Remote intelligent monitoring alarm device and method for drying equipment

Abstract

The present invention discloses a remote intelligent monitoring alarm device and method for drying equipment, which relates to the field related to alarm devices. The device includes: an atomization image acquisition module for acquiring atomization images and identifying and obtaining droplet feature distribution; a feature array acquisition module for acquiring hot air feature arrays, temperature arrays and humidity arrays; a drug atomization drying prediction module for predicting drug atomization drying; a first drug drying quality information acquisition module for obtaining first drug drying quality information; an accumulated iterative prediction module for performing accumulated iterative predictions; and an alarm module for judging whether the accumulated prediction time length is greater than the inertia adjustment time length, and if so, no alarm is issued, and if not, an alarm is issued. The present invention solves the technical problem that the alarm information of the existing remote monitoring of drying equipment is not accurate enough and it is difficult to accurately reflect the actual operating status of the drying equipment, and achieves the technical effect of improving the remote monitoring capability of the drying equipment and the accuracy of the alarm.

Description

Technical Field

[0001] The present application relates to the field of alarm devices, and in particular to a remote intelligent monitoring alarm device and method for drying equipment. Background Art

[0002] In the pharmaceutical industry, drug drying is a crucial process step, and its quality directly affects the purity and stability of the final product. As one of the key production equipment, real-time monitoring and timely alarm of the operating status of drying equipment are crucial to ensuring production safety and improving product quality. However, the current drying equipment monitoring and alarm systems are often limited to monitoring a few parameters such as temperature and humidity in the drying room, lacking comprehensive monitoring of the drug drying process, and the existing alarm systems mostly use a simple threshold judgment method, triggering an alarm when the monitoring parameters exceed the preset threshold. This method is difficult to cope with the complex and changeable drying process.

[0003] In the current related technologies, remote monitoring of drying equipment has technical problems such as inaccurate alarm information and difficulty in accurately reflecting the actual operating status of the drying equipment due to the one-sidedness of monitoring data and the limitations of the alarm mechanism. Summary of the invention

[0004] The present application provides a remote intelligent monitoring and alarm device and method for drying equipment, which uses image recognition to monitor the atomization state of the drug in real time, obtain the characteristic distribution of droplets, collect hot air characteristics, temperature and humidity parameters in the drying chamber, construct a multi-dimensional data array, predict the drying distribution and deterioration distribution of the drug based on multivariate data, analyze the drying quality according to the predicted results, and when the drying quality meets the requirements, continue to predict future changes until the predicted results do not meet the requirements, determine the accumulated prediction time, and dynamically generate and correct the alarm information according to the predicted results and adjust the inertia time. Technical means ensure the timeliness and accuracy of the alarm, achieve the technical effect of improving the remote monitoring capability of the drying equipment and the accuracy of the alarm, and ensuring the quality and safety of the drug drying process.

[0005] The present application provides a remote intelligent monitoring and alarm device for a drying device, the device is used for a drying device, the drying device includes an atomizer, a drying chamber, and a hot air module, the device includes:

[0006] Atomization image acquisition module, the atomization image acquisition module is used to remotely acquire the atomization image of the drug after atomization based on the atomization monitor arranged in the atomizer, and identify and obtain the droplet characteristic distribution; a feature array acquisition module, the feature array acquisition module is used to acquire the hot air feature array based on the hot air monitor arranged in the hot air module, and acquire the temperature array and humidity array in the drying chamber; a drug atomization drying prediction module, the drug atomization drying prediction module is used to perform drug atomization drying prediction based on the droplet characteristic distribution, hot air characteristic array, temperature array and humidity array, and obtain drug drying distribution and drug deterioration distribution; a first drug drying quality information acquisition module, the first drug drying quality information acquisition module is used to analyze and obtain the first drug drying distribution and drug deterioration distribution according to the drug drying distribution and drug deterioration distribution. Drug drying quality information; a cumulative iterative prediction module, the cumulative iterative prediction module is used to predict the change of humidity array and temperature array according to the drug drying distribution when the first drug drying quality information meets the requirements, obtain the predicted humidity array and the predicted temperature array, and perform cumulative iterative prediction until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative predicted time length is obtained; an alarm module, the alarm module is used to determine whether the cumulative predicted time length is greater than the adjusted inertia time length, if so, no alarm is issued, if not, a first alarm information is generated according to the cumulative predicted time length and the adjusted inertia time length, and the first alarm information is corrected according to the cumulative predicted drug drying quality information to obtain a second alarm information and an alarm is issued.

[0007] In a possible implementation, the atomization image acquisition module is used to:

[0008] Based on an atomization monitor arranged on the nebulizer, an atomization image after the drug is atomized is collected, wherein the atomization image includes images of multiple atomized drug droplets; a set of sample atomization images is collected, and the size of the atomized drug droplets in each sample atomization image is identified and marked to obtain a sample droplet feature distribution set; the sample atomization image set and the sample droplet feature distribution set are used to train an atomized droplet feature identifier; the atomized droplet feature identifier is used to identify the atomized image to obtain a droplet feature distribution.

[0009] In a possible implementation, the feature array acquisition module is used to:

[0010] Based on the hot air monitor arranged in the hot air module, the hot air characteristic parameters of the hot air module in multiple directions in the drying chamber are collected to obtain a hot air characteristic array, wherein each hot air characteristic parameter includes wind speed and wind temperature; the temperature and humidity of multiple areas in the drying chamber are collected to obtain a temperature array and a humidity array.

