A method for obtaining a dynamic prediction tool for short-term storage and transportation schemes of a product

Abstract

The application discloses a kind of acquisition methods of product short-term storage and transport scheme dynamic prediction tool, the acquisition method includes the following steps: based on the same batch product construction test group and control group, using the same packaging and sealing mode, the difference is that test group is equipped with humidity control material in packaging;Given external environment temperature and humidity conditions and the target humidity range in packaging within target period;Data acquisition is carried out to test group and control group within target period;According to the data of control group collected, the equivalent exchange coefficient is deduced back;According to the equivalent exchange coefficient and the data of test group collected, the equivalent humidity control coefficient is deduced back, and a migratory prediction model is established according to external environment temperature and humidity conditions and the target humidity range in packaging, to obtain product short-term storage and transport scheme dynamic prediction tool.Can provide a kind of tool for obtaining product short-term storage and transport scheme, it is convenient to design the humidity control material setting scheme of product packaging, namely the actual packaging scheme of product, through the tool.

Description

Technical Field

[0001] This invention relates to the field of product packaging technology, and in particular to a method for obtaining a dynamic prediction tool for short-term product storage and transportation plans. Background Technology

[0002] As the coverage of modern distribution networks continues to expand and the scale of cross-regional commodity trade grows rapidly, the quality assurance requirements for different product categories in the storage and transportation process are constantly upgrading. Among these requirements, humidity control is one of the core indicators for ensuring the quality of most perishable and easily damaged products. Currently, the industry's commonly used product packaging humidity control solutions generally adopt a standardized packaging structure combined with a fixed ratio of humidity-regulating materials, using a generic design logic. This primarily relies on the product's conventional long-term storage parameters and the characteristics of common distribution environments to pre-determine the type, quantity, and packaging method of humidity-regulating materials, without providing targeted adaptation for the differentiated storage and transportation needs of different orders.

[0003] Existing product packaging technologies suffer from several shortcomings in humidity control: First, current solutions lack a coupled correlation model between the barrier properties of packaging materials and the humidity control properties of moisture-regulating materials. Relying solely on industry experience to determine humidity parameters fails to achieve quantitative control of humidity precision. This can lead to either excessive humidity control resulting in unnecessary increases in packaging material and transportation costs, or insufficient humidity control causing products to become damp, deteriorate, and lose value during storage and transportation. Second, existing solutions are not specifically optimized for short-term storage and transportation scenarios. Most directly adopt the humidity control logic of long-term warehousing, resulting in redundant designs for short-term storage and transportation needs ranging from 7 to 90 days. This increases material input in the packaging process and raises the pressure of subsequent packaging waste disposal, failing to meet the requirements of low-carbon distribution. Furthermore, existing solutions lack adaptability and flexibility. They cannot dynamically adjust based on the user-specified target shelf life or the actual environmental temperature and humidity parameters of the distribution path. When the product's storage and transportation cycle is temporarily adjusted, or the distribution path changes, the original packaging solution cannot adapt to the new needs, requiring a complete packaging replacement, further increasing the time and economic costs of the distribution process. However, there is a lack of relevant data and tools to support the dynamic design of targeted short-term storage and transportation solutions.

[0004] In summary, the industry currently lacks forecasting tools that can predict and optimize for short-term storage and transportation scenarios, and accurately customize packaging solutions based on the product's target shelf life and actual storage environment parameters. Corresponding forecasting tools are urgently needed to fill this industry gap. Summary of the Invention

[0005] In view of the above-mentioned shortcomings in the current product packaging technology field, the present invention provides a method for obtaining a dynamic prediction tool for short-term product storage and transportation schemes. This tool can be used to obtain short-term product storage and transportation schemes, and facilitates the flexible design of humidity control material settings for product packaging, i.e., the actual packaging scheme of the product.

[0006] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions: A method for obtaining a dynamic forecasting tool for short-term product storage and transportation plans, the method comprising the following steps: Test and control groups were constructed based on the same batch of products, using the same packaging and sealing methods. The difference was that the test group packaging contained humidity-regulating materials. Given the external environmental temperature and humidity conditions and the target humidity range inside the packaging within the target period; Data were collected from the experimental and control groups within the target timeframe; Based on the data collected from the control group, the equivalent exchange coefficient can be deduced. Based on the equivalent exchange coefficient and the collected test group data, the equivalent humidity regulation coefficient is derived, and a transferable prediction model is established by fitting the external environmental temperature and humidity conditions and the target humidity range inside the packaging, thus obtaining a dynamic prediction tool for short-term product storage and transportation.

[0007] According to one aspect of the present invention, both the experimental group and the control group are equipped with temperature and humidity acquisition devices for acquiring temperature and humidity data within the experimental group and the control group at the same frequency.

[0008] According to one aspect of the invention, the test group includes several combinations of different water activity and dosage of humidifying materials.

[0009] According to one aspect of the present invention, the external environmental temperature and humidity conditions within the given target period include: simulating the external environment with temperature and humidity variations within the target period under different seasons and different logistics routes.

[0010] According to one aspect of the present invention, the data collection of the experimental group and the control group within the target period includes: weighing the experimental group and the control group and calculating the daily weight difference mass to enhance identification and verification.

[0011] According to one aspect of the invention, the equivalent exchange coefficient is a quantification of the airtightness of the product packaging, and is a single fixed value under the packaging and sealing method, the external environmental temperature and humidity conditions within the given target period, and the target period.

[0012] According to one aspect of the present invention, the equivalent humidity regulating coefficient comprehensively reflects the moisture absorption and release kinetics of the humidity regulating material and the moisture permeation limit of the coating material.

[0013] According to one aspect of the present invention, the prediction model is: de_in / dt=k_e·(e_out-e_in)-K_m(T_in,a_in)·(e_in-e_set) Formula (1) Where, a_in = RH_in / 100, a_out = RH_out / 100, e_in = a_in·e_s(T_in), e_out = a_out·e_s(T_out), e_set(T_in) = a_w·e_s(T_in); e_in is the partial pressure of water vapor inside the product packaging; e_out is the partial pressure of water vapor outside the product packaging; T_in is the internal temperature of the product packaging; T_out is the external ambient temperature of the product packaging; e_s(T_out) is the saturated water vapor pressure at the external temperature T_out; e_s(T_in) is the internal temperature of the product packaging. The partial pressure of water vapor inside the product packaging is T_in; e_set(T_in) is the set partial pressure of the material at the internal temperature T_in at time t; a_w is the water activity of the material, a_in is the water activity inside the product packaging, and a_out is the water activity of the external environment of the product packaging; K_m is the equivalent humidity regulation coefficient; k_e is the equivalent exchange coefficient, which is a single fixed value obtained by fitting all valid data points collected by the control group within the target period; de_in / dt is the rate of change of the partial pressure of water vapor inside the product packaging with time t; RH_in is the relative humidity inside the product packaging, and RH_out is the relative humidity outside the product packaging.

[0014] According to one aspect of the invention, the step of deducing the equivalent exchange coefficient based on data from a collected control group includes: Under the external environmental temperature and humidity conditions within the target period and the given target period, k_e is used as a single constant to be estimated. e_in,blank(t) is calculated from RH_in,blank(t) and T_in,blank(t) according to de_in,blank / dt = k_e·(e_out - e_in,blank). e_out(t) is calculated from RH_out(t) and T_out(t). The derivative is approximated using the difference approximation de / dt ≈ [e(t+Δt) - e(t)] / Δt, and de / dt = k_e·(e_out - e_in) is fitted on all valid data points collected within the target period. The output is a single fixed value k_e, and k_e is substituted into the prediction model.

[0015] According to one aspect of the present invention, the method for obtaining the dynamic prediction tool for short-term storage and transportation of products includes: preprocessing the collected humidity and temperature data, specifically by unifying the timestamp and sampling interval, and removing short-term anomalies caused by opening the lid and / or handling.

