A method for determining a target phase change temperature and a target phase change enthalpy value of a phase change material based on deicing and snow melting requirements in different regions

Abstract

This invention discloses a method for determining the target phase change temperature and target phase change enthalpy of phase change materials based on the snow melting and de-icing needs of different regions, belonging to the technical field of road snow melting and de-icing and phase change energy storage parameter design. The method includes: obtaining the latitude, altitude, winter minimum temperature, duration near freezing point, snowfall intensity, and total winter solar radiation of the target region; determining the characteristic winter temperature based on latitude and altitude; automatically identifying the snow melting and de-icing demand type by combining the above parameters; inputting the regional parameters and demand type into a phase change material thermal parameter prediction model to obtain the predicted values ​​of the target phase change temperature and target phase change enthalpy; after engineering constraint correction, matching with a candidate phase change material library, and outputting recommended phase change temperature range, phase change enthalpy range, and candidate material category. This invention can provide a basis for determining phase change snow melting and de-icing parameters for transportation infrastructure such as roads, bridge decks, airport runways, and railway turnouts.

Description

Technical Field

[0001] This invention relates to the field of road snow melting and de-icing and phase change energy storage parameter design technology, specifically to a method for determining the target phase change temperature and target phase change enthalpy value of a phase change material based on the snow melting and de-icing needs of different regions. Background Technology

[0002] In winter, roads, bridges, airport runways, and railway switches are prone to snow accumulation, icing, and freezing damage under low temperatures, snowfall, and freezing rain. This leads to a decrease in the road surface friction coefficient, poor tire adhesion, and further reduces traffic efficiency, increases operational safety risks, and deteriorates the durability of the road surface structure. To address the need for snow melting and de-icing, existing technologies have developed into various approaches, including mechanical removal, chemical de-icing agents, electric heating, fluid heating, geothermal utilization, phase change thermal storage, and intelligent control. Mechanical removal methods are highly dependent on equipment, involve repetitive operations at high costs, and easily cause wear and tear on the road surface. While chemical de-icing agents are fast-acting, long-term use can lead to steel corrosion, reduced concrete durability, asphalt aging, and environmental burden along the route. Active heating systems can achieve relatively direct snow melting and de-icing effects, but their construction and operating costs are high, and they have stringent requirements for energy supply, pipelines, and subsequent maintenance. Phase change materials, due to their ability to absorb or release latent heat during phase change, are considered a potential technological approach for passive snow melting, de-icing, and road surface thermal regulation.

[0003] Existing patent literature has explored road snow and ice removal technology extensively. For example, CN115433544A / CN115433544B discloses a green and environmentally friendly road snow melting and anti-skid composite material and its preparation method, which prepares an environmentally friendly acetate snow melting agent and combines it with volcanic ash particles to achieve both snow melting and anti-skid functions; CN221877933U discloses an electric heating snow melting and de-icing device for high-altitude airport runways. This type of technology generally belongs to the device scheme of arranging electric heating units in the runway structure and cooperating with control devices to achieve active snow melting and de-icing; CN113531919B discloses... A multi-source complementary railway turnout snow melting and de-icing system and method were developed, which combines solar cascade heat storage, geothermal supplementation, and electric heating emergency measures to reduce the power consumption and capacity expansion costs of traditional pure electric heating systems; CN114383341B discloses an active snow and ice melting system for road surfaces and its control method, which controls the start-up, operation, and shutdown process of a ground source heat pump system by acquiring data such as road surface temperature, road ice and snow condition images, regional meteorological feedback information, and ambient temperature; CN115162096 CN115795622B discloses a temperature control method and system for road snow melting and ice removal using fluid heating. This method determines the initial parameters of the fluid heating system by acquiring data on road snow accumulation and meteorological data, and dynamically adjusts the system heating temperature based on real-time road surface temperature during the snow melting process. CN113930220A discloses a design method for an active bridge deck de-icing and snow melting system using energy piles based on machine learning. This method uses a machine learning model to predict the snow melting efficiency and thermal stress of the bridge deck buried pipe structure, and determines the design scheme of the active bridge deck snow melting system accordingly. CN113930220A discloses a steel slag-based binary composite phase change material for road surfaces and its preparation method. It has been recognized that the phase change temperature and latent heat can be controlled by combining two phase change materials to adapt to different environmental temperature conditions. CN116631544A discloses a method for determining the performance parameters of solid-solid phase change materials by constructing a finite element model based on target area climate data and road structure type. This indicates that designing phase change material parameters based on regional environmental conditions has become an important research direction in the field of road functional materials.

[0004] The aforementioned technologies address some issues in road snow and ice removal from the perspectives of device heating, system control, material preparation, parameter design, and data-driven modeling. However, from a higher-level engineering adaptation perspective—the difference in snow and ice removal needs across different regions—the existing technologies are still insufficient. On the one hand, the protection focus of existing technologies is mostly on specific material formulations, heating structures, buried pipe configurations, heat pump systems, or finite element parameter design processes, lacking a universal method that can determine the thermal parameters of phase change materials directly for target regions, independent of specific material systems. On the other hand, even when considering regional differences, existing technologies often rely on empirical ratios, simple climate zoning, or regional finite element trial calculations, failing to establish a reusable mapping mechanism of "geospatial parameters—key regional parameters in the cold season—target phase change temperature / target phase change enthalpy."

[0005] Furthermore, the needs for snow melting and de-icing vary across different regions, and this difference is not determined solely by the "ambient temperature" factor, but rather by the combined effects of multiple regional characteristics. Significant differences exist between regions in terms of latitude, altitude, minimum winter temperatures, sustained temperature characteristics near freezing point, snowfall intensity, and solar radiation conditions. For some regions, the main problem is not heavy snow cover, but rather the persistent fluctuation of ambient temperature around freezing point, leading to the formation of transparent thin ice or black ice on the road surface. These regions require a greater focus on preventing and mitigating icing. For other regions, high snowfall intensity and a large snow melting heat load necessitate attention to whether the heat provided per unit area by phase change materials is sufficient to support continuous snow melting. In high-altitude, high-radiation regions, abundant daytime solar radiation and significant diurnal temperature variations result in snow melting and de-icing needs that differ significantly from those in low-radiation regions, making a synergistic utilization model of daytime heat storage and nighttime heat release more suitable. Therefore, the division of "different regions" should not be based solely on administrative divisions, average annual temperature, or simple cold region classifications. Instead, it should be based on technical analysis and demand identification of multi-source regional characteristics directly related to snow melting and ice removal activities.

