Sample integrity intelligent monitoring and temperature control method based on multi-parameter prediction

Abstract

The invention relates to the technical field of data processing related to cold-chain logistics and the Internet of Things, and discloses a sample integrity intelligent monitoring and temperature control method based on multi-parameter prediction. The method comprises the following steps: periodically acquiring multi-dimensional parameters such as temperature, illumination, vibration, box opening state, battery remaining capacity and phase change material state; carrying out weighted fusion calculation on the temperature, illumination, vibration and box opening states to obtain a sample integrity index SII (t), wherein the temperature accumulation deviation degree is obtained through time weighted integration; based on the change trend of the SII (t), the residual electric quantity of the battery and the state of the phase change material, predicting residual safe transportation time (RSTT) through a minimum bottleneck principle; and the temperature control power is intelligently adjusted according to the comparison result of the RSTT and the predicted transportation duration, and early warning is triggered. According to the method, the technical upgrade from recording-response to evaluation-prediction-optimization is realized, and the sample integrity evaluation accuracy, the temperature control cruising ability and the early warning timeliness are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of data processing technology related to cold chain logistics and the Internet of Things, and in particular to a multi-parameter fusion monitoring, sample integrity assessment and intelligent active temperature control technology for the transportation and preservation of environmental, biological or medical testing samples. Background Technology

[0002] In fields such as environmental monitoring, biomedicine, and scientific research testing, the transportation and preservation of samples have a decisive impact on the accuracy and reliability of the final test results. Taking environmental testing as an example, soil samples, water samples, and atmospheric samples need to be sent to the laboratory for analysis as soon as possible after collection. During transportation, temperature must be strictly controlled, light exposure must be avoided, and severe vibrations must be prevented; otherwise, key components in the samples, such as volatile organic compounds, microorganisms, and heavy metals, may undergo irreversible changes. In the biomedical field, vaccines, blood products, and biological reagents are extremely sensitive to temperature; even short-term temperature deviations can lead to loss of biological activity, posing medical safety risks. In the scientific research field, the transportation of gene samples, cell cultures, and chemical reagents also faces stringent environmental control requirements.

[0003] Currently, cold chain transportation of samples mainly relies on passive insulation (such as ice packs, dry ice, or phase change materials) combined with temperature recorders, or on active temperature-controlled boxes based on thermoelectric cooling (TEC) with integrated GPS and data upload functions. However, existing technologies have revealed many limitations in practical applications. Traditional monitoring systems focus almost exclusively on temperature, neglecting the cumulative impact of non-temperature factors such as vibration, light exposure, and exposure upon opening the box on sample integrity. Taking environmental testing samples as an example, vibration and impact during transportation may damage the structure of soil samples or cause suspended solids to settle in water samples; light exposure can accelerate the volatilization of volatile organic compounds or the degradation of photosensitive components; and frequent opening of the box for inspection can cause temperature fluctuations and the risk of external contamination. The combined effect of these factors is often more complex and insidious than a single temperature deviation.

[0004] Furthermore, most existing systems employ threshold-triggered alarm mechanisms, meaning an alarm is only triggered when the temperature exceeds a set limit. This reactive alarm approach fails to provide transport personnel with sufficient time to intervene or take remedial measures, especially in remote areas or during nighttime transport, where it is often too late by the time the alarm sounds. Regarding temperature control strategies, existing active temperature control systems mostly use fixed power or simple PID control, unable to intelligently adjust based on dynamic parameters such as battery level, phase change material status, and remaining transport time. This results in insufficient battery life or loss of temperature control capability due to battery depletion at critical moments. In terms of data reliability, most cold chain data is stored locally or on centralized servers, lacking effective anti-tampering mechanisms and third-party verification methods, making it difficult to meet the requirements of relevant certification bodies for the authenticity and integrity of cold chain data.

[0005] Therefore, there is an urgent need for a new sample cold chain monitoring and temperature control technology that can comprehensively assess multi-dimensional environmental parameters, achieve predictive decision-making, intelligently optimize energy management, and ensure data reliability. Summary of the Invention

[0006] The purpose of this application is to provide a method for intelligent monitoring and temperature control of sample integrity based on multi-parameter prediction, so as to solve the problems mentioned in the background art.

[0007] This application discloses a method for intelligent monitoring and temperature control of sample integrity based on multi-parameter prediction, including the following steps: S1. Multi-parameter acquisition steps: Periodically acquire multi-dimensional environmental parameters inside and outside the transport and storage box. The multi-dimensional environmental parameters include at least temperature, light intensity, vibration, box opening status, remaining battery power, and the state of phase change material. S2. Sample Integrity Assessment Steps: The temperature, light, vibration, and unpacking status data collected in step S1 are weighted and fused to calculate the sample integrity index. The Calculated using the following formula:

[0008] in, For the current transportation time, This is a function for the cumulative temperature deviation. Let be the cumulative light exposure function. This is the cumulative impact force function of vibration. For the out-of-box cumulative exposure function, , , , For the weighting coefficients, satisfying ; S3, Remaining Safe Transportation Time Prediction Step: Based on the calculation obtained in step S2... Based on the changing trend, the remaining battery charge and the state of the phase change material collected in step S1, the remaining safe transportation time is predicted. The Based on the principle of minimum bottleneck, it is determined as follows:

[0009] in, This is the predicted battery life based on the remaining battery charge. The passive heat preservation time is predicted based on the state of the phase change material. Based on the The sample safety time is predicted by the changing trend, and min() is a function that takes the minimum value; S4. Intelligent Temperature Control Execution Steps: The temperature control prediction obtained in step S3... The power of the active temperature control module of the transport storage box is intelligently adjusted based on the comparison result, compared with the estimated transport time. When the When the product of the preset first threshold and the estimated transportation time is reached, the power is reduced to a low-power mode; When the The product of the first threshold and the estimated transport time and When the product of the preset second threshold and the expected transportation time is reached, the standard power mode is maintained. When the When the product of the second threshold and the expected transportation time is reached, the power is increased to a high-power mode and an early warning is triggered, wherein the first threshold > 1, 0 < the second threshold < 1.

[0010] In a preferred embodiment, the temperature cumulative deviation function By measuring the actual temperature Deviation from target temperature range The degree is obtained by time-weighted cumulative integral; the... The calculation formula is:

[0011] in, Let be the temperature penalty function, when Located within the target temperature range hour, ;when When deviating from the target temperature range The value is proportional to the absolute value of the temperature deviation, and the Includes deviation time weighting factor The Used to illustrate the cumulative effect over time.

[0012] In a preferred embodiment, the temperature penalty function The specific calculation method is as follows:

[0013] in, This is the temperature penalty coefficient. For the target temperature, This refers to the deviation time weighting factor.

[0014] In a preferred embodiment, the deviation time weighting factor A linear growth model or an exponential growth model can be used, wherein the linear growth model is: The exponential growth model is ,in The cumulative duration of temperature deviation from the target range. and This is the time accumulation coefficient.

[0015] In a preferred embodiment, the cumulative light exposure function Characterizes the cumulative effect of light exposure on a sample; The vibration cumulative impact function Characterizes the cumulative effects of vibration and impact on the sample; The open-box cumulative exposure function Characterize the cumulative impact of environmental exposure caused by unpacking the sample.

[0016] In a preferred embodiment, the The passive heat preservation time is predicted by measuring the temperature change rate of the phase change material layer, combining the melting point curve and heat capacity of the phase change material, and then inferring the solid phase percentage of the phase change material.

[0017] In a preferred embodiment, the The average power consumption of the active temperature control module is predicted by fitting historical power consumption using the least squares method and then calculated in conjunction with the remaining battery power.

[0018] In a preferred embodiment, the active temperature control module is a semiconductor refrigeration module.

[0019] In a preferred embodiment, the step of intelligently adjusting the power of the active temperature control module includes: in predictive mode, the power... The adjustment formula introduces a time margin factor. :

[0020] in, The temperature difference between the actual temperature inside the transport and storage box and the target temperature. This is the power regulation coefficient. This represents the maximum power of the active temperature control module. The time margin factor, the The value of and the The margin between the estimated transport time and the estimated transport time is positively correlated.

[0021] In a preferred embodiment, a time margin is defined. ,in For the estimated transit time; when At that time, the The value ranges from 0.1 to 0.3, including the endpoint value, to reduce the power; when At that time, the The value ranges from 0.7 to 1.0, including the endpoint value, to increase the power; wherein and This is a preset time margin percentage threshold, and .

[0022] In a preferred embodiment, it further includes: S5, a blockchain traceability step: periodically transferring the aforementioned... The above The key event records are timestamped and hashed, and the hash values ​​are uploaded to the blockchain network to form an immutable cold chain proof.

[0023] In a preferred embodiment, the critical event log includes unpacking events, impact events, and alarm events; the periodic upload period is 300 to 900 seconds.

[0024] In a preferred embodiment, the hash calculation includes: generating a Merkle root hash from locally collected data packets and uploading it to the blockchain network in the form of a transaction; the blockchain network is a consortium blockchain or a private blockchain.

[0025] In a preferred embodiment, when the When the time margin between the estimated transportation time and the actual transportation time is large, the Approaching 0 reduces the power; when the time margin is small, the Approaching 1 increases the power.

[0026] In a preferred embodiment, the power regulation coefficient The numerical range is .

[0027] In a preferred embodiment, when the If the estimated transportation time is less than the expected time, the triggering of the warning includes suggesting that the transportation personnel find alternative refrigerated facilities or change the route to speed up the process, so as to achieve route optimization and coordination.

[0028] In a preferred embodiment, it also includes: S5. Blockchain traceability steps: Periodically transfer the aforementioned... The above The key event records are timestamped and hashed, and the hash values ​​are uploaded to the blockchain network to form an immutable cold chain proof.