[0016] The advantages of this invention are: it provides a dynamic prediction tool that combines the equivalent exchange coefficient, stored environmental information, target humidity range and shelf life to calculate the equivalent humidity control coefficient, replacing the traditional empirical method of determining humidity control parameters. This effectively avoids the waste of packaging and transportation costs caused by excessive humidity control, or the problems of product deterioration, performance degradation and value loss caused by insufficient humidity control, thus ensuring product quality while controlling costs.

[0017] Adapted to short-term storage and transportation scenarios, this tool can be optimized for short-term storage and transportation needs of ≤90 days, eliminating redundant design logic of long-term warehousing, reducing the input of packaging materials, reducing the pressure of subsequent packaging waste disposal, and meeting the development requirements of low-carbon circulation.

[0018] This tool allows users to preset humidity control material settings and a library of equivalent humidity control coefficients, enabling rapid matching of suitable solutions. It also requires that the humidity control material be within its reversible moisture absorption and release range within the target shelf life and not approach its capacity limit, ensuring stable and reliable humidity control performance.

[0019] This tool can predict and specify that the packaging structure or sealing condition must remain stable within the target shelf life, providing a reliable guarantee for the effective implementation of humidity control measures and ensuring that the product maintains the target humidity throughout the storage and transportation process.

[0020] This tool can be widely applied to the distribution of various humidity-sensitive products, including fresh agricultural and sideline products, precision electronic components, biopharmaceutical reagents, and paper-based cultural and creative collectibles. Detailed Implementation

[0021] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] A method for obtaining a dynamic forecasting tool for short-term product storage and transportation plans, the method comprising the following steps:

[0023] Step S1: Construct test and control groups based on the same batch of products, using the same packaging and sealing methods, the difference being that the test group packaging contains humidity-regulating material;

[0024] The specific experimental and control group settings are as follows:

[0025] Both the experimental group and the control group are equipped with temperature and humidity acquisition devices to collect temperature and humidity data at the same frequency.

[0026] At least 1–2 temperature and humidity sensors should be placed inside the packaging; for large boxes or locations where there may be local risks, it is recommended to place 2–4 sensors.

[0027] T_out / RH_out are recorded synchronously at different locations outside the packaging, near the packaging, or in the environmental chamber to establish realistic boundary conditions.

[0028] For high-risk locations such as bottle caps, seals, bag edges, and cold spots on the inner wall, it is recommended to use paired sensor patches to record both the surface temperature and the temperature and humidity of the air above the surface.

[0029] Sampling intervals of 5–10 min are recommended; if studying respiratory effects or disturbances caused by opening doors / carrying objects, 1 min is recommended.

[0030] The control group and the experimental group have the same packaging structure and no humidity-regulating material; an inert dummy load can be placed.

[0031] The experimental group included several combinations of different water activity and dosage of humidifying materials.

[0032] It may include any one or more of the following combinations:

[0033] Fresh one-way desiccant or existing desiccant from the customer can be used to establish a high initial moisture absorption capacity baseline.

[0034] Pre-treated desiccant to partial saturation can be used to demonstrate the earlier occurrence of RH drift as capacity decreases.

[0035] Fully saturated or desiccant can be used as a failure reference to distinguish between "material present" and "effective humidity control".

[0036] Selecting a_w grade conditioning material, such as 0.45, 0.55, 0.60, 0.65, or 0.75, can be used to verify the bidirectional buffering capability around a set RH.

[0037] Setting at least two delivery values ​​under the same a_w, such as the predicted minimum value and the +30% safety value, can be used to establish the scaling relationship between K_m and M / V or A / V.

[0038] Step S2: Given the external environmental temperature and humidity conditions and the target humidity range inside the packaging within the target period;

[0039] The external environmental temperature and humidity conditions within the given target period include: simulating the external environment with temperature and humidity changes within the target period under different seasons and different logistics routes.

[0040] Step S3: Collect data from the experimental and control groups within the target timeframe;

[0041] In practical applications, this includes: preprocessing the collected humidity and temperature data, specifically by standardizing the timestamps and sampling intervals, and removing short-term anomalies caused by opening the lid and / or handling.

[0042] Step S4: Based on the data collected from the control group, deduce the equivalent exchange coefficient;

[0043] The equivalent exchange coefficient is a quantification of the airtightness of product packaging. In practical applications, the equivalent exchange coefficient k_e is the equivalent exchange rate, and its reciprocal τ_e = 1 / k_e is the time constant.

[0044] The equivalent exchange coefficients derived from the collected control group data include: The expression `de_in,blank / dt = k_e·(e_out - e_in,blank)` is used to calculate `e_in,blank(t)` from `RH_in,blank(t)` and `T_in,blank(t)`. The expression `e_out(t)` is calculated from `RH_out(t)` and `T_out(t)`, and the derivative is approximated using the difference: `de / dt ≈ [e(t+Δt) - e(t)] / Δt`. `k_e` is treated as a single constant to be estimated and fitted over all valid data points: `de / dt = k_e·(e_out - e_in)`. The output is a single fixed value for `k_e`, which is then used to back-calculate and predict parameters for the experimental group with the same packaging and sealing methods. To reduce noise, it is recommended to perform a 30-minute moving average on `e_in` to remove points where `|e_out - e_in|` is too small, due to handling disturbances, or sensor anomalies.

[0045] Step S5: Based on the equivalent exchange coefficient and the collected test group data, the equivalent humidity regulation coefficient is derived, and a transferable prediction model is established by fitting the external environmental temperature and humidity conditions and the target humidity range inside the packaging, thus obtaining a dynamic prediction tool for short-term storage and transportation of products.

[0046] The equivalent humidity control coefficient comprehensively reflects the moisture absorption and release kinetics of the humidity control material and the moisture permeation limit of the coating material.

[0047] The prediction model is as follows:

[0048] de_in / dt=k_e·(e_out-e_in)-K_m(T_in,a_in)·(e_in-e_set) Formula (1) Where, a_in = RH_in / 100, a_out = RH_out / 100, e_in = a_in·es(T_in), e_out = a_out·es(T_out), e_set(T_in) = a_w·es(T_in); e_in is the partial pressure of water vapor inside the product packaging; e_out is the partial pressure of water vapor outside the product packaging; T_in is the internal temperature of the product packaging; T_out is the external ambient temperature of the product packaging; e_s(T_out) is the saturated water vapor pressure at the external temperature T_out; and e_s(T_in) is the saturated water vapor pressure at the internal temperature T_in. e_set(T_in) is the set partial pressure of the material at the internal temperature T_in at time t; a_w is the water activity of the material, a_in is the water activity inside the product packaging, and a_out is the water activity of the external environment of the product packaging; K_m is the equivalent humidity regulation coefficient; k_e is the equivalent exchange coefficient, which is a single fixed value obtained by fitting all effective data points of the control group under the corresponding packaging and sealing method, given external temperature and humidity conditions and target period, and input into the prediction model; de_in / dt is the rate of change of the water vapor partial pressure inside the product packaging with time t; RH_in is the relative humidity inside the product packaging, and RH_out is the relative humidity outside the product packaging.

[0049] Determined by the material system. K_m comprehensively reflects the moisture absorption and release kinetics of the material and the moisture permeation limit of the coating material (which can vary with the coupling of T_in and a_in); k_e is a single equivalent constant obtained by fitting all effective data points of the control group under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions and target period, comprehensively reflecting the packaging airtightness / moisture permeability / leakage / breathing effect, etc.

[0050] e_in(t): partial pressure of water vapor inside the box (kPa); e_out(t): partial pressure of water vapor outside the box (kPa); e_s(T): saturated water vapor pressure at temperature T (kPa); e_set(t): material set partial pressure (determined by material a_w and temperature T_in inside the box): e_set(t)=a_w·e_s(T_in(t)); k_e: equivalent exchange coefficient of the box obtained by fitting all effective data points of the control group (1 / day or 1 / h), which is used as a single equivalent constant within the corresponding calibration conditions and prediction range, reflecting the airtightness of the packaging; K_m: equivalent "pullback coefficient" of the material (1 / day or 1 / h), reflecting the moisture conditioning strength of the material.