[0006] Meanwhile, the actual conditions for road icing are not necessarily equivalent to an ambient temperature reaching 0°C. In practical engineering, road surface icing behavior is usually related to the regional low-temperature boundary, the duration of time near the freezing point, snowfall processes, solar radiation conditions, and the road surface thermal response process. For areas with small fluctuations close to the freezing point, even if the absolute minimum temperature is not extreme, the longer time spent near the freezing point may still result in a higher risk of black ice formation. For areas with heavy snowfall, even if the icing period is short, the higher latent heat support may be required due to the larger snow melting heat load per unit time. For areas with strong solar radiation in winter, the level of external heat input will also affect the heat storage and release process of phase change materials. Therefore, selecting phase change materials based solely on the regional average temperature or a single indicator often fails to accurately reflect the service thermal environment of the target area near the freezing point, and also fails to simultaneously meet multiple requirements such as anti-icing, snow melting, and heat utilization efficiency.

[0007] For phase change materials (PCMs), phase change temperature and phase change enthalpy are two core thermal parameters that determine their engineering suitability. Phase change temperature determines when the PCM begins to absorb or release heat, directly corresponding to the material's response timing to thermal disturbances near the road surface's freezing point. If the phase change temperature is too low, the material may not have started releasing heat during the critical anti-icing stage, making it difficult to suppress initial icing in time. If the phase change temperature is too high, the material may complete the phase change prematurely, resulting in a misalignment between the latent heat release period and the target snow melting and de-icing requirements. Phase change enthalpy determines the total latent heat that the PCM can provide per unit mass or unit volume, directly affecting its ability to support snow melting, temperature maintenance, and thermal fluctuation buffering. When the phase change enthalpy is insufficient, even with a reasonable phase change temperature, it may be difficult to meet the actual snow melting heat load due to insufficient heat release. Conversely, when the phase change enthalpy is too high, it may lead to increased material implementation difficulty, limited encapsulation conditions, increased costs, or incompatibility with road structures. Therefore, the phase transition temperature determines "when it takes effect", and the phase transition enthalpy determines "how much it can take effect". Both directly affect the thermal response characteristics of the road surface near the freezing point and the snow melting and de-icing effect. Moreover, the two are not completely independent, but should be jointly determined according to the regional needs.

[0008] However, existing publicly available technologies typically treat phase change temperature and enthalpy as inherent parameters of a specific phase change material or as results obtained through material blending, lacking a technical approach that jointly optimizes phase change temperature and enthalpy based on regional needs. In particular, existing technologies lack clear, rigorous, and reusable criteria for identifying the need to "focus on preventing black ice, slowing icing, melting snow, or utilizing winter solar radiation for heat compensation in different regions." They also lack a unified methodological framework for determining "which regional parameters should be used as primary input variables and what quantitative relationships exist between these variables and the target phase change temperature and enthalpy." Furthermore, existing technologies are insufficiently disclosed on "how to further match candidate phase change material libraries based on regional conditions to obtain feasible material parameter ranges or candidate material combinations suitable for the region."

[0009] Furthermore, from the perspective of engineering application and patent implementation feasibility, while incorporating wind speed, humidity, structural layer thickness, thermal conductivity, specific heat capacity, exposure status, and various correction coefficients into the main variable system can increase the descriptive dimensions, it also significantly increases the difficulty of data acquisition, sample annotation complexity, and implementation support, hindering the formation of a concise, stable, and easily scalable methodological framework. In contrast, latitude and altitude can characterize the basic geospatial conditions of the target region and can be used to determine the characteristic winter temperatures of the target region; the lowest winter temperature and the duration near the freezing point can be used to characterize the temperature triggering requirements of the phase change material (PCM) for "when it will take effect"; snowfall intensity and total winter solar radiation can be used to characterize the heat requirements of the PCM for "how much heat needs to be released" and "whether there is external heat compensation". This approach can reflect the differences in snow melting and de-icing needs in different regions and is also conducive to model construction, implementation verification, and engineering implementation.

[0010] Therefore, there is an urgent need to provide a method for determining the thermal parameters of phase change materials based on the snow melting and de-icing needs of different regions. This method should first determine the characteristic winter temperature of the target region based on its latitude and altitude, and then use the characteristic winter temperature, the minimum winter temperature, the duration of the freezing point, the snowfall intensity, and the total winter solar radiation as key input parameters to identify and classify the snow melting and de-icing needs of the region. This method should establish a quantitative relationship between regional conditions and the target phase change temperature and the target phase change enthalpy, and further match the material with a candidate phase change material library. This will provide a basis for the design of phase change snow melting and de-icing for roads, bridges, airport runways, and railway-related infrastructure in different regions. Summary of the Invention

[0011] The purpose of this invention is to provide a method for determining the target phase change temperature and target phase change enthalpy of a phase change material based on the snow melting and de-icing needs of different regions, so as to solve the technical problems in the prior art, such as insufficient regional adaptation basis, reliance on experience for parameter selection, non-closed chain of demand identification and thermal parameter determination, unclear engineering correction rules, and lack of quantitative standards for material matching process.

[0012] To achieve the above objectives, this invention provides a method for determining the target phase transition temperature and target phase transition enthalpy of a phase change material based on the snow melting and de-icing needs of different regions, comprising the following steps: S1. Obtain the geospatial parameters and key regional parameters of the target area; wherein, the geospatial parameters include latitude (Lat) and altitude (Alt), and the key regional parameters include the lowest winter temperature. Duration near freezing point Snowfall intensity and total solar radiation in winter .

[0013] S2. Determine the characteristic winter temperature of the target region based on latitude (Lat) and altitude (Alt). .

[0014] Preferably, the characteristic winter temperature The first-stage winter characteristic temperature model was determined to be a linear regression model, and its expression is as follows:

[0015] in, , and The regression coefficients are obtained by fitting historical winter temperature data of the sample area.

[0016] S3. Based on the winter temperature characteristics of the target region Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter In addition to the altitude Alt, the system automatically identifies the demand in the target area and obtains the snow melting and de-icing demand type D for the target area.

[0017] Preferably, the snow melting and de-icing demand type D includes black ice high-risk anti-icing type, light snow rapid melting type, continuous snow accumulation high heat demand type, extreme cold low temperature continuous response type, and high radiation day and night regulation type.

[0018] Furthermore, the automatic demand identification includes calculating scores G1, G2, G3, G4, and G5 for five categories of snow melting and de-icing demands, respectively, wherein: Black ice high-risk anti-icing type score G1 is: G1=2×I( ≥450)+I( ≥-12)+I(-2< ≤2)+I( <0.30) The score for G2 in the light snow rapid melting type is: G2=I(300≤ <550)+I( ≥-10)+I( ≥0)+I( ≥0.20)+I( <1000) The score for G3, representing sustained snow cover and high demand, is: G3=2×I( ≥0.21)+I( ≤-20)+I( ≤-8)+I(100≤ ≤260) The extreme cold sustained low temperature response score G4 is: G4=2×I( ≤-10)+2×I( ≤-28)+I( <150) The score for the high-radiation diurnal regulation type G5 is: G5=3×I( ≥1050)+I(Alt≥2000)+I( ≥350) Where I(·) is the indicator function, which takes the value 1 when the condition is met and 0 otherwise. The demand type corresponding to the maximum value among G1, G2, G3, G4 and G5 is taken as the demand type D of the target area; when there are ties for the maximum value, the priority is determined according to the following: high radiation day and night regulation type, extreme cold and low temperature continuous response type, continuous snow accumulation and high heat demand type, black ice high risk anti-icing type, and light snow rapid melting type.