[0029] In a preferred embodiment, the critical event log includes unpacking events, impact events, and alarm events; the periodic upload period is 300 to 900 seconds.

[0030] In a preferred embodiment, the hash calculation includes: generating a Merkle root hash from locally collected data packets and uploading it to the blockchain network in the form of a transaction; the blockchain network is a consortium blockchain or a private blockchain.

[0031] In a preferred embodiment, a sample transport cold chain integrity report containing a blockchain verification code is automatically generated after the transport is completed.

[0032] In a preferred embodiment, the transport and storage box employs a hybrid active-passive temperature control system, comprising a sandwich structure of a vacuum insulation panel, a phase change material layer, and the semiconductor refrigeration module; the phase change material layer is embedded with sensors for monitoring the state of the phase change material.

[0033] In a preferred embodiment, the multi-parameter acquisition step performs Kalman filtering to denoise the acquired raw data to ensure the accuracy of the input data; the method uses the main control unit in the transport and storage box as an edge computing unit to run the sample integrity assessment step and the remaining safe transport time prediction step in real time.

[0034] In a preferred embodiment, the acquisition period of the multi-parameter acquisition step is 30 seconds, and the prediction period of the remaining safe transportation time prediction step is 5 minutes.

[0035] In a preferred embodiment, the weighting coefficient , , , .

[0036] In a preferred embodiment, the weighting coefficients are configured based on the sample type: for temperature-sensitive biological samples, The values ​​are relatively large; for shock-sensitive soil samples, The value is relatively large.

[0037] In a preferred embodiment, when the When the threshold is lower than the preset warning threshold, a tiered warning mechanism is activated.

[0038] In a preferred embodiment, the method is applied to the monitoring of the transportation and preservation of environmental testing samples, biological samples, or medical testing samples.

[0039] To address the limitation of existing technologies that rely on a single dimension for sample integrity assessment, this application periodically collects multi-dimensional environmental parameters, including temperature, illumination, vibration, open-box status, remaining battery charge, and the state of phase change materials (step S1). This provides a comprehensive data foundation for subsequent integrated assessment and predictive decision-making, overcoming the limitations of traditional single-dimensional temperature monitoring. Based on this, this application calculates the sample integrity index using a weighted fusion algorithm based on these multi-dimensional parameters. (Step S2) This index is obtained through the temperature cumulative deviation function. Cumulative light exposure function Vibration cumulative impact function and open-box cumulative exposure function The weighted linear combination is achieved, thereby unifying and quantifying the dispersed multidimensional risk factors into a single operable evaluation index. This realizes the technical transformation from one-sided transient parameter judgment to comprehensive cumulative risk assessment, effectively avoiding sample quality misjudgment and potential scrapping caused by non-temperature factors such as light and vibration.

[0040] For temperature cumulative deviation function The calculations in this application, based on actual temperature... Deviation from target temperature range The result is obtained by time-weighted cumulative integration of the degree of deviation, and uses a weighting factor that includes deviation from the time. Temperature penalty function This method causes the contribution of the duration of temperature deviation to cumulative damage to increase non-linearly, truly reflecting the irreversibility and time-cumulative nature of sample quality degradation. Compared to traditional instantaneous temperature threshold-triggered alarms, this cumulative integration method can more scientifically quantify the true degree of sample damage under full-time temperature fluctuations. Furthermore, by using a temperature penalty function... Concretize into The form of this method allows the penalty intensity to depend on both the deviation magnitude and the deviation time, thus enabling accurate modeling of the thermal stress history of the sample.

[0041] To address the technical problem of varying sensitivities of different sample types to various risk factors, this application addresses this issue by using weighting coefficients... , , , Configured with an adjustable parameter range and adaptively configured based on sample type, allowing for adjustments to temperature-sensitive biological samples by increasing... The weight of temperature factors can be increased by adjusting the value of the temperature factor; for shock-sensitive soil samples, this can be achieved by increasing the value of the temperature factor. The values ​​are selected to highlight the importance of vibration factors, thereby achieving the wide applicability and flexibility of the monitoring method for various sample types.

[0042] To address the technical problem of lagging alarm mechanisms in existing technologies, this application proposes a solution based on... The changing trend, remaining battery capacity, and phase change material state predict the remaining safe transportation time. (Step S3) represents a technological leap from post-event alarms to predictive early warning. Calculate battery life using the principle of minimum bottleneck. Phase change material heat preservation time and safe time based on SII trend The minimum value of the three factors determines the design, which ensures the conservatism and reliability of the prediction results, can identify the bottleneck factors that will fail first in the system, provides transportation personnel with sufficient time to intervene, and avoids sample loss caused by the response lag of traditional threshold-triggered alarms.

[0043] Regarding the passive heat preservation time of phase change materials This application predicts the remaining passive insulation time by measuring the temperature change rate of the phase change material layer and combining the melting point curve and heat capacity parameters of the phase change material to infer the solid phase percentage of the phase change material. This technology upgrades the phase change material from a simple passive insulation material to a predictable and quantifiable energy buffer unit, significantly improving the robustness and reliability of the system in the event of active temperature control module failure or battery depletion. Regarding battery life... In this application, the average power consumption of the active temperature control module is predicted by fitting historical power consumption using the least squares method, and then calculated in combination with the remaining battery power. This method comprehensively considers the dynamic changes in the power of the temperature control module and the intermittent operation of the communication module, thereby improving the accuracy of the battery life prediction.

[0044] To address the technical problems of rigid temperature control strategies and the inability to dynamically optimize power based on remaining time in existing technologies, this application solves the problem by... By comparing the estimated transport time with the actual transport time, the power of the active temperature control module is intelligently adjusted based on the comparison result (step S4), thus achieving synergy between predictive maintenance and energy optimization management. When the travel time is much longer than expected, reduce power to a low-power mode to extend the range. If the transport time is less than expected, the power is increased to high power mode and an alert is triggered to enhance sample protection. This strategy will transfer information flow ( The predicted results are transformed into closed-loop feedback control of the physical flow (temperature control power) and the decision flow (path optimization suggestions), which significantly improves the system's endurance while ensuring temperature safety.

[0045] Furthermore, by introducing a time margin factor into the power regulation formula... This makes the power Not only in response to the current temperature difference Also under the control of With a time margin between the estimated transport time and the actual transport duration, this two-dimensional coupled power regulation mechanism achieves a dynamic balance between temperature control accuracy and energy efficiency. When the time margin is large... Approaching zero causes power to decrease, entering energy-saving mode. When the time margin is small... Approaching 1 triggers a power increase to enter protection mode. This mechanism demonstrates the proactive guidance of predictive decision-making for physical control, significantly outperforming traditional passive control strategies based solely on temperature difference feedback. This is achieved by adjusting the power regulation coefficient... When configured within a reasonable numerical range, the temperature control system can be optimized and calibrated based on the insulation performance of the storage box and the response characteristics of the temperature control module, further improving its adaptability.

[0046] Regarding the implementation of path optimization early warning, when When the transport time is less than expected, the system not only increases the temperature control power, but also suggests that transport personnel find alternative refrigeration facilities or change the route to speed up the process. This human-machine collaboration mechanism combines the predictive ability of the algorithm with the flexible decision-making of humans, realizing the synergy between temperature control optimization and route optimization, and minimizing the risk of sample loss.

[0047] To address the technical problems of low data reliability and susceptibility to tampering in existing technologies, this application addresses these issues by periodically... , After timestamping and hashing key event records, the data is uploaded to the blockchain network (step S5), forming an immutable cold chain proof, fundamentally solving the credibility problem of traditional centralized data storage. The key events include unpacking events, impact events, and alarm events; the immutable records of these events provide highly credible evidence for sample quality assessment and legal arbitration. By generating Merkle root hashes from locally collected data packets and uploading them to a consortium or private blockchain in the form of transactions, both the decentralization and tamper-proof characteristics of the data are guaranteed, while also considering transaction speed and cost control. After transportation, a sample transportation cold chain integrity report containing a blockchain verification code is automatically generated, allowing any third party to query and verify the authenticity of the data through the verification code, meeting relevant regulatory and certification requirements, and helping to reduce audit costs and data dispute risks.

[0048] This application realizes a hybrid active and passive temperature control system through a sandwich structure of vacuum insulation plate, phase change material layer and semiconductor refrigeration module used in the transport and storage box. The phase change material layer is embedded with a sensor for monitoring the state of the phase change material. This structural design organically combines the high reliability of passive insulation with the flexibility of active temperature control. When active temperature control fails, the phase change material can provide reliable backup power, which greatly improves the robustness of the system.

[0049] By performing Kalman filtering to denoise the acquired raw data and using the main control unit as an edge computing unit to run sample integrity assessment and remaining safe transport time prediction algorithms in real time, the accuracy of the input data and the real-time performance of the algorithm execution were ensured. A good balance was achieved between data real-time performance and system power consumption by setting the multi-parameter acquisition period to 30 seconds and the remaining safe transport time prediction period to 5 minutes.

[0050] In summary, the various technical features of this application do not exist in isolation, but rather form a complete technical chain through multi-dimensional data acquisition in step S1, SII fusion evaluation in step S2, RSTT bottleneck prediction in step S3, predictive power regulation in step S4, and blockchain traceability in step S5. Each step is tightly coupled through data flow and control flow. In particular, the RSTT prediction result in step S3 provides feedforward control for the power regulation in step S4, achieving closed-loop optimization of the physical flow through the information flow and generating a significant synergistic effect. This technical solution upgrades cold chain monitoring from the traditional "record-response" model to an "evaluation-prediction-optimization" model, achieving outstanding technical results in terms of sample integrity assessment accuracy, temperature control system endurance, early warning timeliness, and data reliability. It effectively solves the technical problems of existing technologies, such as single evaluation dimensions, lagging alarm mechanisms, rigid temperature control strategies, and easy data tampering.

