A Novel Twin Support Vector Regression Model for Wind Speed Time-Series Interval Prediction

Although the machine-learning model demonstrates high accuracy in wind speed prediction, it struggles to accurately depict fluctuation range of predicted values due inherent uncertainty sequences. To address this limitation and enhance reliability, we propose an effective interval prediction that combines twin support vector regression (TSVR), variational mode decomposition (VMD), slime mould algorithm (SMA). In our methodology, complex series is decomposed into multiple relatively stable subsequences using VMD method. The principal component residual are then subject TSVR model, while remaining components undergo point prediction. SMA method employed search for optimal parameter combinations. obtained by aggregating forecasting results all models each subseries. Our proposed has demonstrated superior performance various applications. It ensures value falls within designated achieving narrowest interval. For instance, spring dataset with 1-period, a intervals coverage probability (PICP) 0.9791 normalized width (PINRW) 0.0641. This outperforms other comparative significantly enhances its practical application value. After adding reliability improved. As result, study presents novel as valuable approach multi-step