To address frequency drops under load disturbances in power systems with high renewable penetration, this paper proposes a coordinated frequency control strategy combining wind power and variable-speed pumped storage units. Firstly, a system frequency response model is developed, where sliding-mode predictive control enables both wind turbines and pumped storage units to support grid frequency regulation. Subsequently, a controlled rotor kinetic energy release scheme is introduced to manage the wind turbine’s inertial response. Furthermore, the wind turbine’s speed recovery curve is modulated by mapping the rotor kinetic energy variation to the power reference of the variable-speed pumped storage unit, establishing a coordination mechanism. This allows the fast secondary power response of the pumped storage unit to compensate the power deficit during wind turbine speed recovery, effectively mitigating secondary frequency drops that challenge conventional strategies. The proposed approach is validated via Matlab/Simulink simulations, demonstrating superior frequency regulation performance across scenarios, with eliminated secondary frequency drops and synchronous generator power oscillations.

Based on the characteristics of the two-sided operation of source and load, this paper proposes a low-carbon economic scheduling model of multi-energy complementary microgrid, which integrates electric-carbon-green card joint trading and user independent demand response. Firstly, in order to stimulate the enthusiasm of green electricity consumption, a green certificate trading model based on the absorption responsibility weight of renewable energy is established. In order to encourage the enthusiasm of users for carbon emission reduction, a comprehensive demand response mechanism is established, and quantitative indicators are formulated to measure the effect of energy use to promote the autonomy of energy use. Finally, considering carbon trading, absorption responsibility weight, green certificate trading and other factors, the economic low-carbon scheduling model to minimize total operating costs is constructed. The results of the example verify show that the model can significantly improve the economy, take into account the environmental benefits, and improve the utilization rate of renewable energy. Meanwhile, the beneficial effects of green certificate trading, carbon trading and integrated demand response mechanism are analyzed.

This paper firstly analyzes the uncertain factors driving the morphology evolution of distribution network from three aspects： policy, market and technology. Then, the system dynamics method is adopted to construct the power source structure evolution model of distribution network which considers multiple factors such as the growth rate of the installed capacity of new energy resources, further, the electric network topology evolution model is constructed with considering multiple factors such as the cost of grid structure and new energy accessing capability of distribution network topology. Finally, the relevant data of a certain province in the eastern coastal region is collected for simulation analysis. The simulation results show that the model proposed in this paper can both well reflect the proportion of new energy in the future distribution network, and predict the evolving direction of the future power grid topology.

In order to reduce the cost of coal-to-hydrogen production, efficiently utilize the resources of the industrial park, and objectively assess the comprehensive level of the park, an optimized scheduling and evaluation model of the integrated energy system of the industrial park with improved coal-to-hydrogen production is proposed. Firstly, a thermodynamic model is constructed to consider the participation of high-temperature slag, an industrial by-product of the park, in the coal-to-hydrogen process, and the hydrogen and carbon dioxide yields are determined by variable pressure adsorption and calcium oxide reaction to slag. secondly, the coal-to-hydrogen process is coupled with the hydrogen and thermal systems, and carbon dioxide is consumed by the combination of carbon capture, sequestration, and utilization, and electricity-to-gas conversion devices. Then, the tiered carbon trading mechanism is introduced to analyze the relationship between carbon emission and total cost from three perspectives of carbon trading base price, interval length and price growth rate. Finally the solution is based on the objective of minimizing the total operating cost, and the entropy weight-TOPSIS-gray correlation model is used to make a comprehensive evaluation in terms of economy, environmental protection, and equipment wear and tear. The results show that the proposed model reduces carbon emissions and operating costs by 7.4% and 31.3%, respectively, which also proves that operating costs cannot be the sole criterion for evaluating the advantages and disadvantages of the system.

