Exploring Model-based Planning with Policy Networks

2019/06/20 by Tingwu Wang, Jimmy Ba, Wang, Tingwu +1 · 11 citations
Computer Science · Engineering · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Real-time simulation and control systems #Reinforcement Learning in Robotics #Robotic Locomotion and Control #Robotics (cs.RO)

paper · pdf · doi:10.48550/arxiv.1906.08649

openalex publication_date 2019/06/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Model-based reinforcement learning (MBRL) with model-predictive control or online planning has shown great potential for locomotion control tasks in terms of both sample efficiency and asymptotic performance. Despite their initial successes, the existing planning methods search from candidate sequences randomly generated in the action space, which is inefficient in complex high-dimensional environments. In this paper, we propose a novel MBRL algorithm, model-based policy planning (POPLIN), that combines policy networks with online planning. More specifically, we formulate action planning at each time-step as an optimization problem using neural networks. We experiment with both optimization w.r.t. the action sequences initialized from the policy network, and also online optimization directly w.r.t. the parameters of the policy network. We show that POPLIN obtains state-of-the-art performance in the MuJoCo benchmarking environments, being about 3x more sample efficient than the state-of-the-art algorithms, such as PETS, TD3 and SAC. To explain the effectiveness of our algorithm, we show that the optimization surface in parameter space is smoother than in action space. Further more, we found the distilled policy network can be effectively applied without the expansive model predictive control during test time for some environments such as Cheetah. Code is released in https://github.com/WilsonWangTHU/POPLIN.

Citations

• Combining Model-Based and Model-Free Updates for Trajectory-Centric\n Reinforcement Learning
• Model-Ensemble Trust-Region Policy Optimization
• Addressing Function Approximation Error in Actor-Critic Methods
• Model-Based Value Estimation for Efficient Model-Free Reinforcement Learning
• Recurrent World Models Facilitate Policy Evolution

Cited by

• Amortising Trajectory Optimisation for Residual MPC via Implicit Contact Differentiation
• SCOPE: Language Models as One-Time Teacher for Hierarchical Planning in Text Environments
• Bootstrap Off-policy with World Model
• Control as Hybrid Inference
• On the model-based stochastic value gradient for continuous reinforcement learning
• MOPO: Model-based Offline Policy Optimization
• RT-HCP: Dealing with Inference Delays and Sample Efficiency to Learn Directly on Robotic Platforms
• First Order Model-Based RL through Decoupled Backpropagation
• Bounding Distributional Shifts in World Modeling through Novelty Detection
• Model-based Adversarial Meta-Reinforcement Learning
• Model-Predictive Control via Cross-Entropy and Gradient-Based Optimization

Related

• Evaluating model-based planning and planner amortization for continuous control
• Simplified Temporal Consistency Reinforcement Learning
• Theoretically Guaranteed Policy Improvement Distilled from Model-Based Planning
• Bootstrapped Model Predictive Control
• Flexible Locomotion Learning with Diffusion Model Predictive Control