Policy Gradient Reinforcement Learning for Policy Represented by Fuzzy Rules: Application to Simulations of Speed Control of an Automobile

TL;DR

This paper introduces a fuzzy policy gradient reinforcement learning method with a smoothness constraint for automobile speed control, reducing undesirable fluctuations and improving policy quality.

Contribution

It proposes a fusion approach combining fuzzy inference with a smoothness constraint in policy gradient reinforcement learning for better control policies.

Findings

The method suppresses undesirable fluctuations in output speed.

Reward function choice affects learning stability.

The approach improves policy smoothness in speed control.

Abstract

A method of a fusion of fuzzy inference and policy gradient reinforcement learning has been proposed that directly learns, as maximizes the expected value of the reward per episode, parameters in a policy function represented by fuzzy rules with weights. A study has applied this method to a task of speed control of an automobile and has obtained correct policies, some of which control speed of the automobile appropriately but many others generate inappropriate vibration of speed. In general, the policy is not desirable that causes sudden time change or vibration in the output value, and there would be many cases where the policy giving smooth time change in the output value is desirable. In this paper, we propose a fusion method using the objective function, that introduces defuzzification with the center of gravity model weighted stochastically and a constraint term for smoothness of time change, as an improvement measure in order to suppress sudden change of the output value of the fuzzy controller. Then we show the learning rule in the fusion, and also consider the effect by reward functions on the fluctuation of the output value. As experimental results of an application of our method on speed control of an automobile, it was confirmed that the proposed method has the effect of suppressing the undesirable fluctuation in time-series of the output value. Moreover, it was also showed that the difference between reward functions might adversely affect the results of learning.