Robust Sampling Based Model Predictive Control with Sparse Objective Information

Williams, Grady; Goldfain, Brian; Drews, Paul; Saigol, Kamil; Rehg, James M.; Theodorou, Evangelos A.

doi:10.15607/rss.2018.xiv.042

Cited by 58 publications

(51 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [9], a ℒ adaptive control method is combined with MPPI to address this problem and validated in multirotor racing. The Tube-MPPI in [10] utilize tube-based model predictive framework and robustifies MPPI by combine an ancillary controller as the tracking controller of nominal MPPI controller. Still, large difference between deep neural dynamic and true environment will impact central path and deteriorate ancillary controller tracking performance.…”

Section: Model-based Rlmentioning

confidence: 99%

“…Based on the meta-learning neural dynamic model trained above, a tube-MPPI controller can be built for the guidance problem. Tube-MPPI is a variant of tube-MPC which consist of a nominal controller and an ancillary controller [10]. The nominal considers general costs and generates nominal state: the central path, while the ancillary controller tracks the actual system state in a tube centered at the central path.…”

Section: B Tube-mppi Controllermentioning

confidence: 99%

“…The nominal system can ignore system disturbances, like in [10] [31], two copies of the nominal controller are run with one from the actual state and the other one from nominal state. The mechanism accepts the MPPI solution of actual state if its cost is lower than the cost of solution from nominal state plus some threshold.…”

Section: B Tube-mppi Controllermentioning

confidence: 99%

“…Many work try to robustified the MPPI method by ensemble models [8], ℒ adaptive control [9], and so on. Tube-MPPI method is also proposed by combining an ancillary controller to keep the system states in the tube centered at nominal state computed using MPPI as nominal controller [10]. In this work, the deep neural dynamics mismatch problem in Tube-MPPI is further improved using meta-learning to adapt the deep neural dynamics to environment changes online.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Range-Aware Impact Angle Guidance Law With Deep Reinforcement Meta-Learning

et al. 2020

View full text Add to dashboard Cite

In this paper, a new guidance law is proposed for impact angle constrained missile with timevarying velocity against a maneuvering target. The proposed guidance law is based on model-based deep reinforcement learning (RL) technique where a deep neural network is trained to be a predictive model used in model predictive path integral (MPPI) control. Tube-MPPI, a robust approach utilizing ancillary controller for disturbance rejection, is introduced in guidance law design in this work to deal with the MPPI degradation of robustness when the deep predictive model differs with actual environment. To further improve the performance, meta-learning is utilized to enable the deep neural dynamics adapt to environment changes online. With this approach the model mismatch of the nominal controller is reduced to improve tube-MPPI performance. Furthermore, a range-aware hyperbolic function is proposed as an adaptive function in the MPPI performance index design. Thus, reduced initial acceleration command and increased terminal velocity benefit guidance performance. Numerical simulations under various conditions demonstrate the effectiveness of proposed guidance law. INDEX TERMS Missile guidance, tube model predictive control, meta-learning, deep reinforcement learning, impact angle constraint

show abstract

Section: Model-based Rlmentioning

confidence: 99%

Section: B Tube-mppi Controllermentioning

confidence: 99%

Section: B Tube-mppi Controllermentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Range-Aware Impact Angle Guidance Law With Deep Reinforcement Meta-Learning

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Our approach expands upon the free energy formulation of model-predictive path integral control (MPPI) [6], [7], [8], [9] by naturally encoding variations in physical parameters and structure in physics engine into online control synthesis. Any synthesized control signal is able to generalize to variations in the simulated worlds while adapting to the uncertainty to solve the task.…”

mentioning

confidence: 99%

Model-Based Generalization Under Parameter Uncertainty Using Path Integral Control

Abraham

Handa

Ratliff

et al. 2020

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Fig. 1. Environments where we test our method on model-based control for generalizing under parameter uncertainty. Top row illustrates the experimental test beds where we test object reconfiguration, opening drawers, and nonprehensile manipulation under parameter uncertainty. Bottom row illustrates a sample set of the simulated environments where we tested our method. From left to right we have the Shadow hand manipulating a dice, a half-cheetah robot performing a backflip, and the Adroit hand opening a door.Abstract-This work addresses the problem of robot interaction in complex environments where online control and adaptation is necessary. By expanding the sample space in the free energy formulation of path integral control, we derive a natural extension to the path integral control that embeds uncertainty into action and provides robustness for model-based robot planning. Our algorithm is applied to a diverse set of tasks using different robots and validate our results in simulation and real-world experiments. We further show that our method is capable of running in real-time without loss of performance. Videos of the experiments as well as additional implementation details can be found at https://sites.google.com/view/emppi .

show abstract

Connection of nonlinear model predictive controllers for smooth task switching in autonomous driving

Honda

Okuda

Suzuki

et al. 2022

Asian Journal of Control

View full text Add to dashboard Cite

Motion planning, decision making, and control are vital functions in autonomous driving for accomplishing the desired driving task while considering passenger comfort, road infrastructure, and surrounding traffic participants. Model predictive control (MPC) is a promising method for simultaneously realizing these functions. However, formulating a single MPC that can run through all driving scenarios is difficult, and previous research has often been conducted to design an MPC for a specific driving task. To extend the availability of MPC for all driving tasks, smooth switching between different MPCs designed for each driving task must be addressed. One of the difficulties in switching between MPCs is guiding the state to a feasible set of optimization problems after switching. In this paper, we present a new framework to realize the smooth connection of MPCs, that is, to reduce the optimization infeasibility at the time of MPC switching. In our proposed method, two general nonlinear MPCs with different state spaces, cost functions, constraints, and formulations can be systematically switched via automatically generated intermediate‐MPCs without requiring any particular alterations. This can help reduce the system complexity of the hybrid MPC system.

show abstract

Robust Sampling Based Model Predictive Control with Sparse Objective Information

Cited by 58 publications

References 13 publications

Range-Aware Impact Angle Guidance Law With Deep Reinforcement Meta-Learning

Range-Aware Impact Angle Guidance Law With Deep Reinforcement Meta-Learning

Model-Based Generalization Under Parameter Uncertainty Using Path Integral Control

Connection of nonlinear model predictive controllers for smooth task switching in autonomous driving

Contact Info

Product

Resources

About