[0051] The specification of this application contains numerous technical features distributed across various technical solutions. Listing all possible combinations of these technical features (i.e., technical solutions) would make the specification excessively lengthy. To avoid this problem, the various technical features disclosed in the above-described invention, the various technical features disclosed in the following embodiments and examples, and the various technical features disclosed in the accompanying drawings can be freely combined to form various new technical solutions (all of which are considered to have been described in this specification), unless such a combination of technical features is technically infeasible. For example, one example discloses feature A+B+C, and another example discloses feature A+B+D+E. Features C and D are equivalent technical means that serve the same function, and technically only one needs to be used; they cannot be used simultaneously. Feature E can technically be combined with feature C. Therefore, the solution A+B+C+D should not be considered as described because it is technically infeasible, while the solution A+B+C+E should be considered as described. Attached Figure Description

[0052] Figure 1 This is an overall method flowchart of the intelligent monitoring and temperature control method for sample integrity based on multi-parameter prediction according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of an intelligent monitoring transport and storage box system according to an embodiment of this application.

[0053] Explanation of reference numerals in the attached figures: 1: Vacuum insulation panel; 2: Phase change material layer; 3: Semiconductor refrigeration module; 4: Embedded phase change material state sensor; 5: Sample storage area; 6: Internal temperature sensor; 7: Light sensor; 8: External temperature sensor; 9: Triaxial vibration sensor; 10: Door magnetic sensor; 11: Main control unit; 12: Battery management system; 13: Wireless communication module; 14: Third-party sensor. Detailed Implementation

[0054] In the following description, many technical details are presented to help the reader better understand this application. However, those skilled in the art will understand that the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments.

[0055] Explanation of some concepts: Sample integrity refers to the degree to which a sample retains its original physical, chemical, and biological characteristics without significant alteration during collection, transportation, and storage. Sample integrity is a prerequisite for ensuring the accuracy and reliability of test results.

[0056] Sample Integrity Index (SII): This application proposes a quantitative index that comprehensively quantifies sample quality risk into a value of 0-100 by weighted and fused four dimensions: cumulative temperature deviation, cumulative light exposure, cumulative vibration impact, and cumulative exposure after unpacking. The initial value of 100 indicates complete integrity, and a decrease in value indicates a decline in integrity.

[0057] Remaining Safe Transportation Time (RSTT): This refers to the remaining time under current transportation conditions during which the system can continue to maintain the sample integrity and safety. RSTT is determined by the minimum value among battery life, phase change material insulation time, and SII safety time, based on the principle of minimum bottleneck.

[0058] Temperature Cumulative Deviation Function (T_D(t)): This function, obtained by time-weighted cumulative integration of the degree to which the actual temperature deviates from the target temperature range, is used to quantify the cumulative damage caused to the sample by temperature deviation. This function reflects the irreversibility of sample quality damage and the cumulative effect over time.

[0059] The Minimum Bottleneck Principle (RSTT) is a predictive strategy that determines the overall safe time of a system by identifying the bottleneck among multiple limiting factors that is the first to reach a critical state. In this application, RSTT depends on the factor that is the first to be depleted or fail among the battery, phase change material, and SII.

[0060] Active-Passive Hybrid Temperature Control System: Combining an active temperature control module (such as...) Figure 2 The temperature control system consists of an active module (3) for semiconductor cooling and a passive insulation structure (such as a vacuum insulation panel or phase change material). The active module provides precise temperature control, while the passive structure provides energy buffering and backup insulation.

[0061] Phase change materials (PCMs) are a class of materials that can undergo a phase transition (solid-liquid phase transition) within a specific temperature range and absorb or release a large amount of latent heat. In cold chain transportation, PCMs utilize their phase transition process to maintain a stable temperature inside the container.

[0062] Time margin factor (M_RSTT): A power regulation parameter proposed in this application, the value of which reflects the time margin between RSTT and the expected transportation time. When the time margin is large, M_RSTT is close to 0, and the power is reduced to enter the power saving mode; when the time margin is small, M_RSTT is close to 1, and the power is increased to enter the protection mode.

[0063] Predictive maintenance is a maintenance strategy that intervenes proactively based on predictive information, unlike traditional reactive or periodic maintenance. This application achieves predictive maintenance through RSTT prediction, taking action before problems occur.

[0064] Blockchain Traceability: Utilizing the decentralized and immutable characteristics of blockchain technology, key data in the cold chain transportation process is hashed and stored on the chain to form a verifiable and tamper-proof digital evidence chain.

[0065] Merkle Tree: A binary tree data structure where leaf nodes store the hash value of the data, and non-leaf nodes store the hash values ​​of their child nodes. The root hash value can be used to efficiently verify the integrity of the entire dataset.

[0066] Edge computing is a distributed computing architecture that places data processing and algorithm computation on edge devices closer to the data source, rather than relying entirely on the cloud. In this application, the master control unit inside the transport and storage box serves as an edge computing unit, running the SII and RSTT algorithms in real time.

[0067] Kalman Filter: An optimal recursive data processing algorithm that provides the optimal estimate of the state of a dynamic system in the presence of noise through prediction and measurement updates of the system state. This application is used for denoising sensor data.

[0068] The following is a brief summary of some of the innovative aspects of this application: In summary, the technical concept of this application lies in breaking through the linear control paradigm of "single parameter evaluation - threshold-triggered alarm - passive power adjustment" in traditional cold chain monitoring, and constructing a predictive collaborative control architecture that couples a nonlinear fusion evaluation system based on the time-cumulative integral of multi-dimensional risk factors with a minimum bottleneck identification mechanism based on multi-source time-series prediction. Specifically, this architecture does not simply control temperature... ,illumination ,vibration and unboxing status Instead of linearly superimposing or independently monitoring discrete parameters, a deviation-time weighting factor is introduced. By performing a weighted cumulative integral operation over the time domain on the historical deviation trajectories of each risk factor, the transient parameter deviations are transformed into a damage degree function with irreversible cumulative properties. , , and And through sample type adaptive normalized weight coefficients , , , Project these heterogeneous damage functions onto a unified sample integrity index. The evaluation space thus achieves a dimensionality reduction mapping from "multi-parameter discrete monitoring" to "single-index comprehensive quantification." This multi-dimensional fusion evaluation mechanism based on a time integral penalty model not only provides more monitoring dimensions, but also achieves its technical effect through the cumulative integral operator. The introduction of this method enables the evaluation results to possess the time memory characteristic of the irreversible degradation process of sample quality, thereby overcoming the inherent defect that the instantaneous threshold judgment method cannot capture cumulative damage.

[0069] Furthermore, the predictive maintenance mechanism of this application does not... Instead of using the assessment result as an endpoint, it uses it as the remaining safe transport time. One of the key input variables of the prediction algorithm, and through the principle of minimum bottleneck Three time predictions from heterogeneous physical domains (battery domain) Thermodynamic domain and sample integrity domain This method involves cross-domain coupling to identify the weakest link that determines the overall safety margin of the system. The non-obviousness of this multi-source time-series prediction fusion method based on the minimum bottleneck principle lies in the fact that it not only requires separate knowledge of electrochemical power consumption extrapolation predictions (using the least squares method to extrapolate historical power consumption) but also... Perform fitting to obtain ), Inverse prediction of the solid percentage of phase change materials (based on the temperature change rate of the phase change material) Combining the melting point curve to infer the solid phase ratio and extrapolating the sample integrity trend to predict (through...) The three independent prediction sub-problems (predicting the time it takes to fall to the threshold) require establishing a mapping relationship between these three heterogeneous prediction results and a unified time margin concept at the architectural level, and further integrating this time margin information. The real-time power of the semiconductor cooling module is modulated by the feedback control signal. This forms a closed-loop feedback path of "predictive information flow → physical control flow".

[0070] More importantly, this application utilizes a power regulation formula... Introducing a time margin factor This achieves multiplicative coupling between the traditional temperature difference-driven control dimension and the predictive time margin-driven control dimension, enabling the power regulation strategy to no longer be limited to the current temperature deviation. It is not a transient response, but is simultaneously constrained by future time. Feedforward modulation. The non-obviousness of this two-dimensional multiplicative coupled power regulation mechanism lies in the fact that its synergistic effect is not a simple superposition of two independent control strategies, but rather a deep fusion of "current state feedback" and "future constraint feedforward" in the power decision space through the multiplicative operator, thereby achieving a balance between temperature difference and power regulation. Larger but with more time margin When fully charged, it can automatically enter power-saving mode, which is beneficial in temperature differences. Still small but with time leeway When under pressure, the power can be increased in advance to enter the protection mode. This predictive power optimization strategy is fundamentally different from the traditional PID control, which can only passively correct the temperature deviation that has already occurred. It precisely reflects the technical essence of the control paradigm shift from "post-event response" to "pre-intervention" in this application.

[0071] Furthermore, this application elevates the phase change material layer from a traditional passive insulation material to an energy buffer unit with state-aware and remaining capacity-predictable characteristics, and presents its state prediction results. Included The minimum bottleneck identification mechanism ensures that the active-passive hybrid temperature control system is no longer a simple physical superposition of an active cooling module (semiconductor cooling module) and a passive insulation material / phase change material, but rather... The introduction of predictive information enables reverse information coupling between the passive insulation subsystem and the active temperature control decision-making, forming a bidirectional collaborative mechanism between the active and passive subsystems based on predictive information flow. This collaborative mechanism is integrated with the periodically uploaded timestamp hash values ​​in the blockchain traceability process. In conjunction with the Merkle root hash structure, not only is predictive optimization of temperature control achieved at the physical level, but also, at the information level, a decentralized multi-party witnessing mechanism through consortium blockchains or private blockchains provides an immutable data trust anchor for the entire closed-loop process of "collection → evaluation → prediction → control". Thus, in terms of technical implementation path, a two-layer guarantee architecture of "predictive collaborative temperature control in the physical domain" and "tamper-proof traceability proof in the information domain" is formed. The non-obviousness of the overall concept of this systematic technical solution across the physical and information domains far exceeds the scope that can be expected from the simple combination of individual technical features.

