Method and system for predicting quality changes of refrigerated drugs in cold chain transportation

Abstract

The present invention relates to a method and system for predicting changes in the quality of refrigerated drugs during cold chain transportation, and to the technical field of cold chain transportation of drugs, including: obtaining a refrigerated reference temperature interval and a refrigerated temperature fluctuation threshold; collecting refrigerated temperature time series information during cold chain transportation; counting first-class refrigerated temperature time series information, second-class refrigerated temperature time series information, third-class refrigerated temperature time series information, and fourth-class refrigerated temperature time series information; processing through a drug anomaly binary model of the refrigerated drug model, obtaining a drug quality binary identification, and setting it as a quality prediction identification of the refrigerated drug model. The method solves the technical problem that the efficiency of drug quality detection is low in the prior art because a full inspection is usually performed after the end of transportation, and the drug quality cannot be tracked during the transportation process.

Description

Technical Field

[0001] The present invention relates to the technical field of cold chain transportation of medicines, and in particular to a method and system for predicting quality changes of refrigerated medicines during cold chain transportation. Background Art

[0002] In the field of cold chain transportation, especially for the transportation of refrigerated medicines, temperature control is a key factor in ensuring the quality of medicines. Traditional cold chain transportation management mainly relies on temperature recorders to monitor temperature changes during transportation to ensure that the medicines are always within the appropriate temperature range. However, these methods often lack in-depth analysis and real-time monitoring of temperature data, resulting in the inability to detect and respond to temperature anomalies in a timely manner, thus affecting the quality of medicines.

[0003] The existing cold chain transportation quality inspection method usually conducts a full inspection after the transportation is completed, which means that even if the drug has no abnormalities during the entire transportation process, it still needs to be inspected, which not only increases the inspection cost but also reduces the inspection efficiency. Secondly, the traditional inspection method cannot monitor the quality status of the drug in real time, resulting in the inability to detect and deal with the possible deterioration during transportation in a timely manner. Summary of the invention

[0004] The present invention aims to solve the technical problem that in the prior art, since a full inspection is usually carried out after the transportation is completed, the quality of the drugs cannot be tracked during the transportation process, resulting in low efficiency of drug quality inspection. A method and system for predicting the change of refrigerated drug quality during cold chain transportation is provided to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for predicting quality changes of refrigerated medicines during cold chain transportation, including: performing refrigerated sample analysis on refrigerated medicine models to obtain a refrigerated reference temperature range and a refrigerated temperature fluctuation threshold; during cold chain transportation, starting a timer to start timing, and whenever the timing data meets a preset time length, the timer is reset to zero and restarts timing, while collecting refrigerated temperature timing information; counting a type of refrigerated temperature timing information that satisfies the refrigerated reference temperature range and the refrigerated temperature fluctuation threshold; counting a type of refrigerated temperature timing information that satisfies the refrigerated reference temperature range and does not meet the refrigerated temperature fluctuation threshold Two types of refrigerated temperature timing information; counting the three types of refrigerated temperature timing information that do not satisfy the refrigerated reference temperature interval and satisfy the refrigerated temperature fluctuation threshold; counting the four types of refrigerated temperature timing information that do not satisfy the refrigerated reference temperature interval and do not satisfy the refrigerated temperature fluctuation threshold; processing the one type of refrigerated temperature timing information, the two types of refrigerated temperature timing information, the three types of refrigerated temperature timing information and the four types of refrigerated temperature timing information through the drug abnormality binary model of the refrigerated drug model to obtain the drug quality binary identification, and set it as the quality prediction identification of the refrigerated drug model.

[0007] In a second aspect, the present invention provides a system for predicting quality changes of refrigerated drugs during cold chain transportation, including: a refrigerated sample analysis module, used to perform refrigerated sample analysis on refrigerated drug models to obtain a refrigerated reference temperature range and a refrigerated temperature fluctuation threshold; a refrigerated temperature acquisition module, used to start a timer to start timing during cold chain transportation, and whenever the timing data meets a preset time length, the timer is reset to zero and restarts timing, and at the same time, collects refrigerated temperature timing information; a type of refrigerated temperature extraction module, used to count the refrigerated temperature timing information that meets the refrigerated reference temperature range and meets the refrigerated temperature fluctuation threshold; a type of refrigerated temperature extraction module, used to count the refrigerated temperature timing information that meets the refrigerated reference temperature range and does not meet the refrigerated temperature A second type of refrigerated temperature timing information that meets the refrigerated temperature fluctuation threshold; a third type of refrigerated temperature extraction module, used to count the three types of refrigerated temperature timing information that do not meet the refrigerated reference temperature interval and meet the refrigerated temperature fluctuation threshold; a fourth type of refrigerated temperature extraction module, used to count the four types of refrigerated temperature timing information that do not meet the refrigerated reference temperature interval and do not meet the refrigerated temperature fluctuation threshold; a drug quality prediction module, used to process the first type of refrigerated temperature timing information, the second type of refrigerated temperature timing information, the third type of refrigerated temperature timing information and the fourth type of refrigerated temperature timing information through the drug abnormality binary model of the refrigerated drug model, to obtain a drug quality binary identification, and set it as the quality prediction identification of the refrigerated drug model.

[0008] In a third aspect, the present invention provides a computer-readable storage medium having computer-executable instructions stored thereon, which, when executed by a processor, implement the method for predicting quality changes of refrigerated drugs during cold chain transportation as described above.

[0009] The beneficial effect of the present invention is: by analyzing the temperature monitoring data during cold chain transportation, especially the abnormal temperature time series information, including the first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information, and combining the drug abnormality binary model, it is possible to make a real-time prediction of drug quality and provide reference information for drug quality inspection after the end of cold chain transportation, thereby achieving the technical effect of improving inspection efficiency and reducing the need for full inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A schematic diagram of the process of predicting the quality change of refrigerated medicines in cold chain transportation provided by the present invention;

[0011] Figure 2 A schematic diagram of the structure of the prediction of quality changes of refrigerated medicines in cold chain transportation provided by the present invention;

[0012] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention;

[0013] Figure 4 A schematic diagram of the structure of a computer-readable storage medium provided by the present invention.

