SIGNRL: A Population-Based Reinforcement Learning Method for Continuous Control Academic Article in Scopus