In order to solve the problem that the traditional longitudinal differential protection is difficult to ensure reliable operation after IIDG being massively connected to the distribution network by T connection mode, a longitudinal protection scheme of active distribution network based on active injection of characteristic harmonics is proposed. Firstly, a characteristic harmonic current injection method based on harmonic voltage control link is proposed, and the selection of its control parameters is described. Then, the fault characteristics of high-frequency characteristic harmonic voltage and current of distribution networks are analysed. On this basis, the relationship between the calculated voltage of the characteristic harmonic and the actual measured value is derived. Finally, the fault identification criterion is constructed by using the waveform distance coefficient between the characteristic harmonic voltage and the measured value. The simulation results show that the proposed scheme can operate reliably under different fault conditions, without need to obtain the operation information of T-connected IIDG in real time. The communication pressure is small, and it is still applicable in the case of a large number of T-connected IIDGs.

To address the challenge of insufficient accuracy in traditional baseline load estimation methods under dynamic electricity consumption patterns, this study proposes an integrated approach combining the k-shape clustering algorithm with a hybrid model based on temporal convolutional network （TCN）, attention mechanism, and extreme gradient boosting （XGBoost）. Firstly, the TCN-Attention model is employed for time series forecasting to extract temporal features of load profiles. Subsequently, the k-shape clustering algorithm is applied to group user load data into clusters with high similarity in shape and trend, while the XGBoost algorithm is utilized to capture spatial characteristics of load patterns. Finally, the individual estimates from both components are integrated using an inverse error weighting method to generate the final load estimation. The proposed method is validated through real-world case studies and compared other methods. The results demonstrate the effectiveness of the proposed method.

In order to solve the problems of uncertainty and seasonal imbalance of renewable energy output, an optimal allocation method of electric, heat, gas and hydrogen integrated energy system considering stepped carbon trading and seasonal hydrogen storage was proposed. Firstly, in order to reduce the complexity of long-term planning and calculation and maintain the long-term power fluctuation characteristics, in this paper, the K-means++ method based on improved silhouette coefficient clustering method is used to perform cluster analysis of wind and solar load data. Secondly, the seasonal hydrogen storage unit is modeled, and the stepped carbon trading mechanism is introduced into the optimal allocation model, and the optimal allocation model of the park integrated energy system with seasonal hydrogen storage is established with the minimum economic cost and carbon transaction cost as the objective function. Based on the comparison of different scenarios, the impact of different carbon trading mechanisms on the integrated energy system of the park is analyzed, and finally the effectiveness of the proposed model in improving the consumption, economy and carbon emissions of renewable energy is verified.

With the massive popularity and application of electric vehicles, in order to solve the problem of coordination and interaction between integrated energy system and electric vehicles under the most economical scenario for users and to achieve optimal scheduling, this paper innovatively proposes a two-layer co-optimization model of electric vehicle-integrated energy system on the basis of analyzing the charging characteristics of electric vehicles. Firstly, the Monte Carlo method is used to calculate the charging load of common brands of electric vehicles, and the vertical analogy is made to the regional EV charging plan. Secondly, the impact of EV participation in the integrated energy system on the system economy is considered on the basis of time-sharing tariffs, and the optimal scheduling strategy is investigated. Finally, substituting the arithmetic examples for verification, the results show that after using the two-layer model innovatively proposed in this paper for the analysis of the interaction between electric vehicles and the integrated regional energy system, the total amount of power of all types of loads interacting with the network is increased, which greatly reduces the cost of the users.

In this paper, the wind and solar probability prediction models based on Copula theory are established, and a two-level capacity optimization configuration model of the main components for a GW-scale renewable power to ammonia（RePtA） systems is proposed comprehensively considering the impact of installation sites, the strategies of energy systems configuration, and system operation strategies. By the proposed models, the minimum unit ammonia cost is adopted as the objective function considering the operation constraints and strategies of different system configurations and different installation sites to optimize the capacity configurations of main components, such as wind generators, photovoltaic generators, equipment of producing hydrogen by water electrolysis, and energy storages. Taking a GW-level RePtA system as an example, the impact of different energy storage configurations at typical five types of installation sites on the optimal configuration and economic performance of the system are quantitatively analyzed and compared.