[0019] The aforementioned thresholds are not arbitrarily set, but are determined by combining the parameter distribution of the sample area, the experience of snow melting and de-icing projects in the region, and the verification results of typical cities. Without changing the demand identification logic, the thresholds can be adjusted according to the statistical period, data source, or engineering standards.

[0020] S4. Winter characteristic temperature based on sample regions Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter Based on the automatically identified demand type D, supervised output data of target phase change temperature and target phase change enthalpy corresponding to the sample region are constructed, and a prediction model for phase change material thermal parameters is established.

[0021] Preferably, the monitoring output data is determined using a standardized engineering annotation method, including the center value of the phase transition temperature in the engineering annotation. and the central value of phase transition enthalpy marked in the engineering. .

[0022] When the snow melting and de-icing requirement type D is the high-risk black ice prevention type: =1.0+0.5×I( ≥-15) +0.5×I( ≥500) =95+10×I( ≥450) +5×I( ≥0.25) When the snow melting and de-icing requirement type D is light snow rapid melting type: =0.5+0.5×I( ≥600) +0.5×I( ≥0.25) =90+10×I(0.18≤ <0.25) +5×I( ≥500) When the snow melting and de-icing demand type D is the continuous snow accumulation and high heat demand type: =-2.0-0.5×I( ≤-25)-0.5×I( ≥0.22) =125+10×I( ≥0.21) +10×I( ≤-25) When the snow melting and de-icing demand type D is the extreme cold and low temperature continuous response type: =-4.0-0.5×I( ≤-15)-0.5×I( ≤-35) =135+10×I( ≤-15) +10×I( ≤-35) When the snow melting and de-icing demand type D is the high-radiation diurnal regulation type: =1.0+0.5×I( ≥1300) =100+8×I( ≥1300) +5×I(Alt≥3000) Where I(·) is the indicator function, which takes the value 1 when the condition is true, and 0 otherwise.

[0023] S5. Target region's winter characteristic temperature Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter The dummy variable D, representing the snow melting demand type, is input into the phase change material thermal parameter prediction model to obtain the predicted value of the target phase change temperature. Predicted value of target phase transition enthalpy .

[0024] Preferably, the phase change material parameter prediction model is a two-stage thermal parameter prediction model. The first stage determines the characteristic winter temperature based on latitude (Lat) and altitude (Alt). The second stage is characterized by winter-like temperatures. Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter Using the snow melting demand type D as input features, target phase transition temperature prediction sub-model and target phase transition enthalpy prediction sub-model are established to output the target phase transition temperature prediction value. Predicted value of target phase transition enthalpy .

[0025] More preferably, the two-stage thermal parameter prediction model is an automatic demand-assisted two-stage linear / XGBoost joint model, wherein the first stage uses a linear regression model to determine the characteristic winter temperature. In the second stage, extreme gradient boosting regression models were used to establish target phase transition temperature prediction sub-models and target phase transition enthalpy prediction sub-models, respectively.

[0026] S6, predict the target phase transition temperature. Predicted value of target phase transition enthalpy By performing engineering constraint corrections, the corrected target phase transition temperature is obtained. Phase transition enthalpy with target .

[0027] Preferably, the predicted value of the target phase transition temperature is... Limited to The target phase transition temperature was obtained by quantizing the temperature within the range of 6.5℃ to 3.0℃ in 0.5℃ increments. ; Target phase transition enthalpy prediction value The target phase transition enthalpy was obtained by limiting it to the range of 80 J / g to 160 J / g and quantizing it in steps of 5 J / g. .

[0028] S7, the corrected target phase transition temperature and the corrected target phase transition enthalpy It matches the candidate phase change material library and outputs the recommended phase change temperature range, phase change enthalpy range, candidate material category or candidate phase change material combination for the target region.

[0029] Preferably, the candidate phase change material library includes candidate material category number, candidate material family, achievable phase change temperature range, achievable phase change enthalpy range, cycle stability level, pavement structure compatibility level, cost level, and suitable region type. The matching of the candidate phase change material library includes initial screening and comprehensive scoring and ranking.

[0030] The initial screening includes: (1) Retain the target phase transition temperature after correction to achieve phase transition temperature range coverage. Candidate phase change materials for the given region; (2) Retain the target phase transition enthalpy after the phase transition enthalpy range coverage correction is achieved. Candidate phase change materials for the given region; (3) Eliminate candidate phase change materials whose cycle stability level is lower than the preset level or whose compatibility level with the road structure is lower than the preset level.

[0031] After the initial screening, the remaining candidate phase change materials are ranked based on a comprehensive score. Preferably, the comprehensive score ranking is calculated according to the following formula: Score = 0.35 +0.35 +0.15 +0.10 +0.05

[0032] in The score is determined by the temperature range matching. For latent heat matching score, The score is for cycle stability. To score compatibility with road surface structure, This is for economic performance.

[0033] The temperature zone matching score Matching score with latent heat Calculate according to the following formulas respectively: =max(0, 1-| - | / ) =max(0, 1-| - | / ) in This represents the center value of the phase transition temperature range achievable by candidate phase change materials. The value represents the center of the enthalpy range for phase transition that the candidate phase transition materials can achieve. The corrected target phase transition temperature. The corrected target phase transition enthalpy, For temperature tolerance, This refers to the allowable deviation of the enthalpy value. Preferably, Take 2.0℃, Take 20 J / g; The cycle stability score Compatibility score with road surface structure The values ​​are determined according to the grade assignment method, where "high, medium-high, medium, medium-low, and low" correspond to 1.00, 0.85, 0.70, 0.55, and 0.40 respectively; Economic score The values ​​are determined according to the cost level assignment method, where "low cost", "medium cost" and "high cost" correspond to 1.00, 0.75 and 0.50 respectively.

[0034] Compared with the prior art, the present invention has at least the following beneficial effects: First, this invention forms a complete technical chain of "regional parameters - demand identification - thermal parameter prediction - engineering correction - material matching". It does not rely on a specific phase change material formula or a single heating device, and can determine the thermal parameters for phase change snow melting and de-icing needs of transportation infrastructure in different regions, thus having good versatility.

[0035] Second, this invention converges the geospatial parameters into latitude (Lat) and altitude (Alt), and first uses latitude (Lat) and altitude (Alt) to determine the characteristic winter temperatures. Then, the characteristic winter temperatures With the lowest winter temperature Duration near freezing point Snowfall intensity and total solar radiation in winter The shared input of thermal parameters in the prediction process makes the division of labor among variables clearer and reduces the cost of data acquisition and implementation.

