Air pollution forecasting can provide the reliable information about the future pollution situation, which is useful for an efficient operation of air pollution control and helps to plan for prevention. Dynamics of air pollution are usually reflected by various factors, such as the temperature, humidity, wind direction, wind speed, snowfall, rainfall, and so on, which increase the difficulty in understanding the change of air pollutant concentration. In this paper, a short term forecasting model based on deep learning is proposed for PM2.5 (particulate matter with an aerodynamic diameter less than or equal to 2.5 lm) concentration, and the convolutional-based bidirectional gated recurrent unit (CBGRU) method is presented, which combines 1D convnets (convolutional neural networks) and bidirectional GRU (gated recurrent unit) neural networks. The case is carried out by using the Beijing PM2.5 Data Set in UCI Machine Learning Repository. Comparing the prediction results with the traditional ones, it is proved that the error of CBGRU model is lower and the prediction performance is better.

In this paper, a flexible voltage control strategy, which takes good use of the distributed energy storage (DES) units, is proposed to enhance the voltage stability and robustness of dc distribution network. The characteristics of ac/dc interface in network are analyzed, and the virtual inertia and capacitance are given to demonstrate the interactive influence of the ac and the dc systems. The control strategy for DES which is located at the ac microgrid or at the network terminal bus is designed based on the interactive characteristics, enabling the DES to respond to both voltage variation of dc network and frequency change of utility ac grid. A cascading droop control strategy is proposed for DES in dc microgrid to relieve the pressure of voltage deterioration of dc network buses which connect the dc microgrid. The proposed comprehensive flexible control strategy for DESs at different interfaces features independence of communication as well as enhancement of system robustness, and reduces the impact of dc distribution network on utility ac grid. The performance of the proposed control strategy is validated under different operating conditions of the dc distribution network.

The access of large‐scale distributed generation (DG) easily leads to energy imbalance in distribution network. To deal with this issue, this paper proposes an energy optimal schedule method for distribution network considering the participation of source‐load‐storage aggregation groups (SAGs). Firstly, the system model consisting of distribution network layer and SAGs layer is established, and the schedule objectives and constraints of each layer are also given. Secondly, considering the fluctuation on the load side, a forecasting method based on Adaboost integrated convolutional neural networks and bidirectional long‐short term memory is proposed. Then, the improved sparrow search algorithm (ISSA) is proposed by using the tent map and Levy flight on the original sparrow search algorithm. At the same time, by introducing Pareto dominance relation and adaptive grid algorithm, the multi‐objective sparrow search algorithm (MOSSA) is derived. After that, a two‐layer optimization framework (ISSA–MOSSA) is proposed to solve the studied system. The simulation results verify the accuracy of the proposed load forecasting model, the superiority of ISSA as well as MOSSA, and the effectiveness of ISSA–MOSSA in solving the energy optimal schedule problem of the distribution system with the access of DG.