The iron and steel industry faces challenges in recycling solid waste and reducing carbon emissions. This study introduces a composite agglomeration process for producing pre-reduced sinter, enabling the use of carbon-bearing dust and preparing a low-carbon blast furnace burden. Investigations into the high-temperature mineralization mechanism show that controlling CaO levels in both the raw sintering matrix and the pellets enables liquid-phase bonding and the matrix's encapsulation of the pellets. This results in a dense shell and a high-porosity core. This structure not only prevents re-oxidation of the internally pre-reduced pellets but also improves the mechanical strength of the sinter. Semi-industrial tests demonstrate enhanced sintering performance and 4.49 kg t(-1) in fuel savings, with sintering and ironmaking processes saving 22.93 kg of coke and reducing CO2 emissions by 71.46 kg per ton of sinter. With low equipment costs and simple operation, this approach shows strong potential for industrial application.

Efficient cooling of electronic devices is critical for ensuring stable performance, extending service life, and enhancing reliability. This capability is essential for advancing high-power-density and integration technologies, mitigating thermal failure risks, and improving energy efficiency. However, existing miniaturized ionic wind radiators face limitations in both research depth and spatial coupling between corona wind modes and heat load distributions. To address these challenges, this study integrates corona discharge with a heat sink, proposing a novel coreless monolithic corona wind heat sink (CM-CWHS) for electronic cooling. An unsteady electric fieldflow coupling model is developed to characterize complex inter-fin flow dynamics. Key parameters-including operating voltage, discharge distance, emitter longitudinal height, collector width, and fixing plate positionare analyzed via a response surface model. Through multiple regression, a predictive model for the flow enhancement coefficient is established, quantifying the relationship between structural parameters and cooling performance. A four-factor, three-level Box-Behnken experimental design reveals the relative contributions of each parameter to flow enhancement. A composite main-auxiliary fin structure (30 degrees main fin angle, 15-mm auxiliary fin length) is proposed, achieving a flow enhancement coefficient of 247.25 x 10-6 m3/J and a Nusselt number of 46.44-representing improvements of 37.2 % and 46.6 %, respectively, over the baseline design. Flow distribution analysis demonstrates that optimized wall pressure gradients reduce ineffective ionization regions, resulting in a centrosymmetric divergent corona wind pattern between main fins, with maximum velocity gradients aligned spatially with heat source locations. Compared to conventional aluminum substrate heat sinks, the optimized CM-CWHS reduces temperature by 17.9 degrees C under a 15.5-W heat load, improves cooling capacity by 19.51 %, and achieves a convective heat transfer coefficient of 43 W/(m2 & sdot;K) (a 16.1 % increase). This work offers a viable solution to the conflict between electronic miniaturization and high heat flux density.

With the rapid evolution of laptops toward higher performance and lighter designs, conventional single-heat-source cooling solutions are proving increasingly inadequate for meeting modern thermal management demands. This paper presents the design of a 3 mm thick dual-heat-source vapor chamber (DHS-VC) tailored for high-performance laptop cooling. It specifically investigates the effects of liquid filling ratios and inclination angles on the startup characteristics and continuous operation performance of the DHS-VC. The experimental results demonstrate that DHS-VCs at all tested filling ratios successfully achieve startup under thermal loads ranging from 30 W-30 W to 120 W-120 W. Notably, the DHS-VC with an 80% filling ratio exhibits superior startup performance, characterized by a lower startup temperature and a faster response time. At a filling ratio of 100%, the DHS-VC achieved a maximum thermal load of 180 W per source (180 W-180 W) under equal thermal load conditions, resulting in a minimum thermal resistance of 0.028 degrees C/W. In single-source operation modes, the maximum heat dissipation capacities reached 300 W and 270 W, respectively. Furthermore, under unequal thermal load operation, the maximum thermal loads were 280 W-60 W and 60 W-250 W. At an orientation of 90 degrees, the DHS-VC with a 100% filling ratio achieved a maximum thermal load of 170 W-170 W. Notably, the temperature difference across the condenser section remained below 4.1 degrees C across all tested inclination angles. The proposed DHS-VC exhibits ultra-low thermal resistance and superior temperature uniformity, demonstrating its capability for highly efficient centralized thermal management. Consequently, this work provides a solid technical foundation for the thermal design of next-generation mobile computing devices.

