[Output] LogFile = results/sweep_log_{run_id}.csv LogVariables = Time, Vehicle.Speed, Vehicle.YawRate, Vehicle.LateralAccel
Returns: - service_time (float): The time of the service event. """ service_time = current_time + expon.rvs(scale=1 / self.service_rate) self.event_queue.append((service_time, "service")) return service_time autosim
if event_type == "arrival": self.generate_service(current_time) else: # Customer has been served, generate a new arrival self.generate_arrival(current_time) [Output] LogFile = results/sweep_log_{run_id}
This approach leads to . An agent trained via Autosim can often be dropped into the real world without any additional training, having already "seen" every possible variation of the environment in its self-generated dreams. autosim