Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

Online Adaptive Platoon Control for Connected and Automated Vehicles via Physics Enhanced Residual Learning

T0 review · 3 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper argues that platoon control accuracy improves when a small neural network learns only the residual left by a physics-based MPC, retraining online, with simulation and robot tests reporting large error reductions.

desk verdict Useful engineering paper with a sensible hybrid idea, but the headline gains over the NN baseline are unsupported until that baseline is actually defined. read the letter →

arxiv 2412.20680 v1 pith:YN7NG723 submitted 2024-12-30 cs.RO cs.SYeess.SY

Peng Zhang , Heye Huang , Hang Zhou , Haotian Shi , Keke Long , Xiaopeng Li This is my paper · ORCID
classification cs.ROcs.SYeess.SY
keywords physicsenhancedresiduallearningconnectedandautomatedvehiclescentralizedplatooncontrolonlineadaptivemodelpredictivevehicle
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Platoon control for connected and automated vehicles can be made more accurate by letting a small neural network learn, online from a short window of recent data, the residual error left by a physics-based model predictive controller, rather than replacing the physics model with learning or building a hybrid from scratch. In simulation, this physics-enhanced residual learning (PERL) reduces average cumulative absolute position and speed errors by up to 58.5% and 40.1% relative to the physics-only controller, and by 58.4% and 47.7% relative to a neural-network-only controller. On a three-robot reduced-scale platform, the paper reports even larger reductions, with position and speed cumulative errors down by 72.73% and 99.05% versus the physics model. The reason to care is that the approach promises the interpretability and safety constraints of a physics-based controller together with the adaptability of online data-driven correction, at low data cost.

What carries the argument

The load-bearing object is the residual term $R$ in the speed model $v_k^i = \alpha_i u_k^i + \beta_i + R$, learned online by a fully connected neural network with one hidden layer of 64 units and ReLU activation, trained on a short window of recent desired-speed, actual-speed, and residual data. The physics component is a centralized model predictive controller with a finite prediction horizon $N$, a quadratic cost over position, velocity, acceleration, and control-change errors, and constraints on inter-vehicle spacing, speed, and acceleration; the constants $\alpha=615.4$ and $\beta=25$ are calibrated on the robot platform. Algorithm 1 retrains the residual network every 20 simulation steps (2 seconds) or every 0.4 seconds on the robots, so the mechanism is a physics prior plus a small, frequently retrained correction term rather than a learned policy replacing the controller.

What would settle it

Run PERL on a disturbance that did not appear in its online training window, for example a step change in road slope or a motor voltage drop, and measure cumulative position error over the first ten seconds. If the online-learned residual does not reduce error relative to the physics-only controller, or if retraining on different random seeds produces widely scattered results, the generalizable-online-adaptation claim is refuted.

Watch

Extended reading notes

Core claim

The paper's central claim is that the residual between a linear physics model of vehicle speed, $$v_k^i = \alpha_i u_k^i + \beta_i + R,$$ and the actual speed is a learnable quantity, and that feeding the learned correction $R$ back into a centralized model predictive controller yields platoon tracking that is more accurate than either the physics model alone or a standalone neural-network controller. The authors position the physics model as the part that already delivers roughly ninety percent of the desired performance, with the residual learner cleaning up the remaining ten percent caused by friction, back electromotive force, motor resistance, and external disturbances. They support the claim with two simulation scenarios (an OpenACC real-world trajectory and an IDM synthetic trajectory), each under affine and quadratic control-error models, and with single-robot and three-robot platoon experiments on a circular track. Their stated reading of the results is that PERL keeps the transparency of the physics model while adding the precision and adaptability of data-driven learning, and that its online updates are what allow rapid convergence after disturbances.

Load-bearing premise

The whole result depends on the short online data window being representative: the neural network sees only a few seconds of recent data, and the paper assumes that this brief look is enough to learn a residual correction that also reduces error under the next disturbance rather than one that only fits the current conditions.

