Method for judging tunnel portal frost damage in cold region based on multi-source environmental data fusion

Abstract

The application discloses a cold region tunnel portal freeze damage discrimination method based on multi-source environment data fusion, relates to the technical field of disaster warning, and comprises the following steps: for tunnel portal freeze damage discrimination, simultaneously collecting air temperature, ground surface temperature, residual snow evaporation and wind pressure change data in the tunnel portal area, and fusing and analyzing all the collected data according to a unified time sequence to generate a combined change curve reflecting the relationship between inversion temperature and evaporation change. The application establishes a dynamic corresponding relationship between inversion temperature and heat distribution through multi-source environment data fusion, realizes accurate identification of heat transfer reversal and early determination of freezing risk. Through generation of heat reverse guidance instructions, the air flow intensity and direction are dynamically controlled, phased air supply and heat exchange are implemented, the freezing nucleus condition is eliminated, and the long-term stable operation of the tunnel portal is ensured.

Description

Technical Field

[0001] This invention relates to the field of disaster early warning technology, specifically to a method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion. Background Technology

[0002] The multi-source environmental data fusion-based identification of frost damage at tunnel entrances in cold regions refers to a method that, in cold areas, addresses issues such as frost heave, icing, and lining cracking at tunnel entrances by analyzing and determining damage in a unified manner, rather than relying on a single temperature or empirical threshold. This method simultaneously acquires multi-source environmental information at the tunnel entrance, including air temperature, surface and lining temperatures, humidity changes, wind speed and direction, snowfall and melting conditions, groundwater seepage characteristics, and the tunnel entrance's orientation and shielding conditions. This data is then correlated and fused on a unified timescale to characterize the coupled evolution of low temperature, moisture content, and heat exchange, thereby identifying whether the tunnel entrance area has entered a dangerous state with conditions conducive to freezing. This fusion-based identification method can distinguish between short-term low temperatures and continuous freezing, determine whether moisture replenishment is sufficient, and whether freezing has a cumulative effect. This provides a more objective, continuous, and targeted basis for early warning, classification, and selection of prevention and control measures for frost damage at tunnel entrances in cold regions.

[0003] The existing technology has the following shortcomings:

[0004] In existing technologies, the assessment of frost damage at tunnel entrances in cold regions is mostly based on air temperature thresholds or single temperature gradient models, lacking a mechanism to identify the reversal of heat transfer direction under abrupt changes in local microclimate. When temperature inversion occurs at the tunnel entrance area at night, accompanied by the evaporation of residual snow on the surface, a local temperature gradient reversal occurs in the near-surface layer, with the temperature above the air layer being higher than that below the surface, causing a reversal in the direction of heat conduction. This type of temperature inversion superposition is characterized by short duration, narrow spatial range, and slow temperature difference changes. Conventional temperature sensor arrays deployed in existing technologies are unable to capture such local thermal field anomalies with sufficient temporal and spatial resolution, leading to deviations in the direction determination of the heat exchange model. During monitoring, the system mistakenly believes that the temperature difference between the inside and outside of the tunnel entrance has dissipated, while in reality, heat continues to accumulate on the surface of the rock mass, forming a hidden freezing core below the surface. This freezing core forms a local ice expansion center at the bottom of the lining, causing a sudden increase in expansion stress, which in turn induces structural damage problems such as lining cracks, circumferential joint opening, and bottom plate uplift, seriously affecting the operational safety and long-term stability of tunnels in cold regions.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion, comprising the following steps:

[0008] To identify frost damage at tunnel entrances, air temperature, surface temperature, residual snow evaporation, and wind pressure changes were collected simultaneously in the tunnel entrance area. All collected data were then fused and analyzed according to a unified time series to generate a joint change curve reflecting the relationship between temperature inversion and evaporation changes, which was used to provide a time reference for judging heat distribution.

[0009] Based on the joint variation curve, a continuous analysis is performed on the regions where the near-surface temperature difference shows an inverse change in the time series, to identify anomalous regions where heat is transferred from bottom to top, and to determine the range of heat accumulation below the surface based on the spatial distribution of the anomalous regions, which is used as a basis for identifying freezing risk.

[0010] Based on the evolution characteristics of the heat accumulation range in the time series, the starting time of the formation of the freezing core, the temperature change amplitude and distribution direction are extracted to generate a time distribution table reflecting the freezing trend, which can be used to provide a time series reference for active heat intervention.

[0011] Based on the freezing trend time distribution table, and taking into account the ventilation airflow intensity, air supply rhythm and airflow direction angle parameters, heat reverse guidance command is calculated to enable the airflow to dynamically adjust the heat in the potential freezing area in time and direction.

[0012] Based on the instruction of reverse heat guidance, staged air supply, intermittent heat exchange and directional flow operation are implemented. Through layer-by-layer heat conduction and local airflow peak compensation, the heat backflow phenomenon caused by temperature inversion is smoothly dispersed, and the conditions for the formation of freezing cores below the ground surface are effectively eliminated, thereby completing the process of frost damage identification and risk elimination at the entrance of tunnels in cold regions.

[0013] Preferably, the steps for generating the joint change curve are as follows:

[0014] In the tunnel entrance area of ​​cold regions, based on the topographic features, orientation conditions, air flow direction and surrounding cover distribution, air temperature collection points, surface temperature collection points, residual snow evaporation collection points and wind pressure change collection points are set up, and a continuous coverage of air flow field, surface temperature field, latent heat of evaporation field and wind pressure disturbance field is formed along the tunnel axis.

[0015] Maintain a consistent sampling frequency at all collection points and record data on changes in air temperature, surface temperature, residual snow evaporation, and wind pressure at a uniform time step to form a continuous time-series data chain covering the air, surface, and strata inside and outside the cave.

[0016] A corresponding analysis was conducted on the time series of changes in air temperature, surface temperature, residual snow evaporation and wind pressure. The magnitude and direction of changes in different time intervals were compared to establish the heat exchange relationship between the air layer and the surface layer.

[0017] By forming a joint change curve from continuous time data, the time periods in which the direction of temperature change reverses are identified, and the direction and range of heat transfer are determined through spatial mapping, providing a temporal and spatial basis for subsequent heat anomaly identification.

