To realize the synchronous coordination of economy and low-carbon performance in the development of multiple interconnected integrated energy systems (IESs), a low-carbon economic and optimal operation method for multi-energy hub interconnected IESs based on cooperative game is proposed. First, energy hubs are used to uniformly model the main bodies of each energy station, and a reward and punishment coefficient is introduced to establish a model of reward and punishment ladder-type carbon trading mechanism. Second, considering the coordinated operation between various comprehensive energy stations and data privacy, a low-carbon economic and optimal scheduling model of the interconnected system based on the concept of cooperative game is set up, the asymmetric Nash bargaining allocation strategy is adopted to the income distribution according to contribution of each power station, and the alternating direction method of multipliers is combined to solve the model. Finally, simulation results verify the effectiveness of the proposed low-carbon economic operation scheme based on cooperative game.

To address the stacks allocation problem when a single-stack fuel cell system is expanded to a multi-stack fuel cell system (MFCS), an MFCS stacks allocation method based on the minimum life-cycle cost (LCC) of stacks is proposed. First, an MFCS stacks allocation optimization model with the minimum LCC as an optimization index is established. Then, the MFCS stacks allocation results under specific application scenarios are solved, and a stacks demand power allocation strategy under the optimal stacks allocation scheme is analyzed. Finally, the advantages of the optimized MFCS stacks allocation scheme in hydrogen usage cost are analyzed. Compared with the traditional MFCS stacks average allocation scheme and the stacks demand power average allocation strategy, the proposed MFCS stacks allocation scheme based on minimum LCC and the corresponding stacks demand power allocation strategy have lower hydrogen usage cost.

Under special conditions such as rapid changes in solar irradiance, local shading or poor solar irradiance, the traditional method will cause a photovoltaic (PV) power generation system to oscillate near the maximum power point (MPP), and the tracking speed will be slow. In serious cases, the PV power generation system will exhibit severe power fluctuations, which will lead to divergence. To solve this problem, a maximum power point tracking (MPPT) control method for PV power generation systems based on reinforcement learning (RL) is proposed in this paper. The RBF neural network weight coefficient in RL is adjusted online and in real time to adapt to the output control signal, so that the active power output from PV power generation can be quickly tracked to the MPP. A PV power generation system was constructed in MATLAB/Simulink, and a comparison with the conventional RBF neural network control method was performed. Results show that the RL control can make the PV power generation system converge to the MPP stably and accurately in a shorter time. At the same time, the PV power generation system can output the maximum active power.

In recent years, the proportion of DC transmission and renewable energy power generation has increased rapidly, and the requirement that renewable energy units should participate in the primary frequency regulation of power grid has been gradually introduced into the management rules of various localities. However, the existing evaluation methods for the primary frequency regulation performance of a renewable energy station based on simulated power grid frequency disturbance tests focus on the comprehensive evaluation of primary frequency regulation performance of the station and cannot accurately characterize its initial dynamic characteristics, which are not suitable for the analysis of field measured data. Aimed at this problem, a segmented evaluation method for the primary frequency regulation performance of the station is proposed. Based on the description of the existing evaluation methods for the primary frequency regulation performance of renewable energy stations, the shortcomings of the existing evaluation methods are revealed by combining with the analysis results of field fault recording data. Then, according to the dynamic characteristics of primary frequency regulation and the control process of primary frequency regulation of the station, the measured data is analyzed and evaluated in segments to achieve an accurate characterization of primary frequency regulation performance during the whole process. Test results show that the proposed segmented evaluation method can be effectively applied to the analysis of measured data of primary frequency regulation and accurately evaluate the primary frequency regulation performance of the station, providing an alternative scheme for the real-time evaluation of primary frequency regulation performance of renewable energy stations.

To maintain the midpoint potential at the time of resonance of a three-level NPC active damper in medium-voltage distribution network, an active voltage balancing method based on second harmonic is proposed. In a three-phase three-wire system, DC voltage balancing control is generally achieved by injecting zero-sequence component into the modulation signal. However, due to the output of small active current from the active damper, the traditional method is difficult to meet the DC voltage balancing requirements when resonance occurs. Therefore, the generation mechanism of offset adjustment current is analyzed, and the balance of midpoint potential is realized by using the inner product structure of secondary component in the absolute value of the modulation signal and the secondary current. Simulation results verify the effectiveness of the proposed method in a pure reactive environment, as well as the rapidity of DC-side voltage balancing control at the moment of resonance.