Editorial extensions

If this is right

  • A controller can retain physics-based safety constraints and still become more accurate: the residual learner only adds a correction to the MPC output, it does not override the spacing, speed, or acceleration limits.
  • The small size of the residual network (64 units, 100 training epochs) means the added computation is modest relative to solving the MPC quadratic program, so online adaptation is feasible at the reported update rates.
  • Because the benefit shows up in both simulated and physical platoons, with the physical experiments showing larger relative error reductions, the learned residual is not an artifact of the simulator's error model.
  • The largest gains are in cumulative position error, with smaller or even negative gains in maximum speed error in one scenario; the practical benefit is sustained tracking accuracy rather than smaller worst-case speed spikes.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: a natural next test is to expose PERL to disturbances absent from its online window, such as a step change in road slope or a gust, because the paper's disturbance models are affine and quadratic speed-tracking errors; if the residual network does not generalize to a different error structure, the online-adaptation claim weakens.
  • Editorial inference: the residual $R$ is claimed to capture friction, back electromotive force, and motor resistance, but the paper does not directly measure those quantities; comparing the learned $R$ to a separately identified friction model would make the interpretability claim testable.
  • Editorial inference: the same physics-plus-residual architecture could be lifted to distributed or tube-based MPC, since the correction term is agnostic to which optimizer produces the baseline control.
  • Editorial inference: on the robot platform the physical model's speed error is very large, which suggests part of PERL's apparent advantage comes from a deliberately simple physics baseline; a better-calibrated baseline would give a fairer estimate of the residual learner's marginal value.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes a Physics Enhanced Residual Learning (PERL) framework for centralized connected and automated vehicle platoon control. The controller combines a model predictive control (MPC) based on a linear vehicle dynamics model with a small feedforward neural network that learns the residual between the physical model's predicted speed and the actual speed, and updates online. The authors evaluate PERL against a physics-only MPC and an NN-based controller in four simulation tests (two reference trajectories times two actuator error models), reporting up to 58.5%/40.1% reductions in cumulative position/velocity errors versus physics and 58.4%/47.7% versus NN. They also present experiments on three Pololu Zumo robots on a circular track, reporting analogous improvements. The paper includes a GitHub link for code.

Significance. If the reported gains are robust, PERL offers a practical hybrid control design that retains the interpretability of a physics-based MPC while correcting unmodeled effects through residual learning, with modest computational overhead. The framework is simple and reproducible in its core components: the physical model, MPC formulation, network structure (64 units, ReLU, Adam, MSE), and online update schedule are stated, and a code link is provided. The results are, however, not yet fully convincing: the NN baseline is not specified, the simulation results are single-run with stochastic noise, and the physical model constants are calibrated on the same platform used for validation. These gaps limit the strength of the empirical claims and should be addressed before publication.

