Time-Optimal nonlinear model predictive control with minimal control interventions

Rösmann, Christoph; Makarow, Artemi; Hoffmann, Frank; Bertram, Torsten

doi:10.1109/ccta.2017.8062434

Cited by 16 publications

(12 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For each planner, we conduct 30 test runs on each scenario. Finally, we compare our joint systems against conventional ones-TEB [19] and MPC [20]-in terms of safety, robustness and efficiency. For STH-WP, the time and spatial horizon is set to t lim = 4s and d ahead = 1.55m respectively.…”

Section: Results and Evaluationmentioning

confidence: 99%

“…After evaluating our approaches against one another, we compare the navigation behavior of our joint navigation systems against model-based navigation systems MPC [20] and TEB [19]. Our joint systems includes the following: our proposed waypoint generators as intermediate planner, the kino A-star global planner, and a DRL-based local planner of our previous work [21] against the model-based navigation systems MPC [20] and TEB [19]. Throughout this paper, the notion previously used to denote the waypoint generators are used to denote the joint systems.…”

Section: B Comparison With Conventional Navigation Systemsmentioning

confidence: 99%

See 1 more Smart Citation

Connecting Deep-Reinforcement-Learning-based Obstacle Avoidance with Conventional Global Planners using Waypoint Generators

Kästner

Zhao

Buiyan

et al. 2021

2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

Deep Reinforcement Learning has emerged as an efficient dynamic obstacle avoidance method in highly dynamic environments. It has the potential to replace overly conservative or inefficient navigation approaches. However, the integration of Deep Reinforcement Learning into existing navigation systems is still an open frontier due to the myopic nature of Deep-Reinforcement-Learning-based navigation, which hinders its widespread integration into current navigation systems. In this paper, we propose the concept of an intermediate planner to interconnect novel Deep-Reinforcement-Learning-based obstacle avoidance with conventional global planning methods using waypoint generation. Therefore, we integrate different waypoint generators into existing navigation systems and compare the joint system against traditional ones. We found an increased performance in terms of safety, efficiency and path smoothness especially in highly dynamic environments.

show abstract

Section: Results and Evaluationmentioning

confidence: 99%

Section: B Comparison With Conventional Navigation Systemsmentioning

confidence: 99%

Connecting Deep-Reinforcement-Learning-based Obstacle Avoidance with Conventional Global Planners using Waypoint Generators

Kästner

Zhao

Buiyan

et al. 2021

2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

View full text Add to dashboard Cite

show abstract

“…Whereas navigation in static environments can be solved by using traditional approaches such as A-Star or RRT, dynamic obstacle avoidance is still an open frontier. Traditional methods regard the obstacle avoidance task as an optimization control problem [10], [11], [12]. These approaches require high computational calculations and can not cope well with fast-moving obstacles.…”

Section: Related Workmentioning

confidence: 99%

“…with that of DRL-based approaches namely obstacle avoidance, our interface is able to integrate a variety of both planning paradigms. In total, the interface provides the planners MPC [11], TEB [10], DWA [12] as model-based planners, and CADRL [5], RLCA [6] and DRL [9] of our previous work as learning-based planners. The interface can be extended to include more planners.…”

Section: A System Designmentioning

confidence: 99%

See 1 more Smart Citation

All-in-One: A DRL-based Control Switch Combining State-of-the-art Navigation Planners

Kästner¹,

Cox²,

Buiyan³

et al. 2021

Preprint

View full text Add to dashboard Cite

Autonomous navigation of mobile robots is an essential aspect in use cases such as delivery, assistance or logistics. Although traditional planning methods are well integrated into existing navigation systems, they struggle in highly dynamic environments. On the other hand, Deep-Reinforcement-Learningbased methods show superior performance in dynamic obstacle avoidance but are not suitable for long-range navigation and struggle with local minima. In this paper, we propose a Deep-Reinforcement-Learning-based control switch, which has the ability to select between different planning paradigms based solely on sensor data observations. Therefore, we develop an interface to efficiently operate multiple model-based, as well as learning-based local planners and integrate a variety of state-of-the-art planners to be selected by the control switch. Subsequently, we evaluate our approach against each planner individually and found improvements in navigation performance especially for highly dynamic scenarios. Our planner was able to prefer learning-based approaches in situations with a high number of obstacles while relying on the traditional modelbased planners in long corridors or empty spaces.

show abstract

Approximate time optimal control by deep neural networks trained with numerically obtained optimal trajectories

Zauner,

Gattringer,

Müller

2023

Proc Appl Math and Mech

View full text Add to dashboard Cite

This paper focuses on online time optimal control of nonlinear systems. This is achieved by approximating the results of time optimal control problems (TOCP) with deep neural networks (DNN) depending on the initial and terminal system state. In general, solving a TOCP for nonlinear systems is a computationally challenging task. Especially in the context of time optimal nonlinear model predictive control (TMPC) with hard real time constraints successful termination of a TOCP within sample times suitable for controlling mechanical systems cannot be guaranteed. Therefore, our approach is to train three DNNs with different aspects of numerical solutions of TOCPs with random initial and terminal state. These networks can then be used to approximate the TMPC by a one step model predictive control scheme with a significantly simpler structure and decreased calculation time. In order to verify this procedure by simulation, it is applied to a prove of concept example as well as the model of an industrial robot.

show abstract

Time-Optimal nonlinear model predictive control with minimal control interventions

Cited by 16 publications

References 17 publications

Connecting Deep-Reinforcement-Learning-based Obstacle Avoidance with Conventional Global Planners using Waypoint Generators

Connecting Deep-Reinforcement-Learning-based Obstacle Avoidance with Conventional Global Planners using Waypoint Generators

All-in-One: A DRL-based Control Switch Combining State-of-the-art Navigation Planners

Approximate time optimal control by deep neural networks trained with numerically obtained optimal trajectories

Contact Info

Product

Resources

About