[0036] Third, this invention transforms the identification of snow melting and de-icing needs in different regions from empirical descriptions into executable automatic judgment rules. By classifying target regions through five categories of demand scores, it improves the consistency and reproducibility of regional demand identification.

[0037] Fourth, this invention outputs predicted values ​​of the target phase change temperature through a phase change material thermal parameter prediction model. Predicted value of target phase transition enthalpy This allows the phase change temperature and phase change enthalpy to be determined jointly based on regional conditions, which helps to avoid the mismatch between the response timing of the phase change material and the needs of snow melting and de-icing, or the problem of insufficient latent heat release capacity.

[0038] Fifth, after model prediction, this invention further employs engineering constraint correction rules and candidate material library matching rules, enabling the output results to directly serve the screening, compounding, or engineering recommendation of candidate phase change materials, thereby improving the feasibility of the phase change material thermal parameter design results. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings are briefly described below. Obviously, the following drawings are only some embodiments of the present invention; those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0040] Figure 1 This is a schematic diagram of the overall process of the method of the present invention; Figure 2 This is a schematic diagram of the automatic demand identification and thermal parameter regression process of the present invention; Figure 3 This is a flowchart of the engineering constraint correction and candidate phase change material library matching process for this invention. Detailed Implementation The present invention will be further described below with reference to embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. For those skilled in the art, equivalent substitutions made to the order of steps, form of variables, model structure, material library composition, and engineering constraint thresholds without departing from the concept of the present invention should all fall within the scope of protection of the present invention.

[0041] This invention provides a method for determining the target phase change temperature and target phase change enthalpy of a phase change material based on the snow melting and de-icing needs of different regions. This method does not rely on a specific phase change material formulation, encapsulation structure, or road heating device as its core protection mechanism. Instead, it uses geospatial parameter extraction, determination of winter characteristic temperatures, automatic demand identification, thermal parameter labeling, regression prediction, engineering constraint correction, and candidate material matching as its main technical steps. This enables the determination of phase change snow melting and de-icing parameters for roads, bridge decks, airport runways, and railway-related infrastructure in different regions.

[0042] For ease of explanation, the meanings of the main symbols in this manual are as follows:

[0043] Example 1: Overall Method Example like Figure 1 As shown, this embodiment provides a method for determining the target phase transition temperature and target phase transition enthalpy of a phase change material based on the snow melting and de-icing needs of different regions, including the following steps: S1. Obtain the latitude (Lat), altitude (Alt), and lowest winter temperature of the target region. Duration near freezing point Snowfall intensity and total solar radiation in winter ; S2. Determine the characteristic winter temperature of the target region based on latitude (Lat) and altitude (Alt). Preferably, the characteristic winter temperature... The first-stage winter characteristic temperature model was determined to be a linear regression model, and its expression is as follows:

[0044] in, , and The regression coefficients are obtained by fitting historical winter temperature data of the sample area.

[0045] S3, according to , , , , Calculate the five demand scores G1 to G5 using Alt, and determine the demand type D for the target region based on the highest score; S4, based on , , , , Establish a target phase transition temperature prediction sub-model and a target phase transition enthalpy prediction sub-model for demand type D; S5. Input the target region's input parameters into the prediction sub-model to obtain the predicted target phase transition temperature. Predicted value of target phase transition enthalpy ; S6. Predicted value of target phase transition temperature Predicted value of target phase transition enthalpy The target phase transition temperature is obtained by correcting the engineering constraints according to the preset threshold. and the corrected target phase transition enthalpy ; S7. Based on the corrected target phase transition temperature and the corrected target phase transition enthalpy The candidate phase change materials library is screened and sorted to obtain recommended material categories and recommended parameter ranges.

[0046] Example 2: Regional Characteristic Variable System and Regional Division Example In this embodiment, a variable system is first constructed to characterize the snow melting and de-icing needs of different regions, as shown in Table 1.

[0047] Table 1 - Variable System for Snow Melting and Ice Removal Demand Analysis in Different Regions

[0048] After selecting the variables, it is necessary to classify the snow melting and de-icing needs of different regions. Preferably, this invention does not use a rough classification based solely on administrative divisions or single temperature conditions, but rather a comprehensive classification based on anti-icing needs, snow melting heat load needs, and radiation compensation needs.

[0049] Preferably, the demand can be identified based on the automatic judgment scores of the five types of demands, thereby explicitly introducing the demand layer information into the thermal parameter prediction model. The regional snow melting and de-icing demand types classification in this embodiment are shown in Table 2.

[0050] Table 2 - Classification of Regional Snow Melting and Ice Removal Demand Types

[0051] The scores for the five categories of snow melting and de-icing needs are calculated as follows: Black ice high-risk anti-icing type score G1 is: G1=2×I( ≥450)+I( ≥-12)+I(-2< ≤2)+I( <0.30) The score for G2 in the light snow rapid melting type is: G2=I(300≤ <550)+I( ≥-10)+I( ≥0)+I( ≥0.20)+I( <1000) The score for G3, representing sustained snow cover and high demand, is: G3=2×I( ≥0.21)+I( ≤-20)+I( ≤-8)+I(100≤ ≤260) The extreme cold sustained low temperature response score G4 is: G4=2×I( ≤-10)+2×I( ≤-28)+I( <150) The score for the high-radiation diurnal regulation type G5 is: G5=3×I( ≥1050)+I(Alt≥2000)+I( ≥350) Where I(·) is the indicator function, which takes the value 1 when the condition is true and 0 otherwise.

[0052] The black ice high-risk anti-icing type is mainly for scenarios where the ice temperature stays near the freezing point for a long time and the extreme low temperature is not particularly low, mainly addressing the risk of initial thin ice or black ice; the light snow rapid melting type is mainly for scenarios where the winter is not extremely cold but there is light snow and near-zero temperature fluctuations, mainly addressing the risk of light snow rapid melting and secondary freezing; the continuous snow accumulation high heat demand type is mainly for scenarios with high snowfall intensity and high snow melting heat load per unit time; the extreme cold low temperature continuous response type is mainly for scenarios with long-term low temperature and obvious extreme low temperature; the high radiation diurnal regulation type is mainly for scenarios with high altitude, high radiation and obvious diurnal heat exchange.

[0053] Example 3: Sample Region Construction and Model Training Example In this embodiment, a training dataset is constructed by selecting several representative sample regions to cover extremely cold and low-temperature regions, regions with continuous snowmelt in severe cold, high-risk black ice prevention and anti-icing regions in North China, regions with rapid light snow melting in the south, and regions with high radiation and diurnal regulation. The selection of the sample regions takes into account regional differences, completeness of demand types, and engineering representativeness.