[0014] In the accompanying drawings, the components represented by the reference numerals are described as follows:

[0015] Electronic device 500 , memory 510 , processor 520 , first computer program 511 , computer-readable storage medium 600 , second computer program 611 . DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0017] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0018] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention. Embodiment 1:

[0019] like Figure 1 As shown, the embodiment of the present invention provides a method for predicting the quality change of refrigerated medicines in cold chain transportation, comprising the steps of:

[0020] S10: Analyze the refrigerated samples of the refrigerated medicine models to obtain the refrigerated reference temperature range and the refrigerated temperature fluctuation threshold;

[0021] Furthermore, the refrigerated samples of the refrigerated medicine models are analyzed to obtain the refrigerated reference temperature range and refrigerated temperature fluctuation threshold, including:

[0022] Collecting non-deteriorated refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record intervals;

[0023] Performing frequent boundary statistics on the refrigerated temperature record interval set to obtain the refrigerated reference temperature interval;

[0024] Collect spoiled refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record time series data, wherein the spoiled refrigerated samples meet the refrigerated reference temperature range, and the refrigerated temperature record time series data is the time series data from the time when the medicine is not spoiled to the time when the deterioration is detected;

[0025] The refrigerated temperature record time series data set is traversed to perform frequent fluctuation statistics to obtain the refrigerated temperature fluctuation threshold.

[0026] Furthermore, frequent boundary statistics are performed on the refrigerated temperature record interval set to obtain the refrigerated reference temperature interval, including:

[0027] Obtaining an interval upper limit temperature record value set and an interval lower limit temperature record value set of the refrigeration temperature record interval set;

[0028] Traversing the interval upper limit temperature record value set to perform frequency analysis to obtain the interval upper limit temperature record value frequency set;

[0029] According to the interval upper limit temperature record value frequency set, the interval upper limit temperature record values ​​that are less than or equal to the interval upper limit temperature record value frequency threshold are deleted from the interval upper limit temperature record value set to obtain frequent interval upper limit temperature record values, and the maximum value of the frequent interval upper limit temperature record values ​​is extracted and set as the refrigeration reference temperature interval upper limit value;

[0030] Traversing the interval lower limit temperature record value set to perform frequency analysis to obtain the interval lower limit temperature record value frequency set;

[0031] According to the interval lower limit temperature record value frequency set, the interval lower limit temperature record values ​​that are less than or equal to the interval lower limit temperature record value frequency threshold are deleted from the interval lower limit temperature record value set to obtain frequent interval lower limit temperature record values, and the minimum value of the frequent interval lower limit temperature record values ​​is extracted and set as the lower limit value of the refrigeration reference temperature interval;

[0032] The refrigerated reference temperature interval is constructed according to the refrigerated reference temperature interval upper limit value and the refrigerated reference temperature interval lower limit value.

[0033] Further, the interval upper limit temperature record value set is traversed to perform frequency analysis to obtain the interval upper limit temperature record value frequency set, including:

[0034] Extracting a first interval upper limit temperature record value according to the interval upper limit temperature record value set;

[0035] Count the number of interval upper limit temperature record values ​​in the interval upper limit temperature record value set whose temperature deviation from the first interval upper limit temperature record value is less than or equal to the temperature deviation threshold, calculate the ratio of the number of interval upper limit temperature record values ​​to the total number of interval upper limit temperature record value sets, set it as the first interval upper limit temperature record value frequency, and add it to the interval upper limit temperature record value frequency set.

[0036] Further, traversing the refrigerated temperature record time series data set to perform frequent fluctuation statistics to obtain the refrigerated temperature fluctuation threshold includes:

[0037] Extracting first refrigerated temperature record time series data according to the refrigerated temperature record time series data set, calculating the temperature deviation of adjacent moments of the first refrigerated temperature record time series data, and obtaining a first adjacent moment temperature deviation set;

[0038] Traversing the first adjacent time temperature deviation set to perform frequency analysis to obtain the first adjacent time temperature deviation frequency set;

[0039] According to the first adjacent moment temperature deviation frequency set, the first adjacent moment temperature deviation that is less than or equal to the first adjacent moment temperature deviation frequency threshold is deleted from the first adjacent moment temperature deviation set to obtain frequent adjacent moment temperature deviations, and the minimum value of the frequent adjacent moment temperature deviations is extracted and set as the first refrigerated temperature fluctuation threshold, and added to the initial refrigerated temperature fluctuation threshold set;

[0040] The minimum value of the initial refrigeration temperature fluctuation threshold set is extracted and set as the refrigeration temperature fluctuation threshold.

[0041] Specifically, refrigerated drug models refer to drug models that require cold chain transportation; the refrigerated reference temperature range refers to the temperature range required to maintain the stability of drug quality, that is, when the drug is stored within this temperature range, its quality will not change significantly; the refrigerated temperature fluctuation threshold refers to the maximum value of the temperature fluctuation allowed within the refrigerated reference temperature range. Exceeding this threshold may cause damage to the drug quality.

[0042] Furthermore, the detailed process of cold storage sample analysis is as follows:

[0043] The collection of non-spoilage refrigerated samples refers to collecting drug samples that are stored under normal refrigeration conditions and have not deteriorated, obtaining the temperature records of these samples during the storage process, and forming a set of refrigerated temperature record intervals; frequent boundary statistics is a statistical method used to determine the temperature range in which a drug can be safely stored from these temperature records, that is, the refrigerated reference temperature range; the collection of spoiled refrigerated samples refers to collecting drug samples that have deteriorated during storage, and recording the temperature change data of these samples from the beginning of storage to the time when the deterioration is discovered, to form a set of refrigerated temperature record time series data; frequent fluctuation statistics refers to analyzing the temperature records of these spoiled samples to determine the maximum temperature fluctuation allowed under normal storage conditions, that is, the refrigerated temperature fluctuation threshold.

[0044] The specific process is as follows:

[0045] Collection of non-spoiled refrigerated samples: select a certain number of non-spoiled refrigerated drug samples and record their temperature data during the refrigeration process. These data are used to determine the temperature range in which the drugs can be safely stored; frequent boundary statistics: analyze the temperature records of the collected non-spoiled samples to find out the temperature change boundaries of these samples during the storage process, that is, the refrigerated reference temperature range. This range defines the temperature range that the drugs should maintain during storage to ensure the quality of the drugs; collection of spoiled refrigerated samples: collect refrigerated drug samples that have deteriorated during storage, and record the temperature change data of these samples from the beginning of storage to the time when the deterioration is discovered, which is used to determine the maximum temperature fluctuation allowed under normal storage conditions; frequent fluctuation statistics: analyze the temperature records of spoiled samples to find out the temperature fluctuations that occur within the refrigerated reference temperature range, and determine the maximum allowable fluctuation value, that is, the refrigerated temperature fluctuation threshold. This threshold will be used for subsequent temperature monitoring and drug quality prediction.

