Accurately forecasting long-term future wind power is critical to achieve safe power grid integration. This problem is quite challenging due to wind power's high volatility and randomness. In this paper, we propose a novel time series forecasting method, namely Deep Conditional Generative Spatio-Temporal model (DCGST), and its high accuracy is achieved by tackling two critical issues simultaneously: a proper handling of the non-stationarity of multiple wind power time series, and a fine-grained modeling of their complicated yet dynamic spatio-temporal dependencies. Specifically, we first formally define the Spatio-Temporal Concept Drift (STCD) problem of wind power, and then we propose a novel deep conditional generative model to learn probabilistic distributions of future wind power values under STCD. Three different tailored neural networks are designed for distributions parameterization, including a graph-based prior network, an attention-based recognition network, and a stochastic seq2seq-based generation network. They are able to encode the dynamic spatio-temporal dependencies of multiple wind power time series and infer one-to-many mappings for future wind power generation. Compared to existing methods, DCGST can learn better spatio-temporal representations of wind power data and learn better uncertainties of data distribution to generate future values. Comprehensive experiments on real-world datasets including the largest public turbine-level wind power dataset verify the effectiveness, efficiency, generality and scalability of our method. Compared with state-of-the-art methods, DCGST can reduce the lone-term wind power prediction error by up to 22.8% and improve the online prediction efficiency by up to 76.8%.

Time series forecasting is a very vital research topic. The scale of time series in numerous industries has risen considerably in recent years as a result of the advancement of information technology. However, the existing algorithms pay little attention to generating large-scale time series. This article designs a state causality and adaptive covariance decomposition-based time series forecasting method (SCACD). As an observation sequence, the majority of time series is generated under the influence of hidden states. First, SCACD builds neural networks to adaptively estimate the mean and covariance matrix of latent variables; Then, SCACD employs causal convolution to forecast the distribution of future latent variables; Lastly, to avoid loss of information, SCACD applies a sampling approach based on Cholesky decomposition to generate latent variables and observation sequences. Compared to existing outstanding time series prediction models on six real datasets, the model can achieve long-term forecasting while also being lighter, and the forecasting accuracy is improved in the great majority of the prediction tasks.

How to capture dynamic spatial-temporal dependencies remains an open question in multivariate time series (MTS) forecasting. Although recent advanced spatial-temporal graph neural networks (STGNNs) achieve superior forecasting performance, they either consider pre-defined spatial correlations or simply learn static graphs. Some research has tried to learn many adjacent matrices to reveal time-varying spatial correlations, but they generate discrete graphs which cannot encode evolutionary information and also face computational complexity problem. In this article, we propose two significant plugins to help automatically learn enhanced dynamic spatial-temporal embedding of MTS data: (1) a novel neural conditional random field (CRF) layer. We find that the implicit time-varying spatial dependencies are reflected by the explicit changeable links between edges, and we propose the neural CRF to encode such pairwise changeable evolutionary inter-dependencies; (2) a structure adaptive graph convolution (SAGC) that does not require pre-defined graphs to capture semantically richer spatial correlations. Then, we integrate the neural CRF, SAGC with recurrent neural network to develop a new STGNN paradigm termed Adaptive Spatial-Temporal graph neural network with Conditional Random Field (ASTCRF), which can be trained in an end-to-end fashion. We validate the effectiveness, efficiency, and scalability of ASTCRF on five public benchmark MTS datasets.