To solve the problem of strong asymmetry and time-varying harmonic voltage in active distribution networks, which makes it difficult to suppress residual harmonic currents in grounding faults, a tracking control method for active harmonic arc suppression in active distribution networks is proposed, considering the imbalance and fluctuation of harmonic voltage. By analyzing the formation mechanism of harmonic voltage in the distribution network and combining it with measured data, the imbalance characteristics of three-phase harmonic voltage in the distribution network have been confirmed. On this basis, a harmonic fault current component expression considering harmonic voltage imbalance is established, and it is demonstrated that the voltage imbalance component is the main component of harmonic fault current. In order to eliminate the influence of voltage imbalance on harmonic arc suppression, an active flexible arc suppression method considering harmonic voltage imbalance is proposed, which can achieve good compensation effects for low and high resistance grounding faults. At the same time, considering the changes in fault current caused by the time-varying harmonic voltage, a dynamic tracking control strategy for controlling arc suppression current injection is proposed to achieve real-time tracking of fault development changes and dynamic suppression of fault current. The simulation and experimental results show that the proposed method can effectively suppress harmonic fault currents in active distribution networks and achieve reliable arc suppression.

The combined non-isolated three-phase inverter consists of three single-phase inverters using a Heric topology, each of which can operate independently. To solve the problem of unbalanced output voltage of a combined non-isolated three-phase inverter during off grid operation due to asymmetric load carried by the three-phase, an improved droop control strategy based on the combined non-isolated three-phase inverter is proposed. Replace the traditional active-frequency （P-f） droop curve with the inverse square root unit （ISRU） curve of the neural network activation function, search for the active intermediate phase P_mid and the reactive intermediate phase Q_mid to improve the active-frequency （P-f） and reactive-voltage （Q-U） droop curves, thereby obtaining a three-phase balanced reference voltage. Meanwhile, introduce voltage amplitude compensation to further ensure the symmetry of the output voltage. The correctness and feasibility of the proposed control strategy were verified in Matlab/Simulink and StarSim HIL semi-physical platform. The results indicate that the strategy has good control performance and can effectively reduce the unbalance degree of output voltage.

Onder the weak cuvrent network, grid-cornnected inverters need both altive and power to support operation simultaneously. so its operating characteristic and stability issue under different power factors are particularly important. Therefore, from the perspective of system damping ratio, the small signal models of grid-connected inverter under three control methods are respectively constructed： current loop vector control based on phase-locked loop （PLL）, voltage and current double loop vector control and virtual synchronous generator （VSG） control under d-q synchronous rotating reference frame. It is found that when the current vector is a circular trajectory, its damping ratio characteristic is a non-circular trajectory. It means d-q axis present different stability feature, namely stability ‘asymmetry’ phenomenon; Based on that, it is found that the integral term, PLL and VSG angle under the three control methods have the greatest impact on the symmetry of the power stability. Besides, when power factor changes from -90°~0°~90°, the participation factor of the PLL angle increases first and then decreases, and vice versa for VSG angle. In addition, the oscillation modes related to the ‘asymmetrical’ characteristics is more under the double-loop vector control strategy and the participation of the two integral links of the current loop d and q axes on the same oscillation mode is quite different; the range of instability is wider and the phenomenon of ‘asymmetry’ is more obvious. Finally, the accuracy of the analysis is verified by Matlab simulation and experimental platform.

On the basis of studying the maximum power point tracking control strategy, a simulation and experimental model of the tthermoelectric difference power generation system are built to study the actual output power of the thermoelectric difference power generation system under the two control strategies. The results show that the system output under the hybrid constant voltage tracking method & genetic algorithms CVT&GA） is the most stable and the degree of power oscillation is the smallest in the case of significant power oscillation and high pulse voltage. The power generation efficiency of both the constant voltage method and the hybrid CVT&GA algorithm can reach more than 99.2%, and the system output under both control strategies can reach the steady state before 0.1 s.

The off-design characteristics of units in integrated energy systems （IES） leads to errors in the solution results. A low-carbon economic operation method of IES is proposed, which combines off-design characteristics of units and a ladder carbon-green certificate trading mechanism. Based on an energy hub （EH） model, a hybrid energy hub （HEH） model considering off-design characteristics of units is established to describe multi-energy coupling relationships of the IES and enhance the modeling precision of IES during different working conditions. A ladder carbon-green certificate trading mechanism is introduced to effectively couple the carbon and the green certificate market. Meanwhile, the objective function of IES is proposed by minimizing total operation cost considering the economy and low carbon emissions, and numerous simulations verify the effectiveness of the proposed model. Compared with only considering off-design characteristics of units, the total operation cost of proposed system in this paper is retuced by 11.87% under the ladder carbon-green certificate trading mechanism proposed in this paper, which can realize the economic operation of IES.