[0011] In a possible implementation, the drug atomization drying prediction module is used to:

[0012] According to the historical data of drug drying performed by the drying device, a sample droplet feature distribution set, a sample hot air feature array set, a sample temperature array set and a sample humidity array set are collected, and according to the evaporation drying ratio of atomized drug droplets of different sizes and the deterioration ratio of atomized drug droplets of different sizes after drying, a sample drug drying distribution set and a sample drug deterioration distribution set are annotated and obtained; the sample droplet feature distribution set, the sample hot air feature array set, the sample temperature array set and the sample humidity array set are used as input, and the sample drug drying distribution set and the sample drug deterioration distribution set are used as output to train a drug drying predictor; the drug drying predictor is used to predict drug atomization drying for the droplet feature distribution, hot air feature array, temperature array and humidity array, and obtain drug drying distribution and drug deterioration distribution.

[0013] In a possible implementation, the first drug drying quality information acquisition module is used to:

[0014] According to the drug drying experiment data record, a sample drug drying distribution set and a sample drug deterioration distribution set are collected, and a sample drug drying quality information set is collected, wherein each sample drug drying quality information includes a drug drying quality grade; using the sample drug drying distribution set, the sample drug deterioration distribution set and the sample drug drying quality information set, a drug drying quality classifier is constructed based on a decision tree; using the drug drying quality classifier, the drug drying distribution and the drug deterioration distribution are classified to obtain first drug drying quality information.

[0015] In a possible implementation, the cumulative iterative prediction module is used to:

[0016] According to the historical data of drug drying performed by the drying device, a sample initial temperature array set, a sample initial humidity array set, and a sample drug drying distribution set are collected, and a sample change humidity array set and a sample change temperature array set after the influence of the sample drug drying distribution are obtained; the sample initial temperature array set, the sample initial humidity array set, the sample drug drying distribution set, and the sample change humidity array set and the sample change temperature array set are used to train a temperature and humidity influence predictor; it is determined whether the first drug drying quality information meets the drug drying quality level threshold, and if not, a maximum alarm information is generated; if so, the temperature and humidity influence predictor is used to predict the changes of the drug drying distribution, humidity array, and temperature array to obtain a predicted humidity array and a predicted temperature array; based on the predicted humidity array and the predicted temperature array, the drug atomization drying prediction and the temperature and humidity change prediction are iteratively performed in combination with the droplet feature distribution and the hot air feature array until the cumulative predicted drug drying quality information that does not meet the drug drying quality level threshold is obtained; according to the cumulative iterative prediction times of the cumulative iterative prediction, combined with the unit time of each atomization drug drying, the cumulative prediction time length is calculated.

[0017] In a possible implementation, the alarm module is used to:

[0018] Obtaining the adjustment inertia time length of the drying equipment, wherein the adjustment inertia time length includes the reaction time length for maintenance and adjustment of the drying equipment; judging whether the accumulated predicted time length is greater than the adjustment inertia time length, and if so, not giving an alarm; if not, calculating the ratio of the accumulated predicted time length to the adjustment inertia time length, and classifying to obtain a first alarm information, wherein the first alarm information includes an alarm level; setting an alarm correction coefficient according to a deviation between the accumulated predicted drug drying quality information and a drug drying quality level threshold, wherein the size of the alarm correction coefficient is positively correlated to the size of the deviation; using the alarm correction coefficient, performing a correction calculation on the first alarm information to obtain a second alarm information.

[0019] The present application also provides a remote intelligent monitoring and alarm method for drying equipment, including:

[0020] Based on the atomization monitor arranged on the atomizer, the atomization image of the drug after atomization is remotely collected to identify and obtain the droplet characteristic distribution; based on the hot air monitor arranged on the hot air module, the hot air characteristic array is collected, and the temperature array and humidity array in the drying chamber are collected; based on the droplet characteristic distribution, the hot air characteristic array, the temperature array and the humidity array, the drug atomization drying prediction is performed to obtain the drug drying distribution and the drug deterioration distribution; according to the drug drying distribution and the drug deterioration distribution, the first drug drying quality information is analyzed and obtained; when the first drug drying quality information meets the requirements, according to the drug drying distribution, the humidity array and the temperature array change prediction is performed to obtain the predicted humidity array and the predicted temperature array, and the cumulative iterative prediction is performed until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative predicted time length is obtained; it is judged whether the cumulative predicted time length is greater than the inertia adjustment time length, if so, no alarm is issued, if not, the first alarm information is generated according to the cumulative predicted time length and the inertia adjustment time length, and the first alarm information is corrected according to the cumulative predicted drug drying quality information to obtain the second alarm information, and an alarm is issued.

[0021] It is intended to adopt the remote intelligent monitoring alarm device and method for drying equipment proposed in the present application, remotely collect the atomized image of the drug after atomization based on the atomization monitor arranged in the atomizer through the atomization image acquisition module, identify and obtain the droplet feature distribution, collect the hot air feature array based on the hot air monitor arranged in the hot air module through the feature array acquisition module, and collect the temperature array and humidity array in the drying chamber, and predict the drug atomization drying based on the droplet feature distribution, the hot air feature array, the temperature array and the humidity array through the drug atomization drying prediction module to obtain the drug drying distribution and the drug deterioration distribution, analyze and obtain the first drug drying quality information according to the drug drying distribution and the drug deterioration distribution through the first drug drying quality information acquisition module, and identify and obtain the droplet feature distribution through the feature array acquisition module. When the drug drying quality information meets the requirements, the humidity array and temperature array change prediction is performed according to the drug drying distribution, the predicted humidity array and the predicted temperature array are obtained, and cumulative iterative prediction is performed until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative predicted time length is obtained. The alarm module is used to determine whether the cumulative predicted time length is greater than the adjusted inertia time length. If so, no alarm is issued. If not, a first alarm message is generated according to the cumulative predicted time length and the adjusted inertia time length, and the first alarm message is corrected according to the cumulative predicted drug drying quality information to obtain the second alarm message and issue an alarm, thereby achieving the technical effect of improving the remote monitoring capability and alarm accuracy of the drying equipment and ensuring the quality and safety of the drug drying process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the device according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0023] Figure 1 A schematic diagram of the structure of a remote intelligent monitoring and alarm device for drying equipment provided in an embodiment of the present application.