[0046] Furthermore, the frequent boundary statistics processing flow is as follows:

[0047] The interval upper limit temperature record value set and the interval lower limit temperature record value set refer to the set of the highest temperature and the lowest temperature of each record extracted from the refrigerated temperature record interval set; the interval upper limit temperature record value frequency set and the interval lower limit temperature record value frequency set contain the frequency of each upper limit and lower limit temperature value respectively; the frequent interval upper limit temperature record value and the frequent interval lower limit temperature record value refer to those temperature record values ​​whose frequency exceeds the preset threshold value preset by the user, which are considered to be the temperature values ​​required to maintain the stability of drug quality in most cases. Finally, these values ​​are used to determine the upper limit value and the lower limit value of the refrigerated reference temperature interval, thereby constructing the entire refrigerated reference temperature interval.

[0048] In detail, the highest temperature and the lowest temperature of all records are extracted from the set of refrigerated temperature record intervals to form a set of interval upper and lower temperature record values; the set of interval upper and lower temperature record values ​​and the set of interval lower temperature record values ​​are traversed to count the ratio of the frequency of occurrence of each temperature value to the total number of temperature values ​​in the interval to form a frequency set; according to the frequency threshold, those temperature record values ​​with lower frequency, that is, those values ​​with a frequency lower than the threshold, are deleted from the frequency set; after deleting the low-frequency values, what remains are the upper and lower temperature record values ​​of the frequent interval; the maximum value is extracted from the upper temperature record value of the frequent interval as the upper limit value of the refrigerated reference temperature interval, and the minimum value is extracted from the lower temperature record value of the frequent interval as the lower limit value of the refrigerated reference temperature interval; according to the determined upper and lower limits, a refrigerated reference temperature interval is constructed, which defines the temperature range that the drug should maintain during storage and transportation to ensure the quality of the drug.

[0049] Furthermore, frequency analysis refers to the process of statistically analyzing each temperature value in the set of interval upper limit temperature records to determine the frequency of its occurrence and the ratio of the total number of temperature values ​​in the interval. The first interval upper limit temperature record value is an exemplary temperature value extracted from the set, which is used to illustrate the specific process of frequency analysis; temperature deviation refers to the difference between other temperature values ​​and the reference temperature value; temperature deviation threshold is a preset value used to determine the number of temperature values ​​that are close enough to the reference value for subsequent frequency calculation.

[0050] In detail, the first temperature value is extracted from the interval upper limit temperature record value set as a reference value, the interval upper limit temperature record value set is traversed, the number of temperature values ​​whose temperature deviation from the first interval upper limit temperature record value is less than or equal to the temperature deviation threshold is counted, the number of temperature values ​​is compared with the total number of the interval upper limit temperature record value set, and the ratio, that is, the frequency of the first interval upper limit temperature record value is calculated, and the calculated frequency is added to the interval upper limit temperature record value frequency set.

[0051] By analyzing the frequency of temperature recording values, we can help determine which temperature values ​​are common in the refrigeration process, thereby providing data support for determining the refrigeration benchmark temperature range. This can more accurately determine the temperature range that drugs should maintain during cold chain transportation to ensure drug quality.

[0052] The detailed process of frequent fluctuation statistics is as follows:

[0053] The "first refrigerated temperature record time series data" is the temperature change data of a single sample extracted from the set. The temperature deviation at adjacent moments refers to the temperature difference between two adjacent time points in the time series; the "frequency analysis" counts these deviation values ​​to determine their frequency of occurrence; the first adjacent moment temperature deviation frequency set contains the frequency of each deviation value; frequent adjacent moment temperature deviations refer to those deviation values ​​whose frequency is higher than the preset threshold, which are considered to be normal temperature fluctuations.

[0054] In detail: extract the temperature change data of the first sample from the refrigerated temperature record time series data set, calculate the temperature difference between adjacent time points in this sample to form a first adjacent moment temperature deviation set, perform frequency analysis on each deviation value in the first adjacent moment temperature deviation set, count their occurrence frequencies to form a first adjacent moment temperature deviation frequency set, delete those deviation values ​​whose frequency is lower than or equal to the threshold from the first adjacent moment temperature deviation set according to the first adjacent moment temperature deviation frequency threshold, retain the frequent adjacent moment temperature deviations, extract the minimum value from the retained frequent adjacent moment temperature deviations, set it as the first refrigerated temperature fluctuation threshold, and add it to the initial refrigerated temperature fluctuation threshold set, extract the minimum value from the initial refrigerated temperature fluctuation threshold set, and set it as the refrigerated temperature fluctuation threshold.

[0055] By analyzing the temperature changes of refrigerated drugs during storage and determining a reasonable temperature fluctuation range, it can help identify abnormal temperature fluctuations that may cause drug deterioration, thereby improving the accuracy and efficiency of drug quality monitoring.

[0056] S20: During cold chain transportation, a timer is started to count, and whenever the timing data meets the preset time length, the timer is reset to zero and starts counting again, and at the same time, the refrigeration temperature time series information is collected;

[0057] Specifically, starting the timer means turning on a timing device to record the time during cold chain transportation. The preset duration refers to a pre-set time length, which is the standard for determining when to collect temperature data; resetting the timer to zero means that whenever the time recorded by the timer reaches the preset duration, the timer will return to zero and restart the timing; collecting refrigeration temperature timing information means collecting the current refrigeration temperature data while resetting the timer, which will be used for subsequent temperature change analysis and drug quality prediction.

[0058] S30: counting the refrigerated temperature time series information that satisfies the refrigerated temperature reference temperature interval and the refrigerated temperature fluctuation threshold;

[0059] Specifically, a type of refrigerated temperature time series information refers to temperature records that meet both the refrigerated reference temperature range and the refrigerated temperature fluctuation threshold. The detailed process is: from all the collected refrigerated temperature time series information, select those temperature records whose temperature values ​​fall within the refrigerated reference temperature range. For the selected temperature records, further check whether their temperature fluctuations are within the range allowed by the refrigerated temperature fluctuation threshold, and classify the temperature records that meet both of the above conditions as a type of refrigerated temperature time series information.

[0060] S40: counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold;

[0061] Specifically, the second-class refrigerated temperature time series information refers to those temperature records that meet the requirements of the refrigerated reference temperature range but exceed the refrigerated temperature fluctuation threshold. The acquisition process is as follows: filter out those temperature records whose temperature values ​​fall within the refrigerated reference temperature range from all collected refrigerated temperature time series information. Furthermore, for the filtered temperature records, further check whether their temperature fluctuations exceed the range allowed by the refrigerated temperature fluctuation threshold. Those temperature records whose temperature fluctuations exceed the threshold are classified as second-class refrigerated temperature time series information. Although the temperature of the second-class refrigerated temperature time series information is generally kept within an appropriate range, there is a certain degree of temperature fluctuation, which may affect the quality of drugs. Statistics of this type of specific temperature time series information are quantitatively summarized for further analysis and evaluation.