[0072] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0073] Analysis of sample failure cases shows that damage to sample integrity is often the result of the cumulative effect of multiple factors, rather than caused by a single factor. Traditional monitoring systems focus only on temperature parameters, which stems from an insufficient understanding of the mechanisms of sample quality degradation. While temperature is a major factor affecting sample stability, mechanical vibration during actual transportation can lead to changes in the physical structure of some samples (e.g., breakage of aggregates in soil samples, sedimentation and stratification of suspended matter in water samples), light exposure can trigger chemical degradation of photosensitive components (e.g., photolysis of certain environmental pollutants, inactivation of photosensitive proteins in biological samples), and frequent opening of the container can cause instantaneous temperature shocks and the intrusion of external contaminants. More importantly, the effects of these non-temperature factors are irreversible and cumulative—even if subsequent environmental conditions return to normal, the damage already caused cannot be reversed. Experimental verification shows that, in some cases, the damage to samples caused by short-term severe vibration or light exposure can even exceed that caused by slight temperature deviations. However, existing technologies only monitor temperature and are completely unaware of these "invisible killers," resulting in a large number of samples actually failing even when the temperature records show "normal." The inventors estimate that this false positive rate is as high as 25%.

[0074] Further research reveals that the lag problem of existing alarm mechanisms stems from a lack of predictive capability. Traditional threshold-triggered alarms are a classic example of "hindsight bias"—by the time the temperature has deviated from the trigger threshold, the sample has often already been exposed to the adverse environment for some time. Through systematic analysis of transportation routes and time margins, the inventors discovered that this lag is particularly fatal in two scenarios: firstly, during transportation in remote mountainous areas or at night, even if an alarm is received, it is difficult for transport personnel to find backup refrigeration facilities or change routes in a short period; secondly, when the battery is nearly depleted or the phase change material is about to completely melt, the alarm often comes too late, and the system has lost its room for adjustment. This application proposes that instead of waiting for a problem to occur before issuing an alarm, it is better to establish a predictive model to calculate the "remaining safe time" in advance, providing early warnings before the problem actually occurs, leaving sufficient time for human intervention. This shift in thinking from "passive response" to "proactive prediction" is a key technical insight of this application.

[0075] In an in-depth analysis of existing active temperature control systems, the inventors discovered that their energy management strategies are generally rigid and inefficient. Most existing systems' temperature control algorithms are based on simple PID control, adjusting cooling power solely based on the current temperature deviation, completely neglecting "time constraints" and "resource constraints." Analysis of experimental data shows that the temperature control system actually faces a multi-objective optimization problem: ensuring temperature control accuracy (to ensure sample safety), maximizing battery life (to ensure completion of the entire transportation process), and maintaining a power margin to handle unforeseen circumstances. However, these three objectives often conflict—increasing temperature control accuracy requires more power but shortens battery life; reducing power can extend battery life but may sacrifice temperature control accuracy. Therefore, the key to this multi-objective optimization problem lies in introducing the concept of "time margin": if the remaining battery life is predicted to be much greater than the required transportation time, the system can appropriately reduce power to enter a power-saving mode; conversely, if insufficient battery life is predicted, the system should increase power for rapid cooling and provide timely warnings. This closed-loop strategy, which feeds back "predictive information" to "physical control," achieves a qualitative leap from simple feedback control to predictive maintenance.

[0076] Another problem exists with existing applications of phase change materials (PCMs). In hybrid active-passive temperature control systems, PCMs are typically treated as passive insulation materials, their role limited to providing temporary backup insulation in the event of active temperature control failure. However, research on the phase change process of PCMs shows that their cooling capacity is not a fixed value but dynamically changes with the phase change process—the cooling capacity is strongest when the PCM is completely solid, gradually decreases as it melts, and reaches zero after complete liquefaction. More importantly, the phase change state of PCMs can be monitored and quantified in real time using temperature sensors. This application proposes that if the remaining cooling time of the PCM can be predicted in real time and this information is incorporated into the overall system safety time prediction, the PCM can be upgraded from a "passive backup" to an "active and predictable energy buffer unit," significantly improving the robustness and reliability of the system.

[0077] Furthermore, regarding the issue of cold chain data trustworthiness, in traditional centralized data storage models, whether local or cloud storage, the integrity and authenticity of data are difficult to verify effectively by a third party. The root of this problem lies in the lack of "decentralized" and "tamper-proof" mechanisms. The rise of blockchain technology offers a possibility for solving this problem, but how to organically combine blockchain with cold chain monitoring systems to control communication and transaction costs while ensuring data trustworthiness requires careful system design. This application proposes a hybrid scheme based on Merkle trees and consortium blockchains, which ensures data tamper-proofing while controlling implementation costs.

[0078] Based on the aforementioned in-depth research and systematic technical insights, this application proposes a technical solution: comprehensive evaluation is achieved through a multi-parameter fusion Sample Integrity Index (SII) algorithm; predictive decision-making is achieved through a Remaining Safe Transport Time (RSTT) prediction algorithm; closed-loop control of prediction-decision-optimization is achieved through intelligent power management; and data tamper-proofing is achieved through blockchain technology. These four core innovations support and synergize with each other, upgrading the traditional "record-response" model to an "evaluation-prediction-optimization" model.

[0079] The present application will now be described in detail with reference to specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0080] This application provides a multi-parameter prediction-based intelligent monitoring and temperature control method for sample integrity, applied to an intelligent monitoring transport and preservation box. This transport and preservation box possesses a hybrid active and passive temperature control capability for monitoring environmental, biological, or medical testing samples during transport and preservation. The core of this application lies in upgrading traditional cold chain monitoring from a "record / response" model to an "evaluation / prediction / optimization" model through the synergistic cooperation of multi-parameter fusion evaluation, predictive decision-making, and intelligent power management, thereby significantly improving sample validity, transport safety, and regulatory compliance.

[0081] In this embodiment, the transport and storage box adopts a hybrid active-passive temperature control system, such as... Figure 2 As shown, the system includes a vacuum insulation plate 1, a phase change material layer 2, and a semiconductor cooling module 3, arranged in a sandwich structure. The space formed inside the enclosure constitutes a sample storage area 5 for placing samples to be transported. An embedded phase change material state sensor 4 is embedded in the phase change material layer 2 to monitor the state of the phase change material, which can measure the temperature change rate of the phase change material layer 2 in real time. In addition, the transport and storage box is equipped with an internal temperature sensor 6, a light sensor 7, a triaxial vibration sensor 9, and a door magnetic sensor 10, and an external temperature sensor 8 is installed outside the box. In terms of control and communication, the enclosure integrates a main control unit 11, a battery management system 12, and a wireless communication module 13. To expand the monitoring capabilities, the system can also connect or integrate third-party sensors 14 and connect them to the main control unit 11 for data interaction. The main control unit 11 acts as an edge computing unit, running the sample integrity assessment algorithm and the remaining safe transport time prediction algorithm of this application in real time. For example, the main control unit 11 can be an STM32F4 series microcontroller and run the FreeRTOS real-time operating system to ensure the real-time performance of multi-task concurrent processing.

[0082] The method described in this application includes the following key steps, such as Figure 1 As shown: Step 100: Multi-parameter acquisition steps In step 100, a multi-sensor array deployed inside and outside the transport and storage box periodically collects multi-dimensional environmental parameters. These multi-dimensional environmental parameters include at least temperature, light intensity, vibration, box opening status, remaining battery power, and the state of the phase change material.

[0083] Specifically, temperature parameters are acquired through multi-point temperature sensors, which are positioned at different locations inside the chamber (e.g., top, bottom, and center), inside the phase change material layer 2, and in the external environment. Multi-point temperature acquisition allows for a more accurate reflection of the spatial distribution and temporal evolution of temperature within the chamber. Preferably, high-precision digital temperature sensors are used.

[0084] Furthermore, illumination parameters are collected using a light intensity sensor, which monitors whether the enclosure is exposed to light and the intensity of the light. For certain photosensitive samples (such as volatile organic compound samples), light exposure may cause chemical changes in the sample components; therefore, light monitoring is crucial for assessing the integrity of such samples.

[0085] Specifically, vibration parameters are acquired using a triaxial accelerometer, which can monitor vibration and impact events during transportation in real time. Vibration monitoring is particularly important for certain mechanically sensitive samples (such as soil samples and suspended particulate matter samples) because severe vibration or impact can cause irreversible changes in the physical structure of the sample.

[0086] More specifically, the open status is monitored by a door magnetic sensor 10 or a proximity switch. Each time the enclosure is opened, the sensor records the opening event and its duration. Opening the enclosure exposes the sample to the external environment for a short period, which may lead to sudden temperature changes or contamination risks. Therefore, opening monitoring is an important factor in assessing sample integrity.

[0087] It should be noted that the remaining battery power is monitored in real time by the battery management system 12, which provides real-time information on the battery's voltage, current, and remaining capacity. The remaining battery power is fundamental data for predicting battery range.

[0088] Furthermore, the state of the phase change material is monitored by a temperature sensor embedded inside the phase change material layer 2. By measuring the temperature change rate of the phase change material layer 2, and combining it with the melting point curve and heat capacity parameters of the phase change material, the percentage of solid phase in the phase change material can be deduced, thereby predicting the remaining cold storage capacity of the phase change material.

[0089] Step 110: Data Preprocessing Furthermore, step 100 includes a sub-step of preprocessing the acquired raw data. Specifically, the main control unit 11 performs Kalman filtering on the acquired raw data to denoise it, ensuring the accuracy of the input data. Kalman filtering is an optimal recursive data processing algorithm that can effectively remove random noise from sensor signals while preserving the true trend of signal changes. Data preprocessing can avoid misjudgments caused by sensor noise and improve the stability and reliability of subsequent algorithms.