As a key variable-speed component, the super high-speed permanent magnet synchronous motor (SHSPMSM) used in fuel cell vehicles (FCV) generates significant electromagnetic noise due to large-amplitude electromagnetic forces, which seriously affects passenger comfort and restricts further commercial applications. To address this issue, this paper first establishes the numerical equations for the SHSPMSM and systematically analyzes the underlying mechanism of its electromagnetic noise generation through both modeling and experimental validation. Subsequently, a Pareto-based multi-objective optimization method is proposed to minimize electromagnetic forces and enhance both the acoustic and electromagnetic performance of the SHSPMSM, ensuring comprehensive improvement. The results indicate that the maximum electromagnetic noise peak occurs at the low harmonic order of 2. Within the acoustic boundary of the SHSPMSM, electromagnetic noise is reduced by 13.3 %, 8.13 %, and 6.25 % at 34,000 rpm, 86,500 rpm, and 95,000 rpm, respectively. These findings confirm that reducing electromagnetic forces can significantly lower electromagnetic noise without compromising the overall performance or efficiency of the SHSPMSM, providing valuable guidance for the development of low-noise, high-performance drive systems in next-generation fuel cell vehicles.

Electric vehicle headlamp illuminance directly affects the driver’s visibility. Accurately predicting electric vehicle headlamp illuminance is crucial to enhancing driving safety. Existing deep learning models are trained using data collected from real-world road testing, yet external factors may compromise its reliability. Electric vehicle headlamp illuminance prediction primarily relies on data fitting, and such models are prone to overfitting when input data are affected by external disturbances. To solve the problem, we propose a luminancxel properties physical information neural network (LumiPINN) prediction model. Test conditions are designed in accordance with standard. The data was collected in an indoor laboratory to eliminate the influence of external factors, then underwent cleaning and pre-processing to ensure data quality. During the modelling process, the physical model is treated as a constraint, with the loss function to jointly optimise the prediction model. Compared with Deep Neural Network and Artificial Neural Network prediction models, the Mean Absolute Error, Mean Square Error, Root Mean Square Error, Mean Relative Error were reduced by 60.2%, 83.6%, 59.6%, 61.3%, and 71.7%, 90.7%, 69.5%, 71.4%. The Coefficient of Determination improved by 0.0015 and 0.0029. The results show that the LumiPINN prediction model demonstrates higher accuracy in prediction outcomes.

Vehicle speed prediction can improve the energy management strategy for fuel cell vehicles by anticipating future load changes. However, the cumulative and propagated amplification of vehicle speed prediction errors increases their impact on the performance of the strategy. To address this issue, this paper proposes a predictive energy management strategy based on enhanced random exploration and the fusion of internal and external multi-state predictions. Specifically, to mitigate training instability caused by vehicle speed prediction errors and enhance random exploration capabilities, a twin delayed deep deterministic policy gradient agent model with a variance-aware mechanism is constructed. From a data-driven perspective, to prevent external vehicle prediction errors from subjecting the fuel cell and power battery to extreme operating conditions, the future time-domain trends of the internal state of charge are considered. Using the predicted state variables as inputs for the agent, an energy management strategy is formulated that integrates multi-state predictions from both internal and external sources. A hardware-in-the-loop platform is established to evaluate the proposed method under both a standard driving cycle and a real-road random driving cycle. Compared with the baseline strategy, the proposed method reduces equivalent hydrogen consumption by 9.48% under the standard driving cycle and by 5.41% under the random driving cycle, thereby improving the overall performance of the strategy.

Heat exchangers are vital for emission reduction and energy saving. Although theoretical research on electrohydrodynamics (EHD) for heat transfer enhancement is advanced, studies on the role of ionic wind in heat exchanger efficiency are limited. This study presents a novel dividing-wall-type ionic wind heat exchanger. It experimentally explores the effects of parameters such as ground electrode shape, emitting electrode distribution, inlet velocity, and hot-air temperature on ionic wind intensity and heat exchange rate. The drying characteristics of the EHD-based heat exchanger are also investigated. A two-dimensional model is used for internal temperature distribution and flow analysis, and the design is optimized. Results show that optimal air temperature and velocity improve heat transfer. The ionic wind has better heat transfer when the emitting electrode is at the flow channel center. For multiwire electrodes, adjacent emitter barrier effects and energy consumption matter, with three-wire electrodes maximizing performance. The partially grounded flat plate structure has a 5.3% higher thermal rise rate (TRR) compared with a fully grounded one, and the sawtooth-shaped ground electrode design has a 39.8% higher TRR. After 120 minutes of drying, the single-wire EHD heat exchanger reduced corn moisture by 16% compared with hot air only. An optimized three-emitter and partial grounding configuration can further enhance drying capacity.