[0051] Basic assumptions:

[0052] The chamber volume V is fixed, and the effect of changes in the total air volume inside the chamber on the partial pressure of water vapor is negligible; the airtightness is represented by an equivalent exchange coefficient that remains fixed within the corresponding calibration conditions and prediction range (including the combined effects of infiltration, micro-leakage, and breathing effect); the humidity conditioning material can be considered as an equivalent "buffer" that pulls the partial pressure inside the chamber back to its set partial pressure e_set, and its strength is represented by the coefficient K_m; the performance of the chamber and materials does not undergo irreversible degradation in the short term (about one week); if degradation occurs, it can be retested and corrected in the second week.

[0053] The functional relationship between the steady-state solution and the "inverse calculation of material a_w":

[0054] When the external environment can be considered approximately stable within a 5–10 day window (or replaced by a window average), the steady-state partial pressure of equation (1) is:

[0055] (Formula 2) e*=(k_e·e_out+K_m·e_set) / (k_e+K_m);

[0056] If the target RH_tar (e.g., 62% RH) is required inside the chamber, the target partial pressure can be obtained first:

[0057] (Formula 3) e_tar=RH_tar·e_s(T_in);

[0058] The required partial pressure of the material is obtained by inverse solution of equation (2):

[0059] (Formula 4) e_set=e_tar+(k_e / K_m)·(e_tar-e_out);

[0060] Then calculate the water activity of the material:

[0061] (Formula 5) a_w=e_set / e_s(T_in);

[0062] The more humid the outside environment (the larger e_out), the lower a_w is required; the less airtight the enclosure (the larger k_e), the stronger the material is required (a larger K_m, or a lower a_w).

[0063] Parameter inverse calculation: Obtain k_e and K_m through calibration over one week;

[0064] The airtightness, i.e., the equivalent commutation coefficient k_e, is calculated from the control box (Blank);

[0065] If no humidity-regulating material is added to the control box, then K_m≈0, and equation (1) degenerates into:

[0066] (Formula 6) de_in,blank / dt=k_e·(e_out-e_in,blank)

[0067] Fitting method (least squares):

[0068] 1) Calculate e_in,blank(t) from RH_in,blank(t) and T_in,blank(t); Calculate e_out(t) from RH_out(t) and T_out(t);

[0069] 2) Approximate the derivative using the difference method: de / dt≈[e(t+Δt)-e(t)] / Δt;

[0070] 3) Treat k_e as a single constant to be estimated and fit it on all valid data points: de / dt=k_e·(e_out-e_in); 4) Output k_e as a single fixed value and the goodness of fit (R², residual distribution), and substitute the k_e into equation (1).

[0071] Note: The dimension of k_e is 1 / time, and its reciprocal τ_e = 1 / k_e is the "time constant for the partial pressure inside the chamber to approach the external environment". Under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions, and target life, k_e is used as a single fixed value; when the packaging or sealing method is changed, k_e should be recalibrated.

[0072] The material's moisture conditioning strength K_m is calculated from the test chamber (Test) under a known k_e. For the test chamber, e_set(t)=a_w·e_s(T_in(t)) in equation (1) is known (material a_w is known), the single fixed value k_e has been obtained from Step 1 and is regarded as unchanged here, and the unknown quantity is K_m.

[0073] Fitting method:

[0074] 1) Calculate e_in(t) from the Test box RH_in, T_in; calculate e_out(t) from external sources;

[0075] 2) Construct the regression: de_in / dt-k_e(e_out-e_in)=-K_m(e_in-e_set);

[0076] 3) Perform least squares fitting on K_m across all data points and output K_m, R², and confidence interval.

[0077] After obtaining full-time series data such as e_in(t), e_out(t), T(t), and RH(t), e_in(t) or de_in / dt is predicted using dynamic equations, and K_m is selected to minimize the sum of squared residuals (SSE) at all time points. If using derivative form, it can be written as residual_i = [Δe_in,i / Δt] - k_e,i·(e_out,i - e_in,i) + K_m·(e_in,i - e_set,i), and K_m is obtained by minimizing Σresidual_i^2. If using integral form, e_in,pred(t) is obtained by numerical integration using ODE, and Σ[e_in,obs(t_i) - e_in,pred(t_i)]^2 is minimized.

[0078] R² represents the proportion of variation in the observed data explained by the model, typically defined as R² = 1 - SSE / SST, where SST = Σ(e_in,obs - mean(e_in,obs))². If using derivative fitting, it can also be calculated based on the residuals of de_in / dt, but the report must specify the response variable used. Confidence intervals are typically reported as 95% CI, which can be calculated from the parametric covariance matrix Var(θ) ≈ σ²(JᵀJ)^-1, or obtained using bootstrap resampling. Formal reports recommend outputting the K_m point estimate, standard error, 95% CI, R², RMSE, and residual plot.

[0079] If the material exhibits a "response rate decaying over time" (the material gradually approaches equilibrium), K_m can be fitted as a piecewise constant or a first-order decay term can be added; typically, the average K_m over a 5–10 day window is used as the design parameter.

[0080] It provides a dynamic prediction tool that combines the equivalent exchange coefficient, storage environment information, target humidity range and storage period to calculate the equivalent humidity control coefficient, replacing the traditional empirical method of determining humidity control parameters. This effectively avoids the waste of packaging and transportation costs caused by excessive humidity control, or the problems of product deterioration, performance degradation and value loss caused by insufficient humidity control, thus ensuring product quality while controlling costs.

[0081] Adapted to short-term storage and transportation scenarios, this tool can be optimized for short-term storage and transportation needs of ≤90 days, eliminating redundant design logic of long-term warehousing, reducing the input of packaging materials, reducing the pressure of subsequent packaging waste disposal, and meeting the development requirements of low-carbon circulation.

[0082] This tool allows users to preset humidity control material settings and a library of equivalent humidity control coefficients, enabling rapid matching of suitable solutions. It also requires that the humidity control material be within its reversible moisture absorption and release range within the target shelf life and not approach its capacity limit, ensuring stable and reliable humidity control performance.

[0083] This tool can predict and specify that the packaging structure or sealing condition must remain stable within the target shelf life, providing a reliable guarantee for the effective implementation of humidity control measures and ensuring that the product maintains the target humidity throughout the storage and transportation process.

[0084] This tool can be widely applied to the distribution of various humidity-sensitive products, including fresh agricultural and sideline products, precision electronic components, biopharmaceutical reagents, and paper-based cultural and creative collectibles.

[0085] Example 2

[0086] A method for obtaining a dynamic forecasting tool for short-term product storage and transportation plans, the method comprising the following steps:

[0087] Step S1: Construct test and control groups based on the same batch of products, using the same packaging and sealing methods, the difference being that the test group packaging contains humidity-regulating material;

[0088] The specific experimental and control group settings are as follows:

[0089] Both the experimental group and the control group are equipped with temperature and humidity acquisition devices to collect temperature and humidity data at the same frequency.

[0090] At least 1–2 temperature and humidity sensors should be placed inside the packaging; for large boxes or locations where there may be local risks, it is recommended to place 2–4 sensors.

[0091] T_out / RH_out are recorded synchronously at different locations outside the packaging, near the packaging, or in the environmental chamber to establish realistic boundary conditions.

[0092] For high-risk locations such as bottle caps, seals, bag edges, and cold spots on the inner wall, it is recommended to use paired sensor patches to record both the surface temperature and the temperature and humidity of the air above the surface.

[0093] Sampling intervals of 5–10 min are recommended; if studying respiratory effects or disturbances caused by opening doors / carrying objects, 1 min is recommended.

[0094] The control group and the experimental group have the same packaging structure and no humidity-regulating material; an inert dummy load can be placed.

[0095] The experimental group included several combinations of different water activity and dosage of humidifying materials.

[0096] It may include any one or more of the following combinations:

[0097] Fresh one-way desiccant or existing desiccant from the customer can be used to establish a high initial moisture absorption capacity baseline.

[0098] Pre-treated desiccant to partial saturation can be used to demonstrate the earlier occurrence of RH drift as capacity decreases.