[0090] Step 120: Time synchronization and periodic control of data acquisition In this embodiment, the acquisition cycle of the multi-parameter acquisition step is preferably a short cycle. This cycle value is an optimized value chosen to balance data granularity and system power consumption. It should be noted that the main control unit 11 uses a timer interrupt mechanism to ensure that the data acquisition of each sensor remains synchronized in time, preventing calculation errors caused by asynchronous data. After the acquired data is timestamped and uniformly processed, a feature vector is formed for use in subsequent steps.

[0091] Step 200: Sample Integrity Assessment Steps In step 200, the temperature, light, vibration, and unpacking status data collected in step 100 are weighted and fused to calculate the sample integrity index. Sample integrity index This is one of the core innovations of this application. It quantifies the impact of multidimensional environmental parameters on sample quality into a comprehensive index, achieving a technological leap from single-temperature judgment to multi-dimensional comprehensive evaluation. Furthermore, the cumulative temperature deviation, light intensity, cumulative exposure after opening the chamber, and the opening state are each calculated by their respective modules. , , and Then, the weight coefficients are set through the weight configuration module. , , , The sample integrity index is finally output by performing a weighted summation. .

[0092] Step 210: The calculation formula and the meaning of each parameter The Calculated using the following formula:

[0093] in, This is the sample integrity index, with an initial value of 100, indicating that the sample is in a completely intact state; as various risk factors accumulate during transportation, The value gradually decreases. Specifically, For the current transportation time, This is a cumulative temperature deviation function, characterizing the cumulative impact of temperature deviating from the target range; This is a cumulative light exposure function, characterizing the degree of cumulative effect of light exposure on the sample; This is the cumulative impact function, which characterizes the cumulative effect of vibration and impact on the sample. This is the cumulative exposure function after unpacking, characterizing the cumulative impact of environmental exposure caused by the unpacking operation. Furthermore, , , , For the weighting coefficients, satisfying The purpose of these four weighting coefficients is to assign different levels of importance to each risk factor based on the characteristics of different sample types.

[0094] Step 220: Configuration of weighting coefficients In this embodiment, the weighting coefficient , , , It can be adaptively configured based on sample type. For example, the weighting coefficients... The range of values ​​is Weighting coefficient The range of values ​​is Weighting coefficient The range of values ​​is Weighting coefficient The range of values ​​is More specifically, for temperature-sensitive biological samples (such as vaccines and blood products). The value should be relatively large to reflect the importance of temperature control; for shock-sensitive soil samples or suspended particulate matter samples, The value should be relatively large to reflect the importance of vibration control. This flexible configuration mechanism of the weighting coefficients allows the method of this application to adapt to the monitoring needs of different types of samples.

[0095] Step 230: Temperature Cumulative Deviation Function Calculation Temperature cumulative deviation function It is a formula One of the most important components. This function, through the actual temperature... Deviation from target temperature range The degree is obtained by time-weighted cumulative integration. The calculation formula is:

[0096] in, The temperature penalty function is... Let be the integral variable, representing the start of transportation ( ) to the current time ( (Time). Specifically, when the actual temperature Located within the target temperature range At that time, This indicates that the temperature conditions are ideal and do not negatively affect the integrity of the sample; when When deviating from the target temperature range, The value is proportional to the absolute value of the temperature deviation.

[0097] Furthermore, the temperature penalty function The specific calculation method is as follows:

[0098] in, This is the temperature penalty coefficient, the magnitude of which reflects the severity of the impact of temperature deviation on sample quality. This coefficient can be calibrated based on the temperature sensitivity of the sample. The target temperature is usually taken as the midpoint of the target temperature range; The deviation time weighting factor is designed to increase the penalty weight as the deviation time increases, thereby reflecting the cumulative and irreversible effects of sample quality damage over time.

[0099] It should be noted that the formula and formula The design embodies a key technical concept of this application: sample quality degradation depends not only on the degree of temperature deviation but also on the duration of the deviation. This is achieved by introducing a time-weighted cumulative integral and a deviation time weighting factor. This technology can accurately simulate the cumulative damage to samples caused by temperature fluctuations during transportation. For example, short-term, slight temperature deviations can significantly impact... The impact is relatively small, while prolonged severe temperature deviations can lead to... The rapid decline is highly consistent with the actual degradation pattern of sample quality.

[0100] Step 240: Physical meaning and calculation principle of other cumulative functions Similar to the temperature cumulative deviation function Cumulative light exposure function Vibration cumulative impact function and open-box cumulative exposure function They are also calculated using the time accumulation method.

[0101] Specifically, the cumulative light exposure function This system characterizes the cumulative impact of light exposure on a sample. When the light intensity exceeds a preset threshold, the system records the light exposure event and calculates the cumulative exposure based on the light intensity and exposure time.

[0102] Furthermore, the cumulative impact function of vibration It characterizes the cumulative impact of vibration and shock on a sample. When the vibration acceleration detected by the accelerometer exceeds a preset threshold, the system records the impact event and calculates the cumulative impact intensity based on the impact strength and duration.

[0103] More specifically, the open-box cumulative exposure function This characterizes the cumulative impact of environmental exposure caused by opening the container. Each time an opening event occurs, the system records the opening duration and calculates the total exposure time.

[0104] By weighting and integrating four key risk factors—temperature, light, vibration, and unpacking—this application achieves a comprehensive and multi-dimensional assessment of sample integrity, overcoming the limitations of existing technologies that only focus on a single temperature parameter.

[0105] Step 250: SII Tiered Early Warning Mechanism Furthermore, when When the sample level falls below a preset warning threshold, the system activates a tiered warning mechanism. Optionally, multiple warning levels can be set, such as green (sample completely safe), yellow (mild risk), orange (moderate risk), and red (high risk). Different warning levels correspond to different handling recommendations. For example, a yellow warning alerts transport personnel to pay attention; an orange warning suggests accelerating transport or finding alternative refrigerated facilities; and a red warning requires immediate cessation of transport and sample quality testing. This tiered warning mechanism provides transport personnel with clear decision-making guidance, effectively reducing the risk of sample loss.

[0106] Step 300: Remaining Safe Transportation Time Prediction Step In step 300, based on the calculation obtained in step 200 Based on the changing trends, the remaining battery charge and phase change material state collected in step 100, the remaining safe transportation time is predicted. Remaining safe transport time Another core innovation of this application is that it provides time margin information for transportation decisions through predictive algorithms, enabling the system to intervene in advance rather than alerting afterward.

[0107] Step 310: The computational principle and the principle of minimum bottleneck The Based on the principle of minimum bottleneck, it is determined as follows:

[0108] in, This is the predicted battery life based on the remaining battery charge. The passive heat preservation time is based on the prediction of the phase change material state. For based on The changing trend predicts the sample's safe time. It should be noted that the formula... This design embodies the principle of minimum bottleneck, meaning the remaining safe transport time is limited to the shortest of the three time parameters. This design ensures the conservatism and safety of the system's predictions, preventing sample loss due to the failure of any single factor.

[0109] Step 320: Prediction methods Battery life The prediction is obtained by fitting historical power consumption data using the least squares method to predict the average power consumption of the active temperature control module, and then combining this with the remaining battery power. Specifically, the system continuously records the real-time power consumption data of the active temperature control module (semiconductor cooling module 3) under different operating modes, and uses the least squares method to perform linear or nonlinear fitting on this historical data to obtain the extrapolated average power consumption. Then, dividing the remaining battery power by this extrapolated average power consumption yields the predicted battery life. It should be noted that because the power of the semiconductor cooling module 3 dynamically adjusts according to ambient temperature and temperature control requirements, and the communication and positioning modules operate intermittently, the prediction of average power consumption needs to comprehensively consider the influence of these factors. By fitting historical power consumption curves using the least squares method, future power consumption trends can be extrapolated more accurately, thereby improving... The accuracy of the prediction.

[0110] Step 330: Prediction methods Passive heat preservation time The predictions demonstrate the innovation of this application's active-passive hybrid temperature control system. Specifically, The passive insulation time is predicted by measuring the temperature change rate of the phase change material layer 2, combining the melting point curve and heat capacity parameters of the phase change material, and then inferring the percentage of solid phase in the phase change material.

[0111] Furthermore, during the phase transition process, the temperature of the phase change material remains relatively stable (near its melting point), while in a completely solid or completely liquid state, the temperature changes with the inflow or outflow of heat. This is achieved by real-time monitoring of the temperature change rate of the phase change material layer 2. This allows us to determine the current phase state of a phase change material. When When the value is small, it indicates that the phase change material is undergoing a phase change process and still has cold storage capacity; when... A large value indicates that the phase change material (PCM) has nearly completely melted, and its cooling capacity is about to be exhausted. More specifically, by combining the PCM's enthalpy-temperature curve (melting point curve) and specific heat capacity parameter, the percentage of solid phase in the PCM can be quantitatively estimated. For example, if a certain proportion of the PCM has melted, the remaining solid phase can still provide a corresponding proportion of cooling energy. Based on the current heat load (determined by the temperature difference between the ambient temperature and the target temperature inside the chamber), the time required for the PCM to completely melt can be predicted. .

[0112] It should be noted that this phase change material (PCM) state prediction method upgrades PCM from a simple passive insulation material to a predictable and quantifiable energy buffer unit, significantly improving system reliability. In extreme cases such as failure of the active temperature control module (semiconductor cooling module 3) or depletion of battery power, the PCM can still provide passive insulation for a certain period of time, buying valuable time for transport personnel to find alternative solutions.

