This paper explores the effects of attaching radiative cooling films at different positions of solar cells in different seasons on their thermoelectric performance through experiments and simulations. The results show that attaching a radiative cooling film on the front side can significantly reduce temperature, but due to the decrease in transmittance, the electrical performance is lost by nearly 70%; while attaching a film on the back side can reduce the average temperature by 5.08 ℃ in summer, and the power is approximately increased by 2%, maintaining a high photoelectric conversion efficiency while effectively reducing the module temperature, In terms of the comprehensive performance of thermal management and power generation efficiency, it performs even better. In summer, the film on the back side has a good cooling effect, and in a low-temperature environment with an average ambient temperature of -10 ℃ in winter, the temperature difference of the module decreases by 0.59 ℃. By reducing the temperature difference, the service life can be extended. At the same time, the radiative cooling effect is affected by irradiance, ambient temperature and wind speed. The cooling effect of attaching a film is more prominent under high temperature and high irradiance conditions, while enhancing convection will weaken its cooling effect.

To address the problem of poor voltage stability of wind-solar-energy storage DC microgrids in distributed renewable energy systems, an optimized nonlinear active disturbance rejection control strategy based on the SAC algorithm （SAC-ADRC） is proposed. Firstly, the wind-solar-energy storage system is modeled with nonlinear ADRC control. Then, the gain parameters of the nonlinear ADRC are reconstructed by using linear/nonlinear ADRC switching to improve its internal parameters which are more difficult to tune and analyze. Finally, the analysis establishes a mechanism for SAC intelligence to learn interactively with the microgrid environment, enabling the adjustment of non-linear ADRC parameters. Comparative analysis using algorithm convergence curves and simulation of various classical working conditions confirms the superiority of the SAC-ADRC control strategy in terms of interference performance. Thus, it is shown that the organic integration of nonlinear ADRC and deep reinforcement learning improves the stability of the microgrid bus voltage.

In order to study the high dimensional unsteady flow characteristics in the flow field of supercritical carbon dioxide centrifugal compressor reversed as turbine, the research object is centrifugal compressor reversed as turbine, the CFX flow field analysis software was utilized for numerical simulation, the dynamic mode decomposition（DMD） was performed on the simulated results, and the first four order modes and their corresponding spatio-temporal information are obtained. The analysis results show that the DMD method can decompose the unsteady flow field into modes with distinct energy levels and frequencies, including the basic mode with the highest energy contribution （0 Hz）, the dynamic and static interference mode with the second-highest energy （at the blade passing frequency）, and the high order harmonic characteristics of the dynamic and static interference modes with lower energy contributions（at multiples of the blade passing frequency）; The main characteristics of the basic mode are caused by the geometric parameters of the impeller, the main characteristics of the dynamic and static interference modes are caused by the dynamic and static interference between the impeller and the guide vane, and the high order harmonic characteristics of the dynamic and static interference modes show the subtle flow characteristics in the flow field; Under low-flow conditions, the first order mode exhibits a relative liquid flow angle smaller than the inlet blade angle, leading to flow separation on the pressure side of the blade inlet; Under high-flow conditions, the relative liquid flow angle of the first-order mode exceeds the inlet blade angle, resulting in flow impingement on the pressure side and flow separation on the suction side of the blade inlet. The DMD method can effectively decouple the unsteady flow field within the impeller and extract transient flow characteristics.

To achieve accurate load awareness and efficient device control, this study proposes a novel non-intrusive load decomposition （NILD） method by combining multi-modal feature selection techniques with an improved time convolutional network to enhance identification accuracy and efficiency. In this method, the load multi-scale maximum overlapping discrete wavelet transform results are introduced as input features, and the feature is combined with the mRMR algorithm to ensure the effective integration of multi-modal features. At the same time, the adaptive receptive field attention mechanism （RFAM） is added to the TCN method,improving the ability to extract features at different time scales. Finally, the algorithm is validated on datasets such as PLAID. The results show that the proposed method can achieve more accurate and efficient load decomposition with fewer features,which is higher than the regression accuracy of the traditional BiLSTM model of 5.31%,and the detection rate of new energy household equipment such as photovoltaic power has also reached over 96.6%.

When the grid dispatch center issues a reactive power command value, the wind farm needs to rationally allocate this command value to each wind turbine generator （WTG） to meet grid stability requirements. To achieve this objective, this paper proposes a loss-minimization-based reactive power optimization control strategy for onshore wind farms. An improved particle swarm optimization （PSO） algorithm is employed to optimize the reactive power reference value for each WTG, aiming to reduce total losses within the wind farm. These losses encompass generator losses, power converter losses, filter losses, transformer losses, and transmission line losses. To validate the effectiveness of the proposed strategy, a wind farm model with a 5×5 layout is established. Simulation studies compare the conventional reactive power control strategy with the proposed loss-minimization-based reactive power optimization control strategy under different scenarios. The simulation results demonstrate that the proposed optimization strategy significantly reduces power losses in the wind farm.