[0099] Fully saturated or desiccant can be used as a failure reference to distinguish between "material present" and "effective humidity control".

[0100] Selecting a_w grade conditioning material, such as 0.45, 0.55, 0.60, 0.65, or 0.75, can be used to verify the bidirectional buffering capability around a set RH.

[0101] Setting at least two delivery values ​​under the same a_w, such as the predicted minimum value and the +30% safety value, can be used to establish the scaling relationship between K_m and M / V or A / V.

[0102] Step S2: Given the external environmental temperature and humidity conditions and the target humidity range inside the packaging within the target period;

[0103] The external environmental temperature and humidity conditions within the given target period include: simulating the external environment with temperature and humidity changes within the target period under different seasons and different logistics routes.

[0104] Step S3: Collect data from the experimental and control groups within the target timeframe;

[0105] In practical applications, this includes: preprocessing the collected humidity and temperature data, specifically by standardizing the timestamps and sampling intervals, and removing short-term anomalies caused by opening the lid and / or handling.

[0106] In this embodiment, the experimental group and the control group are weighed, and the daily weight difference is calculated to enhance identification and verification.

[0107] In practical applications, the weighing can also be achieved in real time through a weighing sensor. The sampling timestamp of the weighing data is consistent with the temperature and humidity collection timestamps of the test group and the control group, or is unified after data processing.

[0108] In practical applications, weighing sensors are set on the test packaging or packaging support platform, and the data acquisition timestamps of the weighing sensors are made consistent with or unified by the data acquisition timestamps of the temperature and humidity sensors inside and outside the packaging, so as to obtain a synchronized dataset of W(t), T_in(t), RH_in(t), T_out(t), and RH_out(t).

[0109] In this embodiment, the weighing sensor may include a load sensor, an electronic balance module, a pressure sensor array, or other weighing module capable of continuously outputting the total weight W(t) of the packaging.

[0110] The total weight data of the product packaging is collected in real time by a weighing sensor, and the total weight data is correlated with the temperature and humidity data inside and outside the packaging at the same timestamp. This is used to obtain the net water load, the real-time mass flux of water entering or leaving, the remaining capacity of the humidifying material, and the capacity failure time. This upgrades the daily weighing verification of "how much water is left" to dynamic verification of "when water enters, at what speed, by whom, how much capacity is left, and when it fails".

[0111] Step S4: Based on the data collected from the control group, deduce the equivalent exchange coefficient;

[0112] The equivalent exchange coefficient is a quantification of the airtightness of product packaging. In practical applications, the equivalent exchange coefficient k_e is the equivalent exchange rate, and its reciprocal τ_e = 1 / k_e is the time constant.

[0113] The equivalent exchange coefficients derived from the collected control group data include: The expression `de_in,blank / dt = k_e·(e_out - e_in,blank)` is used to calculate `e_in,blank(t)` from `RH_in,blank(t)` and `T_in,blank(t)`. The expression `e_out(t)` is calculated from `RH_out(t)` and `T_out(t)`, and the derivative is approximated using the difference: `de / dt ≈ [e(t+Δt) - e(t)] / Δt`. `k_e` is treated as a single constant to be estimated and fitted over all valid data points: `de / dt = k_e·(e_out - e_in)`. The output is a single fixed value for `k_e`, which is then used to back-calculate and predict parameters for the experimental group with the same packaging and sealing methods. To reduce noise, it is recommended to perform a 30-minute moving average on `e_in` to remove points where `|e_out - e_in|` is too small, due to handling disturbances, or sensor anomalies.

[0114] Step S5: Based on the equivalent exchange coefficient and the collected test group data, the equivalent humidity regulation coefficient is derived, and a transferable prediction model is established by fitting the external environmental temperature and humidity conditions and the target humidity range inside the packaging, thus obtaining a dynamic prediction tool for short-term storage and transportation of products.

[0115] The equivalent humidity control coefficient comprehensively reflects the moisture absorption and release kinetics of the humidity control material and the moisture permeation limit of the coating material.

[0116] The prediction model is as follows:

[0117] de_in / dt=k_e·(e_out-e_in)-K_m(T_in,a_in)·(e_in-e_set) Formula (1) Where, a_in = RH_in / 100, a_out = RH_out / 100, e_in = a_in·es(T_in), e_out = a_out·es(T_out), e_set(T_in) = a_w·es(T_in); e_in is the partial pressure of water vapor inside the product packaging; e_out is the partial pressure of water vapor outside the product packaging; T_in is the internal temperature of the product packaging; T_out is the external ambient temperature of the product packaging; e_s(T_out) is the saturated water vapor pressure at the external temperature T_out; and e_s(T_in) is the saturated water vapor pressure at the internal temperature T_in. e_set(T_in) is the set partial pressure of the material at the internal temperature T_in at time t; a_w is the water activity of the material, a_in is the water activity inside the product packaging, and a_out is the water activity of the external environment of the product packaging; K_m is the equivalent humidity regulation coefficient; k_e is the equivalent exchange coefficient, which is a single fixed value obtained by fitting all effective data points of the control group under the corresponding packaging and sealing method, given external temperature and humidity conditions and target period, and input into the prediction model; de_in / dt is the rate of change of the water vapor partial pressure inside the product packaging with time t; RH_in is the relative humidity inside the product packaging, and RH_out is the relative humidity outside the product packaging.

[0118] When constructing the fitting dynamic prediction model, real-time synchronous weighing data is used as the third observation chain, so that the model can simultaneously use the temperature and humidity curves inside and outside the packaging, the water vapor partial pressure curve obtained by converting temperature and humidity, and the total weight change curve of the packaging for joint back calculation.

[0119] Existing packaging verification methods based on manual or daily weighing can usually only obtain the net change in total packaging weight after one or several days. It is difficult to identify the specific time when moisture enters the packaging, the peak entry rate, the capacity consumption corresponding to short-term high humidity shocks, and the hidden risks that the internal relative humidity has not yet exceeded the limit but the capacity of the humidity-regulating material has been rapidly consumed. Therefore, it is impossible to form a dynamic evidence chain that closes the relationship between temperature and humidity status and moisture mass flux.

[0120] Determined by the material system. K_m comprehensively reflects the moisture absorption and release kinetics of the material and the moisture permeation limit of the coating material (which can vary with the coupling of T_in and a_in); k_e is a single equivalent constant obtained by fitting all effective data points of the control group under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions and target period, and comprehensively reflects the packaging airtightness, moisture permeability, leakage and breathing effect.

[0121] e_in(t): partial pressure of water vapor inside the box (kPa); e_out(t): partial pressure of water vapor outside the box (kPa); e_s(T): saturated water vapor pressure at temperature T (kPa); e_set(t): material set partial pressure (determined by material a_w and temperature T_in inside the box): e_set(t)=a_w·e_s(T_in(t)); k_e: equivalent exchange coefficient of the box obtained by fitting all effective data points of the control group (1 / day or 1 / h), which is used as a single equivalent constant within the corresponding calibration conditions and prediction range, reflecting the airtightness of the packaging; K_m: equivalent "pullback coefficient" of the material (1 / day or 1 / h), reflecting the moisture conditioning strength of the material.

[0122] In this embodiment, a real-time weighing mass conservation equation is introduced:

[0123] ΔW_net(t)=[W_test(t)-W_test(0)]-[W_blank(t)-W_blank(0)], J_net(t)=dΔW_net(t) / dt, where W_test(t) is the total weight of the test package containing the sample and / or conditioning material, W_blank(t) is the total weight of the blank control package, and J_net(t) is the net water mass flux after deducting blank drift.

[0124] When J_net(t) and e_out(t)-e_in(t) change synchronously in time, it can be used to enhance the identification of the equivalent exchange coefficient ke on the packaging side; when RH_in(t) remains stable while ΔW_net(t) continues to increase, it indicates that external moisture is entering the packaging and being absorbed by the humidity conditioning material or product, and the capacity consumption rate of the humidity conditioning material can be further calculated.