[0113] Step 340: Prediction methods Sample safety time Based on The system predicts the changing trends. Specifically, it records... Based on recent change curves, and using trend extrapolation algorithms (such as linear regression or exponential fitting) to predict... The time required for the temperature to drop to the warning threshold. This time is... This indicates the timeframe at which the sample can still be safely transported, given the current rate of risk accumulation. For example, if in the near future... A rapid descent (indicating poor current transportation conditions) then Shorter; conversely, if If it remains stable or declines slowly, then It is relatively long.

[0114] Step 350: Predicting the Cycle In this embodiment, the prediction period for the remaining safe transportation time prediction step is preferably a longer period than the data collection period. This period value strikes a balance between prediction timeliness and computational resource consumption. It is updated periodically. The prediction results can reflect changes in transportation status in a timely manner, while avoiding excessive burden on the main control unit 11 due to overly frequent prediction calculations.

[0115] Step 400: Intelligent Temperature Control Execution Steps In step 400, the prediction obtained in step 300 is... The system compares the estimated transport time with the actual transport time and intelligently adjusts the power of the active temperature control module in the transport storage box based on the comparison result. Compared with the estimated transit time When comparing, Greater than Enter low power mode ,when equal Maintain standard power mode ,when Less than Enter high power mode This will trigger path optimization alerts and warnings.

[0116] This step is crucial for achieving predictive maintenance and intelligent power management in this application, reflecting the information flow ( (Prediction) Closed-loop feedback control of physical flow (temperature control power).

[0117] Step 410: Detailed Implementation of the Active Temperature Control Module In this embodiment, the active temperature control module employs a thermoelectric cooler (TEC) module 3. The thermoelectric cooler (TEC) module 3 has advantages such as small size, light weight, no moving mechanical parts, and fast response speed, making it particularly suitable for portable transport and storage cases. Specifically, the thermoelectric cooler (TEC) module 3 achieves cooling through the Peltier effect. When a direct current passes through the thermoelectric cooler (TEC) module 3, heat is absorbed on one side (cold end) and released on the other side (hot end). The cooling intensity can be flexibly controlled by adjusting the input power.

[0118] Step 420: Based on Intelligent power regulation strategy according to Based on the comparison with the estimated transportation time, the system adopts the following intelligent adjustment strategy: when When the time exceeds the expected transport time, it indicates that the system has sufficient time margin. In this case, the power of the semiconductor cooling module 3 is reduced to a low-power mode. Furthermore, in low-power mode, the semiconductor cooling module 3 operates with lower input power. Although the cooling capacity is reduced, it can significantly extend battery life. Due to the sufficient time margin, even when operating at low power, it can ensure that the task is completed within the expected transport time and that the sample integrity is maintained.

[0119] when When the estimated transport time is approached, it indicates that the system is in a relatively balanced state, at which point the standard power mode of the semiconductor cooling module 3 is maintained. Specifically, the standard power mode is the normal operating power pre-calibrated based on the insulation performance of the enclosure and the ambient temperature conditions, which can maintain a stable internal temperature under most circumstances.

[0120] when If the transport time is less than expected, it indicates that the system faces the risk of insufficient battery life. In this case, the power of the semiconductor cooling module 3 is increased to high-power mode, triggering an early warning. More specifically, in high-power mode, the semiconductor cooling module 3 operates at its maximum or near-maximum input power to achieve rapid cooling and slow down the rate of sample integrity degradation as much as possible. Simultaneously, the system triggers an early warning, suggesting that transport personnel find alternative refrigeration facilities or change the route to expedite transport, thus achieving route optimization. It should be noted that this early warning mechanism provides transport personnel with an opportunity for proactive intervention, rather than passively responding only after the battery is depleted or the sample fails.

[0121] Step 430: Quantization algorithm for power regulation Furthermore, in predictive mode, the semiconductor cooling module 3 power The adjustment is based not only on the current temperature difference, but also incorporates a time margin factor. This enables more intelligent power management. The power regulation formula is:

[0122] in, To adjust the power in real time, This is the maximum power of semiconductor cooling module 3. This refers to the temperature difference between the actual temperature inside the transport and storage box and the target temperature. This is the power regulation coefficient. This is the time margin factor.

[0123] Specifically, the formula In the middle, item This is a traditional power regulation term based on temperature difference, whose function is to increase the power as the temperature difference increases, achieving negative feedback control. When the temperature difference... When the temperature difference is small, this term is close to 0, and the power is low; when the temperature difference is small... When the value is large, this term is close to 1, and the power is close to the maximum value.

[0124] Time margin factor This is the innovative aspect of this application; its value is related to... There is a positive correlation between the margin and the expected transit time. More specifically, when When there is a large time margin between the estimated transit time and the actual transit time (i.e.) (far longer than the estimated transit time) Approaching 0 makes the power Reduced, thus saving power and extending battery life; when the time margin is small (i.e. (close to or less than the estimated transit time) Approaching 1 makes the power This improves the cooling capacity.

[0125] It should be noted that by introducing The power regulation factor is no longer merely a passive response to the current temperature state, but rather comprehensively considers future time and resource constraints, achieving true predictive maintenance. This algorithm can significantly increase system endurance while ensuring temperature safety, and proactively intervene when temperature control capabilities are about to become insufficient.

[0126] Step 440: Power Regulation Coefficient Choice In this embodiment, the power regulation coefficient You can choose from a certain range. The larger the value, the greater the temperature difference. The greater the impact on power, the more sensitive the system is to temperature deviations. Specifically, The value needs to be calibrated based on the insulation performance of the enclosure and the response speed of the semiconductor cooling module 3. For enclosures with good insulation performance, a smaller value can be selected. A higher value makes the system response smoother; for enclosures with poor insulation, a larger value should be selected. This value enables the system to respond quickly to temperature changes.

[0127] Step 500: Blockchain Traceability Steps As an important component of this application, this embodiment also includes a blockchain traceability step to address the issue of cold chain data credibility. In step 500, KT(t), KEK, and key event records are periodically timestamped and hashed, and the hash values ​​are uploaded to the blockchain network to form an immutable cold chain proof. Data collected by data collectors / sensors is aggregated by the data integration module, and a Merkle root hash is generated by the hash calculation module. Then, a transaction is constructed by the transaction construction module and periodically uploaded to the blockchain network (consortium blockchain / private blockchain). After transportation is completed, the system automatically generates a traceability report containing a blockchain verification code, which can be used by any third party for verification.

[0128] Step 510: Definition of Critical Events Specifically, the key event logs include unpacking events, impact events, and alarm events. Whenever the door magnetic sensor 10 detects an unpacking operation, the system records the unpacking time and duration; whenever the accelerometer detects an impact acceleration exceeding a threshold, the system records the time, intensity, and direction of the impact event; whenever… or When an alert is triggered, the system records the time, level, and cause of the alarm event. Furthermore, these critical events serve as important evidence for assessing the quality of sample transportation.

[0129] Step 520: Data on-chain cycle and hash calculation method The periodic upload cycle is between 300 and 900 seconds. In this embodiment, a relatively short upload cycle is preferred, which achieves the best balance between data granularity, communication traffic, and blockchain transaction costs.

[0130] Specifically, the hash calculation includes: converting locally collected data packets (containing...) , Data such as GPS location, temperature data, and key event records are used to generate a Merkle tree root hash, which is then uploaded to the blockchain network as a transaction. It should be noted that a Merkle tree is an efficient data digest structure that can represent the entire data packet with a single root hash value, ensuring efficient data integrity verification while saving storage space on the blockchain.

[0131] Step 530: Blockchain Network Type The blockchain network described herein adopts either a consortium blockchain or a private blockchain. Furthermore, compared to public blockchains, consortium blockchains or private blockchains offer advantages such as faster transaction speeds, lower transaction fees, and better privacy protection, making them more suitable for application scenarios like cold chain logistics that require multi-party collaboration but involve trade secrets. More specifically, the consortium blockchain is jointly maintained by relevant parties such as cold chain transportation companies, sample testing units, and certification bodies, ensuring decentralized data storage and multi-party witnessing, fundamentally eliminating the possibility of data tampering.

[0132] Step 540: Automatic generation of traceability report After transportation is completed, the system automatically generates a sample transportation cold chain integrity report containing a blockchain verification code. This report includes the complete transportation route, temperature profile, and... Change curve The system includes historical forecasts, a list of key events, and blockchain verification codes. For example, any third party (such as certification bodies or forensic institutions) can query and verify the report's authenticity on the blockchain network using the verification code, ensuring the legal validity of the cold chain data. It should be noted that this blockchain-based traceability mechanism can meet the requirements of relevant certification bodies regarding the authenticity and completeness of cold chain data.

[0133] As can be seen from the above embodiments, this application achieves a technological upgrade of cold chain monitoring from "recording / responding" to "evaluation / prediction / optimization" through the multi-parameter fusion sample integrity index (SII) algorithm, the residual safe transport time (RSTT) prediction algorithm based on the minimum bottleneck principle, and the intelligent power management and path optimization collaborative method of predictive maintenance.

[0134] Specifically, the multi-parameter fusion SII algorithm significantly improves the accuracy of sample integrity assessment compared to traditional single temperature monitoring methods by quantifying multi-dimensional risk factors such as temperature, light, vibration, and unpacking. This effectively avoids sample scrapping caused by environmental mechanical or light factors, especially for volatile, photosensitizing, and mechanically sensitive samples.

[0135] Furthermore, predictive maintenance and intelligent power management through Predictive guidance enables the semiconductor cooling module 3 to adjust power and make transportation route decisions, significantly extending the system's runtime while ensuring temperature safety. At the same time, the response time of predictive alarms is significantly earlier than that of traditional threshold alarms, providing transportation personnel with ample time to intervene.

[0136] More specifically, the hybrid active-passive temperature control system and phase change material (PCM) state prediction significantly improve the system's robustness by upgrading PCM from a simple thermal insulation material to a predictable and quantifiable energy buffer unit. Even in extreme cases such as failure of the semiconductor cooling module 3 or battery depletion, the PCM can still provide reliable backup power.