[0018] Preferably, after forming the joint change curve, the continuous time data of air temperature, surface temperature, residual snow evaporation and wind pressure changes are segmented and analyzed. The time intervals in which the temperature change direction reverses are extracted separately, and the temperature values ​​and change rates in each time period are recorded. Based on the time node data, the direction of heat transfer in the cave entrance area is spatially mapped to determine the range of heat transfer from the air layer to the surface or from the surface to the air layer when the inversion is formed, providing an accurate reference for the identification of heat anomaly areas.

[0019] Preferably, the steps for determining the heat accumulation range are as follows:

[0020] By pairing and analyzing air temperature data and surface temperature data within the same time interval in the joint variation curve, a temperature difference change sequence is formed, and the moment when the direction of heat conduction changes is determined by identifying the alternation point where the temperature difference changes from positive to negative or from negative to positive.

[0021] Taking each temperature inversion point as the center, extract air temperature and surface temperature data within the time range before and after, analyze the duration and trend of heat transfer from bottom to top, and determine the time interval of the temperature inversion phenomenon.

[0022] By combining the temperature changes at surface temperature collection points at different depths, a joint analysis of the subsurface temperature in the vertical and horizontal directions is conducted to determine the spatial range of heat accumulation upwards within the strata.

[0023] By overlaying the spatial extent of heat accumulation with the duration of reverse transfer, a distribution map of heat accumulation intensity over time is plotted to identify areas of sustained heat retention and determine areas with high risk of freezing.

[0024] Preferably, the steps for generating the time distribution table are as follows:

[0025] Within the heat accumulation range below the surface, temperature data at different depths and horizontal positions are arranged in time series with time as the horizontal axis to form a time extension series of the heat accumulation range, and time segments in which the temperature continues to decrease and the gradient slows down are identified.

[0026] Based on the time-extended sequence, temperature change curves at different depths and in different directions are compared to determine the starting time and main formation direction of frozen nuclei, and regions where the temperature drops earlier than other locations are identified as the starting point for frozen nuclei formation.

[0027] Continuously track the temperature change process during the formation period of the frozen nucleus, extract the duration and magnitude of the temperature drop, and determine the expansion trend of the frozen nucleus along the depth and horizontal directions;

[0028] The start time, duration, temperature change amplitude, and distribution direction of each time period are summarized to generate a freezing trend time distribution table, which provides a time series reference for active heat intervention.

[0029] Preferably, the freezing trend time distribution table records the formation process of the frozen core based on time. Each time period includes temperature change values ​​at different depths, freezing duration, and information on the expansion of the frozen core along the cave entrance. The intensification and deceleration phases of the freezing trend are determined by the change in the rate of temperature decrease, so as to achieve a precise temporal expression of the frozen core formation process.

[0030] Preferably, the steps for generating the heat reverse guidance command are as follows:

[0031] Using the start time, duration, temperature change amplitude, and spatial expansion direction in the freezing trend time distribution table as the time control benchmark, the speed, flow rate, air supply time, and air flow direction angle of the ventilation airflow are matched in time to form a time framework for airflow regulation.

[0032] Based on the spatial distribution characteristics of the freezing trend in the cave entrance area, a distribution method for airflow intensity in different directions is formulated so that the airflow forms a transmission path along the cave entrance axis, top and ground direction, thereby realizing the directional transfer of heat in space.

[0033] By combining the results of airflow time matching and spatial allocation, the air supply rhythm and airflow direction angle are coordinated and designed, and a time and space matching relationship is formed by intermittent air supply and angle adjustment.

[0034] The airflow speed, flow rate, air supply duration, air supply interval, and airflow direction angle within each time period are summarized to generate a heat reverse guidance command, which is used for dynamic heat regulation of airflow in time and direction.

[0035] Preferably, based on the heat reverse guidance instruction, phased air supply, intermittent heat exchange, and directional recirculation operations are implemented. Through layer-by-layer heat conduction and local airflow peak-shifting compensation, the following steps are taken to dissipate heat and eliminate the conditions of the subsurface freezing core:

[0036] Based on the time sequence of the heat reverse guidance command, the airflow is divided into the start-up stage, the enhancement stage and the decay stage, and the airflow speed, duration and direction are controlled respectively, so that the airflow forms a stable heat transfer channel in different stages.

[0037] During the phased air supply process, air supply intervals are set to allow hot air to diffuse and redistribute within the tunnel opening space, forming a dynamic heat exchange balance between the air layer and the surface layer.

[0038] Based on intermittent heat exchange, directional recirculation operation is implemented to periodically change the direction of some airflow, so that the air forms a stratified and alternating circulation flow inside the cave entrance area, dispersing the cold air on the surface and dissipating the heat at the top.

[0039] By coordinating and combining phased air supply, intermittent heat exchange, and directional recirculation operations, heat is released layer by layer from the inside out, restoring the surface temperature gradient and eliminating the conditions for the formation of a frozen core.

[0040] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0041] This invention establishes a dynamic correlation between temperature inversion and heat distribution over time by integrating multi-source data on changes in air temperature, surface temperature, residual snow evaporation, and wind pressure. This enables accurate identification of heat transfer reversals under local microclimate changes. By correlating the variation curves with the heat accumulation range, this method clearly depicts the spatiotemporal evolution of freezing core formation. This allows for the early identification of risk areas before freezing conditions stabilize, effectively preventing shallow heat retention at tunnel entrances and the formation of potential ice expansion centers. This provides continuous and precise criteria for assessing frost damage in cold-region tunnel operations.