To realize the intelligent interconnection and interaction between the intelligent integrated energy system （IES） and the power supply system of rail transit, a synergistic optimization strategy for the hydrogen-containing IES and the power supply system of rail transit is proposed based on the dynamic ladder-type carbon trading mechanism. This strategy can reduce the cost of energy supply and utilization, while enhancing the low-carbon energy supply of the multi-flow coupling system and the low-carbon energy consumption of the rail transit system. Firstly, a dynamic ladder-type carbon trading model is designed based on the level of renewable energy and loads in IES. Secondly, according to the electric and thermal output characteristics of gas turbine （GT） and hydrogen fuel cell （HFC）, the traditional combined heat and power （CHP） unit and hydrogen energy unit are coupled by the Kalina cycle to develop a hydrogen-containing flexible energy supply unit, which can enhance the flexibility of the energy supply. Then, a hydrogen-containing IES collaborative optimization model connecting to the power supply system of rail transit is established to achieve cross-system joint optimization of rail transit energy consumption and multiple heterogeneous energy supply units, fully exploring the low-carbon economic operation potential of the system. Simulation results show that the proposed strategy can effectively balance the low-carbon economy and flexibility of multi-energy system scheduling, which can provide references for the synergistic operation of IES and power supply system of rail transit.

This paper proposes a detection method for protection mismatch risk settings tailored for high-penetration renewable energy integration scenarios. Using a typical topology of high-penetration distributed renewable energy systems, the degradation mechanism of protection performance is analyzed in depth. Furthermore, the constraints on correct protection operation are revealed by investigating the root causes of protection misoperation and failure to trip. Based on this analysis, the mapping relationship between DG output and protection mismatch risk settings is quantitatively examined. The boundary of protection failure is further determined using a dual-loop iteration method based on the backbone particle swarm optimization （PSO） algorithm. Finally, a comprehensive identification procedure for protection mismatch risk settings is developed. The effectiveness of the proposed approach is validated through simulations using the PSCAD/EMTDC and Matlab/Simulink platforms.

The cascaded H-bridge topology exhibits distinct advantages in the field of photovoltaic （PV） power generation with its modular design, smaller filtering inductance, and simplified layout. However, the three-phase cascaded H-bridge inevitably inherits the inherent issue of the single-phase full-bridge inverter, namely, double-line frequency voltage ripple on the DC-link. This issue leads to an output voltage deviation of photovoltaic array from its maximum power point, thereby reducing the power generation efficiency of system. To address this issue, this paper presents an innovative control method based on adaptive third-harmonic injection, which effectively avoids over-modulation risks under varying power factor angles. The method can calculate the optimal third-harmonic injection based on the real-time operating conditions of the system, thereby achieving optimal suppression of DC-link voltage ripple. Finally, a full-scale experimental platform is constructed, and experimental results confirm the effectiveness and feasibility of the proposed method.

We propose a Mamba-Transformer model integrating physical constraints and multi-scale features for ultra-short-term wind power forecasting. The proposed model employs complete ensemble empirical mode decomposition with adaptive noise to capture nonlinear and nonstationary patterns in wind power data across multiple frequency components. The decomposed modal functions are then fed into the Mamba model alongside meteorological data from the wind farm to uncover local time-varying properties. The self-attention mechanism of the Transformer model is then employed to capture long-range dependencies among multi-scale features. Finally, a one-dimensional Jensen wake model is integrated to establish physical constraints, integrating the prediction results of physical models and data-driven models through an adaptive weighting mechanism. Experimental results demonstrate that compared to the Transformer model, the proposed model reduces root mean square error （RMSE） and mean absolute error （MAE） by 41.29% and 50.26%, respectively. This model enhances wind power forecasting accuracy while exhibiting strong generalization capabilities.

Based on the catenary theory and the lumped mass method, a comparative analysis of the dynamic response characteristics of submarine cables under diverse laying modes is carried out. Special attention is given to the sensitivity analysis of pivotal parameters, including wave direction, laying depth, touchdown point distance, and laying speed during the dynamic laying process of submarine cables. Thereby, the influence mechanism of each key parameter in submarine cable laying is unveiled. By contrasting the maximum effective tension and minimum bending radius of submarine cables under different laying methods, the dynamic response traits of various laying methods are elaborated.