[0137] It should be noted that blockchain-based cold chain traceability achieves tamper-proof cold chain records, greatly reducing audit costs and data dispute risks, and can meet relevant regulatory and certification requirements.

[0138] In summary, the method of this application not only solves the technical problems of single sample integrity assessment, low temperature control efficiency, alarm lag and low data reliability in the prior art, but also generates a significant technical synergy effect through the synergistic cooperation between various innovations, providing an efficient, reliable and trustworthy intelligent monitoring and temperature control solution for cold chain transportation of samples in environmental monitoring, biomedicine and other fields.

[0139] To enable those skilled in the art to more clearly implement the technical solutions of this application, the following provides exemplary supplementary descriptions of the key parameter calculation methods involved in the foregoing embodiments.

[0140] I. Definition of Time Margin and Power Mode Judgment Threshold In the intelligent temperature control execution step S4, the estimated transportation time is recorded as... Time leeway Defined as:

[0141] Preset the first time margin percentage threshold Second time margin percentage threshold And satisfy Based on the above definitions, the rules for determining "much greater than", "close to", and "less than" are as follows: when At that time, it was believed The transit time was much longer than expected, and the system entered low-power mode. when At that time, it was believed As the estimated transit time approaches, the system maintains standard power mode; when At that time, it was believed If the transit time is less than expected, the system enters high-power mode and triggers an alert.

[0142] Preferably, The value range is 0.2 to 0.5. The value range is 0.05 to 0.2. For example, when... , , When you are a child, if If it lasts for 9 hours, it will enter low power mode. During hours, it maintains the standard power mode. It enters high-power mode when the time is short.

[0143] II. Correspondence between Power Mode and Power Range With the rated maximum power of the semiconductor cooling module 3 Based on this, the power ranges for low-power mode, standard power mode, and high-power mode are defined as follows: Real-time power adjustment corresponding to low power mode satisfy: ; The real-time power adjustment corresponding to the standard power mode satisfies: ; The real-time power adjustment required for high-power mode is: .

[0144] III. Time Margin Factor Specific calculation formula The time margin factor The value and time margin The correspondence can be calculated using the following formula:

[0145] This formula ensures The value of is always in Within the interval. When the time margin When it is very large (i.e.) ), When the minimum value of 0.1 is reached, power is reduced and energy-saving mode is entered; when the time margin is negative and large (i.e. Significantly smaller than ), Set the maximum value to 1.0, and the power will increase to enter protection mode.

[0146] Based on the aforementioned power range definition, when lie in When the range is reached, the system operates in low-power mode; when lie in When the range is reached, the system operates in standard power mode; when lie in During the interval, the system operates in high-power mode.

[0147] IV. Deviation Time Weighting Factor Specific calculation form The deviation time weighting factor To represent the cumulative effect of temperature deviation over time, one of the following two models can be used: Linear growth model:

[0148] Exponential growth model:

[0149] in, For the temperature since the first deviation from the target temperature range The cumulative duration of the start (in seconds). For a linear growth model, The linear cumulative coefficient is preferably within the range of values. For the exponential growth model, The cumulative factor is the exponential coefficient, and its preferred range is [value range missing]. .

[0150] This design allows the temperature to start deviating from its normal range as soon as possible. The penalty weight is the baseline value; as the deviation time increases, The penalty weight gradually increases, thus truly reflecting the irreversibility and time-cumulative nature of sample quality degradation. For samples that are extremely sensitive to temperature (such as certain biological products), an exponential growth model is recommended to more strictly penalize long-term temperature deviations.

[0151] V. Cumulative Light Exposure Function Specific calculation formula The cumulative light exposure function Calculated using the following integral formula:

[0152] in, for Light intensity (unit: lux) collected by the constant light sensor 7. The light intensity threshold, This represents the illumination penalty coefficient.

[0153] When the light intensity Below the threshold When the light intensity exceeds the threshold, the max function makes the integrand equal to 0 and does not incur a penalty; when the light intensity exceeds the threshold, the excess portion is integrated over time.

[0154] Preferably, for photosensitizing samples (such as volatile organic compounds or photosensitizing proteins). It can be set to 50-200 lux. Can be set to For general samples, It can be set to 500-1000 lux. Can be set to .

[0155] VI. Vibration Cumulative Impact Function Specific calculation formula The vibration cumulative impact function Calculated using the following integral formula:

[0156] in, for The magnitude of the acceleration vector measured by the triaxial accelerometer at any given time (unit: g, 1g) ), The vibration acceleration threshold, This is the vibration penalty coefficient.

[0157] When vibration acceleration Below the threshold When the integrand is zero, the max function makes the integrand zero and does not impose a penalty; when the vibration acceleration exceeds the threshold, the excess is integrated over time.

[0158] Preferably, for impact-sensitive samples (such as soil aggregate samples, suspended particulate matter samples). It can be set to 0.3–0.5 g. Can be set to For general samples, It can be set to 1.0–2.0 g. Can be set to .

[0159] VII. Unboxing Cumulative Exposure Function Specific calculation formula The open-box cumulative exposure function Calculated using the following cumulative formula:

[0160] Where n is the number of times the box has been opened up to the current time t. The duration of the i-th unpacking event (in seconds). This represents the penalty coefficient for opening loot boxes.

[0161] This formula linearly sums the durations of all unpacking events and multiplies them by a penalty coefficient, reflecting the cumulative impact of unpacking exposure on sample integrity. Preferably, The range of values ​​is For samples that are extremely sensitive to environmental exposure, the concentration can be appropriately increased. The value of .

[0162] eight, Specific prediction methods The sample safety time Based on The system predicts the changing trends. Specifically, it records data within a recent time window (e.g., the last 30 minutes). The change curve was calculated using the linear regression method. average rate of decline Set the sample integrity warning threshold to: (The preferred value range is 60-80), then Calculated using the following formula:

[0163] when When the absolute value approaches zero and is less than 0.1 (i.e.) When remaining stable, Choose a large preset value (e.g., 48 hours) to indicate that the sample can be safely transported for a long time under the current conditions.

[0164] The above embodiments have the following technical effects: To address the limitation of existing technologies that rely on a single dimension for sample integrity assessment, the above embodiment periodically collects multi-dimensional environmental parameters such as temperature, illumination, vibration, open-box status, remaining battery power, and the state of phase change materials (step S1). This provides a comprehensive data foundation for subsequent integrated assessment and predictive decision-making, overcoming the limitations of traditional single-dimensional temperature monitoring. Based on this, the above embodiment calculates the sample integrity index using a weighted fusion algorithm based on these multi-dimensional parameters. (Step S2) This index is obtained through the temperature cumulative deviation function. Cumulative light exposure function Vibration cumulative impact function and open-box cumulative exposure function The weighted linear combination is achieved, thereby unifying and quantifying the dispersed multidimensional risk factors into a single operable evaluation index. This realizes the technical transformation from one-sided transient parameter judgment to comprehensive cumulative risk assessment, effectively avoiding sample quality misjudgment and potential scrapping caused by non-temperature factors such as light and vibration.

[0165] For temperature cumulative deviation function The calculations in the above embodiments are based on the actual temperature. Deviation from target temperature range The result is obtained by time-weighted cumulative integration of the degree of deviation, and uses a weighting factor that includes deviation from the time. Temperature penalty function This method causes the contribution of the duration of temperature deviation to cumulative damage to increase non-linearly, truly reflecting the irreversibility and time-cumulative nature of sample quality degradation. Compared to traditional instantaneous temperature threshold-triggered alarms, this cumulative integration method can more scientifically quantify the true degree of sample damage under full-time temperature fluctuations. Furthermore, by using a temperature penalty function... Concretize into The form of this method allows the penalty intensity to depend on both the deviation magnitude and the deviation time, thus enabling accurate modeling of the thermal stress history of the sample.

[0166] To address the technical problem of varying sensitivity of different types of samples to various risk factors, the above embodiments address this issue by using weighting coefficients. , , , Configured with an adjustable parameter range and adaptively configured based on sample type, allowing for adjustments to temperature-sensitive biological samples by increasing... The weight of temperature factors can be increased by adjusting the value of the temperature factor; for shock-sensitive soil samples, this can be achieved by increasing the value of the temperature factor. The values ​​are selected to highlight the importance of vibration factors, thereby achieving the wide applicability and flexibility of the monitoring method for various sample types.

[0167] To address the technical problem of lagging alarm mechanisms in existing technologies, the above embodiments address this issue by basing them on... The changing trend, remaining battery capacity, and phase change material state predict the remaining safe transportation time. (Step S3) represents a technological leap from post-event alarms to predictive early warning. Calculate battery life using the principle of minimum bottleneck. Phase change material heat preservation time and safe time based on SII trend The minimum value of the three factors determines the design, which ensures the conservatism and reliability of the prediction results, can identify the bottleneck factors that will fail first in the system, provides transportation personnel with sufficient time to intervene, and avoids sample loss caused by the response lag of traditional threshold-triggered alarms.

[0168] Regarding the passive heat preservation time of phase change materials The above embodiment predicts the remaining passive insulation time by measuring the temperature change rate of the phase change material layer 2, combining the melting point curve and heat capacity parameters of the phase change material to infer the solid phase percentage of the phase change material. This technology upgrades the phase change material from a simple passive insulation material to a predictable and quantifiable energy buffer unit, significantly improving the robustness and reliability of the system in the event of active temperature control module failure or battery depletion. Regarding battery life... The above embodiment predicts the average power consumption of the active temperature control module by fitting historical power consumption using the least squares method, and calculates it in combination with the remaining battery power. This method comprehensively considers the dynamic changes in the power of the temperature control module and the actual situation of the intermittent operation of the communication module, thus improving the accuracy of the battery life prediction.