The existing thermal management system in fuel cell heavy-duty trucks fails to achieve efficient waste heat utilization, leading to wastage of hydrogen energy. The fuel cell is controlled within a fixed temperature range in the fuel cell heavy-duty truck thermal management system, which ignores the temperature characteristics of the fuel cell output power and reduces the output efficiency of the fuel cell. The dynamic programming method can improve the utilization efficiency of waste heat and realize the efficient utilization of hydrogen energy. However, dynamic programming needs to be solved when the global working conditions are known, and the excessive amount of computation also limits the application of real-time scenarios. This paper aims to propose a real-time thermal management strategy for fuel cell heavy-duty truck based on velocity prediction-iterative dynamic programming. To establish a method for real-time control, based on the LSTM, establish and analyze vehicle speed prediction models for different prediction time domains; Based on the vehicle speed prediction model, the Velocity Prediction-Iterative Dynamic Programming (VP-IDP) is designed to realize iterative time domain control and iterative scaling control; Based on the VP-IDP, a real-time optimal control strategy for the working temperature of a PTC-free heat exchanger only system (HEOS) is proposed. Simulations were carried out under the two working conditions of NEDC and UDDS. The results showed that, compared with the DP control strategy, the VP-IDP (5s) strategy only increased hydrogen consumption by 2.03% and 1.28%, the total waste heat utilization decreased by 1.38% and 0.97%, and calculation time shortened by 93.7% and 92.9%. Respectively, the VP-IDP (5s) can achieve the waste heat utilization efficiency and hydrogen utilization efficiency close to the global optimum.

As heat flux in high-power electronic devices exceeds 100 W/cm2, conventional thermal management technologies face critical limitations in acoustic noise, physical dimensions, and cooling efficacy. While electrohydrodynamic (EHD) cooling, often manifested as ionic wind, presents a promising solution for managing high heat fluxes, prior research has predominantly focused on enhancing heat transfer while overlooking the concomitant increase in thermodynamic irreversibility. Addressing this gap, this study employs a thermodynamic optimization perspective to investigate the intrinsic trade-off between heat transfer enhancement and entropy generation within a nested annular electrode ionic wind heatsink. A comprehensive coupled multi-physics model is developed, integrating the governing equations for electrostatics, charge transport, fluid dynamics, and heat transfer. This model enables a systematic analysis of the impact of key geometric parameters (discharge gap, collector width, and emitter longitudinal offset) on electric field distribution, ionic wind characteristics, convective heat transfer performance, and entropy generation rate. To pinpoint the dominant sources of irreversible losses under varying operating conditions, the total entropy generation rate is decomposed into its constituent components: entropy generation due to fluid friction and entropy generation due to heat transfer. Leveraging the insights gained from the multiphysics model, a quadratic surrogate model is constructed to approximate the system response. This surrogate model is subsequently combined with the particle swarm optimization (PSO) algorithm to perform a dual-objective optimization, simultaneously targeting the maximization of the Nusselt number (Nu) and the minimization of the total entropy generation rate. The resulting Pareto-optimal solutions reveal that enhancing heat transfer does not monotonically reduce thermodynamic losses; instead, it induces a shift in the dominant mechanisms of irreversibility. Specifically, for the optimized heatsink configuration, the total entropy generation rate is reduced to 2.08 & times; 10-3 W/K. Concurrently, the proportion of entropy generation attributed to fluid friction increases from 38% to 52%, while the proportion due to heat transfer decreases from 61% to 49%. Experimental validation confirms the efficacy of the optimized design: it achieves a 33.2% increase in Nu, a 39.1% rise in Reynolds number (Re), and a significant 13 K reduction in the peak temperature of the heat source. This study establishes a thermodynamically informed design framework for ionic wind heatsinks, elucidates the underlying mechanisms of irreversible energy dissipation, and provides practical guidance for developing high-efficiency EHD cooling systems.

