An extended policy gradient algorithm for robot task learning

In real-world robotic applications, many factors, both at low-level (e. g. , vision and motion control parameters) and at high-level (e. g. , the behaviors) determine the quality of the robot performance. Thus, for many tasks, robots require fine tuning of the parameters, in the implementation of behaviors and basic control actions, as well as in strategic decisional processes. In recent years, machine learning techniques have been used to find optimal parameter sets for different behaviors. However, a drawback of learning techniques is time consumption: in practical applications, methods designed for physical robots must be effective with small amounts of data. In this paper, we present a method for concurrent learning of best strategy and optimal parameters, by extending the policy gradient reinforcement learning algorithm. The results of our experimental work in a simulated environment and on a real robot show a very high convergence rate.

Authors

• Andrea Cherubini
• Francesca Giannone
• Luca Iocchi
• Pier Francesco Palamara

Keywords

• Robot kinematics
• Machine learning
• Cognitive robotics
• Robot vision systems
• Intelligent robots
• Motion control
• Design methodology
• Robot sensing systems
• Learning systems
• Genetic programming
• Learning Algorithms
• Learning Task
• Policy Gradient
• Robotic Tasks
• Policy Gradient Algorithm
• Convergence Rate
• Machine Learning Techniques
• Simulation Environment
• Learnable Parameters
• Small Amount Of Data
• Robotic Applications
• Real Robot
• Implementation Of Behaviors
• Robot Performance
• High Convergence Rate
• Learning Process
• Optimization Problem
• Objective Function
• Learning Outcomes
• Learning Behavior
• Robot Learning
• Choice Of Strategy
• Relevant Parameters
• Strategy Execution
• Policy Gradient Method
• Position Of The Robot
• Combination Of Behaviors
• Set Of Behaviors
• Dotted Curve