[0062] S50: counting three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold;

[0063] Specifically, the three types of refrigerated temperature time series information refer to those temperature records that are within the temperature fluctuation range allowed by the refrigerated temperature fluctuation threshold, but the overall temperature exceeds the refrigerated reference temperature range. The three types of refrigerated temperature time series information indicate that although the temperature fluctuation is controlled within an acceptable range, the average temperature has deviated from the appropriate temperature range required to maintain the quality of the drug. The detailed acquisition process is as follows: from all the collected refrigerated temperature time series information, those temperature records whose average temperature is not within the refrigerated reference temperature range are screened out. For the screened temperature records, their temperature fluctuations are further checked to see if they are within the range allowed by the refrigerated temperature fluctuation threshold. The temperature records whose temperatures exceed the reference range but whose temperature fluctuations are controlled within the threshold are classified into the three types of refrigerated temperature time series information.

[0064] S60: counting four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature range and do not meet the refrigeration temperature fluctuation threshold;

[0065] Specifically, the four types of refrigerated temperature time series information refer to temperature records that neither meet the refrigerated reference temperature range nor the refrigerated temperature fluctuation threshold; the process of obtaining the four types of refrigerated temperature time series information is as follows: filter out those temperature records whose temperatures are not within the refrigerated reference temperature range from all collected refrigerated temperature time series information; for the filtered temperature records, further check whether their temperature fluctuations exceed the range allowed by the refrigerated temperature fluctuation threshold; classify those temperature records whose average temperature exceeds the reference range and whose temperature fluctuations exceed the threshold into four types of refrigerated temperature time series information. The four types of refrigerated temperature time series information are the records with the worst temperature control during the cold chain transportation process. Through the statistics and analysis of the four types of refrigerated temperature time series information, the serious deviations in temperature control during the cold chain transportation process can be evaluated, providing important data support for subsequent drug quality predictions.

[0066] S70: The first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model.

[0067] Specifically, the drug abnormality binary model refers to a classification model used to distinguish whether a drug is abnormal (normal or abnormal). The drug abnormality binary model is trained based on different types of refrigeration temperature time series information to predict the quality status of refrigerated drugs; the drug quality binary identification refers to the result after model processing, which is used to indicate the quality status of the drug, usually represented by a binary identification, for example, 1 may indicate abnormal quality, and 0 indicates normal quality. By setting the drug quality binary identification as the quality prediction identification of the refrigerated drug model, the drug quality can be tracked, providing reference information for subsequent drug quality testing and improving the detection efficiency.

[0068] Further, the spoiled refrigerated samples of the refrigerated medicine models are collected to obtain a refrigerated temperature record time series data set, which includes:

[0069] Infinite refrigeration sample collection is performed according to the refrigerated drug model to obtain an infinite refrigerated sample set, wherein any infinite refrigerated sample has a refrigerated drug deterioration mark or a refrigerated drug health mark;

[0070] Calculate the proportion of the infinite refrigerated samples with the refrigerated drug deterioration mark in the infinite refrigerated sample set, and set it as the refrigerated drug temperature fluctuation sensitivity coefficient;

[0071] When the temperature fluctuation sensitivity coefficient of the refrigerated medicine is greater than or equal to the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine, a deteriorated refrigerated sample is collected for the refrigerated medicine model to obtain the refrigerated temperature record time series data set, wherein the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine is greater than or equal to 0.2;

[0072] When the refrigerated drug temperature fluctuation sensitivity coefficient is less than the refrigerated drug temperature fluctuation sensitivity coefficient threshold, the refrigerated temperature fluctuation threshold is configured to be infinite.

[0073] Specifically, unlimited refrigerated sample collection refers to refrigerated drug samples collected without specific restrictions. These samples may include spoiled and non-spoiled drugs; the refrigerated drug spoilage mark and refrigerated drug health mark are labels used to distinguish whether the samples are spoiled; the refrigerated drug temperature fluctuation sensitivity coefficient is a proportional value, which indicates the proportion of spoiled samples in all samples, and is used to evaluate the sensitivity of drugs to temperature fluctuations; the refrigerated drug temperature fluctuation sensitivity coefficient threshold is a preset threshold used to determine whether it is necessary to collect spoiled samples.

[0074] The detailed process is as follows: collect a certain number of refrigerated drug samples, regardless of whether they are spoiled or not, and attach a spoiled or healthy label to each sample; count the proportion of spoiled samples in all samples, that is, the temperature fluctuation sensitivity coefficient of refrigerated drugs; if the temperature fluctuation sensitivity coefficient is greater than or equal to the preset threshold (for example, 0.2), collect spoiled refrigerated samples to obtain a set of refrigerated temperature record time series data; if the temperature fluctuation sensitivity coefficient is less than the preset threshold, set the refrigerated temperature fluctuation threshold to infinity, which means that the impact of temperature fluctuations will not be considered in subsequent analysis.

[0075] Further, the first type of refrigeration temperature time series information, the second type of refrigeration temperature time series information, the third type of refrigeration temperature time series information, and the fourth type of refrigeration temperature time series information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary mark, which is set as the quality prediction mark of the refrigerated drug model, including:

[0076] Obtain the refrigerated humidity monitoring timing information and refrigerated pH monitoring timing information of the refrigerated medicine model;

[0077] Configure the refrigerated humidity reference interval and refrigerated pH reference interval through the user terminal;

[0078] When the proportion of the first moment quantity of the refrigerated humidity monitoring time series information that does not belong to the refrigerated humidity reference interval is greater than or equal to the first ratio threshold, or / and when the proportion of the second moment quantity of the refrigerated pH monitoring time series information that does not belong to the refrigerated pH reference interval is greater than or equal to the second ratio threshold, the quality prediction mark of the refrigerated drug model is set to a quality abnormality mark;

[0079] Otherwise, the first type of refrigerated temperature timing information, the second type of refrigerated temperature timing information, the third type of refrigerated temperature timing information and the fourth type of refrigerated temperature timing information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model, wherein the drug quality binary identification is a normal quality identification or a quality abnormality identification.