Due to the intermittent and stochastic characteristics of renewable energy power generation, large-scale grid connection can cause significant fluctuations in grid strength, which adversely affects the stability of grid-connected control systems. To address this issue, this paper establishes a system impedance model for an LCL-type three-phase grid-connected inverter, conducts an in-depth analysis of the impact of dynamic grid strength variations on the inverter’s output equivalent impedance, and reveals the result of resonance frequency migration. Based on the above theoretical analysis, this paper proposes an adaptive resonance suppression method with hybrid damping considering the short circuit ratio （SCR）. This method dynamically adjusts the weighted control coefficients according to the real-time SCR value of the power generation system, adaptively responds to system resonance frequency migration, effectively suppresses resonance, and ensures stable grid-connected currents under different grid strengths. This approach can not only operate effectively when the power grid intensity varies over a wide range, but also has strong adaptability and robustness. Simulation and hardware-in-the-loop （HIL） experiments verify that the proposed resonance suppression method exhibits superior performance in high-penetration renewable energy grid-connected systems.

Aiming at the problems of low prediction accuracy and difficult prediction caused by complex coupling relationship between multiple loads in the combination of traditional decomposition technology and artificial intelligence model, a short-term load forecasting model based on Completely Adaptive Ensemble Empirical Mode Decomposition（CEEMDAN）,Whale Optimization Algorithm（WOA）and Gated Recurrent Unit（GRU）is proposed. The Spearman correlation coefficient（SRCC）is used to select different loads as input features to mine the internal relationship of multiple loads in different seasons. The load sequence is decomposed by CEEMDAN to solve the problem of low prediction accuracy caused by too complex load sequence. The WOA algorithm is introduced to optimize the parameters in the GRU network to overcome the randomness of the parameter selection process of the GRU network. Finally,the effectiveness of the proposed model is verified by comparative simulation. The experimental results showed that compared with RNN, TCN, LSTM, and GRU, CEEMDAN-WOA-GRU reduced root mean square error and mean absolute error by 44.5%, 43.1%, 44.4%, 32.2%, and 47.2%, 46.9%, 44.1%, and 36.4%, respectively, and increased coefficient of determination by 27.4%, 34.8%, 25.7%, and 17.7%, respectively.

In order to study the impact of meteorological disasters on PV power generation projects in Inner Mongolia, based on the meteorological disaster risk assessment theory, the comprehensive risk zoning of meteorological disasters in photovoltaic power generation projects in Inner Mongolia was carried out through analytic hierarchy process （AHP） and other research methods from three aspects： Hazard factors, exposure and vulnerability of disaster recipients. A meteorological disaster risk assessment model for photovoltaic power generation projects was established. As the result the overall spatial distribution of meteorological disaster risk for PV power stations in Inner Mongolia showed a slightly decreasing trend from west to east, while the risk degree of snow cover and low temperature in the east increases. The high risk areas are mainly distributed in the central and northern of the Alxa League, the west of Bayan Nur City; in the central and southern of Xilin Gol League. The risk zoning results show that the scope of meteorological disasters in PV power generation projects in Inner Mongolia Autonomous Region is relatively concentrated and small, which is conducive to the safe operation of PV power stations. Meanwhile, during the process of solar resource development and construction, the negative impact of meteorological disasters on project construction and operation, should be considered in advance. The preliminary assessment of meteorological disasters, and the emergency plan for meterological disasters during operation, and minimize the impact of meteorological disasters.