[0042] This invention generates heat-reverse guidance commands and combines them with the intensity, rhythm, and direction parameters of ventilation airflow to achieve active heat conduction and precise regulation of airflow in time and space. This method dynamically organizes the air supply and heat exchange process based on a freezing trend time distribution table, allowing the airflow to directionally dissipate heat along the potential freezing zone, gradually eliminating the heat backflow effect caused by temperature inversion. Through the synergistic effect of staged air supply and directional return flow, the temperature field in the tunnel entrance area is restored to equilibrium, and the conditions for freezing nucleus formation are fundamentally destroyed, thereby significantly improving the freeze-resistance stability and operational safety of the tunnel entrance. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0044] Figure 1 This is a flowchart of the method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion, as described in this invention. Detailed Implementation

[0045] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0046] This invention provides, for example Figure 1 The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion, as shown, includes the following steps:

[0047] To identify frost damage at tunnel entrances, air temperature, surface temperature, residual snow evaporation, and wind pressure changes were collected simultaneously in the tunnel entrance area. All collected data were then fused and analyzed according to a unified time series to generate a joint change curve reflecting the relationship between temperature inversion and evaporation changes, which was used to provide a time reference for judging heat distribution.

[0048] To accurately reflect the temperature variations, surface thermal conductivity response, residual snow evaporation process, and wind pressure disturbance patterns in the cave entrance area in both time and space, a unified acquisition and time-series fusion method for multi-source environmental data was adopted. The specific steps are as follows:

[0049] In the area surrounding tunnel entrances in cold regions, typical cross-sections are selected as data collection zones based on the terrain features, orientation, distribution of surrounding cover, and airflow direction. Using the tunnel entrance centerline as a reference, air temperature collection points are deployed from inside the tunnel outwards, ensuring coverage of the airflow zone inside the entrance, the heat exchange zone at the entrance edge, and the natural air zone outside. Air temperature collection points are distributed along the tunnel axis, with spacing determined by wind speed gradients, typically ranging from 2 to 5 meters, to ensure continuous recording of air temperature changes along the tunnel entrance and exit directions. For surface temperature data collection, collection points are deployed along the tunnel's side slopes, roof, and floor, extending from the surface downwards to approximately 1 to 1.5 meters above the foundation, to capture the thermal conductivity response of the surface layer and shallow soil and rock. Residual snow evaporation collection points are set on typical snow-covered surfaces within a 5 to 15-meter area in front of the tunnel entrance, continuously measuring the mass change caused by snow surface evaporation to reflect the latent heat release process. Wind pressure variation data collection points were deployed at the air outlet inside the tunnel, the upper edge of the tunnel entrance, the air inlet outside the tunnel, and near the ground surface on the slope to reflect the pressure difference variations caused by airflow at different locations. Each collection point spatially forms a continuous coverage of the airflow field, surface temperature field, latent heat of vaporization field, and wind pressure disturbance field, ensuring that the heat transfer environment in the tunnel entrance area is completely captured spatially. All collection points are numbered according to a unified coordinate system to ensure accurate spatial correspondence of subsequent data.

[0050] To ensure the temporal correlation between different physical quantities, the sampling frequency at each collection point remained consistent, with the sampling interval determined based on the rate of microclimate change at the tunnel entrance, typically using a time step of 5 to 10 minutes. During the collection process, air temperature records reflected the dynamic changes between the intrusion of cold air from outside and the outflow of hot air from inside the tunnel; surface temperature records reflected the temperature difference response between radiative cooling and conductive recirculation of the surface; records of residual snow evaporation reflected the latent heat release process of the snow surface under low temperature and high humidity conditions; and records of wind pressure changes reflected the impact of changes in airflow direction on local heat transport. All data were recorded with the same timestamp and identified by the collection point number, ensuring that changes in air temperature, surface temperature, residual snow evaporation, and wind pressure disturbance processes corresponded on the same timeline. This time-consistent recording method ensured the comparability and temporal consistency of data in subsequent analysis phases. Each set of data was archived chronologically after recording, forming a continuous time-series data chain covering air, surface, and strata inside and outside the tunnel.

[0051] In this stage, the data on air temperature, surface temperature, residual snow evaporation, and wind pressure changes collected in the previous step are analyzed point-by-point. By comparing the magnitude and direction of changes in air temperature and surface temperature within the same time interval, the heat exchange state between the air layer and the surface layer is identified. When the air temperature rises while the surface temperature falls, it indicates that there is heat input from the outside air but increased heat loss from the surface, suggesting a possible temperature inversion. When the air temperature falls while the surface temperature rises, it indicates that heat is being transferred from the surface to the air. Subsequently, the changes in residual snow evaporation over the same time period are correlated with changes in air temperature, and the impact of latent heat release on local air temperature is determined by the increase or decrease in the evaporation rate. Wind pressure changes are then compared with air temperature changes; the phenomenon of increased wind pressure accompanied by decreased air temperature is used to determine the intensity and direction of cold air intrusion. Through continuous comparison at consecutive time points, the interaction between air temperature, surface temperature, latent heat release, and wind pressure changes can be observed. All the corresponding relationships are connected in chronological order to form a joint variation curve. This curve can simultaneously reflect the heat exchange between the air and the surface in the cave entrance area, the release of latent heat from the evaporation of residual snow, and the thermal field fluctuations caused by airflow disturbances, thus demonstrating the temporal characteristics and duration of local temperature inversion formation. The formation of the joint variation curve allows for a direct presentation of the dynamic correlation of multi-source data, providing a continuous temporal reference for subsequent determination of the direction of heat transfer and accumulation in the cave entrance area.

[0052] By segmenting and analyzing the fluctuation patterns of the joint variation curve across different time periods, intervals where the direction of temperature change reverses are identified separately. For each time period with a reversal, the values ​​and rates of change of air temperature, surface temperature, residual snow evaporation, and wind pressure are recorded, forming a corresponding time node data table. Based on this time node data table, the direction of heat change at different sampling points in the cave entrance area is spatially mapped during this time period, determining the distribution range of heat transfer from the surface to the air layer or from the air layer to the surface. Time periods where the air layer temperature is higher than the surface temperature are specially marked as key periods where temperature inversions may form, and the upward trend of evaporation and wind pressure fluctuations during these periods are correlated and recorded. By mapping these time nodes to spatial distribution results, a direct relationship between heat distribution and temporal change is established. In this process, the joint variation curve not only provides a temporal reference but also a spatial correlation basis, enabling the subsequent identification of abnormal heat transfer areas within a unified spatiotemporal framework. In this way, the basic data required for the identification of frost damage at the tunnel entrance are completely corresponding in time and space, which fully reveals the coupling relationship between temperature inversion, evaporation and airflow disturbance in the tunnel entrance area, providing continuous, reliable and traceable data support for subsequent analysis.