[0024] Figure 2 A schematic flow chart of a remote intelligent monitoring and alarm method for drying equipment provided in an embodiment of the present application.

[0025] Description of the accompanying drawings: atomization image acquisition module 10, feature array acquisition module 20, drug atomization drying prediction module 30, first drug drying quality information acquisition module 40, cumulative iterative prediction module 50, alarm module 60. DETAILED DESCRIPTION

[0026] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0027] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0028] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, device, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0029] The embodiment of the present application provides a remote intelligent monitoring alarm device for a drying device, the device is used for a drying device, the drying device includes an atomizer, a drying chamber, a hot air module, such as Figure 1 As shown, the device comprises:

[0030] Atomization image acquisition module 10, the atomization image acquisition module 10 is used to remotely collect the atomization image of the drug after atomization based on the atomization monitor arranged in the atomizer, and identify and obtain the characteristic distribution of droplets. Specifically, when the drying device is started or restarted, the atomization image acquisition module 10 is first initialized, including loading the necessary image processing algorithm, setting the parameters of image acquisition, configuring the network connection, etc., wherein the drying device refers to a spray dryer, which is used to spray the liquid material (such as drug solution) into tiny droplets through the atomizer, and contact with hot air, and quickly evaporate the water through heat exchange to achieve rapid drying, and finally obtain dry product particles. Through the atomization monitor (a device for real-time collection of images of drugs after atomization, such as a high-definition camera or other image sensor) arranged near the atomizer, the image of the drug after atomization is captured in real time, and the captured image is preprocessed, including removing noise, adjusting contrast, enhancing edges, etc., to improve image quality. Use image recognition algorithms (such as edge detection, shape recognition, color analysis, etc.) to analyze the pre-processed images and extract the characteristics of the droplets, that is, the physical properties of the atomized drug droplets, including the size (diameter), shape (circularity, aspect ratio, etc.), distribution density (number of droplets per unit area), etc. of the droplets. According to the extracted droplet characteristics, a droplet characteristic distribution map or data table is constructed. This distribution map or data table describes the spatial distribution and characteristic statistical information of the atomized drug droplets in the image, reflecting the quality of the atomization effect. The quality of the atomization effect directly affects the size and distribution uniformity of the drug droplets, and thus affects the subsequent drying process and the final quality and performance of the product.

[0031] In a possible implementation, the atomization image acquisition module 10 is used to: based on an atomization monitor arranged in the nebulizer, collect an atomization image after the drug is atomized, wherein the atomization image includes multiple images of atomized drug droplets; collect a set of sample atomization images, identify and mark the size of the atomized drug droplets in each sample atomization image, and obtain a sample droplet feature distribution set; use the sample atomization image set and the sample droplet feature distribution set to train an atomization droplet feature identifier; use the atomization droplet feature identifier to identify the atomization image and obtain a droplet feature distribution.

[0032] Specifically, when the nebulizer starts working, the atomization image acquisition module 10 automatically starts and connects to the atomization monitor, and the atomization monitor captures the images of the atomized drug in real time. These images contain images of multiple atomized drug droplets, that is, the atomized image is an image captured and formed by the camera of tiny droplets formed by the drug during the atomization process. From a large number of collected atomization images, representative images are selected as samples. These samples cover different atomization conditions, drug types and droplet distributions, and the selected sample images are organized into a sample atomization image set. Professionals use image annotation tools or image processing algorithms to manually or automatically annotate the size and position of the atomized drug droplets in each sample atomization image, extract the features of the droplets from the annotated sample atomization images, such as diameter, area, shape, etc., and construct a sample droplet feature distribution set based on the extracted features. The sample droplet feature distribution set describes the distribution of droplet features under different conditions. Using a set of sample atomized images and a set of sample droplet feature distributions as training data, a machine learning or deep learning algorithm is selected to train the atomized droplet feature identifier. During the training process, the algorithm learns how to identify droplets from the image and extract their features. The performance of the trained atomized droplet feature identifier is evaluated through cross-validation or other evaluation methods. If the performance does not meet the requirements, the algorithm parameters are adjusted or the model is retrained. The trained atomized droplet feature identifier is applied to the atomized image collected in real time, and the droplets in the atomized image are automatically identified and their features are extracted. Based on the identification results, the droplet feature distribution of the current atomized image is constructed. This implementation method uses machine learning or deep learning algorithms to train the atomized droplet feature identifier, which can automatically extract useful feature information from the atomized image, achieving the technical effect of improving the accuracy and efficiency of the droplet feature extraction of the atomized image.

[0033] The feature array acquisition module 20 is used to collect the hot air feature array based on the hot air monitor arranged in the hot air module, and to collect the temperature array and humidity array in the drying chamber. Specifically, when the drying equipment is started, the feature array acquisition module 20 first ensures that the hot air monitor (equipment for real-time monitoring of the hot air state, including anemometers, temperature sensors, etc.) arranged in the hot air module and the sensor for monitoring the temperature and humidity in the drying chamber are in normal working condition. The hot air features generated by the hot air module, such as wind speed, wind direction, temperature distribution, etc., are collected in real time by the hot air monitor, and the collected data are organized into a hot air feature array to record the hot air state at each time point. The temperature sensor is used to collect temperature data in real time at different positions in the drying chamber, and the collected temperature data is organized into a temperature array according to position and time. Similarly, the humidity sensor is used to collect humidity data in real time at different positions in the drying chamber, and the collected humidity data is organized into a humidity array corresponding to the temperature array.