A proper lightning overvoltage protection for photovoltaic (PV) arrays is crucial to their long-term stable operation. The damage caused by lightning induced overvoltage to a PV array is analyzed. The equivalent circuit of the PV array is modelled in the EMTP with the consideration of transient models of PV cells as well as the lightning overvoltage coupling characteristics of PV modules. The values of surge overvoltage across the inverter port and bypass diode and the potential differences between a DC cable and a metallic frame are calculated when lightning stroke the isolated rod, and the protective effects of surge protection device (SPD) and transient voltage suppressor (TVS) under varying waveforms of lightning current are discussed. Results show that if SPDs are only installed at the port of an inverter, induced overvoltage across the inverter port can be suppressed to some extent while the large-amplitude overvoltage may occur across the bypass diode and potential differences may occur between the DC cable and the metallic frame. When SPDs are installed on both sides of the PV string, the overvoltage of all the equipment is significantly suppressed. The amplitude of lightning induced overvoltage is obviously affected by the wave front time of lightning current, and a short wave front time may result in an increase in the amplitude of induced overvoltage. However, a risk of permanent breakdown still exists for the bypass diode in the case of a very short wave front time. It is suggested that SPDs should be installed on both sides of the PV string while adding an extra TVS in parallel with the bypass diode, so as to realize reliable protection for PV array against lightning induced overvoltage.

Thermal management is an important technology for ensuring the normal operation of submarine data centers. Aimed at the special working environment of a submarine data center, the solid oxide fuel cell (SOFC) is used as its emergency power source, and an SOFC thermal management system model with seawater as the only coolant is improved. Then, the total heat dissipation model and dynamic thermal model of SOFC are established, the fuzzy control strategy used for SOFC is optimized by a genetic algorithm, and the implementation effects of non-dominated sorting genetic algorithm-II (NSGA-II) and other algorithms are compared based on MATLAB. Finally, the temperature distribution and average current density of the power stack under different voltages are analyzed in simulation experiments. Experimental results verify that the improved thermal management method can effectively improve the heat dissipation efficiency of the SOFC system in the submarine data center and ensure its safe operation, which is helpful for improving the operation stability of the submarine data center.

An integrated energy microgrid (IEM) source-load coordinated optimal scheduling method based on master-slave game and two-stage robust optimization is proposed. First, a master-slave game model is constructed with the IEM operator as a leader and the users as followers, taking into account the demand response on the load side. Second, considering the uncertainty of wind power output, a two-stage robust game model for IEM source-load is established, in which a master-slave model between source and load is carried out at the first stage and the worst wind power output scenario is solved at the second stage. Finally, the Karush Kuhn Tucker (KKT) condition is used to transform the first-stage two-layer game problem into a single-layer linearized problem, and the overall two-stage robust optimization problem is solved by a nested column and constraint generation algorithm. Results show that the proposed two-stage robust game model can achieve an equilibrium of benefit between source and load, which effectively copes with the impact of new energy uncertainty.

The scheduling model based on a microgrid virtual distributed cloud energy storage ecosystem can solve the problem of random analysis caused by power distribution and effectively reduce the electricity consumption cost, which has distinct advantages compared with the traditional distributed energy storage system. In this paper, through the introduction of a concept of natural ecosystem to the distributed energy storage system, the operators are added to the microgrid system, forming the users under the cloud energy storage system, operators, microgrid system operators, and a dynamic energy storage system with the conventional load agent design. By dispatching the electrical energy in the microgrid and the cloud energy storage system in real-time, the electricity consumption cost in the virtual distributed cloud energy storage system is reduced, and the power load curve is optimized, thus realizing a dynamic multi-entity balance in the game model and benefits as well.

To solve the problems related to the safety and stability of regional power grids which are caused by the start-up of large-capacity motors, the mechanism and influencing factors of voltage sag caused by large-capacity motor start-up are studied. First, the dynamic process of large-capacity motor start-up and the mechanism of voltage sag are analyzed in detail, and the quantitative characterization of the severity of voltage sag is conducted using evaluation indexes. Then, the start-up of large-capacity motors in a petrochemical enterprise’s regional power grid is taken as an example, and a time-domain model based on ETAP is established to verify the theoretical analysis of voltage sag caused by large-capacity motor start-up. Finally, the influencing factors of voltage sag caused by large-capacity motor start-up are quantitatively analyzed. The influences of start-up transformer capacity and grid strength on voltage sag are studied, and a comparative analysis of voltage sag indexes for different combinations of multi-motor simultaneous start-up and different start-up modes is conducted. The research results provide a theoretical basis for reducing the risk of voltage sag accidents caused by the start-up of large-capacity motors.