major comments (3)
  1. [Section 4.1 / Table 2; Section 5.3 / Table 4] The 'MPC with NN' baseline in Table 2 and the 'Online NN model' in Figs. 9-12 and Table 4 are never defined. Section 4.1 only says results are compared with 'the MPC with physical model and the PERL controller,' yet Table 2 includes an NN column and the abstract highlights reductions of 58.4%/47.7% (simulation) and 64.71%/72.58% (robots) against NN models. No architecture, training data, feature set, retraining frequency, or objective function is given for this baseline. Because the central claim is superiority over pure learning controllers, the authors must specify the NN baseline in detail (ideally an MPC-with-NN or NN policy that uses the same network structure and online update rule as PERL's residual network) and provide its performance. This issue is load-bearing and must be fixed before the comparison can be accepted.
  2. [Section 4.1 / Table 2; Section 5.3 / Table 4] The simulation results are based on a single run per scenario/error-type combination, and the error model includes zero-mean Gaussian noise. Consequently, the reported gaps (e.g., 58.5% CAE_p reduction) are point estimates with no measure of variability. The robot experiments (Table 4) similarly report one session per model. To support the quantitative claims, the authors should run multiple independent trials (e.g., different noise seeds) and report means with standard deviations or confidence intervals, and ideally a significance test for the key error metrics.
  3. [Section 3.3 / Section 5.2] The physical model constants alpha=615.4 and beta=25 are calibrated on the same Pololu Zumo platform used for the validation experiments, with beta set from RMSE between desired and actual speeds. This weakens the interpretation that the physics model provides an independent 'prior' from which PERL learns residuals: the baseline physics model is already tuned to the test platform. The authors should state explicitly whether these calibrated constants are used in the simulation, and discuss the sensitivity of the reported gains to the calibration. A comparison with an uncalibrated physics model or a first-principles parameter set would clarify the source of the improvement.
minor comments (6)
  1. [Section 4.2] The text states that 'the results are compared with the MPC with physical model and the PERL controller,' but Table 2 has three columns; please add the NN baseline to the comparison statement.
  2. [Section 5.3 / Table 4] The text reports a single-robot PERL position MSE of 0.0019, while Table 4 lists 0.0017; please reconcile the numbers.
  3. [Algorithm 1] The cost function in line 8 is garbled: 'min sum_{n=0}^{N-1} [sum_{i=1}^{I} q1^2 + q2^2 + q3^2 + q4^2]' lacks state/control variables and indices; it should reference Eq. (19). Also, 'Fast online adaptation' (line 18) is vague and should specify how the model parameters are updated.
  4. [Section 3.4] The phrase 'the action space is discretized within the vehicle's acceleration range' appears once and is not used in the experiments; either remove it or explain its role in the proposed method.
  5. [Section 4.1] Only one OpenACC trajectory and one IDM trajectory are used; given the stochastic error models, the authors should justify that these are representative and consider reporting results over multiple trajectories.
  6. [General] The acronym is written as 'PERL' throughout except in Algorithm 1's caption and line 18 where 'PELR' appears; unify to 'PERL'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the PERL derivation is a standard residual-learning construction with independently stated MPC and physical dynamics.

full rationale

The derivation chain is self-contained in the relevant sense. The MPC cost (Eqs. 19-28), state prediction (Eq. 12), and physical model (Eq. 29) are stated independently of the reported error reductions. The residual R in Eq. 29 is, by definition, the gap between the physical model output and the desired or actual speed, and the neural network in Algorithm 1 is trained online to predict that gap. Reporting that PERL reduces tracking error relative to the physical model is therefore an empirical comparison, not a quantity forced by construction. The calibration of alpha = 615.4 and beta = 25 on the robot platform weakens the physics-first framing but does not make the comparison circular, because the same calibrated physical model serves as the baseline and the residual is not the evaluation metric. The paper's own caveat in Section 4.2 that the simulations may not mirror real-world conditions is a generality limitation, not a circular step. The underspecified NN baseline in Tables 2 and 4 is a reproducibility and fair-comparison concern, not an equation that reduces to its own input. No self-citation is load-bearing: citations to the authors' prior work (e.g., H. Huang et al. 2023) motivate but do not define the PERL construction. No circular step can be exhibited with a specific equation, so the appropriate finding is no significant circularity.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim rests on calibrated constants and unspecified tuning weights rather than on a derivation from first principles, so the ledger has four free parameters. No new physical entities are introduced.

free parameters (4)
  • alpha (motor speed constant) = 615.4
    Section 3.3: treated as known from motor characteristics, but effectively calibrated to the robot platform.
  • beta (motor rotation resistance) = 25
    Section 3.3: chosen via RMSE between desired and actual speeds on the robot platform.
  • MPC cost weights q1, q2, q3, q4 = not reported
    Eq (19) defines the cost with weights q1..q4, but the values are never given; the control behavior depends on them.
  • inertial delay tau = not reported
    Eq (1) includes tau_i, and homogeneity tau_i = tau is assumed, but no numerical value is provided for simulation or robot tests.
assumptions (4)
  • domain assumption Vehicle longitudinal dynamics follow the linear state-space model in Eq (1) with constant sampling interval and inertial delay.
    Section 3.1 uses these equations to build the MPC predictor; real vehicle dynamics are more nonlinear and heterogeneous.
  • domain assumption All platoon vehicles are homogeneous (A_i = A, B_i = B, tau_i = tau).
    Section 3.1, Eq (3) onward; the robot experiments use identical Pololu robots, but the paper generalizes to CAV platoons where this will not hold.
  • ad hoc to paper A small feedforward neural network trained on short online data windows can learn a residual R that captures unmodeled dynamics (friction, back-EMF, motor resistance).
    Section 3.4 and Algorithm 1 assume this without convergence guarantees or analysis of data requirements.
  • ad hoc to paper The affine and quadratic actuator error models in Section 4.1 represent relevant real-world disturbances.
    The paper itself notes the simulations use simplified error forms that may not mirror real-world conditions accurately.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Online Adaptive Platoon Control for Connected and Automated Vehicles via Physics Enhanced Residual Learning." pith.science (2026). https://pith.science/paper/YN7NG723

