Medicine cold chain transportation abnormity identification system and method based on machine learning

Abstract

The invention provides a medicine cold chain transportation abnormity identification system and method based on machine learning, and belongs to the technical field of machine learning, and the system comprises an information obtaining module which is used for analyzing historical collection data according to a predefined index of a cold chain target in the transportation process, and obtaining key information; the model construction module is used for dividing all key information to obtain a training sample and a verification sample, training the initialized neural network model according to the training sample, and verifying and correcting the trained model according to the verification sample to obtain an anomaly recognition model; and the abnormity identification module is used for collecting current information in a cold chain transportation process in real time, inputting the current information into the abnormity identification model to obtain all abnormal factors, and carrying out early warning reminding on each abnormal factor to realize abnormity management and control. The effectiveness of comprehensive analysis of current information and the timeliness of control and early warning are ensured, and abnormities generated in the transportation process are eliminated in time.

Description

Technical Field

[0001] The present invention relates to the technical field of machine learning, and particularly to a system and method for identifying anomalies in the cold chain transportation of pharmaceuticals based on machine learning. Background Art

[0002] In the field of biological cells, cold chain transportation plays a decisive role in ensuring cell viability and the quality of related biological products. Biological cell products, such as CAR-T cell therapy drugs, stem cell preparations, etc., have extremely strict requirements for temperature, time limit, and status during transportation.

[0003] Currently, cold chain transportation data is complex and lacks effective integration and analysis. Each historical cold chain transportation generates a large amount of historical collection data, covering multi-dimensional information such as temperature, humidity, transportation duration, geographical location, etc. However, during the process of anomaly detection, it is based on the comparison of single-dimensional information with the set range, and comprehensive regulation is not carried out based on all factors, resulting in the inability to comprehensively analyze and regulate the currently collected information. That is, the prior art cannot analyze and warn the current information during the cold chain transportation process in a timely and accurate manner.

[0004] Therefore, the present invention proposes a system and method for identifying anomalies in the cold chain transportation of pharmaceuticals based on machine learning. Summary of the Invention

[0005] The present invention provides a system and method for identifying anomalies in the cold chain transportation of pharmaceuticals based on machine learning. By parsing the collected historical data to obtain samples for training and correcting the model, the accuracy of the model is ensured to ensure the effectiveness of comprehensive analysis of the current information and the timeliness of control and warning, and to achieve the timely elimination of anomalies generated during the transportation process.

[0006] The present invention provides a system for identifying anomalies in the cold chain transportation of pharmaceuticals based on machine learning, including: An information acquisition module, configured to acquire the historical collection data of each historical cold chain transportation, and parse the historical collection data according to predefined indicators of the cold chain target during transportation to obtain key information corresponding to the transportation; A model construction module, configured to divide all key information into training samples and validation samples, train the initialized neural network model according to the training samples, and verify and correct the trained model according to the validation samples to obtain an anomaly recognition model; An anomaly recognition module, configured to collect the current information during the cold chain transportation in real time, input it into the anomaly recognition model to obtain all anomaly factors, and give a warning reminder for each anomaly factor to achieve anomaly control.

[0007] Preferably, the historical collected data is multi-dimensional data, including: the transportation temperature, transportation humidity, driving speed of the transportation vehicle, geographical location, and vibration of the cold chain target during transportation at each transportation time point; The number of the predefined indicators is consistent with the dimension, each dimension corresponds to a predefined indicator, and the predefined indicators include: temperature deviation range indicator, humidity qualified interval indicator, transportation duration indicator, and vibration allowable threshold indicator.

[0008] Preferably, the information acquisition module includes: A range comparison unit for analyzing whether the corresponding dimension value at each transportation time point is within the standard range under the corresponding predefined indicator; If the corresponding predefined indicator is given 1, otherwise, give 0 to the corresponding predefined indicator, and construct a transportation rough vector at the corresponding transportation time point; A standardization unit for performing standardization processing on the dimension values under each predefined indicator to obtain standardized values, and constructing a transportation fine vector at the corresponding transportation time point;

[0009] Among them, represents the standardized value of the kth transportation time point under the corresponding predefined indicator, represents the dimension value of the kth transportation time point under the corresponding predefined indicator, represents the mean value of all dimension values under the corresponding predefined indicator, represents the standard deviation of all dimension values under the corresponding predefined indicator; R0 is the reference threshold of the corresponding predefined indicator; ln represents the logarithmic function symbol; 、 are set constants; A control acquisition unit for performing control analysis and alarm analysis on the transportation rough vector and the transportation fine vector to obtain corresponding control information and alarm information; A key information acquisition unit for using the transportation rough vector, the transportation fine vector, the control information, and the alarm information as key information for the corresponding transportation, where the control information includes control instructions and abnormal factors for control adjustment.