[0125] The change in net weight was calculated based on the packaging weight of the experimental group W_test(t) and the packaging weight of the blank control group W_blank(t):

[0126] ΔW_net(t)=[W_test(t)-W_test(0)]-[W_blank(t)-W_blank(0)]

[0127] The real-time mass flux of moisture entering or leaving the packaging system is obtained by J_net(t)=dΔW_net(t) / dt.

[0128] The cumulative moisture load q(t) = ∫J_net(t)dt or the cumulative moisture absorption load q_abs(t) = ∫max[J_net(t),0]dt within the target storage period is calculated based on the real-time quality flux, and the capacity consumption rate q(t) / C_effective or q_abs(t) / C_effective is obtained by combining the effective working capacity C_effective of the moisture conditioning material.

[0129] The remaining capacity of the humidity-regulating material, C_remaining(t) = C_effective - q_abs(t), is obtained based on the capacity consumption rate. The capacity failure time, t_capacity_fail, is determined when C_remaining(t) is lower than a preset threshold or when the equivalent humidity-regulating coefficient Km decays.

[0130] The time when the internal humidity of the product packaging exceeds the target humidity range (t_RH_fail) is compared with the capacity failure time (t_capacity_fail), and the earlier of the two times is taken as the risk failure time of the packaging and storage plan, or as the basis for adjusting the water activity of materials, the amount of materials added, the effective moisture permeability area, or the airtightness of the packaging.

[0131] In this embodiment, the total weight W(t) of the packaging is obtained in real time by a weighing sensor or weighing module set at the bearing position of the product packaging. W(t) is made to have the same or mappable timestamp as the internal temperature and humidity of the product packaging and the external temperature and humidity, so as to synchronously obtain the water vapor partial pressure response and the change in the total moisture mass of the packaging.

[0132] In practical applications, the net water mass flux J_net(t) can be obtained from the real-time total weight difference, and J_net(t) can be compared with the model flux derived from the water vapor partial pressure difference e_out(t)-e_in(t) to perform mass conservation verification, so as to improve the back calculation stability of the equivalent exchange coefficient ke and the equivalent humidity regulation coefficient Km.

[0133] In practical applications, the ratio of the cumulative moisture load q(t) within the target storage period to the effective capacity C_effective of the humidity conditioning material can be used to output the capacity consumption rate, remaining capacity, capacity warning time, and capacity failure time, which can be used to determine whether the material capacity is close to being exhausted when the internal relative humidity is still within the target range.

[0134] In this embodiment, for a 7-day short-term verification window, q_7d=∫_0^7dJ_net(t)dt or q_abs,7d=∫_0^7dmax[J_net(t),0]dt can be calculated, and q_7d / C_effective or q_abs,7d / C_effective can be used as the effective capacity consumption rate of the material.

[0135] Wherein, C_effective is the effective moisture absorption or release capacity of the humidity-regulating material within the product's allowable humidity window, rather than the material's fully saturated capacity; when the capacity consumption rate exceeds the preset threshold, even if RH_in(t) is still within the target range, it should be determined that the packaging scheme has a risk of insufficient capacity margin.

[0136] Basic assumptions:

[0137] The chamber volume V is fixed, and the effect of changes in the total air volume inside the chamber on the partial pressure of water vapor is negligible; the airtightness is represented by an equivalent exchange coefficient that remains fixed within the corresponding calibration conditions and prediction range (including the combined effects of infiltration, micro-leakage, and breathing effect); the humidity conditioning material can be considered as an equivalent "buffer" that pulls the partial pressure inside the chamber back to its set partial pressure e_set, and its strength is represented by the coefficient K_m; the performance of the chamber and materials does not undergo irreversible degradation in the short term (about one week); if degradation occurs, it can be retested and corrected in the second week.

[0138] The functional relationship between the steady-state solution and the "inverse calculation of material a_w":

[0139] When the external environment can be considered approximately stable within a 5–10 day window (or replaced by a window average), the steady-state partial pressure of equation (1) is:

[0140] (Formula 2) e*=(k_e·e_out+K_m·e_set) / (k_e+K_m);

[0141] If the target RH_tar (e.g., 62% RH) is required inside the chamber, the target partial pressure can be obtained first:

[0142] (Formula 3) e_tar=RH_tar·e_s(T_in);

[0143] The required partial pressure of the material is obtained by inverse solution of equation (2):

[0144] (Formula 4) e_set=e_tar+(k_e / K_m)·(e_tar-e_out);

[0145] Then calculate the water activity of the material:

[0146] (Formula 5) a_w=e_set / e_s(T_in);

[0147] The more humid the outside environment (the larger e_out), the lower a_w is required; the less airtight the enclosure (the larger k_e), the stronger the material is required (a larger K_m, or a lower a_w).

[0148] Parameter inverse calculation: Obtain k_e and K_m through calibration over one week;

[0149] The airtightness, i.e., the equivalent commutation coefficient k_e, is calculated from the control box (Blank);

[0150] If no humidity-regulating material is added to the control box, then K_m≈0, and equation (1) degenerates into:

[0151] (Formula 6) de_in,blank / dt=k_e·(e_out-e_in,blank)

[0152] Fitting method (least squares):

[0153] 1) Calculate e_in,blank(t) from RH_in,blank(t) and T_in,blank(t); Calculate e_out(t) from RH_out(t) and T_out(t);

[0154] 2) Approximate the derivative using the difference method: de / dt≈[e(t+Δt)-e(t)] / Δt;

[0155] 3) Treat k_e as a single constant to be estimated and fit it on all valid data points: de / dt=k_e·(e_out-e_in); 4) Output k_e as a single fixed value and the goodness of fit (R², residual distribution).

[0156] Note: The dimension of k_e is 1 / time, and its reciprocal τ_e = 1 / k_e is the "time constant for the partial pressure inside the chamber to approach the external environment". Under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions, and target life, k_e is used as a single fixed value; when the packaging or sealing method is changed, k_e should be recalibrated.

[0157] The material's moisture conditioning strength K_m is calculated from the test chamber (Test) under a known k_e. For the test chamber, in equation (1), e_set(t)=a_w·e_s(T_in(t)) is known (material a_w is known), the single fixed value k_e has been obtained from Step 1 and is regarded as remaining unchanged, and the unknown quantity is K_m.

[0158] Fitting method:

[0159] 1) Calculate e_in(t) from the Test box RH_in, T_in; calculate e_out(t) from external sources;

[0160] 2) Construct the regression: de_in / dt-k_e(e_out-e_in)=-K_m(e_in-e_set);

[0161] 3) Perform least squares fitting on K_m across all data points and output K_m, R², and confidence interval.

[0162] After obtaining full-time series data such as e_in(t), e_out(t), T(t), and RH(t), e_in(t) or de_in / dt is predicted using dynamic equations, and K_m is selected to minimize the sum of squared residuals (SSE) at all time points. If using derivative form, it can be written as residual_i = [Δe_in,i / Δt] - k_e·(e_out,i - e_in,i) + K_m·(e_in,i - e_set,i), and K_m is obtained by minimizing Σresidual_i^2. If using integral form, e_in,pred(t) is obtained by numerical integration using ODE, and Σ[e_in,obs(t_i) - e_in,pred(t_i)]^2 is minimized.

[0163] R² represents the proportion of variation in the observed data explained by the model, typically defined as R² = 1 - SSE / SST, where SST = Σ(e_in,obs - mean(e_in,obs))². If using derivative fitting, it can also be calculated based on the residuals of de_in / dt, but the report must specify the response variable used. Confidence intervals are typically reported as 95% CI, which can be calculated from the parametric covariance matrix Var(θ) ≈ σ²(JᵀJ)^-1, or obtained using bootstrap resampling. Formal reports recommend outputting the K_m point estimate, standard error, 95% CI, R², RMSE, and residual plot.

[0164] If the material exhibits a "response rate decaying over time" (the material gradually approaches equilibrium), K_m can be fitted as a piecewise constant or a first-order decay term can be added; typically, the average K_m over a 5–10 day window is used as the design parameter.

