Enhancing Exploration in Actor-Critic Algorithms:An Approach to Incentivize Plausible Novel States