[0054] The preferred geospatial parameters are latitude (Lat) and altitude (Alt) to determine the characteristic winter temperatures of the target region. The core area parameters are preferably selected based on the lowest winter temperature. Duration near freezing point Snowfall intensity and total solar radiation in winter To characterize "when the phase change material takes effect" and "how much heat needs to be released and whether there is external heat compensation", respectively.

[0055] The sample region data table for this embodiment is shown in Table 3.

[0056] Table 3 - Sample Area Data Table

[0057] Latitude (Lat) and altitude (Alt) in the table are used as geospatial parameters to first determine the characteristic winter temperatures of the sample area. The characteristic winter temperature And the lowest winter temperature Duration near freezing point Snowfall intensity and total solar radiation in winter A common input phase change material thermal parameter regression model is used to predict the target phase change temperature. Phase transition enthalpy with target .

[0058] Among them, the The example values ​​of winter characteristic temperatures determined based on latitude and altitude in this embodiment are multi-year averages of DJF winter average temperatures; the minimum winter temperature... Example values ​​for the lowest winter temperatures are defined as the boundary values ​​representing the winter low temperatures of the target region within the statistical period, preferably using multi-year statistical values ​​representing the extreme minimum winter temperatures of meteorological stations; and Together they are used to describe the winter service thermal environment of the target area, among which The reaction background is characterized by low temperature. The extreme low-temperature boundary of the reaction, together with the target phase transition temperature and target phase transition enthalpy, are used to determine the duration of the period near the freezing point. The snowfall intensity is defined as the cumulative duration of winter temperatures in the target region falling within the range of -2°C to 2°C during the statistical period; Defined as the hourly average snowfall intensity corresponding to winter snowfall events in the target region during the statistical period; the total winter solar radiation Defined as the cumulative downward-flowing shortwave solar radiation reaching the Earth's surface in the target region during the winter of the statistical period. Preferably, the statistical period is 1991-2020.

[0059] Preferably, the Chinese surface climate standard value dataset can be used as... The extraction sources include basic surface meteorological observation data and historical surface meteorological observation data of China. and The statistical basis of data, NASA POWERsolar values ​​can be used as... This table contains supplementary data. The values ​​in this table are example statistics for sample regions used in drafting patent embodiments. When formally training and modeling, data from sites downloaded in batches can be used to replace or correct these values ​​while maintaining the same statistical caliber.

[0060] Preferably, the sample area uses the national surface meteorological station with the same name in the corresponding city in the China Meteorological Data Network as the representative station; for cities with different station numbers in different business systems, it is preferred to use the station table in the selected dataset page, and keep the same station number for the same city in the subsequent parameter extraction process.

[0061] After completing the processing of input parameters for the sample regions, the target phase transition temperature and target phase transition enthalpy for each sample region were further labeled to form the supervised output data, as shown in Table 4.

[0062] Table 4 - Target Phase Transition Temperatures Corresponding to Sample Regions Phase transition enthalpy with target labeling table

[0063] The target phase change temperature center value and target phase change enthalpy center value listed in this table are engineering-annotated center values ​​obtained by the sample area based on demand type, key regional parameters and engineering rules. They are used as supervisory outputs for constructing the phase change material thermal parameter regression model, and are not regarded as the only inherent parameters of a specific phase change material. They can be further corrected in the future by combining batch data, heat load calculation and engineering constraint correction rules.

[0064] Among them, the engineering label of the target phase change temperature is mainly determined based on the characteristic winter temperature, the lowest winter temperature, and the duration of the freezing point, which is used to characterize "when the phase change material takes effect"; the engineering label of the target phase change enthalpy is mainly determined based on the snowfall intensity and the total solar radiation in winter, which is used to characterize "how much heat the phase change material needs to release and whether there is external heat compensation".

[0065] Furthermore, for regions with a sustained response to extreme cold and low temperatures, it is recommended that the phase change temperature zone be located in the negative temperature zone significantly below freezing point, and that the enthalpy value be medium to high. For regions with sustained snow accumulation and high heat demand, it is recommended that the phase change temperature zone cover the typical low-temperature snow melting demand range, and that the enthalpy value be relatively high to meet the sustained snow melting heat load. For regions with a high risk of black ice and ice prevention, it is recommended that the phase change temperature zone be located in the range close to or slightly above 0°C, and that the enthalpy value be medium to medium to high. For regions with rapid light snow melting, it is recommended that the phase change temperature zone be located in a small range above freezing point, and that the enthalpy value be medium to medium to low. For regions with high radiation and diurnal regulation, the target phase change enthalpy should be corrected based on the total solar radiation in winter, while maintaining the response capability near freezing point.

[0066] For example, in extremely cold and continuously low-temperature response areas such as Mohe and Harbin, the target phase change temperature is more inclined towards the low-temperature range, and the target phase change enthalpy is also relatively high, in order to balance the buffering of low-temperature fluctuations and the continuous supply of heat. In areas with continuous snow accumulation and high heat demand such as Changchun and Urumqi, the emphasis is on continuously providing snow melting heat in the lower temperature range, so the enthalpy value is set at a higher level. In areas with high black ice risk and ice prevention such as Beijing, Zhangjiakou and Taiyuan, the emphasis is on triggering the heat response as early as possible above the ice risk temperature range. In areas with rapid light snow melting such as Nanjing, Wuhan and Guiyang, the focus is on rapid light snow melting and thin ice suppression, and the target phase change temperature is closer to a small range above freezing point. In areas with high radiation diurnal regulation such as Xining, Lhasa and Shigatse, the nighttime heat release demand is considered when the target phase change temperature is determined, and the available radiation input during the day is considered when the target phase change enthalpy is determined.

[0067] In this embodiment, the characteristic winter temperature is first determined based on the latitude (Lat) and altitude (Alt) of the sample area. Preferably, the method is determined by regression; in a preferred embodiment, a linear regression method is used. The characteristic winter temperatures It can be represented as:

[0068] in, , and The regression coefficients are obtained by fitting historical winter temperature data of the sample area.

[0069] In a preferred embodiment, a linear fit is performed based on reference values ​​for latitude, altitude, and characteristic winter temperatures of 15 regions to obtain:

[0070] The model has an R² of 0.930848 under full-sample fitting and an R² of 0.895489 under strict leave-one-out cross-validation, indicating that it can be used as a prerequisite model for subsequent thermal parameter prediction.

[0071] In obtaining characteristic winter temperatures Then, using the sample area input parameters in Table 3 as the input vector X, and the target phase transition temperature in Table 4 as the input vector X... Phase transition enthalpy with target As the output vector Y, a regression model for the thermal parameters of the phase change material is constructed. The input vector X includes characteristic winter temperatures. Winter minimum temperature Duration near freezing point Snowfall intensity and total solar radiation in winter .