[0053] Based on the joint variation curve, a continuous analysis is performed on the regions where the near-surface temperature difference shows an inverse change in the time series, to identify anomalous regions where heat is transferred from bottom to top, and to determine the range of heat accumulation below the surface based on the spatial distribution of the anomalous regions, which is used as a basis for identifying freezing risk.

[0054] After the combined variation curves were formed, a continuous analysis process was conducted to reveal the changing characteristics of heat transfer direction in the near-surface strata of the tunnel portal area from both temporal and spatial dimensions, and to further determine the anomalous range of heat transport from the strata upwards. The specific steps are as follows:

[0055] In the combined change curve, air temperature and surface temperature are the core elements reflecting the direction of heat exchange. To accurately describe their coupled changes, air temperature data and surface temperature data within the same time interval are analyzed in pairs. Using time as the primary sequence parameter, air layer temperature is used as the upper boundary condition, and surface temperature as the lower boundary condition. The difference between the two in each time interval is calculated, and the changes in the difference are arranged over time to form a continuous temperature difference change sequence. In this sequence, each data point represents the vertical temperature difference state within a specific time period. A positive temperature difference in the sequence indicates that heat conduction from the surface to the air layer is dominant, while a negative temperature difference indicates that the air layer temperature is higher than the surface temperature, and the direction of heat conduction may be reversed. Through the continuous temperature difference change sequence, the overall evolution of the heat flow direction can be observed over a long period of time. To facilitate the identification of local abrupt changes, the trend of temperature difference over time is smoothed, and the alternation points where the temperature difference changes from positive to negative or from negative to positive are identified. Each alternation point corresponds to a moment when the direction of heat conduction changes. This process establishes a clear correspondence between the direction of temperature change and the change over time, providing a temporal basis for further identification of heat reverse transfer zones.

[0056] After identifying the reversal points in the temperature difference, continuous change data of air temperature and surface temperature are extracted within a certain time range before and after each reversal point. Generally, a one-hour time interval is selected before and after each reversal point to analyze the continuous trend of heat transfer direction within a short period. The rates of change of air temperature and surface temperature are compared: when air temperature rises while surface temperature falls simultaneously, it indicates that the air layer receives heat input while surface heat loss is enhanced, suggesting heat transfer from bottom to top; when air temperature falls while surface temperature rises, it indicates that heat is flowing back from the air layer to the surface. This continuous comparison over time can identify whether the direction of heat transfer remains stable. If, within several adjacent time intervals, the air temperature remains consistently higher than the surface temperature, and the rate of change of air temperature is significantly less than the rate of decrease in surface temperature, it indicates that heat is trapped in the air layer while surface heat is continuously replenished, forming a stable temperature inversion phenomenon. By continuously tracking multiple time intervals, the complete duration and trend of heat transfer from bottom to top can be identified, providing a temporal range for subsequent spatial identification.

[0057] After determining the duration of the reverse heat transfer, a combined vertical and horizontal analysis was conducted on the temperature variation patterns below the surface, using data from surface temperature collection points at different depths. Vertically, four depths of 0.2 meters, 0.5 meters, 1 meter, and 1.5 meters were measured downwards from the surface, and the temperature variation curves at each depth during the reverse heat transfer period were analyzed. If the shallow layer temperature remained constant or decreased slowly during the same time period, while the deep layer temperature continued to rise, it indicates a trend of heat accumulation upwards within the strata. In this case, the heat did not immediately dissipate to the surface air but instead formed a localized high-temperature zone in the shallow strata. Surface temperature data from collection points at the left and right slope toes, the tunnel roof, and the tunnel floor were compared along the transverse direction of the tunnel entrance. When multiple areas showed a rise in below-surface temperature during the same time period, it indicated that the heat accumulation was not limited to a single point but formed a banded or planar accumulation area along the tunnel entrance edge or slope toe. Through a comprehensive comparison of the depth and horizontal directions, the vertical thickness and horizontal range of the heat accumulation could be determined. Typically, the heat accumulation zone is located 0.5 to 1 meter below the surface, extending horizontally to about 10 meters from the outer edge of the cave entrance. Determining this range clearly defines the heat accumulation state below the surface, laying the foundation for subsequent assessment of freezing risks.

[0058] After determining the spatial extent of heat accumulation, it is overlaid with the duration of the aforementioned reverse transfer, using time as the horizontal dimension and surface depth and horizontal position as the vertical dimensions to create a distribution map of heat accumulation intensity over time. This distribution map reflects the dynamic changes in energy accumulation and release during the transfer of heat from the Earth's interior to the atmosphere. When the heat accumulation intensity continuously increases within a certain time period and the corresponding temperature inversion persists, it indicates that heat below the surface is continuously transported upwards but fails to be released, creating potential freezing risk conditions. By overlaying the heat accumulation ranges of different time periods, persistent areas of heat retention can be identified. Furthermore, the accumulation range is correlated with changes in atmospheric wind pressure and residual snow evaporation. When atmospheric wind pressure is low and residual snow evaporation increases, air convection weakens, latent heat release increases, and heat is difficult to diffuse effectively, thus exacerbating heat accumulation below the surface. The resulting time-space distribution map illustrates the evolution path of heat anomalies, providing a complete reference for freezing risk identification. This reference can be used to determine whether heat is being transferred from bottom to top in the tunnel entrance area and the persistence of heat accumulation, thereby identifying areas with a high risk of freezing.

[0059] Based on the evolution characteristics of the heat accumulation range in the time series, the starting time of the formation of the freezing core, the temperature change amplitude and distribution direction are extracted to generate a time distribution table reflecting the freezing trend, which can be used to provide a time series reference for active heat intervention.

