With the high percentage access of photovoltaic (PV) power generation, accurate and stable short-term PV power generation forecasting has become popular to the existing power system planning and operation. This paper proposes an ensemble learning method based on signal decomposition, deep learning, and optimization strategy for forecasting short-term PV power. At first, the original PV series is decomposed by utilizing the complete ensemble empirical mode decomposition with adaptive noises (CEEMDAN). Then, the decomposed PV series are separately allocated to different deep temporal convolutional networks (DeepTCNs) for forecasting. Finally, the multi-verse optimizer strategy based on no-negative constraint theory (NNCT) is introduced to integrate the weight coefficients of the ensemble DeepTCNs strategy and reconstruct eventual forecasting results. The case studies on real-time PV data from Alice Springs, Australia, present that the proposed method is superior to other benchmark methods in four conventional performance indexes and two statistical tests, demonstrating the validity of the proposed method in forecasting PV power.

Thanks to its excellent nonlinear representation ability, deep neural network (DNN) has been designed to aid canonical correlation analysis (CCA) for stable kernel representation. These DNN-aided CCA methods make use of correlation as the optimization objective and exploit its change to distinguish system states. However, the maximum correlation-based training strategy lacks robustness to effectively tackle system false alarms caused by changes in operating conditions. To this end, this work proposes a self-attention sliding window enhanced CCA (SaECCA) for incipient fault detection in dynamic industrial processes. The main novelties of this work include the following: first, with the aid of DNNs, an adaptive weighting mechanism with self-attention is developed to amplify the incipient fault information; second, residual estimation-based, a novel and robust optimization objective for nonlinear CCA is formulated; third, an SaECCA-based fault detection algorithm is designed, whose convergence and detectability are illustrated via theoretical analysis. Studies on a three-tank system simulation and a multiphase flow industrial process are presented to verify the effectiveness of the proposed SaECCA method.

With the increasing penetration of renewable energy sources, coal-fired units are facing growing challenges in meeting the rapid ramping demands of modern power systems. To address the limited flexibility and adaptability of coordinated ramping of coal-fired units with battery energy storage system (BESS), this work proposes a key event-triggered adaptive control strategy. The primary contributions of this work include: 1) by combining improved rain-flow algorithm with battery degradation soft measurements, a precise remaining useful life prediction model for BESS is developed, based on which an adaptive ramping factor is presented to optimize BESS scheduling; 2) designed a density peak clustering algorithm driven by heat storage coefficient to evaluate unit ramping performance and determine the BESS action boundary, which flexibly drive the BESS to quickly reach peak output and effectively mitigate the quadratic drop in ramping rate; 3) a key event-triggered adaptive control strategy is presented to accurately predict the optimal activation time of the controller and solve the corresponding optimal control problem within the prediction window. The studies on a numerical simulation and a 300-MW coal-fired unit equipped with BESS for automatic generation control variable load test are conducted to verify the effectiveness and superiority of the proposed control strategy.

With the large-scale application of electric vehicle and grid-scale energy storage systems, accurate and reliable fault detection (FD) of lithium-ion (Li-ion) battery packs is critical to the safe operation of such devices. As the key feature of battery faults, voltage correlation reflects the abnormal state of the battery by measuring the correlation change between data. Inspired by this, this brief proposes a canonical variate autoencoder for online FD in battery packs. First, the original voltages are reconstructed to obey the Gaussian distribution. Then, the canonical variable model is constructed to assess the synchronization between past and future voltages. Finally, a voltage correlation-based robust optimization objective is formulated to improve FD performance. Studies on a real lithium-ion battery experimental rig verify that the proposed method has reliable detection performance and clear physical interpretability.