[0072] To evaluate the gain effect of automatic demand identification on thermal parameter prediction and the applicability of different regression models under small sample conditions, this embodiment constructs the following three types of models: A two-stage multiple linear regression model without automatic demand identification is used as a control model. A two-stage multiple linear regression model with automatic demand identification; A two-stage linear / XGBoost joint model with automatic demand identification.

[0073] In the first stage of all three types of models, latitude (Lat) and altitude (Alt) are used to measure winter characteristic temperatures. The first stage involves linear regression prediction; the second stage involves calculating the target phase transition temperature. Phase transition enthalpy with target Establish a predictive sub-model.

[0074] For model (1), the second-stage input variables include Their expressions can be represented as follows:

[0075]

[0076] For model (2), the automatic demand identification result D is further introduced as a category variable based on model (1). Preferably, the type I demand group is used as the baseline group to construct... , , and Dummy variables, and their relationship with The common input is the second-stage multiple linear regression model; its expression can be represented as follows:

[0077]

[0078] For model (3), the first stage still uses linear regression to determine... The second stage, based on the results of automatic demand identification, establishes target phase transition temperature prediction sub-models and target phase transition enthalpy prediction sub-models using the extreme gradient boosting regression method, respectively. Their expressions can be represented as follows:

[0079]

[0080] in, This represents the target phase transition temperature prediction sub-model. The target phase transition enthalpy prediction sub-model is represented by both models, which are constructed using the extreme gradient boosting regression method and together constitute an automatic demand-assisted two-stage linear / XGBoost joint model.

[0081] Furthermore, the prediction output of the model (3) can be expressed as follows:

[0082]

[0083] in, Let be the input feature vector for the i-th sample region. For the k-th regression tree in the target phase transition temperature prediction sub-model, The k-th regression tree in the target phase transition enthalpy prediction sub-model. and These represent the number of regression trees in the two sub-models, respectively. The target phase transition temperature is the predicted value. The target phase transition enthalpy is the predicted value. The two sub-models are formed by integrating multiple regression trees after training. Their specific tree structure, split nodes, and leaf node values ​​are automatically determined during the training process and will not be elaborated on in detail in the specification. Instead, their usability is characterized by the model training and evaluation results.

[0084] Among them, model (1) is used as a benchmark when demand type information is not introduced; model (2) is used to reflect the impact of automatic demand identification on the prediction effect of linear regression model; model (3) is used to characterize the improvement of the thermal parameter prediction ability of nonlinear ensemble learning method on the basis of introducing automatic demand identification.

[0085] Given the small sample size, this embodiment preferably uses strict leave-one-out cross-validation as the main evaluation method for model selection, while using the full sample fitting results as an auxiliary reference to characterize the model's ability to fit the existing samples, rather than as the sole basis for model selection.

[0086] Furthermore, the prediction performance of the target phase transition temperature and target phase transition enthalpy was evaluated using the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²). The comparison results of leave-one-out cross-validation for different models are shown in Table 5A, and the full-sample fitting results are shown in Table 5B.

[0087] Table 5A - Comparison of Leave-one-out Cross-validation Results for Regression Models of Thermal Parameters of Different Phase Change Materials

[0088] Table 5B - Full-sample fitting results of regression models for thermal parameters of different phase change materials

[0089] As shown in Table 5A, the two-stage multiple linear regression model without automatic demand identification can be used as a control model, but its predictive ability for target phase transition temperature and target phase transition enthalpy is relatively limited.

[0090] After introducing automatic demand identification, the two-stage multiple linear regression model significantly improved the prediction accuracy of the target phase transition temperature, indicating that automatic demand identification can enhance the expressive power of regional demand difference information for the prediction of the target phase transition temperature.

[0091] Further comparison reveals that the automatic demand-assisted two-stage linear / XGBoost joint model, under strict leave-one-out cross-validation, exhibits superior overall predictive performance for both the target phase transition temperature and the target phase transition enthalpy. Specifically, this model achieves an R² of 0.814468 for the target phase transition temperature prediction, outperforming both the control model and the automatic demand-assisted two-stage multiple linear regression model; its R² for the target phase transition enthalpy prediction reaches 0.854187, also surpassing the other two models, indicating that this model is better able to characterize the nonlinear relationship between regional parameters, demand type, and target thermal parameters.

[0092] As shown in Table 5B, under the condition of full-sample fitting, the two-stage linear / XGBoost joint model with automatic demand assistance also exhibits high fitting accuracy, especially for the target phase transition enthalpy. The model showed the best performance in fitting the data. Combining the leave-one-out cross-validation results and the full-sample fitting results, this embodiment ultimately determined the two-stage linear / XGBoost joint model with automatic demand assistance as the preferred regression model for phase change material thermal parameters, and used it for subsequent validation in typical cities and parameter recommendation for target areas.

[0093] Meanwhile, the comparison results between the control model and the automatic demand-assisted two-stage multiple linear regression model show that automatic demand identification has a gain effect on thermal parameter prediction; and the automatic demand-assisted two-stage linear / XGBoost joint model further improves the model's ability to express complex nonlinear relationships compared with the automatic demand-assisted two-stage multiple linear regression model.

[0094] Example 4: Typical City Validation Example In this embodiment, to verify the rationality and applicability of the method of the present invention, five typical sample cities with different needs were selected from the training samples as the verification objects of the leave-one-out method. Preferably, the verification areas include extremely cold and low-temperature areas, high-risk black ice prevention areas, high-radiation diurnal regulation areas, and southern areas with rapid light snow melting, so as to cover different types of snow melting and de-icing thermal response needs.

[0095] In this embodiment, the geospatial parameters and core area parameters corresponding to each typical city are first obtained. The geospatial parameters include latitude (Lat) and altitude (Alt), and the core area parameters include the lowest winter temperature. Duration near freezing point Snowfall intensity and total solar radiation in winter Subsequently, the characteristic winter temperatures of typical cities were calculated based on latitude (Lat) and altitude (Alt). Then based on , , , and Determine the demand type of typical cities according to the rules in Example 2, and input the above parameters into the thermal parameter regression model trained in Example 3 to obtain the predicted value of the target phase transition temperature. Predicted value of target phase transition enthalpy The preferred method is to use strict leave-one-out cross-validation to obtain the predicted values ​​of the target phase transition temperature and the target phase transition enthalpy.

[0096] Furthermore, on and The target phase transition temperature was obtained by processing according to the engineering constraint correction rules in Example 1. and the corrected target phase transition enthalpy Then and Inputting the candidate phase change material library matching module yields recommended phase change temperature ranges, recommended phase change enthalpy ranges, and candidate material categories. Table 6 shows the validation results using the leave-one-out method for typical sample cities.