This study systematically examines the dense gas-solid two-phase flow behavior and conjugate heat transfer characteristics in an air-sand moving bed heat exchanger for tower-type solar thermal power generation systems. A mesoscopic-scale coupled lattice Boltzmann method-discrete element method （LBM-DEM） framework is developed to characterize particle-fluid hydrodynamic interactions and associated thermal transport phenomena under high solid-phase volume fraction conditions. Numerical simulations of heat transfer in the air-sand moving bed were conducted with different mass flow ratios of the air and sand. The paper studies the changes in the temperature fields of both the gas and solid phases and compares the proportions and the trends of different heat transfer modes throughout the process. The results indicate that the dense sand which hinders heat transfer along the airflow direction significantly affects the gas phase temperature field. After a period of heating, the gas temperature distribution in the bed stabilizes with a diagonal temperature gradient. It conforms to the characteristics of staggered-flow heat transfer in air-sand moving beds. The simulation reveals a significant variation in temperature changes of sands at different positions in the bed. At the initial stage, only the sand on the left side of the bed is heated. After the temperature field stabilized, the temperature gradient of sand along the direction of airflow incidence is substantial. The main mode of heat transfer in the air-sand moving bed is convective tranfer rate. At a heating air temperature of 500 ℃, the convective tranfer rate accounts for 98.50% of the total heat transfer. The radiation tranfer rate and conduction tranfer rate are positively correlated with the incident air velocity. The air-sand flow ratio is negatively correlated with the gas phase efficiency, but positively correlated with the solid phase efficiency. In addition, the air-sand flow ratio is positively correlated with the exergy efficiency.

In order to improve the system performance of PHOTOVOLTAIC-THERMAL coupled air source heat pump （PV/T-ASHP） heating system, a solal Photovoltaic-Thermal coupled air source heat pump heating （PV/T-P-ASHP） system with an added preheating method is presented in this paper. The system model is built based on TRNSYS, and after the accuracy of the model is verified, the relevant performance enhancement of the PV/-P-ASHP system and the PV/T-ASHP system is compared and analyzed, and the applicability of the systems in different regions is explored. The results show that the average relative errors of the system model validation for PV/T electrical efficiency, thermal efficiency, and COP of the heat pump unit are 5.09%,9.06%, and 5.25%, respectively, and the experimental model and the results coincide accurately. Taking Beling as an example, it is found that the PV/T-P-ASHP system saves 4.38% of the annual operating energy consumption compared to the PV/-ASHP system, the COP of the heat pump unit is increased by 36.28%, the integrated efficiency of the PV/T is increased by 4.32%, the COP of the system is increased by 4.18% during the heating season, and the primary energy saving rate is increased by 6.47%. Simulation comparisons in different regions show that the highest system COP improvement of 4.18% is observed in Beling, A 5.78% increase in PV/T overall efficiency, a 6.58% increase in primary energy savings, and a 2.3% increase in system COP is observed in Dalian, a 4.24% increase in PV/T overall eficiency. a 5.81% increase in primary energy savings, and a 2.29% increase in system COP is observed in Shenyang, and a 1.4% increase in system COP with the addition of preheating is observed in Shanghai, with a smaller overall system performance improvement.

A new combined heating method integrating a solar heating system with the original gas boiler and other heat sources is proposed in this paper, in response to the issue that the heating system cannot satisfy the demand generated by the capacity increase of the towns in West China. The goal of the capacity matching optimization for the combination solar and gas boiler heating system is to minimize annual cost while accounting for the capacity matching issue between the two systems. The TRNSYS platform is employed for simulation, and the Hooke-Jeeves function is used for calculation. A suggested connection scheme for the combined heating system of solar energy and the existing heat source is provided through the case study calculation. The system optimization configuration methods are proposed, and the value ranges of the scale of the solar collection system and the capacity of the heat storage system are given under various conditions of capacity increase rate. In terms of system energy efficiency, the total energy consumption and the natural gas consumption of the combined heating system are 45.9% to 81.5% and 46.9% to 83.1%, respectively, lower than that of the gas boiler system.

This study used self-developed heliostat field efficiency analysis software to calculate and compare the annual cosine efficiency, shadowing and blocking efficiency, atmospheric attenuation efficiency, interception efficiency, and comprehensive efficiency of 4 types of heliostat fields： radial cornfield, radial staggered, Campo-type, and bio-mimetic type. The calculation results show that when the number and size of heliostats, tower height, and available area of heliostat field are basically equal, the shadowing and blocking loss of radial staggered layout is the smallest, and it has a significant advantage in the remote heliostat field; The annual cosine efficiency of the radial cornfield heliostat field is the highest, and it has a significant advantage in the near field heliostat field; The bio-mimetic heliostat fields have certain advantages in the mid-range due to their tendency to form higher densities; The Campo-type heliostat field has no significant advantages in all aspects.