[0165] Forming a closed chain of evidence for mass conservation: By synchronously collecting the total weight of the packaging in real time through weighing sensors, the cumulative verification of daily weighing is upgraded to continuous moisture flux monitoring. This can determine when moisture enters, the rate of entry, the proportion borne by the humidification materials or products, the remaining capacity, and the expected failure time, thereby improving the prediction accuracy and anomaly diagnosis capabilities of short-term storage and transportation plans.

[0166] It provides a dynamic prediction tool that combines the equivalent exchange coefficient, storage environment information, target humidity range and storage period to calculate the equivalent humidity control coefficient, replacing the traditional empirical method of determining humidity control parameters. This effectively avoids the waste of packaging and transportation costs caused by excessive humidity control, or the problems of product deterioration, performance degradation and value loss caused by insufficient humidity control, thus ensuring product quality while controlling costs.

[0167] Adapted to short-term storage and transportation scenarios, this tool can be optimized for short-term storage and transportation needs of ≤90 days, eliminating redundant design logic of long-term warehousing, reducing the input of packaging materials, reducing the pressure of subsequent packaging waste disposal, and meeting the development requirements of low-carbon circulation.

[0168] This tool allows users to preset humidity control material settings and a library of equivalent humidity control coefficients, enabling rapid matching of suitable solutions. It also requires that the humidity control material be within its reversible moisture absorption and release range within the target shelf life and not approach its capacity limit, ensuring stable and reliable humidity control performance.

[0169] This tool can predict and specify that the packaging structure or sealing condition must remain stable within the target shelf life, providing a reliable guarantee for the effective implementation of humidity control measures and ensuring that the product maintains the target humidity throughout the storage and transportation process.

[0170] This tool can be widely applied to the distribution of various humidity-sensitive products, including fresh agricultural and sideline products, precision electronic components, biopharmaceutical reagents, and paper-based cultural and creative collectibles.

[0171] Example 3

[0172] A week-long calibration experiment was conducted to build and validate a dynamic prediction tool for short-term product storage and transportation. The specific experimental process included the following:

[0173] S01: Prepare the test chamber and control chamber;

[0174] (A) Test Box:

[0175] Box: 65L storage box, sealing method as described above;

[0176] The contents consist of several combinations of water activity and dosage of humidifying tablets, as shown in Table 1 below:

[0177] Table 1:

[0178]

[0179] Sensors: Two temperature and humidity loggers are placed in each box (sampling every 5–10 minutes);

[0180] (B) Blank box (control box):

[0181] Chamber: Same model, volume, and sealing method as the test chamber;

[0182] Contents: Inert load / dummy load (i.e., no moisture absorption or release), weight and heat capacity are similar to those of the Test chamber;

[0183] Sensors: Two temperature and humidity recorders inside the box (at the same sampling interval);

[0184] S02: Select or arrange the external environment;

[0185] Set up the external environment according to the actual storage location of the product (for example, it can be set up according to the customer's logistics route through a climate chamber): and record T_out and RH_out (an environmental recorder is placed near the enclosure).

[0186] S03: Daily Weighing Standard Operating Procedure;

[0187] SOP (Execute at the same time window every day):

[0188] 1) Simultaneously remove several Test and Blank boxes;

[0189] 2) Wipe the outer surface dry in the same way (using the same material wiping cloth, the same path, and the same pressure);

[0190] 3) Allow the surface to stand for a fixed time (2–3 minutes recommended) to bring the surface temperature close to the weighing environment;

[0191] 4) Weigh and record the results separately: M_test(d) and M_blank(d), where d is the date;

[0192] 5) Record the temperature T_w and relative humidity RH_w of the weighing environment;

[0193] 6) Return the enclosure to its original environment for continued exposure.

[0194] In practical applications, weighing readings can be affected by changes in air buoyancy, surface condensation / adsorbed water films, etc., easily resulting in drifts of several grams. By using the Test / Blank simultaneous weighing process and performing differential weighing, these systematic errors can be significantly offset.

[0195] S04: Set data fields, units, and sampling;

[0196]

[0197] S05: Perform the following preprocessing on the raw relative humidity (RH) / temperature (T) data:

[0198] Unify timestamps and sampling intervals (interpolate or resample to 10 minutes);

[0199] Eliminate short-term anomalies caused by opening / handling;

[0200] The mean / quantile of the external environment can be calculated within a 5-10 day window for steady-state design and risk assessment.

[0201] S06: Construct a dynamic prediction model;

[0202] 601. Selection of modeling variables (partial pressure of water vapor e);

[0203] To avoid the problem of RH non-conservation caused by temperature fluctuations, water vapor partial pressure e or humidity ratio ω is used as a state variable. This scheme uses water vapor partial pressure e (kPa) for fitting and prediction, and the final output is expressed in RH. At the same time, water vapor activity a (a=RH / 100, dimensionless) is introduced as a humidity state variable to calculate water vapor partial pressure and describe the temperature and humidity coupling input of K_m.

[0204] 602. Core Dynamic Equation (First-Order Equivalent Model of Packaging-Material Coupling): de_in / dt=k_e·(e_out-e_in)-K_m(T_in,a_in)·(e_in-e_set) Formula (1) Where, a_in = RH_in / 100, a_out = RH_out / 100, e_in = a_in·es(T_in), e_out = a_out·es(T_out), e_set(T_in) = a_w·es(T_in); e_in is the partial pressure of water vapor inside the product packaging; e_out is the partial pressure of water vapor outside the product packaging; T_in is the internal temperature of the product packaging; T_out is the external ambient temperature of the product packaging; e_s(T_out) is the saturated water vapor pressure at the external temperature T_out; and e_s(T_in) is the saturated water vapor pressure at the internal temperature T_in. e_set(T_in) is the set partial pressure of the material at the internal temperature T_in at time t; a_w is the water activity of the material, a_in is the water activity inside the product packaging, and a_out is the water activity of the external environment of the product packaging; K_m is the equivalent humidity regulation coefficient; k_e is the equivalent exchange coefficient, which is a single fixed value obtained by fitting all effective data points of the control group under the corresponding packaging and sealing method, given external temperature and humidity conditions and target period, and input into the prediction model; de_in / dt is the rate of change of the water vapor partial pressure inside the product packaging with time t; RH_in is the relative humidity inside the product packaging, and RH_out is the relative humidity outside the product packaging.

[0205] a_w is determined by the material system. K_m comprehensively reflects the material's moisture absorption and release kinetics and the moisture permeation limit of the coating material (which can vary with the coupling of T_in and a_in); k_e is a single equivalent constant obtained by fitting all effective data points of the control group under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions and target period, and comprehensively reflects the packaging's airtightness, moisture permeability, leakage and breathing effect.

[0206] e_in(t): partial pressure of water vapor inside the box (kPa); e_out(t): partial pressure of water vapor outside the box (kPa); e_s(T): saturated water vapor pressure at temperature T (kPa); e_set(t): material set partial pressure (determined by material a_w and temperature T_in inside the box): e_set(t)=a_w·e_s(T_in(t)); k_e: equivalent exchange coefficient of the box obtained by fitting all effective data points of the control group (1 / day or 1 / h), which is used as a single equivalent constant within the corresponding calibration conditions and prediction range, reflecting the airtightness of the packaging; K_m: equivalent "pullback coefficient" of the material (1 / day or 1 / h), reflecting the moisture conditioning strength of the material.

[0207] 603. Basic Assumptions; The chamber volume V is fixed, and the effect of changes in the total air volume inside the chamber on the partial pressure of water vapor is negligible; the airtightness is represented by an equivalent exchange coefficient that remains fixed within the corresponding calibration conditions and prediction range (including the combined effects of infiltration, micro-leakage, and breathing effect); the humidity conditioning material can be considered as an equivalent "buffer" that pulls the partial pressure inside the chamber back to its set partial pressure e_set, and its strength is represented by the coefficient K_m; the performance of the chamber and materials does not undergo irreversible degradation in the short term (about one week); if degradation occurs, it can be retested and corrected in the second week.