[0034] In a possible implementation, the feature array acquisition module 20 is used to: based on a hot air monitor arranged in the hot air module, collect hot air feature parameters of the hot air module in multiple directions in the drying chamber to obtain a hot air feature array, wherein each hot air feature parameter includes wind speed and wind temperature; collect the temperature and humidity of multiple areas in the drying chamber to obtain a temperature array and a humidity array.

[0035] Specifically, sampling points are set around the hot air module and in multiple directions in the drying room to ensure that the characteristics of the hot air in all directions can be captured. The wind speed is measured in real time at each sampling point using an anemometer and the data is recorded; the wind temperature is measured in real time at each sampling point using a wind temperature sensor and the data is recorded. The collected wind speed and wind temperature data are sorted into a hot air feature array according to the sampling point and time sequence, that is, the hot air feature array is a collection of wind speed and wind temperature data at multiple sampling points and multiple time points, which is used to describe the hot air characteristics of the hot air module in the drying room. Temperature sensors and humidity sensors are installed in multiple areas (such as the top, bottom, middle and surroundings) in the drying room. The temperature is measured in real time in each area using a temperature sensor and the data is recorded; the humidity is measured in real time in each area using a humidity sensor and the data is recorded. The collected temperature and humidity data are sorted into a temperature array and a humidity array according to the area and time sequence, that is, the temperature array is a collection of temperature data at multiple areas and multiple time points, which is used to describe the temperature distribution in the drying room. The humidity array is a collection of humidity data at multiple areas and multiple time points, which is used to describe the humidity distribution in the drying room. This implementation method sets sampling points and sensors in multiple directions and areas around the hot air module and in the drying room to comprehensively capture the hot air characteristics and the environmental conditions in the drying room, ensuring the integrity and accuracy of the data and achieving the technical effect of providing reliable data support for subsequent drug atomization drying predictions.

[0036] The drug atomization drying prediction module 30 is used to predict drug atomization drying based on the droplet characteristic distribution, hot air characteristic array, temperature array and humidity array, and obtain drug drying distribution and drug deterioration distribution. Specifically, the droplet characteristic distribution data is received from the atomization image acquisition module 10, and the hot air characteristic array, temperature array and humidity array data are received from the characteristic array acquisition module 20. The received data is verified to ensure the integrity and accuracy of the data. For abnormal or missing data, it is marked or processed by interpolation, filtering and other methods. Data from different sources are integrated according to time and space dimensions to form a unified data set for analysis.

[0037] Select a prediction model, such as a machine learning model (e.g., random forest, neural network), a physical model, or a hybrid model. Set the model parameters, including learning rate, number of iterations, regularization parameters, etc., to ensure that the model can accurately predict. If a machine learning model is used, use historical data to train the model to optimize the model parameters and improve prediction accuracy.

[0038] The pre-processed droplet feature distribution, hot air feature array, temperature array and humidity array are input into the prediction model as input data. The prediction model simulates and predicts the drug atomization drying process based on the input data, combined with its internal algorithms and parameters, and outputs the drug drying distribution and drug deterioration distribution. Among them, the drug drying distribution describes the distribution of the degree of drying of the drug in the drying room, which is used to reflect the uniformity and effect of drug drying, such as which areas have been dried and which have not been dried; the drug deterioration distribution describes the deterioration of the drug due to overheating or other factors during the drying process, which is used to evaluate the stability and safety of the drug, such as which areas of the drug may have deteriorated.

[0039] In a possible implementation, the drug atomization drying prediction module 30 is used to: collect a sample droplet feature distribution set, a sample hot air feature array set, a sample temperature array set and a sample humidity array set according to historical data of drug drying performed by the drying device, and annotate and obtain a sample drug drying distribution set and a sample drug deterioration distribution set according to the evaporation drying ratio of atomized drug droplets of different sizes and the deterioration ratio of atomized drug droplets of different sizes after drying; use the sample droplet feature distribution set, the sample hot air feature array set, the sample temperature array set and the sample humidity array set as inputs, and use the sample drug drying distribution set and the sample drug deterioration distribution set as outputs to train a drug drying predictor; use the drug drying predictor to predict drug atomization drying on the droplet feature distribution, hot air feature array, temperature array and humidity array to obtain drug drying distribution and drug deterioration distribution.

[0040] Specifically, historical data of drug drying are retrieved from the database of the drying device, which contains detailed records in multiple drying cycles. A sample droplet feature distribution set, a sample hot air feature array set, a sample temperature array set, and a sample humidity array set are extracted from the historical data. These sets represent the characteristics of the atomized drug droplets, the hot air characteristics, the temperature distribution, and the humidity distribution in different drying cycles, respectively. According to the drying results recorded in the historical data, the evaporation drying ratio and the deterioration ratio after drying of atomized drug droplets of different sizes are marked, and the evaporation drying ratio is used as the drug drying distribution, and the deterioration ratio after drying is used as the drug deterioration distribution, so as to obtain a sample drug drying distribution set and a sample drug deterioration distribution set. For example, within an appropriate temperature range, as the temperature increases, the evaporation rate of the atomized drug droplets increases, thereby increasing the drying ratio, but for heat-sensitive drugs, too high a temperature will cause the drug to deteriorate. Therefore, at a higher but not excessive temperature, the smaller the atomized drug droplets, the larger their surface area, the more fully they are in contact with the hot air, and the greater the drying ratio; however, since small droplets are more susceptible to high temperatures, their deterioration ratio will also increase accordingly.