[0169] To address the technical problems of rigid temperature control strategies and the inability to dynamically optimize power based on remaining time in existing technologies, the above embodiments address these issues by... By comparing the estimated transport time with the actual transport time, the power of the active temperature control module is intelligently adjusted based on the comparison result (step S4), thus achieving synergy between predictive maintenance and energy optimization management. When the travel time is much longer than expected, reduce power to a low-power mode to extend the range. If the transport time is less than expected, the power is increased to high power mode and an alert is triggered to enhance sample protection. This strategy will transfer information flow ( The predicted results are transformed into closed-loop feedback control of the physical flow (temperature control power) and the decision flow (path optimization suggestions), which significantly improves the system's endurance while ensuring temperature safety.

[0170] Furthermore, by introducing a time margin factor into the power regulation formula... This makes the power Not only in response to the current temperature difference Also under the control of With a time margin between the estimated transport time and the actual transport duration, this two-dimensional coupled power regulation mechanism achieves a dynamic balance between temperature control accuracy and energy efficiency. When the time margin is large... Approaching zero causes power to decrease, entering energy-saving mode. When the time margin is small... Approaching 1 triggers a power increase to enter protection mode. This mechanism demonstrates the proactive guidance of predictive decision-making for physical control, significantly outperforming traditional passive control strategies based solely on temperature difference feedback. This is achieved by adjusting the power regulation coefficient... When configured within a reasonable numerical range, the temperature control system can be optimized and calibrated based on the insulation performance of the storage box and the response characteristics of the temperature control module, further improving its adaptability.

[0171] Regarding the implementation of path optimization early warning, when When the transport time is less than expected, the system not only increases the temperature control power, but also suggests that transport personnel find alternative refrigeration facilities or change the route to speed up the process. This human-machine collaboration mechanism combines the predictive ability of the algorithm with the flexible decision-making of humans, realizing the synergy between temperature control optimization and route optimization, and minimizing the risk of sample loss.

[0172] To address the technical problems of low data reliability and susceptibility to tampering in existing technologies, the above embodiments address these issues by periodically... , After timestamping and hashing key event records, the data is uploaded to the blockchain network (step S5), forming an immutable cold chain proof, fundamentally solving the credibility problem of traditional centralized data storage. The key events include unpacking events, impact events, and alarm events; the immutable records of these events provide highly credible evidence for sample quality assessment and legal arbitration. By generating Merkle root hashes from locally collected data packets and uploading them to a consortium or private blockchain in the form of transactions, both the decentralization and tamper-proof characteristics of the data are guaranteed, while also considering transaction speed and cost control. After transportation, a sample transportation cold chain integrity report containing a blockchain verification code is automatically generated, allowing any third party to query and verify the authenticity of the data through the verification code, meeting relevant regulatory and certification requirements, and helping to reduce audit costs and data dispute risks.

[0173] The above embodiment realizes a hybrid active and passive temperature control system through a sandwich structure of vacuum insulation plate 1, phase change material layer 2 and semiconductor refrigeration module 3 used in the transport and storage box. The phase change material layer 2 is embedded with a sensor for monitoring the state of the phase change material. This structural design organically combines the high reliability of passive insulation with the flexibility of active temperature control. When active temperature control fails, the phase change material can provide reliable backup power, which greatly improves the robustness of the system.

[0174] By performing Kalman filtering to denoise the acquired raw data and using the main control unit 11 as an edge computing unit to run the sample integrity assessment and remaining safe transportation time prediction algorithms in real time, the accuracy of the input data and the real-time performance of the algorithm execution were ensured. By setting the multi-parameter acquisition cycle to 30 seconds and the remaining safe transportation time prediction cycle to 5 minutes, a good balance was achieved between data real-time performance and system power consumption.

[0175] In summary, the technical features of the above embodiments are not isolated, but rather form a complete technical chain through multi-dimensional data acquisition in step S1, SII fusion evaluation in step S2, RSTT bottleneck prediction in step S3, predictive power regulation in step S4, and blockchain traceability in step S5. Each step is tightly coupled through data flow and control flow. In particular, the RSTT prediction result in step S3 provides feedforward control for the power regulation in step S4, achieving closed-loop optimization of the physical flow through the information flow and generating a significant synergistic effect. This technical solution upgrades cold chain monitoring from the traditional "record-response" model to an "evaluation-prediction-optimization" model, achieving outstanding technical results in terms of sample integrity assessment accuracy, temperature control system endurance, early warning timeliness, and data reliability. It effectively solves the technical problems of single evaluation dimensions, lagging alarm mechanisms, rigid temperature control strategies, and easy data tampering in existing technologies.

[0176] It should be noted that in the application documents of this application, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In the application documents of this application, if it refers to performing an action according to an element, it means performing the action at least according to that element, including two cases: performing the action only according to that element, and performing the action according to that element and other elements. Expressions such as "multiple," "repeatedly," and "various" include two, two times, two kinds, and more than two, more than two times, and more than two kinds.

[0177] All documents mentioned in this application are considered to be incorporated in their entirety into the disclosure of this application so that they can serve as a basis for modifications if necessary. Furthermore, it should be understood that after reading the foregoing disclosure of this application, those skilled in the art can make various alterations or modifications to this application, and these equivalent forms also fall within the scope of protection claimed in this application.

Claims

1. A multi-parameter prediction based intelligent monitoring and temperature control method for sample integrity, characterized in that, The method comprises the following steps: S1, a multi-parameter acquisition step: periodically acquiring multi-dimensional environmental parameters inside and outside the transport storage box, the multi-dimensional environmental parameters at least including temperature, illumination, vibration, opening box state, battery remaining capacity and state of phase change material; S2, sample integrity evaluation step: the temperature, light, vibration and unpacking state collected in step S1 are weighted and fused to calculate a sample integrity index , the is calculated by the following formula: wherein, is the current transport time, is a temperature cumulative deviation function, is a light cumulative exposure function, is a vibration cumulative shock function, is an unpacking cumulative exposure function, , , , is a weight coefficient, satisfying ; S3, remaining safe transportation time prediction step: based on the change trend of the battery capacity, the remaining battery capacity and the state of the phase change material collected in step S1, and the change trend calculated in step S2, the remaining safe transportation time is predicted determined as:​​ wherein, is a battery endurance time predicted based on the battery remaining capacity, is a passive insulation time predicted based on the phase change material state, is a sample safety time predicted based on the trend of change of the battery remaining capacity; min() is a function of taking minimum value. S4, intelligent temperature control execution step: intelligently adjusting the power of the active temperature control module of the transport preservation box according to the comparison result of the predicted transport duration and the actual transport duration. comparing the predicted transport duration with the actual transport duration, and intelligently adjusting the power of the active temperature control module of the transport preservation box according to the comparison result. when the a product of a first threshold value and the estimated transport duration, the power is reduced to a low power mode; when the first threshold value is multiplied by the estimated transportation time length and the product of the first threshold value and the estimated transportation time length is less than the second threshold value and maintain the standard power mode when the product of the preset second threshold value and the estimated transportation time length is reached. when the product of the second threshold value and the estimated transportation duration is greater than the first threshold value, the power is increased to a high power mode and a warning is triggered, wherein the first threshold value > 1.0 and the second threshold value < 1.

0. when the product of the second threshold value and the estimated transportation duration is greater than the first threshold value, the power is increased to a high power mode and a warning is triggered, wherein the first threshold value > 1.0 and the second threshold value < 1.

0.

2. The method of claim 1, wherein, The temperature cumulative deviation function By time-weighted cumulative integration of the degree of deviation from the target temperature range The calculation formula of the temperature cumulative deviation function is as follows: The calculation formula of the temperature cumulative deviation function is as follows: The calculation formula of the temperature cumulative deviation function is as follows: wherein, is a temperature penalty function, when is located in the target temperature range , is located outside the target temperature range, the value of is proportional to the absolute value of the temperature deviation, and the includes a deviation time weighting factor , which is used to embody the time accumulation effect.

3. The method of claim 2, wherein, The temperature penalty function The specific form of calculation is: wherein, is a temperature penalty coefficient, is a target temperature, is the deviation time weight factor.

4. The method of claim 3, wherein, the deviation time weight factor using a linear growth model or an exponential growth model, the linear growth model being , and the exponential growth model being wherein is a cumulative duration of temperature deviation from the target range, and is a time cumulative coefficient.

5. The method of claim 1, wherein: The light exposure accumulation function is calculated by time-cumulative integration of the portion of the light intensity that exceeds the threshold value; The vibration cumulative impact degree function is calculated by time-cumulative integration of the portion where the vibration acceleration exceeds the threshold value; The opening box cumulative exposure function is calculated by accumulating the duration of each opening box event.

6. The method of claim 1, wherein, The The passive insulation time is predicted by measuring the temperature change rate of the phase change material layer, combining the melting point curve and heat capacity of the phase change material, and deducing the solid phase percentage of the phase change material.

7. The method of claim 1, wherein, The is predicted by fitting historical power consumption using a least squares method, and is calculated in combination with the battery remaining power.

8. The method of claim 1, wherein, The active temperature control module is a semiconductor refrigeration module.

9. The method of claim 1, wherein, The step of intelligently adjusting the power of the active temperature control module includes: in the prediction mode, the power adjustment formula introduces a time margin factor : wherein, is a temperature difference between an actual temperature inside the transport preservation case and a target temperature, is a power adjustment coefficient, is a maximum power of the active temperature control module, is the time margin factor, the is positively correlated with a margin between the and the predicted transport duration.

10. The method of claim 9, wherein, Defining time margin wherein is the predicted transport time length; when said takes a value comprised between 0.1 and 0.3, inclusively, to decrease said power; when said takes a value comprised between 0.7 and 1.0, inclusively, to increase said power; wherein and are preset time margin percentage threshold values, and .

Images