Nanofluids are widely used in solar photothermal conversion systems to improve the efficiency of the system. The radiation characteristics of nanoparticles are the key to influence solar energy absorption. In this paper, the effects of TiO₂ nanoparticle clusters on the radiation properties of nanofluids were studied. The scattering and absorption factors of individual nanoparticles and clusters were calculated by FDTD method, and the relevant radiation characteristic parameters of nanofluids were calculated according to the independent scattering theory. The multispectral radiation properties of single-cluster and multi-cluster nanofluids are discussed. It is found that the absorption coefficient of single clusters is much higher than that of single nanoparticles without clusters when the incident wavelength is 0.40-1.00 μm. For water-based nanofluids, the clusters of nanoparticles have a great influence on the radiation properties of the nanofluids. When the incident wavelength is 0.30-0.40 μm, it is found that the different proportion of clusters have little effect on the absorption coefficient of the multi-cluster nanofluids. In addition, it is found that the albedo of the nanofluid and the number of nanoparticles contained in a single cluster changes inversely at short and long incident wavelengths.

To address the mismatch between the heat injection rate and heat diffusion rate in conventional thermo-activated walls（CTAW） during heat injection, a modular thermo-activated wall （MTAW） with specific internal channels is proposed. The effects of filler cavity size （a∶b） and thermal conductivity of the filler （λ_f） on dynamic thermal behavior, energy consumption, and thermal comfort are investigated based on a validated numerical model. Results indicate that the dynamic thermal behavior of MTAW can be significantly enhanced, with improvement correlating positively with increases in the a∶b and λ_f. Besides, the enhanced heat injection design can ensure effective radial heat transfer from the pipes to the far boundary, resulting in better continuity and stability of the invisible thermal barrier compared to CTAW. Moreover, increasing a∶b and λ_f can positively impact radiant thermal comfort and energy savings. Compared to CTAW, the energy-saving efficiency of MTAW can be enhanced by up 42.3%. The improvement effect becomes more obvious as the spacing of the embedded tubes decreases.

This study delves into the application of the backward-gazing method in heliostat surface shape detection. It addresses issues in traditional methods, such as significant fitting errors between the solar shape model and the actual solar shape, and the inapplicability of algorithms for complex, irregular heliostat reflection images. Using image processing techniques, the study analyzes the fitting errors of five solar shape models compared to the actual solar shape. Additionally, it proposes an algorithm for tilt correction and registration suitable for irregular heliostat images, based on techniques like statistical optimization and Hough transform. Experimental results indicate that the Buie model has the highest fitting accuracy and better matches the actual solar brightness distribution. Furthermore, the correction algorithm achieves a tilt correction accuracy of 0.03° and a registration accuracy within ±1 pixel. The findings of this study provide a more accurate solar shape model and heliostat images that can be directly used for surface shape calculations in backward-gazing heliostat shape detection methods.

This research develops a device that uses a Fresnel lens to focus sunlight to heat a porous media drill pipe to extract water from the soil. After proving the feasibility, the device is optimized to increase the temperature of the high-temperature area of the drill pipe and strengthen the drill pipe water permeability, and studies the effect of soil moisture content and drill pipe water permeability on water extraction capacity. The results show that under the condition of 27.5% moisture content and 152 W solar spot power, the optimized device can take 45 g of water per hour. The experimental device can operate solely on sunlight without inputting other forms of external energy, and its energy utilization efficiency is as high as 20.41%.

A high gain quasi-Z-source DC/DC converter with double-switch and coupled inductor is proposed to output high DC voltage stably in photovoltaic, fuel cell and other new energy power generation systems. Based on traditional quasi-Z-source structure, an extra switching tube and diode are added. Coupled inductance and switching capacitance unit are used to achieve high voltage gain at lower duty cycle. The converter possesses numerous advantages, including a high voltage gain, continuous input current, reduced voltage stress on switching tubes and diodes, and common input and output. The operating principle, steady-state performance and parameter design of the converter are analyzed in detail, and compared with other quasi-Z-source high-gain DC/DC converters. Ultimately, a prototype delivering 120 W is constructed to demonstrate the validity of the theoretical analysis.