[0010] Preferably, the model construction module includes: A coefficient determination unit for determining the difference coefficient of the trained model based on the verification sample;

[0011] Among them, represents the difference coefficient between the actual output and the ideal output of the trained model based on the verification sample, represents the ideal output of the j-th verification sample, represents the actual output of the j-th verification sample, represents the rough transport vector of the j-th verification sample, represents the fine transport vector of the j-th verification sample, represents the deviation influence coefficient based on the rough transport vector, and its value range is (0, 1), represents the deviation influence coefficient based on the fine transport vector, and its value range is (0, 1); represents the j-th verification sample based on the difference function of, and , 、 respectively represent the regularization coefficients; N represents the number of verification samples; The first judgment unit is used to, when the difference coefficient is greater than or equal to the preset coefficient, adopt the adaptive learning rate optimization algorithm and introduce the regularization technology to update the model parameters and continue to train the model; The second judgment unit is used to, when the difference coefficient is less than the preset coefficient, obtain new samples according to the difference coefficient, train the model once and then stop training to obtain the anomaly recognition model.

[0012] Preferably, the second judgment unit includes: An update subunit is used to update the standardized values involved in each verification sample based on the difference coefficient;

[0013] Among them, represents the o-th standardized value in the corresponding verification sample; 、 respectively represent the boundary range values of the o-th standardized value in the corresponding verification sample; represents the learning rate of the trained model; A new sample acquisition unit is used to construct new samples of the corresponding verification samples based on the updated values, train the model once and then stop training.

[0014] Preferably, the first judgment unit includes: An update subunit is used to introduce an L2 regularization term into the loss function, constrain the model parameters, calculate the gradient Ft of the loss function based on all model parameters through the backpropagation algorithm, and update the first moment and the second moment;

[0015] Among them, represents the regularized loss function; L represents the loss function before regularization; represents the regularization parameter; Represents a set of model parameters of the trained model; Represents a model parameter; A quantity determination subunit, configured to perform bias correction on the updated first moment and second moment, and determine the update quantity of the corresponding model parameter, so as to update each model parameter until the model meets a preset stop condition.

[0016] Preferably, the quantity determination subunit includes: A calculation block, configured to calculate the update quantity of the corresponding model parameter:

[0017]

[0018]

[0019] Wherein, Represents the update quantity of the corresponding model parameter after the t-th iteration; Represents the learning rate; , Respectively represent the first moment and the second moment under the t-th iteration; , Represents the attenuation coefficient under the t-th iteration; Represents a constant, and the value is ; , Respectively represent the bias correction results of the first moment and the second moment under the t-th iteration.

[0020] The present invention provides a method for identifying anomalies in pharmaceutical cold chain transportation based on machine learning, including: Step 1: Obtain historical collection data of each historical cold chain transportation, and parse the historical collection data according to predefined indicators during the transportation process of the cold chain target to obtain key information corresponding to the transportation; Step 2: Divide all the key information into training samples and validation samples, train the initialized neural network model according to the training samples, and verify and correct the trained model according to the validation samples to obtain an anomaly identification model; Step 3: Real-time collect current information during the cold chain transportation process, input it into the anomaly identification model to obtain all anomaly factors, and give a warning reminder for each anomaly factor to achieve anomaly control.

[0021] Compared with the prior art, the beneficial effects of the present application are as follows: The historical data collected is parsed to obtain samples for training and correcting the model, ensuring the accuracy of the model to guarantee the effectiveness of comprehensive analysis of current information and the timeliness of control and warning, and achieving the timely elimination of anomalies generated during the transportation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] To more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0023] Figure 1 It is a structural diagram of a pharmaceutical cold chain transportation anomaly recognition system based on machine learning in an embodiment of the present invention; Figure 2 It is a flowchart of a pharmaceutical cold chain transportation anomaly recognition method based on machine learning in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention fall within the scope of protection of the present invention.