[0080] Specifically, the refrigerated humidity monitoring time series information and the refrigerated pH monitoring time series information refer to the data on humidity and pH changes collected during the cold chain transportation process; the refrigerated humidity benchmark interval and the refrigerated pH benchmark interval are standard ranges set by the user according to the drug storage requirements; the first proportion threshold and the second proportion threshold are proportion standards used to determine whether the humidity and pH data exceed the benchmark interval; the quality prediction mark is set according to the model processing results, and is used to indicate the quality status of the drug, which may be a normal quality mark or an abnormal quality mark.

[0081] In detail, the refrigerated humidity monitoring time series information and the refrigerated pH monitoring time series information of the refrigerated drug model are obtained, and the refrigerated humidity benchmark interval and the refrigerated pH benchmark interval are configured through the user end. These intervals define suitable storage conditions. The first moment ratio of the humidity monitoring time series information that does not belong to the benchmark interval is calculated and compared with the first proportion threshold; the second moment ratio of the pH monitoring time series information that does not belong to the benchmark interval is calculated and compared with the second proportion threshold; if the proportion of humidity or pH data exceeding the benchmark interval reaches or exceeds the corresponding proportion threshold, the quality prediction mark is set as the quality abnormality mark; if both humidity and pH data do not exceed the proportion threshold of the benchmark interval, the drug abnormality binary model is used to process the first to fourth category refrigerated temperature time series information to obtain the drug quality binary mark.

[0082] When the humidity and pH are abnormal, the abnormality is directly marked. When there is no abnormality, the drug abnormality binary model of the refrigerated drug model is dispatched to process the first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information to obtain the drug quality binary mark, which is set as the quality prediction mark of the refrigerated drug model, wherein the drug quality binary mark is a normal quality mark or a quality abnormal mark, to ensure that there is no redundant data interference when analyzing temperature abnormalities.

[0083] Furthermore, the step of constructing the drug abnormality binary model includes:

[0084] Based on the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold, the first type of refrigerated temperature recording timing information, the second type of refrigerated temperature recording timing information, the third type of refrigerated temperature recording timing information and the fourth type of refrigerated temperature recording timing information are randomly configured in the preset time length;

[0085] Collect the refrigerated sample set of the refrigerated medicine model based on the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information;

[0086] Counting the proportion of samples with abnormal quality in the refrigerated sample set, and setting it as a quality abnormality probability identification value;

[0087] Based on the binary probability threshold, a binary fully connected layer is constructed, wherein when the abnormal quality probability identification value is greater than or equal to the binary probability threshold, the drug quality binary identification is 1, otherwise it is 0, 1 represents the abnormal quality identification, and 0 represents the normal quality identification;

[0088] According to the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information, and the quality abnormality probability identification value, training a quality abnormality probability assessment channel;

[0089] The output node of the quality anomaly probability assessment channel is merged with the input node of the binary fully connected layer to obtain the drug anomaly binary model.

[0090] Specifically, the drug abnormality binary model refers to a classification model used to distinguish whether a drug is abnormal. Different types of refrigerated temperature record time series information are configured based on the refrigerated reference temperature range and the refrigerated temperature fluctuation threshold; the first to fourth types of refrigerated temperature record time series information are data classified according to whether the temperature is within the reference range and whether its fluctuation is within the threshold; the refrigerated sample set refers to the sample set collected for training the model; the quality abnormality probability identification value refers to the proportion of quality abnormality samples in the sample set; the binary fully connected layer refers to the fully connected layer used for binary classification in the neural network, and the output result is a binary identification, 1 indicates quality abnormality, and 0 indicates normal quality; the quality abnormality probability assessment channel refers to the part of the neural network that assesses the probability of drug quality abnormality.

[0091] In detail, based on the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold, the refrigerated temperature record time series information of Class I to Class IV is randomly configured within the preset time length; with the configured temperature record time series information as a constraint, a refrigerated sample set of refrigerated drug models is collected; the proportion of quality abnormal samples in the sample set is counted and set as the quality abnormality probability identification value; based on the binary probability threshold, a binary fully connected layer in the neural network is constructed, and the drug quality binary identification is output according to the quality abnormality probability identification value; based on the refrigerated temperature record time series information of Class I to Class IV as input data and the quality abnormality probability identification value as supervision data, an evaluation channel is trained, and when the output deviation of at least 950 times out of 1000 consecutive training times is less than or equal to the output deviation threshold, a quality abnormality probability evaluation channel is generated to accurately evaluate the quality abnormality probability; the output node of the quality abnormality probability evaluation channel is merged with the input node of the binary fully connected layer to complete the construction of the drug abnormality binary model.

[0092] The method for predicting the quality change of refrigerated medicines in cold chain transportation provided by the embodiment of the present invention has at least the following technical effects:

[0093] By analyzing the temperature monitoring data during cold chain transportation, especially the abnormal temperature time series information, including the first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information, and combining the drug abnormality binary model, it is possible to make real-time predictions on drug quality and provide reference information for drug quality testing after the end of cold chain transportation, thereby achieving the technical effect of improving detection efficiency and reducing the need for full inspection. Embodiment 2:

[0094] like Figure 2 As shown, based on the same inventive concept as the method for predicting quality changes of refrigerated drugs in cold chain transportation provided in Example 1, an embodiment of the present invention also provides a system for predicting quality changes of refrigerated drugs in cold chain transportation, including:

[0095] The cold storage sample analysis module is used to analyze the cold storage samples of the cold storage medicine models and obtain the cold storage reference temperature range and the cold storage temperature fluctuation threshold;

[0096] The refrigerated temperature collection module is used to start the timer to start timing during cold chain transportation. Whenever the timing data meets the preset time length, the timer is reset to zero and starts timing again, and at the same time, the refrigerated temperature timing information is collected;

[0097] A type of refrigerated temperature extraction module, used for counting the refrigerated temperature time series information that satisfies the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold;

[0098] A second type of refrigeration temperature extraction module, used for counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold;

[0099] Three types of refrigeration temperature extraction modules, used for counting the three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold;

[0100] Four types of refrigeration temperature extraction modules, used to count four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and do not meet the refrigeration temperature fluctuation threshold;

[0101] The drug quality prediction module is used to process the first type of refrigeration temperature time series information, the second type of refrigeration temperature time series information, the third type of refrigeration temperature time series information and the fourth type of refrigeration temperature time series information through the drug abnormality binary model of the refrigerated drug model, obtain the drug quality binary identification, and set it as the quality prediction identification of the refrigerated drug model.