[0041] A machine learning model or a deep learning model is selected as a drug drying predictor, and the input data is preprocessed, such as normalization, standardization or feature selection, to improve the learning efficiency and prediction accuracy of the model. The drug drying predictor is trained using a set of sample droplet feature distributions, a set of sample hot air feature arrays, a set of sample temperature arrays and a set of sample humidity arrays as inputs, and a set of sample drug drying distributions and a set of sample drug deterioration distributions as outputs. During the training process, the model learns the mapping relationship between the input data and the output data. During or after the training process, the model is evaluated using a validation set to check its prediction accuracy and generalization ability. The droplet feature distribution, hot air feature array, temperature array and humidity array in the current drying process are used as input data and input into the trained drug drying predictor. The drug drying predictor predicts drug atomization drying based on the input data and the learned mapping relationship, and outputs the current drug drying distribution and drug deterioration distribution. This implementation method utilizes the historical data of the drying equipment, annotates the drying results in the historical data, and uses supervised learning to train the drug drying predictor, so that the module can learn the complex laws of the drug drying process, achieving the technical effect of improving the accuracy and reliability of drug atomization drying prediction.

[0042] The first drug drying quality information acquisition module 40 is used to analyze and obtain the first drug drying quality information according to the drug drying distribution and drug deterioration distribution. Specifically, based on the preset quality evaluation standards such as the properties of the drug itself, the requirements of the drying process or the industry specifications, the quality evaluation standards are a series of standards or rules used to judge whether the drug drying quality meets the standards. Based on the preset quality evaluation standards, according to the drug drying distribution and drug deterioration distribution, the relevant indicators of drug drying quality are calculated, including drying uniformity, deterioration ratio, effective drying area ratio, etc. The specific calculation method is determined according to statistics, mathematical models or expert experience knowledge, and is intended to quantitatively evaluate the drying quality of the drug. The drying quality is used to characterize the degree of dryness achieved by the drug during the drying process and the comprehensive ability to avoid deterioration. The calculated quality index is compared with the preset quality threshold to evaluate whether the drug drying quality meets the requirements. The evaluation results may include categories such as "qualified", "unqualified" or "need further observation". According to the evaluation results, the first drug drying quality information is generated, and the first drug drying quality information includes detailed results of the quality evaluation, specific indicators that do not meet the requirements and their values, possible suggestions for improvement measures, etc.

[0043] In a possible implementation, the first drug drying quality information acquisition module 40 is used to: collect a sample drug drying distribution set and a sample drug deterioration distribution set according to drug drying experiment data records, and collect a sample drug drying quality information set, wherein each sample drug drying quality information includes a drug drying quality grade; use the sample drug drying distribution set, the sample drug deterioration distribution set and the sample drug drying quality information set to construct a drug drying quality classifier based on a decision tree; use the drug drying quality classifier to classify the drug drying distribution and the drug deterioration distribution to obtain the first drug drying quality information.

[0044] Specifically, the drug drying experiment records the drying process and results of drugs under different conditions (such as temperature, humidity, atomized droplet size, etc.), extracts the drying distribution and deterioration distribution of each sample drug after drying from the experimental data, and assigns a drying quality grade to each sample drug based on the experimental test results, where the drying quality grade is evaluated according to the drying effect, deterioration degree, and drug efficacy retention. The decision tree algorithm is selected to construct a drug drying quality classifier. The decision tree constructs a tree model by recursively dividing the data set into smaller subsets. The collected sample drug drying distribution set, sample drug deterioration distribution set, and sample drug drying quality information set are used as training data to train the decision tree classifier. During the training process, the algorithm learns how to infer the output (drying quality grade) from the input features (drying distribution and deterioration distribution). The real-time drug drying distribution and drug deterioration distribution information are input into the trained drug drying quality classifier, and the corresponding drying quality grade, i.e., the first drug drying quality information, is output. This implementation method combines experimental data and decision tree algorithm to construct a classifier that can accurately predict the quality of drug drying, achieving the technical effect of improving the accuracy and efficiency of drug drying quality assessment.

[0045] The cumulative iterative prediction module 50 is used to predict the change of the humidity array and the temperature array according to the drug drying distribution when the first drug drying quality information meets the requirements, obtain the predicted humidity array and the predicted temperature array, and perform cumulative iterative prediction until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative prediction time length is obtained. Specifically, if the first drug drying quality information meets the preset quality requirements, that is, it is considered that the current drying conditions can produce qualified drug products, then the cumulative iterative prediction module 50 starts the cumulative iterative prediction process. According to the influence of the current drug drying distribution on the temperature and humidity in the drying room, the change of the humidity array and the temperature array in the drying room is predicted, and the preset model or algorithm (such as a physical model, a machine learning model, etc.) is used to predict the humidity and temperature changes in the future period of time based on the initial humidity array and temperature array, as well as the drug drying distribution, that is, the predicted humidity array and the predicted temperature array. Next, using these predicted values, combined with the droplet characteristic distribution and the hot air characteristic array, and utilizing the functions of the drug atomization drying prediction module 30 and the first drug drying quality information acquisition module 40, it is predicted how the drug drying distribution and deterioration distribution will change under the predicted humidity and temperature conditions, as well as the change in the first drug drying quality information. When the predicted drug drying quality information meets the requirements, the result of the above single prediction is used as a new input condition, and the above prediction process is repeated, and each iteration updates the humidity array, temperature array, drug drying distribution, drug deterioration distribution and drug drying quality information. The iterative process continues until the predicted cumulative drug drying quality information no longer meets the preset quality requirements. At this time, it is believed that under the current adjustment strategy, after a certain time point in the future, the drying equipment will no longer be able to continue to produce drugs that meet the quality requirements. The total time length from the initial prediction to the failure to meet the quality requirements is recorded, which is the cumulative prediction time length. This time length represents the estimated time that the drying equipment can continue to produce qualified products under the current conditions.