[0025] The present invention provides a pharmaceutical cold chain transportation anomaly recognition system based on machine learning, as Figure 1 shown, including: An information acquisition module, configured to acquire historical acquisition data of each historical cold chain transportation, and parse the historical acquisition data according to predefined indicators of the cold chain target during the transportation process to obtain key information corresponding to the transportation; A model construction module, configured to divide all key information into training samples and verification samples, train the initialized neural network model according to the training samples, and verify and correct the trained model according to the verification samples to obtain an anomaly recognition model; An anomaly recognition module, configured to collect current information during the cold chain transportation in real time, input it into the anomaly recognition model to obtain all anomaly factors, and give a warning reminder for each anomaly factor to achieve anomaly control.

[0026] Preferably, the historical collected data is multi-dimensional data, including: the transportation temperature, transportation humidity, driving speed of the transportation vehicle, geographical location, and vibration of the cold chain target during transportation at each transportation time point; The number of the predefined indicators is consistent with the dimension, each dimension corresponds to a predefined indicator, and the predefined indicators include: temperature deviation range indicator, humidity qualified interval indicator, transportation duration indicator, and vibration allowable threshold indicator.

[0027] In this embodiment, the temperature, humidity, driving speed, geographical location, and vibration are respectively measured by a temperature sensor, a humidity sensor arranged inside the cold chain carriage, a speed sensor, a GPS, and a vibration sensor arranged on the cold chain vehicle.

[0028] In this embodiment, for some immune cell products, the predefined indicators may stipulate that the transportation temperature needs to be strictly controlled between 2-8°C, the maximum transportation time cannot exceed 72 hours, and the temperature fluctuation range is within ±1°C, etc.; for stem cell products, it may be required to be transported in a cryogenic environment of -150°C to -196°C, and the duration of the temperature deviation from this range shall not exceed 30 minutes.

[0029] In this embodiment, the key information includes: the rough judgment result, the detailed judgment result, the abnormal factors, and the comprehensive control and early warning result of the collected information corresponding to the predefined range.

[0030] In this embodiment, such as temperature abnormality, transportation overtime, route deviation, etc.

[0031] In this embodiment, the process of dividing the training samples and the validation samples generally follows a certain ratio and principle, usually in a ratio of 7:3 or 8:2.

[0032] In this embodiment, during the training process of the neural network model, by learning the key information in the training samples, the model continuously adjusts its own parameters (such as weights and biases) to adapt to the data characteristics in different transportation situations, and gradually masters the laws of normal transportation and abnormal transportation. For example, the model learns that when the temperature exceeds the predefined range for a certain period of time, it may affect the biological cell activity, so that similar abnormal situations can be identified in subsequent predictions. After training, the validation samples are used to verify the model. The validation samples are input into the trained model to obtain the prediction results of the model. Then, the prediction results are compared with the actual abnormal situations in the validation samples, and evaluation indicators such as the prediction accuracy rate and recall rate of the model are calculated. If there is a large deviation between the prediction results of the model and the actual situation, it indicates that there is a problem with the model and it needs to be corrected. The correction methods include adjusting the model structure, reselecting training parameters, increasing or decreasing training samples, etc. Through continuous adjustment and optimization, the performance of the model is optimized, and finally an abnormal recognition model is obtained.

[0033] Based on the previously learned rules and characteristics, the model analyzes and judges the current information to identify all possible abnormal factors. For example, when the model detects that the current transportation temperature exceeds the predefined index range, "temperature anomaly" is identified as an abnormal factor; if it is found that the transportation duration is about to exceed the specified time, "risk of transportation overtime" is identified as an abnormal factor.

[0034] For each abnormal factor identified by the model, the system will immediately issue a warning reminder. The warning methods can include various forms such as sound alarms, text message notifications, and APP push messages to ensure that transportation management personnel can obtain abnormal information in a timely manner. At the same time, the system will also provide corresponding abnormal control suggestions according to the type and severity of the abnormal factors. For example, for temperature anomalies, it is recommended to immediately adjust the operating parameters of the refrigeration equipment; for the risk of transportation overtime, it is recommended to plan a better transportation route, etc., so as to effectively control the abnormal situation of biological cell cold chain transportation.

[0035] The beneficial effects of the above technical solution are: by analyzing the collected historical data to obtain samples for training and correcting the model, ensuring the accuracy of the model to ensure the effectiveness of the comprehensive analysis of the current information and the timeliness of control and warning, and realizing the timely elimination of abnormalities generated during the transportation process.