[0102] Furthermore, the steps of executing the refrigerated sample analysis module include:

[0103] Collecting non-deteriorated refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record intervals;

[0104] Performing frequent boundary statistics on the refrigerated temperature record interval set to obtain the refrigerated reference temperature interval;

[0105] Collect spoiled refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record time series data, wherein the spoiled refrigerated samples meet the refrigerated reference temperature range, and the refrigerated temperature record time series data is the time series data from the time when the medicine is not spoiled to the time when the deterioration is detected;

[0106] The refrigerated temperature record time series data set is traversed to perform frequent fluctuation statistics to obtain the refrigerated temperature fluctuation threshold.

[0107] Furthermore, the steps of executing the refrigerated sample analysis module include:

[0108] Obtaining an interval upper limit temperature record value set and an interval lower limit temperature record value set of the refrigeration temperature record interval set;

[0109] Traversing the interval upper limit temperature record value set to perform frequency analysis to obtain the interval upper limit temperature record value frequency set;

[0110] According to the interval upper limit temperature record value frequency set, the interval upper limit temperature record values ​​that are less than or equal to the interval upper limit temperature record value frequency threshold are deleted from the interval upper limit temperature record value set to obtain frequent interval upper limit temperature record values, and the maximum value of the frequent interval upper limit temperature record values ​​is extracted and set as the refrigeration reference temperature interval upper limit value;

[0111] Traversing the interval lower limit temperature record value set to perform frequency analysis to obtain the interval lower limit temperature record value frequency set;

[0112] According to the interval lower limit temperature record value frequency set, the interval lower limit temperature record values ​​that are less than or equal to the interval lower limit temperature record value frequency threshold are deleted from the interval lower limit temperature record value set to obtain frequent interval lower limit temperature record values, and the minimum value of the frequent interval lower limit temperature record values ​​is extracted and set as the lower limit value of the refrigeration reference temperature interval;

[0113] The refrigerated reference temperature interval is constructed according to the refrigerated reference temperature interval upper limit value and the refrigerated reference temperature interval lower limit value.

[0114] Furthermore, the steps of executing the refrigerated sample analysis module include:

[0115] Extracting a first interval upper limit temperature record value according to the interval upper limit temperature record value set;

[0116] Count the number of interval upper limit temperature record values ​​in the interval upper limit temperature record value set whose temperature deviation from the first interval upper limit temperature record value is less than or equal to the temperature deviation threshold, calculate the ratio of the number of interval upper limit temperature record values ​​to the total number of interval upper limit temperature record value sets, set it as the first interval upper limit temperature record value frequency, and add it to the interval upper limit temperature record value frequency set.

[0117] Furthermore, the steps of executing the refrigerated sample analysis module include:

[0118] Infinite refrigeration sample collection is performed according to the refrigerated drug model to obtain an infinite refrigerated sample set, wherein any infinite refrigerated sample has a refrigerated drug deterioration mark or a refrigerated drug health mark;

[0119] Calculate the proportion of the infinite refrigerated samples with the refrigerated drug deterioration mark in the infinite refrigerated sample set, and set it as the refrigerated drug temperature fluctuation sensitivity coefficient;

[0120] When the temperature fluctuation sensitivity coefficient of the refrigerated medicine is greater than or equal to the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine, a deteriorated refrigerated sample is collected for the refrigerated medicine model to obtain the refrigerated temperature record time series data set, wherein the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine is greater than or equal to 0.2;

[0121] When the refrigerated drug temperature fluctuation sensitivity coefficient is less than the refrigerated drug temperature fluctuation sensitivity coefficient threshold, the refrigerated temperature fluctuation threshold is configured to be infinite.

[0122] Furthermore, the steps of executing the refrigerated sample analysis module include:

[0123] Extracting first refrigerated temperature record time series data according to the refrigerated temperature record time series data set, calculating the temperature deviation of adjacent moments of the first refrigerated temperature record time series data, and obtaining a first adjacent moment temperature deviation set;

[0124] Traversing the first adjacent time temperature deviation set to perform frequency analysis to obtain the first adjacent time temperature deviation frequency set;

[0125] According to the first adjacent moment temperature deviation frequency set, the first adjacent moment temperature deviation that is less than or equal to the first adjacent moment temperature deviation frequency threshold is deleted from the first adjacent moment temperature deviation set to obtain frequent adjacent moment temperature deviations, and the minimum value of the frequent adjacent moment temperature deviations is extracted and set as the first refrigerated temperature fluctuation threshold, and added to the initial refrigerated temperature fluctuation threshold set;

[0126] The minimum value of the initial refrigeration temperature fluctuation threshold set is extracted and set as the refrigeration temperature fluctuation threshold.

[0127] Furthermore, the drug quality prediction module executes the steps including:

[0128] Obtain the refrigerated humidity monitoring timing information and refrigerated pH monitoring timing information of the refrigerated medicine model;

[0129] Configure the refrigerated humidity reference interval and refrigerated pH reference interval through the user terminal;

[0130] When the proportion of the first moment quantity of the refrigerated humidity monitoring time series information that does not belong to the refrigerated humidity reference interval is greater than or equal to the first ratio threshold, or / and when the proportion of the second moment quantity of the refrigerated pH monitoring time series information that does not belong to the refrigerated pH reference interval is greater than or equal to the second ratio threshold, the quality prediction mark of the refrigerated drug model is set to a quality abnormality mark;

[0131] Otherwise, the first type of refrigerated temperature timing information, the second type of refrigerated temperature timing information, the third type of refrigerated temperature timing information and the fourth type of refrigerated temperature timing information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model, wherein the drug quality binary identification is a normal quality identification or a quality abnormality identification.