[0046] In a possible implementation, the cumulative iterative prediction module 50 is used to: collect a sample initial temperature array set, a sample initial humidity array set, and a sample drug drying distribution set according to historical data of drug drying performed by the drying device, and obtain a sample change humidity array set and a sample change temperature array set after the influence of the sample drug drying distribution; use the sample initial temperature array set, the sample initial humidity array set, the sample drug drying distribution set, and the sample change humidity array set and the sample change temperature array set to train a temperature and humidity influence predictor; determine whether the first drug drying quality information meets the drug drying quality standard. If not, the highest alarm information is generated; if so, the temperature and humidity impact predictor is used to predict the changes of the drug drying distribution, humidity array and temperature array to obtain a predicted humidity array and a predicted temperature array; based on the predicted humidity array and the predicted temperature array, the droplet characteristic distribution and the hot air characteristic array are combined to iteratively perform drug atomization drying prediction and temperature and humidity change prediction until the cumulative predicted drug drying quality information that does not meet the drug drying quality level threshold is obtained; according to the cumulative iterative prediction times of the cumulative iterative prediction and the unit time of each atomized drug drying, the cumulative prediction time length is calculated.

[0047] Specifically, historical data of drug drying by the drying device is collected, and a sample initial temperature array set, a sample initial humidity array set, a sample drug drying distribution set, a sample change humidity array set, and a sample change temperature array set are collected. Using the above collected data (initial and change temperature and humidity, drug drying distribution) as input, a model is trained through machine learning or statistical methods, which can predict how the temperature and humidity change during the drying process under given drug drying distribution and initial temperature and humidity conditions. Before starting the cumulative iterative prediction, it is first determined based on the current conditions whether the first drug drying quality information meets the preset drug drying quality level threshold. If not, it is considered that the current drying conditions have produced unqualified drug products, and the highest alarm information is immediately generated, and an instruction is sent to the control system of the drying device, requiring it to immediately stop the current drying process to prevent continued drying from causing further deterioration or complete deterioration of the drug quality. If the first drug drying quality information meets the requirements, the trained temperature and humidity impact predictor is used for iterative prediction to predict the changes in temperature and humidity during the drug drying process, and obtain the predicted humidity array and the predicted temperature array. Based on these predicted values, combined with the droplet characteristic distribution (particle size and distribution density of the atomized drug, etc.) and the hot air characteristic array (wind speed, wind temperature distribution, etc.), the drug atomization drying effect is predicted, and the above prediction is iterated. Each iteration is updated based on the previous result until it is predicted that after a certain time point, the drug drying quality will not meet the set drug drying quality level threshold, that is, at a certain time point in the future, the drying quality will begin to decline. The number of cumulative iterative predictions is recorded, and each iteration corresponds to a unit time (such as minutes, hours, etc.). The cumulative number of iterative predictions is multiplied by the unit time of atomized drug drying in each iteration (such as the fixed time interval set by the drying equipment) to obtain the total time required from the beginning to the prediction that the drug drying quality level threshold is not met. This implementation method determines whether a prediction is needed based on the first drug drying quality information before performing cumulative iterative prediction, thereby avoiding unnecessary waste of computing resources. By issuing the highest alarm information when the first drug drying quality information does not meet the drug drying quality level threshold, it achieves the technical effect of ensuring that relevant personnel can be notified immediately when serious problems occur, thereby reducing losses.

[0048] Alarm module 60, the alarm module 60 is used to determine whether the cumulative predicted time length is greater than the adjusted inertia time length. If so, no alarm is issued. If not, a first alarm message is generated based on the cumulative predicted time length and the adjusted inertia time length, and the first alarm message is corrected based on the cumulative predicted drug drying quality information to obtain a second alarm message and issue an alarm. Specifically, the preset adjusted inertia time length is obtained. The adjusted inertia time length is a time threshold set according to equipment characteristics, operating experience or process requirements, and refers to the time required from receiving the alarm message to adjusting the drying parameters and taking effect. It is used to evaluate whether the operator has enough time to intervene and adjust the impending problem. Compare the cumulative prediction time length with the adjustment inertia time length. If the cumulative prediction time length is greater than the adjustment inertia time length, it means that the operator has enough time to respond and adjust, so there is no need to generate an alarm message immediately, because the operator can take corresponding measures before the problem actually occurs. In the actual production or operation process, the drying equipment may encounter various small fluctuations or abnormal situations. If each small fluctuation triggers an alarm, it will interrupt the normal operation process and reduce the overall production efficiency. When the cumulative prediction time length is greater than the adjustment inertia time length, no alarm is issued, which can allow the system to self-adjust and recover to a certain extent, thereby maintaining the continuity and stability of the operation process; if the cumulative prediction time length is not greater than the adjustment inertia time length, it means that the situation is urgent and an alarm message needs to be generated immediately to remind the operator. When it is determined that an alarm message needs to be generated, a preliminary first alarm message is generated according to the difference between the cumulative prediction time length and the adjustment inertia time length, as well as a preset alarm template or rule, including the expected time of the problem, the type of problem (such as substandard drying quality), and the degree of urgency. The accumulated predicted drug drying quality information provided by the accumulated iterative prediction module 50 is obtained, and the information describes in detail the decline in the predicted drug drying quality and the possible reasons. Based on this information, the first alarm information is corrected and supplemented to make the alarm information more specific and accurate, which helps the operator to quickly locate the problem and take corresponding measures. The corrected second alarm information is sent to relevant personnel, such as operators, maintenance personnel or management personnel, through a preset alarm channel (such as sound and light alarm, SMS notification, email notification, etc.), so that they can quickly respond and deal with the upcoming problem.The embodiment of the present application adopts technical means such as real-time monitoring of the drug atomization state through image recognition, obtaining droplet characteristic distribution, collecting hot air characteristics, temperature and humidity parameters in the drying chamber, constructing a multi-dimensional data array, predicting the drug drying distribution and deterioration distribution based on multivariate data, analyzing the drying quality according to the predicted results, and continuing to predict future changes when the drying quality meets the requirements until the predicted results do not meet the requirements, determining the accumulated prediction time, dynamically generating and correcting the alarm information according to the predicted results and adjusting the inertia time, thereby ensuring the timeliness and accuracy of the alarm, and achieving the technical effect of improving the remote monitoring capability of the drying equipment and the alarm accuracy, and ensuring the quality and safety of the drug drying process.