@misc{pith2026241220680,
  author       = {Pith},
  title        = {Pith review of: Online Adaptive Platoon Control for Connected and Automated Vehicles via Physics Enhanced Residual Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YN7NG723}},
  note         = {Machine review of arXiv:2412.20680}
}
read the original abstract

This paper introduces a physics enhanced residual learning (PERL) framework for connected and automated vehicle (CAV) platoon control, addressing the dynamics and unpredictability inherent to platoon systems. The framework first develops a physics-based controller to model vehicle dynamics, using driving speed as input to optimize safety and efficiency. Then the residual controller, based on neural network (NN) learning, enriches the prior knowledge of the physical model and corrects residuals caused by vehicle dynamics. By integrating the physical model with data-driven online learning, the PERL framework retains the interpretability and transparency of physics-based models and enhances the adaptability and precision of data-driven learning, achieving significant improvements in computational efficiency and control accuracy in dynamic scenarios. Simulation and robot car platform tests demonstrate that PERL significantly outperforms pure physical and learning models, reducing average cumulative absolute position and speed errors by up to 58.5% and 40.1% (physical model) and 58.4% and 47.7% (NN model). The reduced-scale robot car platform tests further validate the adaptive PERL framework's superior accuracy and rapid convergence under dynamic disturbances, reducing position and speed cumulative errors by 72.73% and 99.05% (physical model) and 64.71% and 72.58% (NN model). PERL enhances platoon control performance through online parameter updates when external disturbances are detected. Results demonstrate the advanced framework's exceptional accuracy and rapid convergence capabilities, proving its effectiveness in maintaining platoon stability under diverse conditions.

Figures

Figures reproduced from arXiv: 2412.20680 by the authors.

Figure 1
Figure 1. Comparison of different model structures. (a) Fixed physics model, (b) NN-based learning model, (c) PINN-based hybrid model and (d) PERL-based adaptive model. (1) We introduce a novel PERL framework that seamlessly integrates a physical model of vehicle dynamics with a neural network-based residual learning approach. This integration allows for the correction of residuals induced by complex vehicle dynamics, enhanci… view at source ↗
Figure 2
Figure 2. The integrated CAVs platoon control framework. 3. Adaptive platoon control The PERL controller framework streamlines into two core components: (1) The foundational physical-based controller that accounts for platoon dynamics in centralized control. (2) A supplementary learning module that embeds a residual feedback mechanism into the physical model, employing NN for adaptive online correction of model inaccuracies a… view at source ↗
Figure 3
Figure 3. The control command for vehicle platoon system. Based on the Root Mean Square Error (RMSE) calculated between the desired and actual speeds, the value of β should be set to 25. Generally, the residual between the physical model and desired speed is most noticeable in the low-speed and high-speed ranges, which is represented by R. This residual can be learned and corrected using PERL, improving the accuracy of the ph… view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: The dynamic residuals between reference and predicted trajertory control in CAV platoon. 4. Simulation experiment and results In this section, we assess the performance of the proposed PERL controller model within a simulation environment by conducting a comparative an…
Figure 5
Figure 5. Figure 5: Example of the OpenACC and IDM trajectories. The parameters in the MPC constraints are set as follows: 𝑑min = 5m, 𝑑max = 8 m, 𝑎max = 5m/s 2 , 𝑎min = 5m/s 2 , 𝑣max = 5 m/s, 𝑣min = 5m/s. To account for control output errors during RPM and velocity transfer, two error mod…
Figure 6
Figure 6. Figure 6: The comparison of experimental results the three control models. The variations in velocity error across all trajectories are depicted in [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Reduced-Scale Platform and the Test track. As shown in [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Robot platoon circle trajectories for 10 laps with a target velocity. 5.3 Experiment results and analysis [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]
Figure 9
Figure 9. Figure 9: Speed and position comparison in single robort control [PITH_FULL_IMAGE:figures/full_fig_p019_9.png]
Figure 10
Figure 10. Figure 10: Speed and position comparison for robot 1 in platoon control [PITH_FULL_IMAGE:figures/full_fig_p019_10.png]
Figure 11
Figure 11. Figure 11: Speed and position comparison for robot 2 in platoon control [PITH_FULL_IMAGE:figures/full_fig_p020_11.png]
Figure 12
Figure 12. Figure 12: Speed and position comparison for robot 3 in platoon control. The comparisons in Figs. 10-12 indicate that the online PERL framework significantly outperforms both the fixed physical model and the online NN model in both speed and position control, showcasing exceptio…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Theory Foundation of Physics-Enhanced Residual Learning