[0036] The present invention provides an abnormal recognition system for pharmaceutical cold chain transportation based on machine learning. The information acquisition module includes: A range comparison unit for analyzing whether the corresponding dimension value at each transportation time point is within the standard range under the corresponding predefined index; If the corresponding predefined index is given 1, otherwise, 0 is given to the corresponding predefined index, and a rough transportation vector at the corresponding transportation time point is constructed; A standardization unit for performing standardization processing on the dimension values under each predefined index to obtain the standardized values and constructing a fine transportation vector at the corresponding transportation time point;

[0037] Among them, represents the standardized value of the kth transportation time point under the corresponding predefined index, represents the dimension value of the kth transportation time point under the corresponding predefined index, represents the mean value of all dimension values under the corresponding predefined index, represents the standard deviation of all dimension values under the corresponding predefined index; R0 is the reference threshold of the corresponding predefined index; ln represents the logarithmic function symbol; 、 are set constants; A control acquisition unit, configured to perform control analysis and alarm analysis on the transportation coarse vector and the transportation fine vector to obtain corresponding control information and alarm information; A key information acquisition unit, configured to use the transportation coarse vector, the transportation fine vector, the control information, and the alarm information as key information under corresponding transportation. Wherein, the control information includes control instructions and abnormal factors for control adjustment.

[0038] In this embodiment, the transportation time point is a specific moment when data is recorded at regular time intervals (such as every minute, every five minutes, etc.) during the cold chain transportation of biological cells, and is used to identify the time node of data collection during transportation. The corresponding dimension value is the data value related to transportation collected by various sensors and devices at each transportation time point. If the temperature value collected at a certain transportation time point is 9°C, while the predefined temperature range is 2 - 8°C, it is determined that the temperature value is not within the standard range; if the humidity value collected is 50%, which is within the range of 40% - 60% specified by the predefined index, it is determined that the humidity value is within the standard range. In this way, a comprehensive judgment is made on all dimension values at each transportation time point, providing a basis for subsequent data processing.

[0039] The transportation coarse vector is a vector composed of 0 or 1 values assigned after judging whether the corresponding dimension values at each transportation time point meet the predefined index. It reflects the compliance of each key index during transportation and is a preliminary quantitative description of the transportation state. Suppose there are three predefined indexes of temperature, humidity, and geographical location at a certain transportation time point. After judgment, the temperature value is not within the standard range, and the humidity value and geographical location value are within the standard range. Then the assignments corresponding to the three predefined indexes at this transportation time point are 0, 1, and 1 respectively. Arranging these assignments in the order of the predefined indexes in sequence to construct a vector, which is the transportation coarse vector at this transportation time point.

[0040] The transportation fine vector is a vector constructed by the values obtained after standardizing the dimension values under each predefined index. It retains the original characteristics of the data and unifies the data with different dimensions to the same scale, facilitating more accurate data analysis and comparison. By comparing the dimension values at each transportation time point with the predefined index, 10,000 transportation coarse vectors are successfully constructed. After statistics, among them, all the predefined index assignments in 8,500 transportation coarse vectors are 1, indicating that the transportation state is normal at these transportation time points; there are 1,500 transportation coarse vectors with at least one predefined index assignment of 0, that is, there are abnormal situations.

[0041] Control analysis is to comprehensively analyze the transportation rough vector and the transportation fine vector, judge whether there are abnormal situations during the transportation process, and formulate corresponding control strategies for the abnormal situations to ensure that the transportation process meets the requirements. During the control analysis process, the system will judge whether there are abnormal situations according to the distribution of 0 values in the transportation rough vector and the deviation degree of the values in the transportation fine vector. For example, if the assigned value corresponding to the temperature index in the transportation rough vector is 0, and the standardized value of the temperature in the transportation fine vector exceeds the normal range by a large margin, the system judges that there is a temperature abnormality. At this time, a control instruction will be generated according to the preset strategy, such as "increase the power of the refrigeration equipment by 20%", and the abnormal factor for the control adjustment will be determined as "too high temperature". These information together constitute the control information. During the alarm analysis process, the system will set different alarm levels according to the severity of the abnormal situation. For example, when the temperature is abnormal and the deviation from the standard range is small, it may be set as a low-level alarm, sending an alarm message of "slight temperature abnormality, please pay attention"; when the temperature is abnormal and the deviation from the standard range is large, it may be set as a high-level alarm, sending an alarm message of "serious temperature abnormality, take measures immediately". Through the comprehensive analysis of the transportation rough vector and the transportation fine vector, the effective monitoring and timely early warning of the transportation process are realized.