[0097] Table 6 - Validation Table of the Leave-One-Out Method in Typical Sample Cities

[0098] Predicted target phase transition temperatures in Table 6 Predicted value of target phase transition enthalpy These are the model prediction results obtained for typical sample cities under strict leave-one-out cross-validation conditions. Specifically, when predicting a typical city, the city is removed from the training samples, and the city is predicted using the remaining samples after training an optimized thermal parameter regression model, in order to avoid directly using the full sample back-substitution values, which would weaken the validation significance.

[0099] When making engineering corrections, it is preferable to first use the predicted value of the target phase transition temperature. Limited to The target phase transition temperature was obtained by quantizing the temperature within the range of 6.5℃ to 3.0℃ in 0.5℃ increments. Then, the target phase transition enthalpy prediction value is... The target phase transition enthalpy was obtained by limiting it to the range of 80 J / g to 160 J / g and quantizing it in steps of 5 J / g. Subsequently, and Input the candidate phase change material library matching module. First, the candidate materials are initially screened based on their achievable temperature range and latent heat range. Then, they are comprehensively ranked based on cycle stability, structural compatibility, and economy to obtain the recommended candidate material categories.

[0100] For regions with a sustained response to extreme cold and low temperatures It is advisable to choose a lower value. A higher value should be chosen to ensure the material continues to respond and provides sufficient heat in sub-zero temperatures significantly below freezing; for regions with persistent snow cover and high heat demand, It should cover the typical low-temperature snow melting demand range. A higher value should be chosen to meet the larger snowmelt heat load under continuous snow accumulation scenarios; for high-risk black ice prevention areas, It is advisable to select a temperature range close to or slightly above 0°C. Medium or medium-high values ​​are recommended to enhance early triggering and initial icing suppression near the freezing point; for areas with rapid light snow melting, It is advisable to select a small area above freezing point. A moderate or low-to-medium value is preferable to balance rapid snowmelt and economic efficiency; for regions with high radiation diurnal regulation, the response capability near freezing point should be maintained, while also considering solar radiation compensation conditions. Make corrections.

[0101] As shown in Table 6, the method of the present invention can output targeted target phase transition temperature, target phase transition enthalpy and candidate material categories according to the differences in the dominant demand in different regions, indicating that the established "regional parameter - demand identification - thermal parameter prediction - engineering correction - material matching" technology chain has good engineering adaptability and application value.

[0102] Example 5: Matching of candidate phase change material library In this embodiment, a candidate phase change material library is established to map the target phase change temperature and target phase change enthalpy, predicted by the model and corrected by engineering, to the actual selectable material categories. It should be noted that the core of this invention is not a specific phase change material, but rather a method for determining target thermal parameters based on regional needs and further matching materials. Therefore, the candidate material library can include organic phase change materials, inorganic phase change materials, solid-solid phase change materials, and composite phase change materials.

[0103] In the candidate phase change material library, the following information is preferably recorded: candidate material category number, candidate material family, achievable phase change temperature range, achievable phase change enthalpy range, cycle stability, compatibility with pavement structure, cost level, and suitable region type. The candidate phase change material library matching table is shown in Table 7.

[0104] Table 7 - Matching Table of Candidate Phase Change Materials Library

[0105] In the material library matching process, it is preferable to perform a preliminary screening step first. The preliminary screening step includes: (1) Retain the target phase transition temperature after correction to achieve phase transition temperature range coverage. Candidate materials within the specified interval; (2) Retain the target phase transition enthalpy after the phase transition enthalpy range coverage correction is achieved. Candidate materials within the specified interval; (3) Eliminate candidate materials whose cyclic stability is lower than the preset level or whose compatibility with the road structure is lower than the preset level.

[0106] After the initial screening, the remaining candidate materials are ranked and scored comprehensively. Preferably, the ranking score of the candidate materials is calculated using the following formula: Score = 0.35 +0.35 +0.15 +0.10 +0.05

[0107] in The score is determined by the temperature range matching. For latent heat matching score, The score is for cycle stability. To score compatibility with road surface structure, This is for economic performance.

[0108] Preferably, the temperature zone matching score Matching score with latent heat Calculate according to the following formulas respectively: =max(0, 1 - | - | / ) =max(0, 1 - | - | / ) in This represents the center value of the phase transition temperature range achievable by candidate phase change materials. The value represents the center of the enthalpy range for phase transition that the candidate phase transition materials can achieve. The corrected target phase transition temperature. The corrected target phase transition enthalpy, For temperature tolerance, To allow for deviation in enthalpy; preferably, Take 2.0℃, Take 20 J / g.

[0109] Preferably, the cycle stability score Compatibility score with road surface structure The scores are determined according to a rating system, with "High," "High-Medium," "Medium," "Low-Medium," and "Low" corresponding to 1.00, 0.85, 0.70, 0.55, and 0.40 respectively; Economic score. The values ​​are determined according to the cost level assignment method, where "low cost", "medium cost" and "high cost" correspond to 1.00, 0.75 and 0.50 respectively.

[0110] For example, when the target region is predicted and corrected by the model, it becomes... 2.0℃ When the value is 110 J / g, candidate material categories A2, A5, and A7 that can achieve a temperature range of 0–4℃ and an enthalpy range of 80–140 J / g can be prioritized for screening; then, the recommended priority is obtained by combining cycle stability, structural compatibility, and cost level.

[0111] Therefore, after engineering constraint correction and quantitative sorting, the output results of this invention can directly serve the screening and engineering implementation of phase change materials.

Claims

1. A method for determining the target phase transition temperature and target phase transition enthalpy of a phase change material based on the snow melting and de-icing needs of different regions, characterized in that, Includes the following steps: S1. Obtain the input parameters for the target region, including latitude (Lat), altitude (Alt), and lowest winter temperature. Duration near freezing point Snowfall intensity and total solar radiation in winter ; S2. Determine the characteristic winter temperature of the target region based on the latitude (Lat) and altitude (Alt). ; S3. Based on the aforementioned winter characteristic temperature Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter In addition to altitude Alt, automatic demand identification is performed on the target area to obtain the demand type D of the target area; S4. The winter characteristic temperature Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter The automatically identified demand type D is input into the trained phase change material thermal parameter prediction model to obtain the predicted value of the target phase change temperature. Predicted value of target phase transition enthalpy ; S5. The predicted target phase transition temperature value Predicted value of target phase transition enthalpy By performing engineering constraint corrections, the corrected target phase transition temperature is obtained. and the corrected target phase transition enthalpy ; S6. The corrected target phase transition temperature Phase transition enthalpy with target It matches the candidate phase change material library and outputs the recommended phase change temperature range, phase change enthalpy range, candidate phase change material category or candidate phase change material combination for the target region.

2. The method according to claim 1, characterized in that, In step S2, the winter characteristic temperature of the target region is... The first-stage winter characteristic temperature model was determined to be a linear regression model, and its expression is as follows: Among them, a0, a1, and a2 are regression coefficients obtained by fitting historical winter temperature data of the sample area.