    cs.LG 2025-08 reject novelty 4.0 of 10

    A set of conditional bounds shows PERL's advantages follow from assumed smaller Lipschitz constant and loss ceiling, without proving those assumptions or connecting them correctly to neural network training.

Reference graph

Works this paper leans on

31 extracted references · 15 canonical work pages · cited by 1 Pith paper

  1. [5]

    Engineering S2095809923004605

    General Optimal Trajectory Planning: Enabling Autonomous Vehicles with the Principle of Least Action. Engineering S2095809923004605. https://doi.org/10.1016/j.eng.2023.10.001 Huang, H., Wang, J., Fei, C., Zheng, X., Yang, Y., Liu, J., Wu, X., Xu, Q.,

    work page doi:10.1016/j.eng.2023.10.001 2023

  2. [7]

    CV2X-LOCA: Roadside Unit-Enabled Cooperative Localization Framework for Autonomous Vehicles

    CV2X-LOCA: Roadside Unit- Enabled Cooperative Localization Framework for Autonomous Vehicles (No. arXiv:2304.00676). arXiv. https://doi.org/10.48550/arXiv.2304.00676 Huang, Z., Sheng, Z., Ma, C., Chen, S.,

    work page Pith review arXiv doi:10.48550/arxiv.2304.00676

  3. [8]

    Presented at the 2023 23rd International Conference on Control, Automation and Systems (ICCAS), IEEE, Yeosu, Korea, Republic of, pp

    PIGD -TL: Physics -Informed Generative Dynamics with Transfer Learning, in: 2023 23rd International Conference on Control, Automation and Systems (ICCAS). Presented at the 2023 23rd International Conference on Control, Automation and Systems (ICCAS), IEEE, Yeosu, Korea, Republic of, pp. 590 –595. https://doi.org/10.23919/ICCAS59377.2023.10316820 Jiang, S....

    work page arXiv 2023

  4. [9]

    Hybrid perimeter control with real -time partitions in heterogeneous urban networks: An integration of deep learning and MPC. Transp. Res. Part C Emerg. Technol. 154, 104240. https://doi.org/10.1016/j.trc.2023.104240 Kennedy, J. M., Heinovski, J., Quevedo, D.E., Dressler, F.,

    arXiv 2023

  5. [10]

    IEEE Trans

    Centralized Model Predictive Control With Human-Driver Interaction for Platooning. IEEE Trans. Veh. Technol. 72, 12664– 12680. https://doi.org/10.1109/TVT.2023.3277451 Kianfar, R., Falcone, P., Fredriksson, J.,

    arXiv 2023

  6. [12]

    Geoenergy Sci

    A critical review of physics - informed machine learning applications in subsurface energy systems. Geoenergy Sci. Eng. 239, 212938. https://doi.org/10.1016/j.geoen.2024.212938 Li, D., Wu, J., Zhu, F., Chen, T., Wong, Y.D.,

    work page arXiv 2024

  7. [13]

    Modeling adaptive platoon and reservation-based intersection control for connected and autonomous vehicles employing deep reinforcement learning. Comput. -Aided Civ. Infrastruct. Eng. 38, 1346 –1364. https://doi.org/10.1111/mice.12956 Li, Q., Chen, Z., Li, X.,

    work page doi:10.1111/mice.12956

  8. [15]