In order to study the wind load characteristics of PV modules on the highway slopes, this paper analyzes and studies the wind load characteristics of PV modules on fill-cut highway slopes based on the wind tunnel test method, and reveals the effects of wind direction angle, slope angle, and module inclination angle on the wind load of PV modules on the slopes, and the results show that： 1） the wind load form factor of the PV modules on the slopes located in different positions are not the same, but they all change from large to small with the increase of the wind direction angle, the upper limit extreme value of the wind load form factor of PV modules under unfavorable wind direction angle appears in 0°-40°, and the lower limit extreme value appears in 110°-160°. Through the cloud diagram of the wind load form factor, it can be found that the components located in the lower part of the slope have similar the surface wind load distribution with the ground-mounted PV module, while the surface wind load form factor of the components in the upper part of the slope is more disordered; 2） the impact of changing the module tilt angle under the same slope but different locations on the components is different, the larger tilt angle increases the surface pressure of the module, but the increase of the two components is different, and for components A and B with the same inclination Angle at different slope angles, the load characteristics they exhibit are also different; 3） The frequency domain analysis of fluctuating wind pressure on the slope PV modules and ground-mounted PV modules reveals that the low-frequency spectra at the leading edge of the modules are relatively similar, and due to the influence of slope topography, the lower slope modules are affected by high-frequency pulsations more than that of the upper slope modules and the ground-mounted modules.

To further improve the accuracy of ultra-short-term photovoltaic power forecasting, a forecasting method based on multi-modal multi-scale features is proposed. Firstly, multi-scale historical cloud images and PV power are obtained based on different sampling intervals as the input data of the forecasting model. Secondly, the spatial and temporal features of multi-scale cloud image data and multi-scale power data are extracted using self-attention convolutional long short-term memory neural network and long short-term memory neural network respectively, to obtain multi-modal multi-scale features. Then, a fusion-attention mechanism based on multi-head self-attention and multi-head cross-attention is proposed to integrate multi-modal multi-scale feature information deeply. Finally, the multi-modal multi-scale fusion features are used as the input of multi-layer perceptron to achieve ultra-short-term photovoltaic power forecasting. Experimental results show that this method can effectively improve the accuracy of ultra-short-term photovoltaic power forecasting.

Addressing the issues of significant differences in data quality under different meteorological conditions and the high volatility in photovoltaic power which make it difficult to predict, a combined model is proposed that incorporates data augmentation （DA） and optimized deep hybrid kernel extreme learning machine （DHKELM）. At the outset, the spectral clustering algorithm is applied to categorize photovoltaic data based on varying meteorological conditions. Subsequently, the data set is expanded and its quality is enhanced by adding random noise of the same shape as the input data. Considering the numerous hyperparameters of DHKELM, a multi-strategy improved IPOA is proposed that integrates good point set initialization, golden sine update strategy, nonlinear perturbation, and adaptive perturbation of the optimal individual for hyperparameter optimization. Employing data from a photovoltaic station in Gonghe county photovoltaic park, Qinghai as a case study, the results demonstrate that the DA-IPOA-DHKELM model minimizes prediction errors under different weather and seasonal conditions and achieves fitting accuracy exceeding 90%, significantly enhancing the precision of photovoltaic power predictions and the applicability of the algorithm.

In order to optimize the lifespan of solar cell modules through thermal mechanical analysis, this paper establishes a thermal-mechanical model with fine structure, considering the glass transition of the intermediate film material at low temperatures and the tin solder plastic model. On this basis, this study investigates the thermal-mechanical stress caused by the mismatch in thermal expansion coefficients of the materials during the high-altitude environmental thermal cycling process. The study also compares the differences in tin solder joint fatigue life when using Ethylene-Vinyl Acetate（EVA）、Polyolefin Elastomer（POE） and Polyvinyl Butyral（PVB） as intermediate films. The results show that when POE is used as encapsulation material, the solder joint will have a longer fatigue life.