[0060] After clarifying the extent and spatial distribution of subsurface heat accumulation, to further understand the formation process of the frozen core, it is necessary to continuously track the evolution characteristics of the heat accumulation range over time. This will allow us to identify the initial period of frozen core formation, the magnitude of temperature changes, and the direction of distribution, and generate a time distribution table that reflects the freezing trend. The specific steps are as follows:

[0061] Within the heat accumulation area identified in the previous stage, temperature data from different depths and horizontal locations are arranged as a continuous time series, with time as the horizontal axis, forming a time-extended sequence of the heat accumulation range. This sequence reflects the temperature change trend over time at different locations below the surface. In this process, the focus is on analyzing the temperature change curve over time within the 0.5-meter to 1.5-meter depth range below the surface. When the curve continuously decreases over a certain period with a slowing gradient, it indicates that the stratum is beginning to enter a state of heat dissipation, potentially forming the initial conditions for a freezing core. When the temperature decreases rapidly and remains stable at a low temperature, it indicates that the freezing core is already in the formation stage. By continuously arranging these trends, segments with sustained low-temperature characteristics in the temperature field can be identified. Each segment represents a time node where heat within the stratum transitions from accumulation to solidification, providing a temporal clue for subsequent freezing trend extraction. The key to this stage is to identify the critical transition zone from stable to solidified heat accumulation range through temporal continuity analysis.

[0062] Based on the time-extended sequence of heat accumulation range, temperature change curves collected at different depths and directions were compared. When the temperature drop in shallow subsurface layers precedes that in deeper layers and lasts for a shorter period, it indicates that the cold source is penetrating rapidly from the surface into the interior. Conversely, when the temperature drop in deeper layers is delayed and lasts for a longer period, it indicates that cold energy is gradually accumulating underground and forming a frozen core. In the lateral direction, if the temperature drop at the leading edge of the cave entrance is faster than that at the toes of the slopes on both sides of the entrance, it indicates that the influence of cold air propagating along the axial direction of the entrance is stronger, and the formation of the frozen core has a clear directionality. By comparing the magnitude and duration of temperature drops in different directions, the starting time and main formation direction of the frozen core can be determined. For example, the period during which the temperature continues to drop and remain low on the windward side of the entrance is usually earlier than on the leeward side, indicating that the frozen core begins to form on the windward side and gradually expands into the cave. Through this multi-point, multi-directional comparative analysis, not only can the starting time of frozen core formation be clearly identified, but its spatial formation direction can also be determined, providing accurate time and directional data support for the establishment of a freezing trend table.

[0063] After determining the initial formation period of the frozen core, the temperature change curves after that period are tracked to analyze its temporal growth process. When the temperature at a certain depth remains below zero degrees Celsius for more than 12 hours, and the rate of temperature decrease gradually slows down, it indicates that the frozen core at that depth has stabilized. At this point, the temperature change amplitude before and after the formation of the frozen core at different depths is quantified to obtain the degree of temperature reduction within the strata. If the temperature decrease exceeds 3 degrees Celsius at a depth of 1 meter below the surface, while the decrease is smaller at 1.5 meters, it indicates that the frozen core is concentrated in the shallow region; if the temperature decrease amplitudes at multiple depths are similar, it indicates that the frozen core is expanding as a whole in the vertical direction. In the horizontal direction, if the freezing duration on the windward side of the cave entrance is longer and the temperature recovery is slow, it indicates a stronger trend of the frozen core extending into the cave. Through synchronous analysis of time and temperature changes, the complete time process of the frozen core from formation to stabilization can be obtained. After organizing these time periods and corresponding temperature change amplitudes, the intensity change of the freezing trend over time can be clarified, providing a detailed data basis for generating a time distribution table of freezing trends.

[0064] After completing the temporal evolution analysis of the frozen core, the onset time, duration, temperature decrease rate, and spatial distribution direction of the frozen core in each time period were summarized to form a time distribution table reflecting the freezing trend. The time distribution table records the formation process of the frozen core in hourly or minute-based time steps, with each row corresponding to a specific time period, including temperature changes at different depths below the surface, freezing duration, and information on the expansion of the frozen core along the cave entrance or slope direction. The time distribution table visually displays the persistence and diffusion of the freezing trend. For example, when the frozen core forms earlier and for a longer duration on the windward side of the cave entrance, while forming later and for a shorter duration on the leeward side, it indicates that the freezing trend mainly develops along the direction of airflow inflow. The time distribution table also shows the growth rate of the frozen core in different time periods; a faster temperature decrease corresponds to a freezing trend intensification phase, while a more stable temperature change corresponds to a freezing trend deceleration phase. This time distribution table provides a precise temporal reference for subsequent active thermal regulation, enabling thermal intervention to be carried out before the frozen core is fully formed, thereby reducing the accumulation of freezing risks. Through this process, the temporal evolution of freezing trends is systematically expressed, temperature changes at various depths below the surface are quantitatively recorded, and the spatial development direction of freezing cores is specifically presented, providing a reliable time control basis for the prevention and control of freezing damage at tunnel entrances in cold regions.

[0065] Based on the freezing trend time distribution table, and taking into account the ventilation airflow intensity, air supply rhythm and airflow direction angle parameters, heat reverse guidance command is calculated to enable the airflow to dynamically adjust the heat in the potential freezing area in time and direction.

[0066] After obtaining the freezing trend time distribution table, in order to actively adjust the airflow to the potential freezing area in terms of time and direction, it is necessary to integrate the freezing trend information with the control parameters of the ventilation airflow in both time and space dimensions. By comprehensively coordinating airflow intensity, air supply rhythm, and airflow direction angle, a heat-reverse guidance command can be generated to guide the operation of the ventilation control device. The specific steps are as follows:

[0067] The freezing trend time distribution table records the onset time, duration, temperature change amplitude, and spatial expansion direction of the freezing core formation. To ensure that airflow regulation behavior aligns with the changes in the freezing trend, this time data is used as a time control benchmark to time-match various parameters of ventilation airflow. Specifically, the initial time period of freezing core formation is mapped to the initial response phase of airflow regulation, the period of rapid increase in the freezing trend is mapped to the airflow enhancement phase, and the period of stabilization or weakening of the freezing trend is mapped to the airflow recovery phase. Within each time period, basic ranges for airflow velocity, airflow rate, air supply duration, and airflow direction angle are determined. For example, when the freezing core is in its initial stage, the airflow velocity is set to a medium range, and the air supply duration is relatively short; when the freezing core enters the enhancement phase, the airflow velocity increases, and the air supply duration is extended; when the freezing trend enters the weakening phase, the airflow velocity gradually decreases, and the air supply interval is extended. By mapping the airflow rhythm to the time rhythm of the freezing trend, subsequent airflow regulation can precisely target the critical periods of freezing core formation and development, avoiding a disconnect between ventilation behavior and the rhythm of heat change. This step forms the overall framework for airflow control in the time dimension, providing a time coordinate for reverse heat guidance.