In response to the challenge of submicron particle collection efficiency in electrostatic precipitators (ESPs) under increasingly stringent industrial emission standards, this study proposes a novel approach based on magnetic field spatial modulation. Unlike traditional wire-plate ESPs, five axisymmetric magnetic confinement topologies were designed in this study, and their interplay in charge transport, magnetohydrodynamics (MHD), and particle dynamics was systematically analyzed through a validated 2D coupled model. The results indicate that the application of a symmetric uniform magnetic field in the corona discharge region increases the spatial charge density by up to 38 %. Flow and particle field simulations further reveal that the magnetic field restructures the ionic wind velocity field, generating vortex structures that extend particle residence time. Notably, magnetic prepolarization in the inlet region effectively enhances charge migration efficiency under low voltage. Additionally, experimental results show that the magnetic field has a more significant impact on the collection efficiency of smaller particles, with the efficiency for 0.1 mu m particles improving by up to 29.1 %. At lower magnetic flux densities or voltages, the contribution of voltage to dust removal efficiency becomes more pronounced. SEC analysis reveals that configurations M2 and M4 achieve minimal energy consumption at 10 kV. At 20 kV, SEC values rise universally, with M2 even exceeding the no-field reference. This demonstrates superior energy efficiency at lower voltages, though M4 maintains exceptional performance (low SEC) even at 20 kV, highlighting its enhanced voltage tolerance. By optimizing the magnetic field configuration and matching the magnetic and electric field strengths, this study achieved a total efficiency of 98.7 % at lower specific energy consumption. The optimized magnetic field configuration provides a feasible theoretical framework for the application of ESPs in complex industrial environments.

In this work, an ionic wind pump with saw-toothed emitters is developed for high-power LED chip thermal control. The multi-objective optimization approach is used to determine the optimal structure. The spatial ionic wind distribution may be efficiently regulated by adjusting the heat sink's installation location, which also changes the flow from vortex to wall flow and increases the system's heat exchange capacity. The system's maximum mean heat transfer coefficient is 74.49 W/(K center dot m2) when the optimized pump is used for thermal management. The chip's case temperature drop reaches its maximum value of 51.41 K when the chip power is greater than 9 W. After multi-objective optimization, the pump's power consumption is reduced by 2.6 % and the chip's case temperature is reduced by 15.07 %. The optimized pump has a better heat transfer coefficient to the energy consumption ratio, up to 196.03 K- 1 center dot m- 2, in comparison to the other two reported ionic wind pumps. This successfully settles the dispute between cooling capacity and energy consumption.

When the on-board hydrogen system (OBHS) releases hydrogen rapidly, the gas expands and the cylinder temperature drops quickly. Under normal release, the final temperature is 236 K, within the recommended safe operating window of 233-358 K, which is defined to protect OBHS materials and ensure vehicle safety. If temperatures fall below this limit, components may be damaged and system reliability compromised. Existing studies on OBHS temperature behavior often rely on simplified models and incomplete failure analysis. To address this, a detailed two-dimensional CFD model was developed to simulate hydrogen release and evaluate temperature variations under three fault modes: infinite flow, excessive restriction, and downstream blockage. The results show that the final cylinder temperatures were 228 K, 253 K, and 241 K, respectively, compared with 236 K under normal release. These correspond to 3.4 % lower, 7.2 % higher, and 2.1 % higher than the baseline. Among them, only infinite-flow failure caused the final temperature to fall below the safety threshold of 233 K, indicating a critical risk. This finding highlights the importance of incorporating design measures, such as redundant flow limiters and thermal buffering, to prevent unsafe cooling and improve future OBHS safety.

Predictive energy management strategy for intelligent fuel cell vehicle low energy consumption relies on data-driven vehicle speed prediction to provide accurate short-term speed. The traditional data-driven vehicle speed prediction based on the vehicle speed sequence lacks consideration of the influence of the power sequence of the power source on the prediction accuracy in fuel cell vehicle. Most data-driven vehicle speed prediction is based on deep learning methods, which have poor generalization capabilities when predicting across working conditions. Based on this, this paper proposes a transfer learning method for vehicle speed prediction that improves across working conditions adaptability. First, Obtain vehicle data under urban, congestion, suburban and expressway conditions through real vehicle road testing. Secondly, a vehicle speed prediction model based on hybrid deep learning is established. And the impact of dual power source power sequence on vehicle speed prediction performance is analyzed under urban conditions. Finally, a vehicle speed transfer prediction method that integrates hybrid deep learning and transfer learning is proposed. And the prediction results of vehicle speed transfer from urban conditions to congestion, suburban and expressway conditions are analyzed. The results show that compared without transfer learning. The vehicle speed transfer prediction method proposed reduces mean absolute error by 21.3%, 24.8%, and 24.9%, respectively, and mean square error by 35.3%, 44.6%, and 37.2% respectively when the prediction time domain is 5 seconds. Both mean absolute error and mean square error are reduced. This method has higher vehicle speed prediction accuracy and better adaptability across working conditions.

