Day-Ahead Spatiotemporal Wind Speed Forecasting Based on a Hybrid Model of Quantum and Residual Long Short-Term Memory Optimized by Particle Swarm Algorithm

Fluctuations in wind speed result in intermittent wind power generation. In a power grid, wind power intermittency has serious repercussions, including poor system reliability, increased reserve capacity requirement, and increased operating costs. Wind speed must be accurately predicted to enable the day-ahead power market to schedule dispatchable generation resources and determine the market prices. This article proposes a novel hybrid model of quantum and residual long short-term memory (LSTM) optimized by particle swarm optimization (PSO) for day-ahead spatiotemporal wind speed forecasting. The hyperparameters (time series, time lag, dropout rate, and learning rate) and the structure parameter of the residual LSTM are tuned by PSO. To improve the accuracy of the proposed model, a quantum embedding layer is added to the optimized residual-LSTM neural network. According to the test results, the proposed model is highly accurate and outperforms numerous machine learning methods and deep learning algorithms.

Automated summary

This article proposes a novel hybrid model of quantum and residual long short-term memory (LSTM) optimized by particle swarm optimization (PSO) for day-ahead spatiotemporal wind speed forecasting. The proposed model outperforms numerous machine learning methods and deep learning algorithms in terms of accuracy.