[0068] After clarifying the temporal correspondence, an airflow intensity distribution method was formulated based on the spatial distribution characteristics of the freezing trend in the tunnel entrance area. When the freezing core extends mainly along the tunnel axis, the airflow direction is tilted forward axially, creating a longitudinal transmission path along the tunnel entrance, allowing heat to be directionally transferred from inside the tunnel to the frozen area. When the freezing core is concentrated at the bottom of the tunnel entrance, the near-surface airflow intensity is increased, causing the airflow to flow close to the ground, enhancing the exchange of heat between the surface and the air. When the freezing trend is mainly distributed at the top of the tunnel entrance, the airflow angle is adjusted upwards, concentrating heat upwards and preventing continuous cooling in the top area. When the freezing core is asymmetrically distributed, the airflow intensity is adjusted proportionally according to the temperature drop rate, increasing the airflow intensity on the side with a faster temperature drop and maintaining a lower airflow on the side with a slower temperature drop, thus creating lateral airflow compensation. In this way, the airflow speed and direction are spatially matched with the distribution direction of the freezing trend, enabling heat to be accurately transferred to the area with the fastest temperature drop and the highest freezing risk in a short time. Each adjustment of wind distribution is based on the freezing trend time distribution table to ensure that airflow regulation and the spatial evolution of freezing phenomena remain synchronized.

[0069] After clarifying the time period and spatial distribution, the airflow rhythm within each time period needs to be specifically designed to ensure that the heat delivered by the airflow matches the rhythm of the freezing core formation. To avoid overshoot or lag in the heat output and cold air exchange process, the airflow cycle and pause time should be rationally arranged. When the freezing core is just beginning to form, a short-cycle intermittent airflow method is used to allow heat to enter the cave entrance area in stages, preventing excessive cold air inflow from causing reverse temperature fluctuations. When the freezing trend enters a rapid growth phase, a long-cycle continuous airflow method is used to continuously push heat flow towards the potential freezing area. When the freezing trend weakens, the short-cycle airflow mode is restored, allowing heat to form a temperature buffer zone around the freezing core, preventing heat loss due to excessive airflow. The airflow angle is adjusted in real time according to the expansion direction of the freezing core. For example, when the freezing core extends into the cave, the airflow angle is tilted inward by 10 to 15 degrees, allowing heat to penetrate deep into the cave entrance; when the freezing core expands outward along the slope, the airflow angle rises, allowing heat to dissipate along the slope. When the frozen core is located below the opening, the airflow angle shifts downwards, causing warm air to flow close to the ground and couple with the surface heat for transfer. In this way, the airflow rhythm and direction angle are matched in time and space, giving the heat guidance process a dynamic control characteristic.

[0070] After completing airflow timing matching, wind force distribution, and direction control, the airflow parameters determined for each time period are summarized to form a time series of heat guidance instructions. The records for each time period include airflow velocity, flow rate, air supply duration, air supply interval, airflow direction angle, and the corresponding rate of change in freezing trend. By continuously arranging these parameters, a time-correlated instruction table can be formed. This instruction table is based on the freezing trend time distribution table and uses airflow regulation parameters as the main body, reflecting how airflow gradually achieves heat guidance and freezing suppression over time. For example, in the initial stage of freezing core formation, the corresponding instruction table records a moderate airflow velocity, an air supply duration of 5 minutes, and an upward flow direction angle; in the stage of increasing freezing trend, the instruction table records an airflow velocity increase of 30%, an air supply duration extended to 15 minutes, and a flow direction angle tilted along the direction inside the cave; in the stage of weakening freezing trend, the airflow velocity decreases by 20%, the air supply duration is shortened, and the angle returns to a horizontal position. The heat guidance instructions formed in this way not only have temporal continuity but also spatial specificity, clearly defining the airflow control methods adopted at different freezing stages. The resulting heat reverse guidance command can directly instruct the ventilation equipment to carry out air supply operations according to the specified time and direction, realizing dynamic heat regulation of airflow in time and space.

[0071] Based on the instruction of reverse heat guidance, phased air supply, intermittent heat exchange and directional flow operation are implemented. Through layer-by-layer heat conduction and local airflow peak compensation, the heat backflow phenomenon caused by temperature inversion is smoothly dispersed, and the formation conditions of the freezing core below the ground surface are effectively eliminated, thereby completing the process of frost damage identification and risk elimination at the entrance of the tunnel in the cold region.

[0072] Upon receiving the instruction to reverse the heat flow, in order to enable the airflow within the tunnel entrance area to dynamically respond to the freezing trend in time and direction, it is necessary to carry out phased air supply, intermittent heat exchange, and directional recirculation operations according to the instruction. Through continuous airflow organization and heat regulation, the heat backflow phenomenon caused by temperature inversion is gradually dispersed, and the conditions for the formation of a freezing core below the surface are eliminated. The specific steps are as follows:

[0073] In the heat guidance command, different time periods correspond to different airflow intensities, ventilation rhythms, and flow angles. To ensure the phased nature of the airflow regulation process, the airflow is controlled in three stages according to the time command sequence: the initiation stage, the enhancement stage, and the decay stage. The initiation stage mainly corresponds to the initial formation stage of the freezing core. During this stage, the ventilation speed is maintained within a medium range, the ventilation duration is short, and the airflow direction is slightly inclined towards the inside of the tunnel, allowing air to flow into the tunnel entrance and carry heat from the near-surface layer outward. The enhancement stage corresponds to the rapid growth stage of the freezing trend. At this time, the ventilation speed is increased and the ventilation time is extended to form a stable mainstream heat transport zone, allowing the heat accumulated inside the tunnel entrance area to diffuse outward through air convection. The decay stage corresponds to the weakening stage of the freezing trend. The ventilation speed is gradually reduced, and the intermittent ventilation mode is restored, allowing heat to form a temperature equilibrium layer around the tunnel entrance, avoiding sudden drops in heat or excessive airflow disturbances that could cause the re-formation of temperature inversion. Through this phased ventilation method, the airflow intensity is synchronized with the changes in the freezing trend, making heat transfer directional and hierarchical, laying the foundation for subsequent intermittent heat exchange and directional return flow.