    IEEE Trans

    Multiple -Model Switching Control of Vehicle Longitudina l Dynamics for Platoon -Level Automation. IEEE Trans. Veh. Technol. 65, 4480 –4492. https://doi.org/10.1109/TVT.2016.2541219 Li, S.E., Zheng, Y., Li, K., Wang, L. -Y., Zhang, H.,

    work page arXiv 2016

Show all 31 references
  1. [17]

    IFAC -Pap., 21st IFAC World Congress 53, 15241–15246

    Platoon control of connected autonomous vehicles: A distributed reinforcement learni ng method by consensus. IFAC -Pap., 21st IFAC World Congress 53, 15241–15246. https://doi.org/10.1016/j.ifacol.2020.12.2310 Luo, X., Chen, T., Li, M., Li, S.,

    2020 doi

  2. [19]

    An open database of car - following experiments to study the properties of commercial ACC systems

    OpenACC. An open database of car - following experiments to study the properties of commercial ACC systems. Transp. Res. Part C Emerg. Technol. 125, 103047. https://doi.org/10.1016/j.trc.2021.103047 Mao, X., Li, P., Weng, Z., Zhao, J.,

    2021

  3. [20]

    Distributed tube model predictive control for string stability of heterogeneous vehicle platoons. Proc. Inst. Mech. Eng. Part J. Syst. Control Eng. 237, 095965182311737. https://doi.org/10.1177/09596518231173763 Mohd Zaihidee, F., Mekhilef, S., Mubin, M.,

    doi

  4. [22]

    Neural-Fly enables rapid learning for agile flight in strong winds. Sci. Robot. 7, eabm6597. https://doi.org/10.1126/scirobotics.abm6597 Rubió-Massegú, J., Palacios -Quiñonero, F., Rossell, J.M., Karimi, H.R.,

    doi

  5. [25]

    IEEE Trans

    A Distributed Platoon Control Framework for Connected Automated Vehicles in an Urban Traffic Network. IEEE Trans. Control Netw. Syst. 9, 1717 –1730. https://doi.org/10.1109/TCNS.2022.3181522 Wang, H., Yuan, Y., Yang, X.T., Zhao, T., Liu, Y.,

    2022

  6. [27]

    Learning -based modeling of human -autonomous vehicle interaction for improved safety in mixed -vehicle platooning control. Transp. Res. Part C Emerg. Technol. 162, 104600. https://doi.org/10.1016/j.trc.2024.104600 Xu, Y., Shi, Y., Tong, X., Chen, S., Ge, Y.,

    2024

  7. [28]

    IEEE Trans

    A Multi -agent Reinforcement Learning Based Control Method for Connected and Autonomous Vehicles in A Mixed Platoon. IEEE Trans. Veh. Technol. 1–14. https://doi.org/10.1109/TVT.2024.3415660 Yan, R., Jiang, R., Jia, B., Huang, J., Yang, D.,

    2024

  8. [29]

    IEEE Trans

    Hybrid Car-Following Strategy Based on Deep Deterministic Policy Gradient and Cooperative Adaptive Cruise Control. IEEE Trans. Autom. Sci. Eng. 19, 2816–2824. https://doi.org/10.1109/TASE.2021.3100709 25 Zhang, H., Du, L.,

    2021

  9. [30]

    Platoon-centered control for eco-driving at signalized intersection built upon hybrid MPC system, online learning and distributed optimization part I: Modeling and solution algorithm design. Transp. Res. Part B Methodol. 172, 174 –198. https://doi.org/10.1016/j.trb.2023.02.006...

    doi

  10. [31]

    Impact of CAV platoon management on traffic flow considering degradation of control mode. Phys. Stat. Mech. Its Appl. 581, 126193. https://doi.org/10.1016/j.physa.2021.126193

    2021

  11. [334]

    https://doi.org/10.1080/15472450.2021.2023016 Wang, J., Pant, Y.V., Jiang, Z.,

    2021

  12. [1669]

    https://doi.org/10.3390/en12091669 O’Connell, M., Shi, G., Shi, X., Azizzadenesheli, K., Anandkumar, A., Yue, Y., Chung, S. -J.,

    doi

  13. [2000]

