headway, velocity, and driving force. Polytopic templates are obtained through backward-reachable-set recursions, and the DCIS synthesis problem is formulated as a linear program. The LP is solved for different numbers of vehicles, and the CPU time records are used to generate the log-scale curve in PlatoonData.
The dataset consists of computational-performance records from numerical experiments rather than field observations. All data points are generated under the same hardware environment and algorithmic pipeline. Quality control includes consistency checks against the model-parameter ranges specified in the paper, feasibility checks of the LP and DCIS constraints, unified CPU-time units and vehicle-count indices, and validation that the records contain no missing, duplicated, or apparent anomalous values. The dataset is suitable for reproducing the trend shown in PlatoonData, but the absolute runtime should not be directly generalized to different hardware environments.
| # | number | name | type |
| 1 | 2022ZD0120000 | Game intelligent scene application |
This work is licensed under
CC BY 4.0 (Creative Commons Attribution 4.0 International License).
CSTR:11738.11.ncdc.db7479.2026
10.12072/ncdc.db7479.2026
