Abstract
This paper introduces a new genetic fuzzy based paradigm for developing scalable set of decentralized homogenous robots for a collaborative task. In this work, the number of robots in the team can be changed without any additional training. The dynamic problem considered in this work involves multiple stationary robots that are assigned with the goal of bringing a common effector, which is physically connected to each of these robots through cables, to any arbitrary target position within the workspace of the robots. The robots do not communicate with each other. This means that each robot has no explicit knowledge of the actions of the other robots in the team. At any instant, the robots only have information related to the common effector and the target. Genetic Fuzzy System (GFS) framework is used to train controllers for the robots to achieve the common goal. The same GFS model is shared among all robots. This way, we take advantage of the homogeneity of the robots to reduce the training parameters. This also provides the capability to scale to any team size without any additional training. This paper shows the effectiveness of this methodology by testing the system on an extensive set of cases involving teams with different number of robots. Although the robots are stationary, the GFS framework presented in this paper does not put any restriction on the placement of the robots. This paper describes the scalable GFS framework and its applicability across a wide set of cases involving a variety of team sizes and robot locations. We also show results in the case of moving targets.
Highlights
We introduce a new scalable framework of Genetic Fuzzy System (GFS) for training a distributed system of robots to work collaboratively to achieve a common goal
We extend on our previous work (Sathyan and Ma, 2019) to design a scalable GFS framework for a reinforcement learning problem involving a team of stationary, homogenous robots
Regarding rulebase of the GFS, since each of the four inputs are defined using three membership functions, there will be 34 = 81 rules to represent all possible combinations of the membership functions across all inputs
Summary
We introduce a new scalable framework of Genetic Fuzzy System (GFS) for training a distributed system of robots to work collaboratively to achieve a common goal. Through the operations of crossover and mutation, the individuals in GA are modified over several generations to search for the optimal set of GFS parameters— membership functions and rules—that minimize the defined cost function This makes it a form of reinforcement learning as the robots are trained to take the optimal control actions to maximize a reward or in this case, minimize the cost function. We extend on our previous work (Sathyan and Ma, 2019) to design a scalable GFS framework for a reinforcement learning problem involving a team of stationary, homogenous robots. Each GFS is trained to model the control actions of a robot in the team at each instant, such that the series of collaborative actions taken by the robot in the team helps with bringing the common effector to the desired target position quickly. This makes the training process very efficient and provides much needed scalability to the system, as the same GFS can be used when adding more agents to the team without additional training
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