    Congested traffic states in empirical observations and microscopic simulations. Phys. Rev. E 62, 1805 –1824. https://doi.org/10.1103/PhysRevE.62.1805 Wang, B., Su, R.,

    doi

  14. [2006]

    Presented at the 2006 9th International Conference on Control, Automatio n, Robotics and Vision, IEEE, Singapore, pp

    Multi-Agent Reinforcement Learning: A Survey, in: 2006 9th International Conference on Control, Automation, Robotics and Vision. Presented at the 2006 9th International Conference on Control, Automatio n, Robotics and Vision, IEEE, Singapore, pp. 1–6. https://doi.org/10.1109/I...

    2006

  15. [2013]

    Static Output - Feedback Control for Vehicle Suspensions: A Single-Step Linear Matrix Inequality Approach. Math. Probl. Eng. 2013, e907056. https://doi.org/10.1155/2013/907056 Shi, H., Chen, D., Zheng, N., Wang, X., Zhou, Y., Ran, B., 2023a. A deep reinforcement learning based...

    2013

  16. [2015]

    Control Eng

    A control matching model predictive control approach to string stable vehicle platooning. Control Eng. Pract. 45, 163 –173. https://doi.org/10.1016/j.conengprac.2015.09.011 Latrach, A., Malki, M.L., Morales, M., Mehana, M., Rabiei, M.,

    2015 doi

  17. [2016]

    Presented at the 2016 20th International Conference on System Theory, Control and Computing (ICSTCC), IEEE, Sinaia, Romania, pp

    Distributed model predictive control for vehicle platooning: A brief survey, in: 2016 20t h International Conference on System Theory, Control and Computing (ICSTCC). Presented at the 2016 20th International Conference on System Theory, Control and Computing (ICSTCC), IEEE, Si...

    2016

  18. [2018]

    IEEE Trans

    Platoon Control of Connected Vehicles from a Networked Control Perspective : Literature Review, Component Modeling, and Controller Synthesis. IEEE Trans. Veh. Technol. 1 –1. https://doi.org/10.1109/TVT.2017.2723881 Liu, B., Ding, Z., Lv, C.,

    2017

  19. [2020]

    A probabilistic risk assessment framework considering lane -changing behavior interaction. Sci. CHINA Inf. Sci. 63, 190203. https://doi.org/10.1007/s11432-019-2983-0 23 Huang, Z., Chen, S., Pian, Y., Sheng, Z., Ahn, S., Noyce, D.A.,

    doi

  20. [2021]

    Presented at 24 the 2021 40th Chinese Control Conference (CCC), IEEE, Shanghai, China, pp

    Platoon Control of Automatic Vehicles Based on Deep Deterministic Policy Gradient, in: 2021 40th Chinese Control Conference (CCC). Presented at 24 the 2021 40th Chinese Control Conference (CCC), IEEE, Shanghai, China, pp. 6154 –6159. https://doi.org/10.23919/CCC52363.2021.9549...

    2021

  21. [2022]

    IEEE Trans

    A Review of Connecte d and Automated Vehicle Platoon Merging and Splitting Operations. IEEE Trans. Intell. Transp. Syst. 23, 22790 –22806. https://doi.org/10.1109/TITS.2022.3193278 Li, S.E., Gao, F., Cao, D., Li, K.,

    2022

  22. [2023]

    IEEE Trans

    Distributed Coordinated Brake Control for Longitudinal Collision Avoidance of Multiple Connected Automated Vehicles. IEEE Trans. Intell. Veh. 8, 745–755. https://doi.org/10.1109/TIV.2022.3197951 Huang, H., Liu, Y., Liu, J., Yang, Q., Wang, J., Abbink, D., Zgonnikov, A.,

    2022

  23. [2024]

    DRL based platooning control with traffic signal synchronization for delay and fuel optimization. Transp. Res. Part C Emerg. Technol. 163, 104655. https://doi.org/10.1016/j.trc.2024.104655 Hu, M., Li, J., Bian, Y., Wang, J., Xu, B., Zhu, Y.,

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.