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.

Integrated energy system is important for the high-efficient utilization of multitype energy systems. In this paper, the stochastic optimal operation is investigated for the micro integrated electric power, natural gas, and heat delivery system (IPGHS). First, a low-carbon micro-IPGHS is proposed with the comprehensive consideration of renewable generation, carbon-capture-based power-to-gas technology, and the combined power and heat units. Second, a scenario-based optimal operation model for micro-IPGHS is proposed to handle uncertainties in energy demand and renewable generation. In the proposed model, energy transactions between micro-IPGHS and upstream energy systems as well as constraints for battery storage, natural gas storage, and heat storage systems are considered. Finally, a case study is used for the proposed low-carbon micro-IPGHS to validate the optimal stochastic operation approach. The proposed integrated system can effectively utilize the variable clean energy for optimizing the delivery of the green operation in micro-IPGHS.

Accurate load forecasting is the key to economic dispatch and efficient operation of Multi-Energy System (MES). This paper proposes a combined load forecasting method for MES based on Bi-directional Long Short-Term Memory (BiLSTM) multi-task learning. Firstly, this paper investigates the multi-energy interaction mechanism and multi-loads characteristics and analyzes the correlation of multi-loads in different seasons. Then, a combined load forecasting method is proposed, which focuses on making full use of the coupling relationship among multiple loads. In the forecasting model, the different loads are selected combinedly as the input features according to the Maximum Information Coefficient (MIC). The multi-task learning is adopted to construct the cooling, heating and electric combined load forecasting model based on the BiLSTM algorithm, which can effectively share the coupling information among the loads. Finally, case studies verify the effectiveness and superiority of the proposed method in both learning speed and forecasting accuracy.

This paper aims to model interdependencies between power systems and dispatching data networks, and to analyze the intricate impacts on cascading failures. The functions of communication networks are embedded into dispatching data networks in China, thus we use dispatching data networks in the paper. The structures of dispatching data networks are generally categorized into two types: 1) double-star; and 2) mesh. The correlation of nodes in double-star networks and power systems is “degree to degree,” whereas “degree to betweenness” is the correlation for mesh networks. Furthermore, the interactive model between power grids and dispatching data networks is presented by a dynamic power flow model. Taking the IEEE 39-bus system and China's Guangdong 500-kV system as examples, in the case of random attacks on the interdependent system, simulation results show that the power grid coupled with double-star dispatching data networks has lower probability of catastrophic failures than with the mesh structure, because the double-star dispatching data network has outstanding capability of delivering information even though some communication nodes are out of order. In contrast, under intentional attacks, the decrement of the transmission performance of the double-star network is more serious than that in the mesh network. Therefore, the power system exhibits much higher vulnerability when coupled with the double-star network.

This paper proposes an adaptive droop control (ADC) strategy for the voltage source converter (VSC) based multiterminal high voltage direct current (VSC-MTDC) system, which enables the VSC-station to provide frequency regulation for onshore ac grids. A V-I-f characteristic is derived to establish the relationship between the frequency deviation and dc voltage. A VSC-station working with the V-I-f characteristic is able to adjust its dc voltage reference autonomously according to the grid frequency deviation. Thus, the power flow of the VSC-MTDC system will be redistributed and more power is going to be injected into the adjacent grid of which the frequency deviation exceeds the threshold value. The ADC strategy inherits the advantage of the traditional droop control (TDC) that there is no need for communication system. An ac/dc hybrid system that includes VSC-MTDC system, onshore grids, and offshore wind farms is built in DIgSILENT/Powerfactory. Modal analysis and nonlinear simulations are performed to obtain the optimal control parameters and demonstrate the performance of the ADC strategy.

With the rapid increase of renewable energy sources and power electronic converters in the power system, the inertia of the power system is declined and the stability problem is more serious than ever before. In this paper, a virtual synchronous generator (VSG) control strategy is proposed to enable the voltage-source converter-based multiterminal high-voltage direct current (VSC-MTDC) system to damp the low-frequency oscillation modes. Both rotational inertia part and governor part are modeled in the VSG controller. The V ² -P-ω characteristic is developed based on the V ² -P droop control to emulate the action of the speed governor when the frequency of ac grid deviates from the nominal value. The virtual inertia part is used to emulate the inertia response of the synchronous generator. The communication between VSC stations is unnecessary with the autonomous power allocation ability of the V ² -P-ω characteristic. Modal analysis is utilized to test the system stability and optimize the controller parameters. Two test systems which include offshore wind farms, VSC-MTDC systems, and onshore power grids are built in the DIgSIENT/PowerFactory. The simulation results verify the effectiveness of the proposed VSG strategy.