[0208] 604. The functional relationship between the steady-state solution and the "inverse calculation of material a_w";

[0209] When the external environment can be considered approximately stable within a 5–10 day window (or replaced by a window average), the steady-state partial pressure of equation (1) is:

[0210] (Formula 2) e*=(k_e·e_out+K_m·e_set) / (k_e+K_m);

[0211] If the target RH_tar (e.g., 62% RH) is required inside the chamber, the target partial pressure can be obtained first:

[0212] (Formula 3) e_tar=RH_tar·e_s(T_in);

[0213] The required partial pressure of the material is obtained by inverse solution of equation (2):

[0214] (Formula 4) e_set=e_tar+(k_e / K_m)·(e_tar-e_out);

[0215] Then calculate the water activity of the material:

[0216] (Formula 5) a_w=e_set / e_s(T_in);

[0217] The more humid the outside environment (the larger e_out), the lower a_w is required; the less airtight the enclosure (the larger k_e), the stronger the material is required (a larger K_m, or a lower a_w).

[0218] S07: Parameter inverse calculation: Obtain k_e and K_m through one week of calibration;

[0219] 701: Step 1: Calculate the airtightness, i.e., the equivalent commutation coefficient k_e, from the control box (Blank);

[0220] If no humidity-regulating material is added to the control box, then K_m≈0, and equation (1) degenerates into:

[0221] (Formula 6) de_in,blank / dt=k_e·(e_out-e_in,blank)

[0222] Fitting method (least squares):

[0223] 1) Calculate e_in,blank(t) from RH_in,blank(t) and T_in,blank(t); Calculate e_out(t) from RH_out(t) and T_out(t);

[0224] 2) Approximate the derivative using the difference method: de / dt≈[e(t+Δt)-e(t)] / Δt;

[0225] 3) Treat k_e as a single constant to be estimated and fit it on all valid data points: de / dt=k_e·(e_out-e_in); 4) Output k_e as a single fixed value and the goodness of fit (R², residual distribution).

[0226] Note: The dimension of k_e is 1 / time, and its reciprocal τ_e = 1 / k_e is the "time constant for the partial pressure inside the chamber to approach the external environment". Under the corresponding packaging and sealing methods, given external environmental temperature and humidity conditions, and target life, k_e is used as a single fixed value; when the packaging or sealing method is changed, k_e should be recalibrated.

[0227] 702: Step 2: Calculate the material's moisture conditioning strength K_m from the test chamber (Test) under the known k_e; For the test chamber, in equation (1), e_set(t)=a_w·e_s(T_in(t)) is known (material a_w is known), the single fixed value k_e has been obtained from Step 1 and is regarded as remaining unchanged, and the unknown quantity is K_m.

[0228] Fitting method:

[0229] 1) Calculate e_in(t) from the Test box RH_in, T_in; calculate e_out(t) from external sources;

[0230] 2) Construct the regression: de_in / dt-k_e(e_out-e_in)=-K_m(e_in-e_set);

[0231] 3) Perform least squares fitting on K_m across all data points and output K_m, R², and confidence interval.

[0232] After obtaining full-time series data such as e_in(t), e_out(t), T(t), and RH(t), e_in(t) or de_in / dt is predicted using dynamic equations, and K_m is selected to minimize the sum of squared residuals (SSE) at all time points. If using derivative form, it can be written as residual_i = [Δe_in,i / Δt] - k_e·(e_out,i - e_in,i) + K_m·(e_in,i - e_set,i), and K_m is obtained by minimizing Σresidual_i^2. If using integral form, e_in,pred(t) is obtained by numerical integration using ODE, and Σ[e_in,obs(t_i) - e_in,pred(t_i)]^2 is minimized.

[0233] R² represents the proportion of variation in the observed data explained by the model, typically defined as R² = 1 - SSE / SST, where SST = Σ(e_in,obs - mean(e_in,obs))². If using derivative fitting, it can also be calculated based on the residuals of de_in / dt, but the report must specify the response variable used. Confidence intervals are typically reported as 95% CI, which can be calculated from the parametric covariance matrix Var(θ) ≈ σ²(JᵀJ)^-1, or obtained using bootstrap resampling. Formal reports recommend outputting the K_m point estimate, standard error, 95% CI, R², RMSE, and residual plot.

[0234] If the material exhibits a "response rate decaying over time" (the material gradually approaches equilibrium), K_m can be fitted as a piecewise constant or a first-order decay term can be added; typically, the average K_m over a 5–10 day window is used as the design parameter.

[0235] Example 4

[0236] The implementation process for obtaining short-term product storage and transportation plans based on the dynamic prediction tool for short-term product storage and transportation plans described in this invention is as follows:

[0237] enter:

[0238] External conditions: T_out(t), RH_out(t) (or window statistics provided for 5–10 days);

[0239] Target: RH_tar (e.g., 62%) and allowable fluctuation range ±ΔRH;

[0240] Time window: 5–10 days;

[0241] Parameters: k_e (box airtightness, also known as equivalent exchange coefficient, a single fixed value obtained by calibrating all valid data points of the control group), K_m (material equivalent humidity regulation coefficient, derived from the experimental group calibration), material dosage M, and product packaging volume V.

[0242] step:

[0243] 1) Calculate the partial pressure of water vapor outside the product packaging at time t, e_out(t) = RH_out·e_s(T_out), and the representative value of the internal temperature of the product packaging, T_in (which can be the time average of Test or the critical operating point).

[0244] 2) The material strength, i.e. the equivalent moisture regulation coefficient K_m, is calculated using K_m=α(M / V);

[0245] 3) The water activity a_w of the candidate material is obtained by using the steady-state inverse solution (Equation 4–5);

[0246] 4) Use the dynamic model (Equation 1) to perform a 5-10 day simulation. Input the partial pressure of water vapor inside the product packaging e_out(t) and the temperature inside the product packaging T_in(t) at time t, and check whether the relative humidity RH_in(t) inside the product packaging at time t falls within the target range.

[0247] 5) If not satisfied: then select the corresponding humidity control material setting scheme or product packaging scheme according to the preset humidity control material setting scheme and equivalent humidity control coefficient reference library.

[0248] In practical applications, when the constructed fitting dynamic prediction model is used to perform fitting dynamic simulation prediction to obtain the equivalent humidity control coefficient, the judgment criteria for whether the simulation prediction meets the requirements are designed and formulated according to the actual product. This embodiment is only an example. For example, the judgment criteria shown in Table 1 below can be used to determine whether the requirements are met:

[0249] Table 1:

[0250]

[0251] In practical applications, different product judgment indicators can be set according to different product types, as shown in Table 2 below:

[0252] Table 2

[0253]

[0254] Example 5

[0255] The implementation process for obtaining short-term product storage and transportation plans based on the dynamic prediction tool for short-term product storage and transportation plans described in this invention is as follows:

[0256] enter:

[0257] External conditions: T_out(t), RH_out(t) (or window statistics provided for 5–10 days);

[0258] Target: RH_tar (e.g., 62%) and allowable fluctuation range ±ΔRH;

[0259] Time window: 5–10 days;

[0260] Parameters: k_e (box airtightness, also known as equivalent exchange coefficient, a single fixed value obtained by calibrating all valid data points of the control group), K_m (material equivalent humidity regulation coefficient, derived from the experimental group calibration), material dosage M, and product packaging volume V.

[0261] step:

[0262] 1) Calculate the partial pressure of water vapor outside the product packaging at time t, e_out(t) = RH_out·e_s(T_out), and the representative value of the internal temperature of the product packaging, T_in (which can be the time average of Test or the critical operating point).

[0263] 2) The material strength, i.e. the equivalent moisture regulation coefficient K_m, is calculated using K_m=α(M / V);

[0264] 3) The water activity a_w of the candidate material is obtained by using the steady-state inverse solution (Equation 4–5);

[0265] 4) Use the dynamic model (Equation 1) to perform a 5-10 day simulation. Input the partial pressure of water vapor inside the product packaging e_out(t) and the temperature inside the product packaging T_in(t) at time t, and check whether the relative humidity RH_in(t) inside the product packaging at time t falls within the target range.