[0132] Furthermore, the drug quality prediction module executes the steps including:

[0133] Based on the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold, the first type of refrigerated temperature recording timing information, the second type of refrigerated temperature recording timing information, the third type of refrigerated temperature recording timing information and the fourth type of refrigerated temperature recording timing information are randomly configured in the preset time length;

[0134] Collect the refrigerated sample set of the refrigerated medicine model based on the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information;

[0135] Counting the proportion of samples with abnormal quality in the refrigerated sample set, and setting it as a quality abnormality probability identification value;

[0136] Based on the binary probability threshold, a binary fully connected layer is constructed, wherein when the abnormal quality probability identification value is greater than or equal to the binary probability threshold, the drug quality binary identification is 1, otherwise it is 0, 1 represents the abnormal quality identification, and 0 represents the normal quality identification;

[0137] According to the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information, and the quality abnormality probability identification value, training a quality abnormality probability assessment channel;

[0138] The output node of the quality anomaly probability assessment channel is merged with the input node of the binary fully connected layer to obtain the drug anomaly binary model. Embodiment three:

[0139] See also Figure 3 , Figure 3 Schematic diagram of an electronic device provided by an embodiment of the present invention. Figure 3 As shown, an embodiment of the present invention provides an electronic device 500, including a memory 510, a processor 520, and a first computer program 511 stored in the memory 510 and executable on the processor 520. When the processor 520 executes the first computer program 511, the following steps are implemented:

[0140] Conduct refrigerated sample analysis on refrigerated drug models to obtain the refrigerated reference temperature range and refrigerated temperature fluctuation threshold;

[0141] During cold chain transportation, a timer is started to count, and whenever the timing data meets the preset time length, the timer is reset to zero and starts counting again, and the refrigeration temperature time series information is collected at the same time;

[0142] Counting a type of refrigerated temperature time series information that satisfies the refrigerated temperature reference temperature interval and the refrigerated temperature fluctuation threshold;

[0143] Counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold;

[0144] Counting three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold;

[0145] Counting four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and do not meet the refrigeration temperature fluctuation threshold;

[0146] The first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model. Embodiment 4:

[0147] See also Figure 4 , Figure 4 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. Figure 4 As shown, this embodiment provides a computer-readable storage medium 600, on which a second computer program 611 is stored. When the second computer program 611 is executed by a processor, the following steps are implemented:

[0148] Conduct refrigerated sample analysis on refrigerated drug models to obtain the refrigerated reference temperature range and refrigerated temperature fluctuation threshold;

[0149] During cold chain transportation, a timer is started to count, and whenever the timing data meets the preset time length, the timer is reset to zero and starts counting again, and the refrigeration temperature time series information is collected at the same time;

[0150] Counting a type of refrigerated temperature time series information that satisfies the refrigerated temperature reference temperature interval and the refrigerated temperature fluctuation threshold;

[0151] Counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold;

[0152] Counting three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold;

[0153] Counting four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and do not meet the refrigeration temperature fluctuation threshold;

[0154] The first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model.

[0155] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0156] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0157] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0158] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0159] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0160] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0161] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A method for predicting the quality change of refrigerated drugs during cold chain transportation, characterized in that: include: Conduct refrigerated sample analysis on refrigerated drug models to obtain the refrigerated reference temperature range and refrigerated temperature fluctuation threshold; During cold chain transportation, a timer is started to count, and whenever the timing data meets the preset time length, the timer is reset to zero and starts counting again, and the refrigeration temperature time series information is collected at the same time; Counting a type of refrigerated temperature time series information that satisfies the refrigerated temperature reference temperature interval and the refrigerated temperature fluctuation threshold; Counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold; Counting three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold; Counting four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and do not meet the refrigeration temperature fluctuation threshold; The first type of refrigerated temperature time series information, the second type of refrigerated temperature time series information, the third type of refrigerated temperature time series information and the fourth type of refrigerated temperature time series information are processed by the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model; The steps of constructing the drug abnormality binary model include: Based on the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold, the first type of refrigerated temperature recording timing information, the second type of refrigerated temperature recording timing information, the third type of refrigerated temperature recording timing information and the fourth type of refrigerated temperature recording timing information are randomly configured in the preset time length; Collect the refrigerated sample set of the refrigerated medicine model based on the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information; Counting the proportion of samples with abnormal quality in the refrigerated sample set, and setting it as a quality abnormality probability identification value; Based on the binary probability threshold, a binary fully connected layer is constructed, wherein when the abnormal quality probability identification value is greater than or equal to the binary probability threshold, the drug quality binary identification is 1, otherwise it is 0, 1 represents the abnormal quality identification, and 0 represents the normal quality identification; According to the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information, and the quality abnormality probability identification value, training a quality abnormality probability assessment channel; The output node of the quality anomaly probability assessment channel is merged with the input node of the binary fully connected layer to obtain the drug anomaly binary model.

2. The method according to claim 1, characterized in that Analyze the refrigerated samples of refrigerated medicine models to obtain the refrigerated reference temperature range and refrigerated temperature fluctuation threshold, including: Collecting non-deteriorated refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record intervals; Performing frequent boundary statistics on the refrigerated temperature record interval set to obtain the refrigerated reference temperature interval; Collect spoiled refrigerated samples of the refrigerated medicine model to obtain a set of refrigerated temperature record time series data, wherein the spoiled refrigerated samples meet the refrigerated reference temperature range, and the refrigerated temperature record time series data is the time series data from the time when the medicine is not spoiled to the time when the deterioration is detected; The refrigerated temperature record time series data set is traversed to perform frequent fluctuation statistics to obtain the refrigerated temperature fluctuation threshold.

3. The method according to claim 2, characterized in that Performing frequent boundary statistics on the refrigerated temperature record interval set to obtain the refrigerated reference temperature interval includes: Obtaining an interval upper limit temperature record value set and an interval lower limit temperature record value set of the refrigeration temperature record interval set; Traversing the interval upper limit temperature record value set to perform frequency analysis to obtain the interval upper limit temperature record value frequency set; According to the interval upper limit temperature record value frequency set, the interval upper limit temperature record values ​​that are less than or equal to the interval upper limit temperature record value frequency threshold are deleted from the interval upper limit temperature record value set to obtain frequent interval upper limit temperature record values, and the maximum value of the frequent interval upper limit temperature record values ​​is extracted and set as the refrigeration reference temperature interval upper limit value; Traversing the interval lower limit temperature record value set to perform frequency analysis to obtain the interval lower limit temperature record value frequency set; According to the interval lower limit temperature record value frequency set, the interval lower limit temperature record values ​​that are less than or equal to the interval lower limit temperature record value frequency threshold are deleted from the interval lower limit temperature record value set to obtain frequent interval lower limit temperature record values, and the minimum value of the frequent interval lower limit temperature record values ​​is extracted and set as the lower limit value of the refrigeration reference temperature interval; The refrigerated reference temperature interval is constructed according to the refrigerated reference temperature interval upper limit value and the refrigerated reference temperature interval lower limit value.

4. The method according to claim 3, characterized in that The frequency analysis is performed on the set of upper temperature record values ​​of the interval to obtain a frequency set of upper temperature record values ​​of the interval, including: Extracting a first interval upper limit temperature record value according to the interval upper limit temperature record value set; Count the number of interval upper limit temperature record values ​​in the interval upper limit temperature record value set whose temperature deviation from the first interval upper limit temperature record value is less than or equal to the temperature deviation threshold, calculate the ratio of the number of interval upper limit temperature record values ​​to the total number of interval upper limit temperature record value sets, set it as the first interval upper limit temperature record value frequency, and add it to the interval upper limit temperature record value frequency set.

