Standardized patients (SPs) are indispensable for clinical skills training but remain expensive and difficult to scale.Although large language model (LLM)-based virtual standardized patients (VSPs) have been proposed as an alternative, their behavior remains unstable and lacks rigorous comparison with human standardized patients.We propose EasyMED, a multi-agent VSP framework that separates case-grounded information disclosure from response generation to support stable, inquiry-conditioned patient behavior.We also introduce SPBench, a human-grounded benchmark with eight expertdefined criteria for interaction-level evaluation.Experiments show that EasyMED more closely matches human SP behavior than existing VSPs, particularly in case consistency and controlled disclosure.A four-week controlled study further demonstrates learning outcomes comparable to human SP training, with stronger early gains for novice learners and improved flexibility, psychological safety, and cost efficiency.

Accurate forecasting of wind speed is vital in renewable power system management. However, wind speed series is an extremely complex system with outliers. Considering the dilemma, we propose a robust extreme learning machine algorithm where a huber loss works as the optimized function for extreme learning machine training. And a decomposition-ensemble method is developed in modelling wind speed. In our hybrid system, the proposed robust extreme learning machine is employed to model high-frequent sub-signals, while least square extreme learning machine is used to model low-frequent sub-signals. Validated by forecasting a 5-minutely wind speed in China, our proposed forecasting framework can provide more accurate predictions.