[0049] In a possible implementation, the alarm module 60 is used to: obtain the adjustment inertia time length of the drying equipment, wherein the adjustment inertia time length includes the reaction time length for maintenance and adjustment of the drying equipment; determine whether the cumulative predicted time length is greater than the adjustment inertia time length, and if so, do not issue an alarm; if not, calculate the ratio of the cumulative predicted time length to the adjustment inertia time length, and classify to obtain a first alarm message, wherein the first alarm message includes an alarm level; set an alarm correction coefficient according to the deviation between the cumulative predicted drug drying quality information and the drug drying quality level threshold, wherein the size of the alarm correction coefficient is positively correlated with the size of the deviation; use the alarm correction coefficient to correct the first alarm message and obtain a second alarm message.

[0050] Specifically, the adjustment inertia time length is obtained from the control system or preset parameters of the drying equipment. The adjustment inertia time length refers to the total time required from the discovery of the problem to the completion of the adjustment, including the time for all necessary steps such as identifying the problem, preparing tools, and performing the adjustment. The cumulative predicted time length is compared with the adjustment inertia time length. If the cumulative predicted time length is greater than the adjustment inertia time length, it means that the operator has enough time to intervene before the problem occurs, so no alarm is issued; if the cumulative predicted time length is not greater than the adjustment inertia time length, it means that the situation is more urgent. At this time, the ratio of the cumulative predicted time length to the adjustment inertia time length is calculated, and the alarm information is classified into different levels according to this ratio, that is, a series of ratio thresholds are set, each threshold corresponds to an alarm level (such as low, medium, and high), and according to the calculated ratio, it is classified into the corresponding alarm level, and a first alarm information is generated, and the first alarm information at least includes the alarm level.

[0051] According to the deviation between the accumulated predicted drug drying quality information and the preset drug drying quality level threshold, an alarm correction coefficient is set. The larger the deviation, the worse the predicted drug drying quality. Therefore, the larger the alarm correction coefficient is set to reflect the urgency and importance, that is, the size of the correction coefficient is positively correlated with the size of the deviation. The first alarm information is corrected and calculated using the alarm correction coefficient to obtain the second alarm information. The correction includes adjusting the alarm level, adding a specific description of the quality deviation, adjusting the urgency or priority of the alarm information, etc. The purpose is to make the alarm information more accurate and detailed, which helps operators to quickly understand the problem and take appropriate measures. This implementation method obtains the alarm information according to the ratio and corrects it according to the deviation, which improves the accuracy and timeliness of the alarm system, enables it to more accurately reflect the actual operation of the drying equipment, and achieves the technical effect of providing operators with more targeted and instructive alarm information.

[0052] In the above, refer to Figure 1 The remote intelligent monitoring alarm device for drying equipment according to the embodiment of the present invention is described in detail. Figure 2 A remote intelligent monitoring and alarm method for drying equipment according to an embodiment of the present invention is described.

[0053] The remote intelligent monitoring and alarm method for drying equipment according to the embodiment of the present invention is used to solve the technical problem that the alarm information is not accurate enough and it is difficult to accurately reflect the actual operating status of the drying equipment due to the one-sidedness of the monitoring data and the limitations of the alarm mechanism in the existing remote monitoring of the drying equipment, so as to improve the remote monitoring capability and alarm accuracy of the drying equipment and ensure the quality and safety of the drug drying process. The remote intelligent monitoring and alarm method for drying equipment includes:

[0054] Based on the atomization monitor arranged on the atomizer, the atomization image of the drug after atomization is remotely collected to identify and obtain the droplet characteristic distribution; based on the hot air monitor arranged on the hot air module, the hot air characteristic array is collected, and the temperature array and humidity array in the drying chamber are collected; based on the droplet characteristic distribution, the hot air characteristic array, the temperature array and the humidity array, the drug atomization drying prediction is performed to obtain the drug drying distribution and the drug deterioration distribution; according to the drug drying distribution and the drug deterioration distribution, the first drug drying quality information is analyzed and obtained; when the first drug drying quality information meets the requirements, according to the drug drying distribution, the humidity array and the temperature array change prediction is performed to obtain the predicted humidity array and the predicted temperature array, and the cumulative iterative prediction is performed until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative predicted time length is obtained; it is judged whether the cumulative predicted time length is greater than the inertia adjustment time length, if so, no alarm is issued, if not, the first alarm information is generated according to the cumulative predicted time length and the inertia adjustment time length, and the first alarm information is corrected according to the cumulative predicted drug drying quality information to obtain the second alarm information, and an alarm is issued.