To address issues in hydrogen fuel cell hybrid power systems such as passive power allocation, multi-time scale dynamic mismatches, and conflicts between economy and durability, This paper integrates the collaborative architecture of Dynamic Programming and Model Predictive Control (DP-MPC) to achieve energy optimization. A hybrid power system model consisting of a hydrogen fuel cell, lithium battery, and supercapacitor is constructed. The fuel cell provides the basic power, the lithium battery balances medium - and long - term power fluctuations, and the supercapacitor handles instantaneous power demands and energy recovery. Simulation results under the WLTP standard driving cycle (low-speed segment) show that, compared with the Equivalent Consumption Minimization Strategy (ECMS), the proposed strategy reduces hydrogen consumption by 9.97%, effectively decreases the power fluctuations of the fuel cell, and improves the State of Charge (SOC) recovery ability of the lithium battery. While reducing energy consumption, it also delays the fuel cell performance degradation. This strategy offers an efficient solution that balances economy and durability for the energy management of hydrogen fuel cell vehicles.

Lithium-ion batteries are widely used in electric vehicles and energy storage, and the inherent sheet stacking structure of conventional graphite leads to its limited lithium storage performance. In this study, N-doped graphene microsheets were in situ formed on the surface of graphite by Ar/N2 plasma technology. The changes of physicochemical properties and lithium storage performance of graphite anode before and after modification were analyzed. The results show that the plasma modification increases the structural defects of graphite, expands the interlayer spacing, and nitrogen atoms is successfully doped. In comparison to pristine graphite, the lithium storage capacity of N-doped graphene microsheets modified graphite (P-G) was increased by 31.2 % (470.63 mA h center dot g- 1 at 100 mA center dot g- 1). Additionally, it exhibited remarkable cycle stability, demonstrating a capacity retention of 83.5 % after 1000 cycles at 1 A center dot g- 1. This performance was significantly superior to that of graphite (G), which retained only 43.6 % of its capacity under identical conditions. The LiFePO4||P-G full cell delivered a reversible specific capacity of 120.35 mA h center dot g- 1 at 1C rate, exhibiting a capacity retention of 91.40 % after 100 cycles. N-doped graphene microsheets increase the lithium adsorption sites, expand the lithium ion diffusion channels, and enhance the charge storage capacity of graphite through the pseudocapacitance effect. This study provides a new idea for expanding efficient and environmentally friendly graphite anode modification technology.

Recent breakthroughs in artificial intelligence are accelerating the intelligent transformation of vehicles. Vehicle electronic and electrical architectures are converging toward centralized domain controllers. Deep learning, reinforcement learning, and deep reinforcement learning now form the core technologies of domain control. This review surveys advances in deep reinforcement learning in four vehicle domains: intelligent driving, powertrain, chassis, and cockpit. It identifies the main tasks and active research fronts in each domain. In intelligent driving, deep reinforcement learning handles object detection, object tracking, vehicle localization, trajectory prediction, and decision making. In the powertrain domain, it improves power regulation, energy management, and thermal management. In the chassis domain, it enables precise steering, braking, and suspension control. In the cockpit domain, it supports occupant monitoring, comfort regulation, and human–machine interaction. The review then synthesizes research on cross-domain fusion. It identifies transfer learning as a crucial method to address scarce training data and poor generalization. These limits still hinder large-scale deployment of deep reinforcement learning in intelligent electric vehicle domain control. The review closes with future directions: rigorous safety assurance, real-time implementation, and scalable on-board learning. It offers a roadmap for the continued evolution of deep-reinforcement-learning-based vehicle domain control technology.