[0042] In this embodiment, the transportation rough vector, the transportation fine vector, the control information, and the alarm information include important contents such as the key states, detailed data characteristics, strategies for dealing with abnormalities, and early warning prompts during the transportation process. Integrating these information together serves as the key information corresponding to the transportation. These key information can be used for various purposes such as subsequent optimization of the transportation process, tracing of abnormal situations, and model training, providing strong support for ensuring the safety and quality of the cold chain transportation of biological cells. The system accurately identified 1450 abnormal situations (including 800 temperature abnormalities, 400 humidity abnormalities, and 250 geographical location abnormalities), and the identification accuracy rate reached 96.7%. For these abnormal situations, the system generated corresponding control information and alarm information, and the rationality and effectiveness of the control instructions were verified manually and met the actual transportation requirements.

[0043] The beneficial effects of the above technical solution are as follows: The transportation rough vector can intuitively reflect the compliance of each key index at the transportation time point, facilitating a quick understanding of the general situation of the transportation state. The transportation fine vector retains the original characteristics of the data and eliminates the influence of dimensions, making the data in different dimensions comparable and providing the possibility for more accurate data analysis. The determination of the control information and the alarm information is to ensure the rationality and effectiveness of the control instructions obtained after analyzing the current information, and to ensure compliance with the actual transportation requirements.

[0044] The present invention provides an abnormal recognition system for pharmaceutical cold chain transportation based on machine learning. The model construction module includes: A coefficient determination unit, configured to determine a difference coefficient of a trained model based on verification samples;

[0045] wherein, represents a difference coefficient between the actual output and the ideal output of the trained model based on the verification samples, represents the ideal output of the j-th verification sample, represents the actual output of the j-th verification sample, represents the transportation rough vector of the j-th verification sample, represents the transportation fine vector of the j-th verification sample, represents a deviation influence coefficient based on the transportation rough vector, and its value range is (0, 1), represents a deviation influence coefficient based on the transportation fine vector, and its value range is (0, 1); represents the difference function of the j-th verification sample based on , and , , respectively represent regularization coefficients; N represents the number of verification samples; A first judgment unit, configured to, when the difference coefficient is greater than or equal to a preset coefficient, adopt an adaptive learning rate optimization algorithm and introduce a regularization technique to update model parameters and continue to train the model; A second judgment unit, configured to, when the difference coefficient is less than the preset coefficient, obtain new samples according to the difference coefficient, train the model once, and then stop training to obtain an anomaly recognition model.

[0046] In this embodiment, the initial learning rate is set to 0.001. At this time, , take the value of 0.001, select 200 historical cold chain transportation data, and divide them into training samples and verification samples according to a ratio of 7:3. The difference coefficient of the model after initial training is 0.35, and the preset coefficient is set to 0.25. Since the difference coefficient is greater than the preset coefficient, the operation of the first judgment unit is triggered.

[0047] The beneficial effects of the above technical solution are: The model optimization method based on the difference coefficient can reasonably select an optimization strategy according to the performance of the model on the verification samples, effectively improve the performance of the biological cell cold chain transportation anomaly recognition model, and has high feasibility and practical value in practical applications.

[0048] The present invention provides a cold chain transportation anomaly recognition system for medicine based on machine learning. The second judgment unit includes: An update subunit, configured to update the standardized values involved in each verification sample based on the difference coefficient;

[0049] Among them, represents the o-th standardized value in the corresponding verification sample; , respectively represent the boundary range values of the o-th standardized value in the corresponding verification sample; represents the learning rate of the trained model; The new sample acquisition unit is used to construct a new sample of the corresponding verification sample based on the updated values, and stop training after training the model once.

[0050] In this embodiment, based on experiments, it is shown that after training the model once, the comprehensive recognition accuracy of the model for temperature anomalies, humidity anomalies, and geographical location anomalies is 80%, the false negative rate is 12%, and the false positive rate is 8%. After sample update and optimization based on the coefficient of variation, the comprehensive recognition accuracy of the model is increased to 93%, the false negative rate is reduced to 4%, and the false positive rate is reduced to 2%.