3. The method according to claim 1, characterized in that, In step S3, the snow melting and de-icing demand type D includes black ice high-risk anti-icing type, light snow rapid melting type, continuous snow accumulation high heat demand type, extreme cold low temperature continuous response type and high radiation day and night regulation type. The automatic demand identification includes calculating scores G1, G2, G3, G4, and G5 for five categories of snow melting and de-icing demand, respectively: Black ice high-risk anti-icing type score G1 is: G1=2×I( ≥450)+I( ≥-12)+I(-2< ≤2)+I( <0.30) The score for G2 in the light snow rapid melting type is: G2=I(300≤ <550)+I( ≥-10)+I( ≥0)+I( ≥0.20)+I( <1000) The score for G3, representing sustained snow cover and high demand, is: G3=2×I( ≥0.21)+I( ≤-20)+I( ≤-8)+I(100≤ ≤260) The extreme cold sustained low temperature response score G4 is: G4=2×I( ≤-10)+2×I( ≤-28)+I( (<150) The score for the high-radiation diurnal regulation type G5 is: G5=3×I( ≥1050)+I(Alt≥2000)+I( ≥350) Where I(·) is the indicator function, which takes the value 1 when the condition is met and 0 otherwise. The demand type corresponding to the maximum value among G1, G2, G3, G4 and G5 is taken as the demand type D of the target area; when there are ties for the maximum value, the priority is determined according to the following: high radiation day and night regulation type, extreme cold and low temperature continuous response type, continuous snow accumulation and high heat demand type, black ice high risk anti-icing type, and light snow rapid melting type.

4. The method according to claim 1, characterized in that, The training samples for the phase change material thermal parameter prediction model include the characteristic winter temperatures of the sample region. Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter And snow melting and de-icing demand type D; The supervised output data of the training samples includes the engineered labeled center value of the phase transition temperature. and the central value of phase transition enthalpy marked in the engineering. The supervisory output data is determined according to the rule-based engineering annotation method.

5. The method according to claim 4, characterized in that, The standardized engineering annotation methods include: When the snow melting and de-icing demand type D is the high-risk black ice prevention type: =1.0+0.5×I( ≥-15) +0.5×I( ≥500) =95+10×I( ≥450) +5×I( ≥0.25) When the snow melting and de-icing requirement type D is light snow rapid melting type: =0.5+0.5×I( ≥600) +0.5×I( ≥0.25) =90+10×I(0.18≤ <0.25) +5×I( ≥500) When the snow melting and de-icing demand type D is the continuous snow accumulation and high heat demand type: =-2.0-0.5×I( ≤-25)-0.5×I( ≥0.22) =125+10×I( ≥0.21) +10×I( ≤-25) When the snow melting and de-icing demand type D is the extreme cold and low temperature continuous response type: =-4.0-0.5×I( ≤-15)-0.5×I( ≤-35) =135+10×I( ≤-15) +10×I( ≤-35) When the snow melting and de-icing demand type D is the high-radiation diurnal regulation type: =1.0+0.5×I( ≥1300) =100 +8×I( ≥1300) +5×I(Alt≥3000) Where I(·) is the indicator function, which takes the value 1 when the condition is true and 0 otherwise.

6. The method according to claim 1, characterized in that, The phase change material thermal parameter prediction model in step S4 is a two-stage thermal parameter prediction model. The first stage determines the characteristic winter temperatures based on latitude (Lat) and altitude (Alt). ; The second stage is characterized by winter-like temperatures. Winter minimum temperature Duration near freezing point Snowfall intensity Total solar radiation in winter Using the snow melting demand type D as input features, target phase transition temperature prediction sub-model and target phase transition enthalpy prediction sub-model are established to output the target phase transition temperature prediction value. Predicted value of target phase transition enthalpy .

7. The method according to claim 6, characterized in that, The target phase transition temperature prediction sub-model and the target phase transition enthalpy prediction sub-model in the second stage are regression models; The regression model is selected from at least one of the following: multiple linear regression model, random forest regression model, gradient boosting regression model, extreme gradient boosting regression model, support vector regression model, or neural network regression model. Preferably, the two-stage thermal parameter prediction model is a two-stage linear / extreme gradient boosting joint model with automatic demand assistance, wherein the first stage uses a linear regression model to determine the characteristic winter temperature. In the second stage, extreme gradient boosting regression models were used to establish target phase transition temperature prediction sub-models and target phase transition enthalpy prediction sub-models, respectively.

8. The method according to claim 1, characterized in that, The engineering constraint correction in step S5 includes: Predicted target phase transition temperature The target phase transition temperature was obtained by limiting it to the range of -6.5℃ to 3.0℃ and quantizing it in 0.5℃ increments. ; Predicted value of target phase transition enthalpy The target phase transition enthalpy was obtained by limiting it to the range of 80 J / g to 160 J / g and quantizing it in steps of 5 J / g. .

9. The method according to claim 1, characterized in that, The candidate phase change material library in step S6 includes candidate material category number, candidate material family, achievable phase change temperature range, achievable phase change enthalpy range, cycle stability level, road structure compatibility level, cost level, and suitable region type. The matching of the candidate phase change material library includes initial screening and comprehensive scoring and ranking. The initial screening includes: (1) Retain the target phase transition temperature after correction that can achieve phase transition temperature range coverage. Candidate phase change materials for the given region; (2) Retain the target phase transition enthalpy after the phase transition enthalpy range coverage correction is achieved. Candidate phase change materials for the given region; (3) Eliminate candidate materials whose cyclic stability is lower than the preset level or whose compatibility with the road structure is lower than the preset level.

10. The method according to claim 9, characterized in that, The comprehensive scoring and ranking of candidate materials is calculated using the following formula: Score=0.35 +0.35 +0.15 +0.10 +0.05 in The score is determined by the temperature range matching. For latent heat matching score, The score is for cycle stability. To score compatibility with road surface structure, For economic reasons; The temperature zone matching score Matching score with latent heat Calculate according to the following formulas respectively: =max(0,1-| - | / ) =max(0,1-| - | / ) in This represents the center value of the phase transition temperature range achievable by candidate phase change materials. The value represents the center of the enthalpy range for phase transition that the candidate phase transition materials can achieve. The corrected target phase transition temperature. The corrected target phase transition enthalpy, For temperature tolerance, This refers to the allowable deviation of the enthalpy value. Preferably, Take 2.0℃, Take 20 J / g; The cycle stability score Compatibility score with road surface structure The values ​​are determined according to the level assignment method, where "high, medium-high, medium, medium-low, and low" correspond to 1.00, 0.85, 0.70, 0.55, and 0.40 respectively; Economic score The values ​​are determined according to the cost level assignment method, where "low cost", "medium cost" and "high cost" correspond to 1.00, 0.75 and 0.50 respectively.

Images