The suboptimal photoelectric conversion efficiency of light-emitting diodes (LEDs) leads to increased temperature. There is a growing interest in using microstructure ionic wind pumps to regulate the chip temperature. But the ionic wind flow and thermal transfer characteristics of thin-plate electrode pumps used for cooling LED chips is unclear. This study proposes ionic wind pumps equipped with wedged and zigzag emitters to effectively manage the heat generated by high-power LED chips. Experimental investigations were conducted to analyze the electrohydrodynamic characteristics of pumps with different emitter types. A two-dimensional model with a wedged electrode and a three-dimensional model with a zigzag electrode were developed for flow distribution analysis and energy efficiency comparison. The cooling capacity of pumps with different configurations was examined. The results show that the pump equipped with a zigzag electrode exhibits improved stability in corona discharge and approximately 1.53 times higher energy efficiency compared to the pump with a wedged electrode. Moreover, the pump with the zigzag electrode covers a larger ionic wind flow area, generating a higher intensity of ionic wind. The angle between the emitter and the grounding electrode significantly affects the ionic wind flow characteristics. The optimal angle is 70 degrees for pumps with wedged emitters and 30 degrees for those with zigzag emitters. Both pumps can produce a steady wall jet at their optimal angle, causing significant disruption in the surrounding area. The pump with a zigzag electrode exhibits superior cooling performance and is more effective with low power consumption.

Hydrogen consumption is not only an evaluation metric for the economic performance of fuel cell vehicles (FCVs), but also one of the key optimization objectives in energy management strategies (EMS). However, EMS that rely on cumulative hydrogen consumption (HC-C) or hydrogen consumption per 100 km (HC-P100) as optimization objectives are limited by the inaccuracies and time delays in hydrogen consumption data. This study proposes an algorithmic framework called Proximal Policy Optimization with Hydrogen Consumption Prediction (PPO-HCP) to optimize energy management strategies for FCVs. First, establishing a dynamic system model for FCVs. Then, innovatively designing a novel reward function to enhance the adaptability of deep reinforcement learning (DRL)-based EMS under complex and dynamic conditions. This includes weighting different terms in the reward function, such as short-term real-time hydrogen consumption (HC-RT), long-term HC-C and HC-P100, fuel cell power, and battery state of charge (SOC). Finally, the proposed PPO-HCP algorithm is evaluated and compared with the conventional PPO algorithm under both training and random conditions. The results show that the energy consumption optimization effect of the PPO-HCP algorithm is more significant, with HC-P100 reduced by 5.3 % under training conditions and 7.9 % under random conditions.

Particulate matter (PM) is harmful to the environment and human health. It is of practical significance to explore green and sustainable technologies for efficient removal of PM by studying the mechanism of PM oxidation and decomposition. A visualization test system for oxidatively decomposing diesel PM using non-thermal plasma (NTP) was developed to perform oxidative decomposition tests on PM samples at different lengths of time at 120 degrees C. The physical and chemical properties of primary particles, elemental carbon (EC) oxidation activity, and surface functional groups of the PM at different oxidation stages were analyzed to investigate how micronanostructure changes of PM affected the oxidation characteristics and elemental occurrence. The structure of primary particles was divided into three parts: in-core, ex-core, and ex-core edge. A dynamic nanostructure model of primary particles was established during the reaction process. The primary particles experienced the evolution process of 'massive removal of crystalline at the outer edge of the nucleus', 'disordered arrangement of crystalline' and 'hollow crystalline in the nucleus after erosion ', corresponding to the early, middle and late stages of oxidation, respectively. The nanostructure properties of PM changed non-linearly during the decomposition of PM by NTP, and there was a linkage effect among the microcrystalline parameters, oxidation activity, and elemental distribution of PM.

The damage to the sealing rings of high-pressure hydrogen valves caused by extreme environmental temperatures, leading to minor hydrogen leakage, is a significant issue in the field of hydrogen safety. However, current understanding is still limited regarding the initial leakage of hydrogen through damaged gaps in sealing rings. This paper aims to explore the distribution of hydrogen leakage in sealing rings with various-shaped damage gaps under extreme environmental temperatures. Firstly, the real damage gaps of the sealing rings are obtained on an extreme temperature experimental platform. Then, a coupled hydrogen leakage model is established, which calculates the viscosity of hydrogen in real-time as temperature changes. Based on these models, the concentration, temperature, and velocity distribution of hydrogen within the sealing ring gaps under extreme environmental temperatures are studied. The results indicate that the high-pressure hydrogen within the gaps, influenced by viscosity and wall adhesion effects, tends to diffuse along the wall surfaces. The high-temperature zone along the edge lines of the gaps lasts longer than that on the centerline. Under extreme high temperature conditions, the average velocity at the inlet and outlet of the gap is 21.6% higher than under extreme low temperatures.