[0074] After each stage of air supply lasts for a certain period, a short interval is set to allow the airflow to form a local stagnation layer in the tunnel opening space. During the interval, hot air undergoes natural diffusion and heat redistribution at different height levels around the tunnel opening. Through the alternation of air supply and pauses, heat is no longer concentrated in a single direction of flow, but rather a relatively balanced temperature gradient is formed between the tunnel opening edge, the ground surface, and the sidewalls. During the air supply interval, the heat released from the ground surface and the lining surface can be slowly conducted to the air layer under conditions without strong airflow interference, compensating for the local cooling effect caused by continuous air supply. When air supply is restarted, the new airflow transports the accumulated heat out of the tunnel, completing a full heat exchange cycle. Each air supply interval cycle is set according to the rate of change of the freezing trend: when the freezing trend intensifies, the interval time is shortened to ensure continuous heat output; when the freezing trend weakens, the interval time is extended to prevent overheating accumulation. Through intermittent heat exchange operations, the heat conduction and dissipation processes of the airflow in the tunnel opening space are kept in dynamic equilibrium, forming a continuous heat dissipation mechanism.

[0075] Directional recirculation refers to the periodic alteration of part of the airflow direction while maintaining a generally consistent airflow direction. This creates a layered, alternating circulation of air within the tunnel entrance area. In practice, part of the airflow angle is adjusted from a forward-leaning direction to an upward-leaning direction, causing the air to flow along the top of the tunnel entrance and recirculate back to the area above the leading edge. Simultaneously, another portion of the airflow diffuses outward along the ground, forming vertical convection channels. When hot air recirculates along the top, it helps to dissipate heat from the tunnel roof and upper lining area, preventing condensation and freezing due to excessively low temperatures. When the airflow expands outward along the bottom, it disperses cold air from the surface and bottom of the tunnel, allowing the temperature of lower areas to recover. The frequency of directional recirculation is directly proportional to the airflow velocity. When the airflow velocity is high, the recirculation switching cycle is shortened to maintain the stability of the airflow circulation; when the airflow velocity is low, the recirculation cycle is extended to maintain the continuity of heat diffusion. By using directional flow reversal, the airflow forms a multi-layered heat conduction structure in the cave entrance area. The upper airflow is responsible for dissipating the heat from the top, while the lower airflow is responsible for dissipating the cold from the ground surface, thus eliminating the heat backflow phenomenon caused by temperature inversion as a whole.

[0076] During execution, following the time sequence of the heat reverse guidance command, the three operating modes were combined in stages: In the early stage of frozen core formation, short-cycle air supply combined with long-intermittent heat exchange allowed the airflow to gradually dissipate shallow cold in a gentle manner; during the rapid growth stage of the freezing trend, continuous air supply combined with frequent directional recirculation created a stable upward convection channel in the cave entrance area, continuously releasing the heat accumulated below the surface into the air layer; during the weakening stage of the freezing trend, low-speed air supply combined with extended intermittent time slowed airflow, resulting in a uniform heat distribution around the cave entrance. Through coordinated control of air supply, heat exchange, and recirculation, the airflow was made periodic in time and stratified in space, thereby eliminating sudden changes in local temperature differences. Heat was released layer by layer from the inside out in the cave entrance area, the temperature gradient below the surface gradually returned to normal, and the frozen core naturally melted due to the loss of sustained low temperature and moisture conditions. The entire process formed a closed-loop control path from heat identification to heat guidance to heat release. Ultimately, the heat backflow caused by the temperature inversion was completely dissipated, the temperature field in the tunnel entrance area returned to stability, and the risk of frost damage to tunnels in cold regions was fundamentally eliminated.

[0077] This invention establishes a dynamic correlation between temperature inversion and heat distribution over time by integrating multi-source data on changes in air temperature, surface temperature, residual snow evaporation, and wind pressure. This enables accurate identification of heat transfer reversals under local microclimate changes. By correlating the variation curves with the heat accumulation range, this method clearly depicts the spatiotemporal evolution of freezing core formation. This allows for the early identification of risk areas before freezing conditions stabilize, effectively preventing shallow heat retention at tunnel entrances and the formation of potential ice expansion centers. This provides continuous and precise criteria for assessing frost damage in cold-region tunnel operations.

[0078] This invention generates heat-reverse guidance commands and combines them with the intensity, rhythm, and direction parameters of ventilation airflow to achieve active heat conduction and precise regulation of airflow in time and space. This method dynamically organizes the air supply and heat exchange process based on a freezing trend time distribution table, allowing the airflow to directionally dissipate heat along the potential freezing zone, gradually eliminating the heat backflow effect caused by temperature inversion. Through the synergistic effect of staged air supply and directional return flow, the temperature field in the tunnel entrance area is restored to equilibrium, and the conditions for freezing nucleus formation are fundamentally destroyed, thereby significantly improving the freeze-resistance stability and operational safety of the tunnel entrance.

[0079] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion, characterized in that, Includes the following steps: To identify frost damage at tunnel entrances, air temperature, surface temperature, residual snow evaporation, and wind pressure changes were simultaneously collected in the tunnel entrance area. All collected data were then fused and analyzed according to a unified time series to generate a joint change curve. Based on the joint variation curve, a continuous analysis is performed on the regions where the near-surface temperature difference shows an inverse change in the time series, identifying anomalous regions where heat is transferred from bottom to top, and determining the range of heat accumulation below the surface based on the spatial distribution of the anomalous regions. Based on the evolution characteristics of the heat accumulation range over time, the starting time of freezing core formation, temperature change amplitude and distribution direction are extracted to generate a time distribution table reflecting the freezing trend. Based on the time distribution table of freezing trend, the heat reverse guidance command is calculated by taking into account the parameters of ventilation airflow intensity, air supply rhythm and airflow direction angle. Based on the instruction to guide heat in the opposite direction, phased air supply, intermittent heat exchange and directional flow operation are implemented. Through layer-by-layer heat conduction and local airflow peak compensation, the heat backflow phenomenon caused by temperature inversion is smoothly dispersed, and the conditions for the formation of frozen cores below the surface are effectively eliminated.

2. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion as described in claim 1, characterized in that, The steps for generating the joint change curve are as follows: In the tunnel entrance area of ​​cold regions, based on the topographic features, orientation conditions, air flow direction and surrounding cover distribution, air temperature collection points, surface temperature collection points, residual snow evaporation collection points and wind pressure change collection points are set up, and a continuous coverage of air flow field, surface temperature field, latent heat of evaporation field and wind pressure disturbance field is formed along the tunnel axis. Maintain a consistent sampling frequency at all collection points and record data on changes in air temperature, surface temperature, residual snow evaporation, and wind pressure at a uniform time step to form a continuous time-series data chain covering the air, surface, and strata inside and outside the cave. A corresponding analysis was conducted on the time series of changes in air temperature, surface temperature, residual snow evaporation and wind pressure. The magnitude and direction of changes in different time intervals were compared to establish the heat exchange relationship between the air layer and the surface layer. By forming a joint change curve from continuous time data, the time periods in which the direction of temperature change reverses are identified, and the direction and range of heat transfer are determined through spatial mapping.

3. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 2, characterized in that, After forming the joint change curve, the continuous time data of air temperature, surface temperature, residual snow evaporation and wind pressure changes are segmented and analyzed. The time intervals in which the temperature change direction reverses are extracted separately, and the temperature values ​​and change rates in each time period are recorded. Based on the time node data, the direction of heat transfer in the cave entrance area is spatially mapped to determine the range of heat transfer from the air layer to the surface or from the surface to the air layer when the inversion is formed.

4. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 2, characterized in that, The steps for determining the heat accumulation range are as follows: By pairing and analyzing air temperature data and surface temperature data within the same time interval in the joint variation curve, a temperature difference change sequence is formed, and the moment when the direction of heat conduction changes is determined by identifying the alternation point where the temperature difference changes from positive to negative or from negative to positive. Taking each temperature inversion point as the center, extract air temperature and surface temperature data within the time range before and after, analyze the duration and trend of heat transfer from bottom to top, and determine the time interval of the temperature inversion phenomenon. By combining the temperature changes at surface temperature collection points at different depths, a joint analysis of the subsurface temperature in the vertical and horizontal directions is conducted to determine the spatial range of heat accumulation upwards within the strata. By overlaying the spatial extent of heat accumulation with the duration of reverse transfer, a distribution map of heat accumulation intensity over time is plotted to identify areas of sustained heat retention and determine areas with high risk of freezing.

5. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 4, characterized in that, The steps to generate a time distribution table are as follows: Within the heat accumulation range below the surface, temperature data at different depths and horizontal positions are arranged in time series with time as the horizontal axis to form a time extension series of the heat accumulation range, and time segments in which the temperature continues to decrease and the gradient slows down are identified. Based on the time-extended sequence, temperature change curves at different depths and in different directions are compared to determine the starting time and main formation direction of frozen nuclei, and regions where the temperature drops earlier than other locations are identified as the starting point for frozen nuclei formation. Continuously track the temperature change process during the formation period of the frozen nucleus, extract the duration and magnitude of the temperature drop, and determine the expansion trend of the frozen nucleus along the depth and horizontal directions; The start time, duration, temperature change amplitude, and distribution direction of each time period are summarized to generate a freezing trend time distribution table.

6. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 5, characterized in that, The freezing trend time distribution table records the formation process of the frozen core based on time. Each time period includes temperature change values ​​at different depths, freezing duration, and information on the expansion of the frozen core along the cave entrance. The intensification and deceleration phases of the freezing trend are determined by the change in the rate of temperature decrease.

7. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 5, characterized in that, The steps for generating the heat reverse guidance command are as follows: Using the start time, duration, temperature change amplitude, and spatial expansion direction in the freezing trend time distribution table as the time control benchmark, the speed, flow rate, air supply time, and air flow direction angle of the ventilation airflow are matched in time to form a time framework for airflow regulation. Based on the spatial distribution characteristics of the freezing trend in the cave entrance area, a distribution method for airflow intensity in different directions is formulated so that the airflow forms a transmission path along the cave entrance axis, top and ground direction, thereby realizing the directional transfer of heat in space. By combining the results of airflow time matching and spatial allocation, the air supply rhythm and airflow direction angle are coordinated and designed, and a time and space matching relationship is formed by intermittent air supply and angle adjustment. The airflow speed, flow rate, air supply duration, air supply interval, and airflow direction angle for each time period are summarized to generate a heat reverse guidance command.

8. The method for identifying frost damage at tunnel entrances in cold regions based on multi-source environmental data fusion according to claim 7, characterized in that, Based on the heat reverse guidance instruction, phased air supply, intermittent heat exchange, and directional recirculation operations are implemented. Through layer-by-layer heat conduction and local airflow peak-shifting compensation, the following steps are taken to dissipate heat and eliminate the conditions of the subsurface freezing core: Based on the time sequence of the heat reverse guidance command, the airflow is divided into the start-up stage, the enhancement stage and the decay stage, and the airflow speed, duration and direction are controlled respectively, so that the airflow forms a stable heat transfer channel in different stages. During the phased air supply process, air supply intervals are set to allow hot air to diffuse and redistribute within the tunnel opening space, forming a dynamic heat exchange balance between the air layer and the surface layer. Based on intermittent heat exchange, directional recirculation operation is implemented to periodically change the direction of some airflow, so that the air forms a stratified and alternating circulation flow inside the cave entrance area, dispersing the cold air on the surface and dissipating the heat at the top. By coordinating and combining phased air supply, intermittent heat exchange, and directional recirculation operations, heat is released layer by layer from the inside out, restoring the surface temperature gradient and eliminating the conditions for the formation of a frozen core.

Images