[0051] The beneficial effects of the above technical solution are: by updating the samples to ensure the accuracy of the subsequent anomaly recognition of the model, and the model can better adapt to the changes in the data, improving the recognition ability of the abnormal situation of the biological cell cold chain transportation.

[0052] The present invention provides a cold chain transportation anomaly recognition system for pharmaceuticals based on machine learning. The first judgment unit includes: The update subunit is used to introduce an L2 regularization term into the loss function, constrain the model parameters, calculate the gradient Ft of the loss function based on all model parameters through the backpropagation algorithm, and update the first moment and the second moment;

[0053] Among them, represents the regularized loss function; L represents the loss function before regularization; represents the regularization parameter; represents the set of model parameters of the trained model; represents the model parameter; The quantity determination subunit is used to correct the deviation of the updated first moment and second moment, and determine the update quantity of the corresponding model parameter, so as to update each model parameter until the model meets the preset stop condition.

[0054] Preferably, the quantity determination subunit includes: The calculation block is used to calculate the update quantity of the corresponding model parameter:

[0055]

[0056]

[0057] wherein, represents the update amount after the t-th iteration of the corresponding model parameter; represents the learning rate; and respectively represent the first moment and the second moment at the t-th iteration; and represent the decay coefficient at the t-th iteration; represents a constant, and its value is ; and respectively represent the bias correction results for the first moment and the second moment at the t-th iteration.

[0058] In this embodiment, a common training method without introducing L2 regularization and adaptive update strategy is used. The recognition accuracy of the model on the training set reaches 85%, but the accuracy on the validation set is only 72%, showing an obvious overfitting phenomenon. Compared with the common training method, the method in this paper has a faster decreasing loss function and reaches a stable state earlier under the same number of training iterations. For example, at the 100th iteration, the loss function value of the common method is 0.8, while the loss function value of the method in this paper has dropped to 0.5.

[0059] The beneficial effects of the above technical solution are as follows: The model parameter optimization method based on L2 regularization and adaptive update can effectively constrain model parameters, avoid overfitting, and at the same time accelerate model convergence by adaptively adjusting the learning rate, significantly improving the performance and reliability of the biological cell cold chain transportation anomaly recognition model, and having high practical value in practical applications.

[0060] The present invention provides a method for identifying anomalies in pharmaceutical cold chain transportation based on machine learning, including: Step 1: Obtain the historical collection data of each historical cold chain transportation, and analyze the historical collection data according to the predefined indicators during the transportation of the cold chain target to obtain the key information corresponding to the transportation; Step 2: Divide all the key information into training samples and validation samples, train the initialized neural network model according to the training samples, and verify and correct the trained model according to the validation samples to obtain an anomaly recognition model; Step 3: Real-time collect the current information during the cold chain transportation, input it into the anomaly recognition model to obtain all anomaly factors, and give early warning reminders for each anomaly factor to achieve anomaly control.

[0061] The beneficial effects of the above technical solution are as follows: By analyzing the collected historical data to obtain samples for training and correcting the model, the accuracy of the model is ensured to ensure the effectiveness of comprehensive analysis of current information and the timeliness of control and warning, and timely elimination of abnormalities generated during the transportation process is achieved.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A pharmaceutical cold chain transportation anomaly identification system based on machine learning, characterized in that: include: The information acquisition module is used to obtain the historical collected data of each historical cold chain transportation, and analyze the historical collected data according to the predefined indicators of the cold chain target during the transportation process to obtain the key information under the corresponding transportation; A model building module is used to divide all key information into training samples and verification samples, and train the initialized neural network model according to the training samples, and verify and correct the trained model according to the verification samples to obtain an anomaly recognition model; The abnormality recognition module is used to collect the current information of the cold chain transportation process in real time, and input it into the abnormality recognition model to obtain all abnormal factors, and issue early warning reminders for each abnormal factor to achieve abnormality management and control.

2. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 1 is characterized in that: The historically collected data is multi-dimensional data, including: transportation temperature, transportation humidity, driving speed of the transportation vehicle, geographical location, and vibration of the cold chain target during transportation at each transportation time point; The number of the predefined indicators is consistent with the dimensions, each dimension corresponds to a predefined indicator, and the predefined indicators include: a temperature deviation range indicator, a humidity qualified interval indicator, a transportation time indicator, and a vibration allowable threshold indicator.

3. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 1 is characterized in that: The information acquisition module comprises: The range comparison unit is used to analyze whether the corresponding dimension value at each transportation time point is within the standard range of the corresponding predefined indicator; If the corresponding predefined index is assigned 1, otherwise, the corresponding predefined index is assigned 0, and the transportation rough vector at the corresponding transportation time point is constructed; The standardization unit is used to standardize the dimension values ​​under each predefined indicator, and the obtained standardized values ​​are used to construct the transportation detailed vector at the corresponding transportation time point; in, represents the standardized value of the kth transportation time point under the corresponding predefined indicator, Indicates the dimension value of the kth transportation time point under the predefined indicator. Represents the mean value of all dimension values ​​under the corresponding predefined indicator. Represents the standard deviation of all dimension values ​​under the corresponding predefined indicator; R0 corresponds to the reference threshold of the predefined indicator; ln represents the sign of the logarithmic function; , To set a constant; A control acquisition unit, used for performing control analysis and alarm analysis on the transport coarse vector and the transport fine vector to obtain corresponding control information and alarm information; The key information acquisition unit is used to use the transport coarse vector, transport fine vector, control information and alarm information as key information for corresponding transport, wherein the control information includes control instructions and abnormal factors of control adjustment.

4. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 1 is characterized in that: The model building module comprises: A coefficient determination unit, used to determine the difference coefficient of the trained model based on the validation sample; in, It represents the difference coefficient between the actual output and the ideal output of the trained model based on the validation sample. represents the ideal output of the j-th validation sample, represents the actual output of the j-th validation sample, represents the transport rough vector of the jth validation sample, represents the transport refinement vector of the jth validation sample, It represents the deviation influence coefficient based on the transport rough vector, and its value range is (0,1). It represents the deviation influence coefficient based on the transport fine vector, and its value range is (0,1); Indicates that the jth validation sample is based on The difference function of , , Respectively represent the regularization coefficient; N represents the number of validation samples; A first judgment unit is used to adopt an adaptive learning rate optimization algorithm and introduce a regularization technique to update the model parameters and continue to train the model when the difference coefficient is greater than or equal to a preset coefficient; The second judgment unit is used to obtain a new sample according to the difference coefficient to train the model once and then stop training to obtain an abnormality recognition model when the difference coefficient is less than a preset coefficient.

5. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 4 is characterized in that: The second judging unit includes: An updating subunit, used for updating the standardized value involved in each validation sample based on the difference coefficient; in, Indicates the oth standardized value in the corresponding validation sample; , They respectively represent the boundary range values ​​of the oth standardized value in the corresponding validation sample; Represents the learning rate of the trained model; The new sample acquisition unit is used to construct a new sample corresponding to the verification sample based on the updated value, and stop training after training the model once.

6. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 5 is characterized in that: The first judging unit includes: The update subunit is used to introduce the L2 regularization term into the loss function, constrain the model parameters, calculate the gradient Ft of the loss function based on all model parameters through the back propagation algorithm, and update the first-order moment and second-order moment; in, represents the loss function after regularization; L represents the loss function before regularization; represents the regularization parameter; Represents the model parameter set of the trained model; represents the model parameters; The amount determination subunit is used to perform deviation correction on the updated first-order moment and second-order moment, and determine the update amount of the corresponding model parameters to update each model parameter until the model meets the preset stop condition.

7. The machine learning-based pharmaceutical cold chain transportation anomaly identification system according to claim 6 is characterized in that: The quantity determination subunit includes: The calculation block is used to calculate the update amount of the corresponding model parameters: in, Represents the update amount of the corresponding model parameters after the tth iteration; represents the learning rate; , Respectively represent the first-order moment and second-order moment at the t-th iteration; , represents the attenuation coefficient at the tth iteration; Represents a constant, whose value is ; , They represent the deviation correction results of the first-order moment and the second-order moment at the t-th iteration respectively.

8. A method for identifying abnormalities in pharmaceutical cold chain transportation based on machine learning, characterized in that: include: Step 1: Obtain the historical collected data of each historical cold chain transportation, and parse the historical collected data according to the predefined indicators of the cold chain target during the transportation process to obtain the key information under the corresponding transportation; Step 2: Divide all key information into training samples and verification samples, train the initialized neural network model according to the training samples, and verify and correct the trained model according to the verification samples to obtain an anomaly recognition model; Step 3: Collect the current information of the cold chain transportation process in real time, and input it into the abnormality recognition model to obtain all abnormal factors, and issue early warning reminders for each abnormal factor to achieve abnormality management and control.

Images