[0266] 5) If not satisfied: First adjust the water activity of the material a_w; if still insufficient, adjust the amount of material M (improved by K_m), or change the packaging structure or sealing method to improve airtightness, and recalibrate k_e.

[0267] In this embodiment, steps 4) and 5) above can be used multiple times to simulate and predict whether the requirements are met under different packaging scheme changes. That is, a_w and K_m are adjusted respectively, or the packaging structure or sealing method is changed and k_e is recalibrated to obtain a packaging scheme that meets the requirements. Then, based on the simulation prediction results of different packaging schemes, suggestions for packaging improvement are given.

[0268] In practical applications, when the constructed fitting dynamic prediction model is used to perform fitting dynamic simulation prediction to obtain the equivalent humidity control coefficient, the judgment criteria for whether the simulation prediction meets the requirements are designed and formulated according to the actual product. This embodiment is only an example. For example, the judgment criteria shown in Table 1 of Embodiment 4 can be used to determine whether the requirements are met.

[0269] In practical applications, different product judgment indicators can be set according to different product types, as shown in Table 2 of Example 4.

[0270] Example 6

[0271] The implementation process for obtaining short-term product storage and transportation plans based on the dynamic prediction tool for short-term product storage and transportation plans described in this invention is as follows:

[0272] enter:

[0273] External conditions: T_out(t), RH_out(t) (or window statistics provided for 5–10 days);

[0274] Target: RH_tar (e.g., 62%) and allowable fluctuation range ±ΔRH;

[0275] Time window: 5–10 days;

[0276] Parameters: k_e (box airtightness, also known as equivalent exchange coefficient, a single fixed value obtained by calibrating all valid data points of the control group), K_m (material equivalent humidity regulation coefficient, derived from the experimental group calibration), material dosage M, and product packaging volume V.

[0277] step:

[0278] 1) Calculate the partial pressure of water vapor outside the product packaging at time t, e_out(t) = RH_out·e_s(T_out), and the representative value of the internal temperature of the product packaging, T_in (which can be the time average of Test or the critical operating point).

[0279] 2) The material strength, i.e. the equivalent moisture regulation coefficient K_m, is calculated using K_m=α(M / V);

[0280] 3) The water activity a_w of the candidate material is obtained by using the steady-state inverse solution (Equation 4–5);

[0281] 4) Use the dynamic model (Equation 1) to perform a 5-10 day simulation. Input the partial pressure of water vapor inside the product packaging e_out(t) and the temperature inside the product packaging T_in(t) at time t, and check whether the relative humidity RH_in(t) inside the product packaging at time t falls within the target range.

[0282] 5) If not satisfied: First adjust the water activity of the material a_w; if still insufficient, adjust the amount of material M (improved by K_m), or change the packaging structure or sealing method to improve airtightness, and recalibrate k_e.

[0283] 6) Equipped with weighing sensors, simultaneously calculating ΔW_net(t) and J_net(t) to identify the time period during which moisture enters the packaging and the peak throughput. and cumulative capacity consumption ;

[0284] 7) Based on C_remaining(t)=C_effective−q_abs(t), predict the capacity warning time and capacity failure time, and use this result together with RH_in(t) failure time to generate packaging improvement suggestions.

[0285] In this embodiment, steps 4) and 5) above can be used multiple times to simulate and predict whether the requirements are met under different packaging scheme changes. That is, a_w and K_m are adjusted respectively, or the packaging structure or sealing method is changed and k_e is recalibrated to obtain a packaging scheme that meets the requirements. Then, based on the simulation prediction results of different packaging schemes, suggestions for packaging improvement are given.

[0286] In practical applications, when the constructed fitting dynamic prediction model is used to perform fitting dynamic simulation prediction to obtain the equivalent humidity control coefficient, the judgment criteria for whether the simulation prediction meets the requirements are designed and formulated according to the actual product. This embodiment is only an example. For example, the judgment criteria shown in Table 3 below can be used to determine whether the requirements are met:

[0287] Table 3:

[0288]

[0289] In practical applications, different product judgment indicators can be set according to different product types, as shown in Table 4 below:

[0290] Table 4:

[0291]

[0292] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for obtaining a dynamic forecasting tool for short-term product storage and transportation plans, characterized in that, The acquisition method includes the following steps: Test and control groups were constructed based on the same batch of products, using the same packaging and sealing methods. The difference was that the test group packaging contained humidity-regulating materials. Given the external environmental temperature and humidity conditions and the target humidity range inside the packaging within the target period; Data were collected from the experimental and control groups within the target timeframe; Based on the data collected from the control group, the equivalent exchange coefficient can be deduced. Based on the equivalent exchange coefficient and the collected test group data, the equivalent humidity regulation coefficient is derived, and a transferable prediction model is established by fitting the external environmental temperature and humidity conditions and the target humidity range inside the packaging, thus obtaining a dynamic prediction tool for short-term product storage and transportation.

2. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, Both the experimental group and the control group are equipped with temperature and humidity acquisition devices to collect temperature and humidity data at the same frequency.

3. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The experimental group included several combinations of different water activity and dosage of humidifying materials.

4. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The external environmental temperature and humidity conditions within the given target period include: simulating the external environment with temperature and humidity changes within the target period under different seasons and different logistics routes.

5. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The data collection for the experimental and control groups within the target period includes: weighing the experimental and control groups and calculating the daily weight difference mass to enhance identification and verification.

6. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The equivalent exchange coefficient is a quantification of the airtightness of the product packaging, and is a single fixed value under the packaging and sealing method, the external environmental temperature and humidity conditions within the given target period, and the target period.

7. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The equivalent humidity control coefficient comprehensively reflects the moisture absorption and release kinetics of the humidity control material and the moisture permeation limit of the coating material.

8. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to any one of claims 1 to 7, characterized in that, The prediction model is as follows: de_in / dt=k_e·(e_out-e_in)-K_m(T_in,a_in)·(e_in-e_set) Formula (1) Where, a_in = RH_in / 100, a_out = RH_out / 100, e_in = a_in·e_s(T_in), e_out = a_out·e_s(T_out), e_set(T_in) = a_w·e_s(T_in); e_in is the partial pressure of water vapor inside the product packaging; e_out is the partial pressure of water vapor outside the product packaging; T_in is the internal temperature of the product packaging; T_out is the external ambient temperature of the product packaging; e_s(T_out) is the saturated water vapor pressure at the external temperature T_out; e_s(T_in) is the internal temperature of the product packaging. The partial pressure of water vapor inside the product packaging is T_in; e_set(T_in) is the set partial pressure of the material at the internal temperature T_in at time t; a_w is the water activity of the material, a_in is the water activity inside the product packaging, and a_out is the water activity of the external environment of the product packaging; K_m is the equivalent humidity regulation coefficient; k_e is the equivalent exchange coefficient, which is a single fixed value obtained by fitting all valid data points collected by the control group within the target period; de_in / dt is the rate of change of the partial pressure of water vapor inside the product packaging with time t; RH_in is the relative humidity inside the product packaging, and RH_out is the relative humidity outside the product packaging.

9. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 8, characterized in that, The step of deriving the equivalent exchange coefficient based on the collected control group data includes: under the external environmental temperature and humidity conditions within the target period and the given target period, taking k_e as a single constant to be estimated, calculating e_in,blank(t) from RH_in,blank(t) and T_in,blank(t) according to de_in,blank / dt=k_e·(e_out-e_in,blank); calculating e_out(t) from RH_out(t) and T_out(t), approximating the derivative with the difference de / dt≈[e(t+Δt)-e(t)] / Δt, and fitting de / dt=k_e·(e_out-e_in) on all valid data points collected within the target period; outputting as a single fixed value k_e, and substituting k_e into the prediction model.

10. The method for obtaining the dynamic forecasting tool for short-term product storage and transportation schemes according to claim 1, characterized in that, The method for obtaining the dynamic prediction tool for short-term storage and transportation of the product includes: preprocessing the collected humidity and temperature data, specifically by unifying the timestamp and sampling interval, and removing short-term anomalies caused by opening the lid and / or handling.

Images