5. The method according to claim 2, characterized in that The deteriorated refrigerated samples of the refrigerated medicine model are collected to obtain a refrigerated temperature record time series data set, which includes: Infinite refrigeration sample collection is performed according to the refrigerated drug model to obtain an infinite refrigerated sample set, wherein any infinite refrigerated sample has a refrigerated drug deterioration mark or a refrigerated drug health mark; Calculate the proportion of the infinite refrigerated samples with the refrigerated drug deterioration mark in the infinite refrigerated sample set, and set it as the refrigerated drug temperature fluctuation sensitivity coefficient; When the temperature fluctuation sensitivity coefficient of the refrigerated medicine is greater than or equal to the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine, a deteriorated refrigerated sample is collected for the refrigerated medicine model to obtain the refrigerated temperature record time series data set, wherein the temperature fluctuation sensitivity coefficient threshold of the refrigerated medicine is greater than or equal to 0.2; When the refrigerated drug temperature fluctuation sensitivity coefficient is less than the refrigerated drug temperature fluctuation sensitivity coefficient threshold, the refrigerated temperature fluctuation threshold is configured to be infinite.

6. The method according to claim 2, characterized in that Traversing the cold storage temperature record time series data set to perform frequent fluctuation statistics to obtain the cold storage temperature fluctuation threshold, including: Extracting first refrigerated temperature record time series data according to the refrigerated temperature record time series data set, calculating the temperature deviation of adjacent moments of the first refrigerated temperature record time series data, and obtaining a first adjacent moment temperature deviation set; Traversing the first adjacent time temperature deviation set to perform frequency analysis to obtain the first adjacent time temperature deviation frequency set; According to the first adjacent moment temperature deviation frequency set, the first adjacent moment temperature deviation that is less than or equal to the first adjacent moment temperature deviation frequency threshold is deleted from the first adjacent moment temperature deviation set to obtain frequent adjacent moment temperature deviations, and the minimum value of the frequent adjacent moment temperature deviations is extracted and set as the first refrigerated temperature fluctuation threshold, and added to the initial refrigerated temperature fluctuation threshold set; The minimum value of the initial refrigeration temperature fluctuation threshold set is extracted and set as the refrigeration temperature fluctuation threshold.

7. The method according to claim 1, characterized in that The first type of refrigeration temperature time series information, the second type of refrigeration temperature time series information, the third type of refrigeration temperature time series information, and the fourth type of refrigeration temperature time series information are processed by the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary mark, which is set as the quality prediction mark of the refrigerated drug model, including: Obtain the refrigerated humidity monitoring timing information and refrigerated pH monitoring timing information of the refrigerated medicine model; Configure the refrigerated humidity reference interval and refrigerated pH reference interval through the user terminal; When the proportion of the first moment quantity of the refrigerated humidity monitoring time series information that does not belong to the refrigerated humidity reference interval is greater than or equal to the first ratio threshold, or / and when the proportion of the second moment quantity of the refrigerated pH monitoring time series information that does not belong to the refrigerated pH reference interval is greater than or equal to the second ratio threshold, the quality prediction mark of the refrigerated drug model is set to a quality abnormality mark; Otherwise, the first type of refrigerated temperature timing information, the second type of refrigerated temperature timing information, the third type of refrigerated temperature timing information and the fourth type of refrigerated temperature timing information are processed through the drug abnormality binary model of the refrigerated drug model to obtain a drug quality binary identification, which is set as the quality prediction identification of the refrigerated drug model, wherein the drug quality binary identification is a normal quality identification or a quality abnormality identification.

8. The system for predicting the quality change of refrigerated medicines in cold chain transportation is characterized by: include: The cold storage sample analysis module is used to analyze the cold storage samples of the cold storage medicine models and obtain the cold storage reference temperature range and the cold storage temperature fluctuation threshold; The refrigerated temperature collection module is used to start the timer to start timing during cold chain transportation. Whenever the timing data meets the preset time length, the timer is reset to zero and starts timing again, and at the same time, the refrigerated temperature timing information is collected; A type of refrigerated temperature extraction module, used for counting the refrigerated temperature time series information that satisfies the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold; A second type of refrigeration temperature extraction module, used for counting the second type of refrigeration temperature time series information that satisfies the refrigeration reference temperature interval and does not satisfy the refrigeration temperature fluctuation threshold; Three types of refrigeration temperature extraction modules, used for counting the three types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and meet the refrigeration temperature fluctuation threshold; Four types of refrigeration temperature extraction modules, used to count four types of refrigeration temperature time series information that do not meet the refrigeration reference temperature interval and do not meet the refrigeration temperature fluctuation threshold; a drug quality prediction module, for processing the first type of refrigeration temperature time series information, the second type of refrigeration temperature time series information, the third type of refrigeration temperature time series information, and the fourth type of refrigeration temperature time series information through a drug anomaly binary model of the refrigerated drug model, and obtaining a drug quality binary identification, which is set as a quality prediction identification of the refrigerated drug model; The steps of constructing the drug abnormality binary model include: Based on the refrigerated reference temperature interval and the refrigerated temperature fluctuation threshold, the first type of refrigerated temperature recording timing information, the second type of refrigerated temperature recording timing information, the third type of refrigerated temperature recording timing information and the fourth type of refrigerated temperature recording timing information are randomly configured in the preset time length; Collect the refrigerated sample set of the refrigerated medicine model based on the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information; Counting the proportion of samples with abnormal quality in the refrigerated sample set, and setting it as a quality abnormality probability identification value; Based on the binary probability threshold, a binary fully connected layer is constructed, wherein when the abnormal quality probability identification value is greater than or equal to the binary probability threshold, the drug quality binary identification is 1, otherwise it is 0, 1 represents the abnormal quality identification, and 0 represents the normal quality identification; According to the first type of refrigerated temperature record time series information, the second type of refrigerated temperature record time series information, the third type of refrigerated temperature record time series information and the fourth type of refrigerated temperature record time series information, and the quality abnormality probability identification value, training a quality abnormality probability assessment channel; The output node of the quality anomaly probability assessment channel is merged with the input node of the binary fully connected layer to obtain the drug anomaly binary model.

9. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the executable instructions are executed by the processor, the method for predicting the quality change of refrigerated medicines in cold chain transportation as described in any one of claims 1 to 7 is implemented.

Images