[0055] In a possible implementation, based on the atomization monitor arranged on the atomizer, the atomization image of the drug after atomization is remotely collected, the characteristic distribution of the droplets is identified, and the following processing is performed:

[0056] Based on an atomization monitor arranged on the nebulizer, an atomization image after the drug is atomized is collected, wherein the atomization image includes images of multiple atomized drug droplets; a set of sample atomization images is collected, and the size of the atomized drug droplets in each sample atomization image is identified and marked to obtain a sample droplet feature distribution set; the sample atomization image set and the sample droplet feature distribution set are used to train an atomized droplet feature identifier; the atomized droplet feature identifier is used to identify the atomized image to obtain a droplet feature distribution.

[0057] In a possible implementation, based on a hot air monitor arranged in the hot air module, a hot air feature array is collected, as well as a temperature array and a humidity array in the drying chamber, and the following processing is performed:

[0058] Based on the hot air monitor arranged in the hot air module, the hot air characteristic parameters of the hot air module in multiple directions in the drying chamber are collected to obtain a hot air characteristic array, wherein each hot air characteristic parameter includes wind speed and wind temperature; the temperature and humidity of multiple areas in the drying chamber are collected to obtain a temperature array and a humidity array.

[0059] In a possible implementation, based on the droplet characteristic distribution, the hot air characteristic array, the temperature array and the humidity array, the drug atomization drying prediction is performed to obtain the drug drying distribution and the drug deterioration distribution, and the following processing is performed:

[0060] According to the historical data of drug drying performed by the drying device, a sample droplet feature distribution set, a sample hot air feature array set, a sample temperature array set and a sample humidity array set are collected, and according to the evaporation drying ratio of atomized drug droplets of different sizes and the deterioration ratio of atomized drug droplets of different sizes after drying, a sample drug drying distribution set and a sample drug deterioration distribution set are annotated and obtained; the sample droplet feature distribution set, the sample hot air feature array set, the sample temperature array set and the sample humidity array set are used as input, and the sample drug drying distribution set and the sample drug deterioration distribution set are used as output to train a drug drying predictor; the drug drying predictor is used to predict drug atomization drying for the droplet feature distribution, hot air feature array, temperature array and humidity array, and obtain drug drying distribution and drug deterioration distribution.

[0061] In a possible implementation, according to the drug drying distribution and the drug deterioration distribution, the first drug drying quality information is analyzed and obtained, and the following processing is performed:

[0062] According to the drug drying experiment data record, a sample drug drying distribution set and a sample drug deterioration distribution set are collected, and a sample drug drying quality information set is collected, wherein each sample drug drying quality information includes a drug drying quality grade; using the sample drug drying distribution set, the sample drug deterioration distribution set and the sample drug drying quality information set, a drug drying quality classifier is constructed based on a decision tree; using the drug drying quality classifier, the drug drying distribution and the drug deterioration distribution are classified to obtain first drug drying quality information.

[0063] In a possible implementation, when the first drug drying quality information meets the requirements, a humidity array and a temperature array change prediction is performed according to the drug drying distribution, a predicted humidity array and a predicted temperature array are obtained, and cumulative iterative prediction is performed until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative prediction time length is obtained, and the following processing is performed:

[0064] According to the historical data of drug drying performed by the drying device, a sample initial temperature array set, a sample initial humidity array set, and a sample drug drying distribution set are collected, and a sample change humidity array set and a sample change temperature array set after the influence of the sample drug drying distribution are obtained; the sample initial temperature array set, the sample initial humidity array set, the sample drug drying distribution set, and the sample change humidity array set and the sample change temperature array set are used to train a temperature and humidity influence predictor; it is determined whether the first drug drying quality information meets the drug drying quality level threshold, and if not, a maximum alarm information is generated; if so, the temperature and humidity influence predictor is used to predict the changes of the drug drying distribution, humidity array, and temperature array to obtain a predicted humidity array and a predicted temperature array; based on the predicted humidity array and the predicted temperature array, the drug atomization drying prediction and the temperature and humidity change prediction are iteratively performed in combination with the droplet feature distribution and the hot air feature array until the cumulative predicted drug drying quality information that does not meet the drug drying quality level threshold is obtained; according to the cumulative iterative prediction times of the cumulative iterative prediction, combined with the unit time of each atomization drug drying, the cumulative prediction time length is calculated.

[0065] In a possible implementation, it is determined whether the accumulated predicted time length is greater than the adjusted inertia time length. If so, no alarm is issued. If not, a first alarm message is generated according to the accumulated predicted time length and the adjusted inertia time length, and the first alarm message is corrected according to the accumulated predicted drug drying mass information to obtain a second alarm message, and the following processing is performed:

[0066] Obtaining the adjustment inertia time length of the drying equipment, wherein the adjustment inertia time length includes the reaction time length for maintenance and adjustment of the drying equipment; judging whether the accumulated predicted time length is greater than the adjustment inertia time length, and if so, not giving an alarm; if not, calculating the ratio of the accumulated predicted time length to the adjustment inertia time length, and classifying to obtain a first alarm information, wherein the first alarm information includes an alarm level; setting an alarm correction coefficient according to a deviation between the accumulated predicted drug drying quality information and a drug drying quality level threshold, wherein the size of the alarm correction coefficient is positively correlated to the size of the deviation; using the alarm correction coefficient, performing a correction calculation on the first alarm information to obtain a second alarm information.

[0067] The remote intelligent monitoring and alarm device for drying equipment provided by the embodiment of the present invention can execute the remote intelligent monitoring and alarm method for drying equipment provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0068] Although the present application makes various references to certain modules in the device according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0069] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A remote intelligent monitoring and alarm device for drying equipment, characterized in that: The device is used for a drying device, the drying device includes an atomizer, a drying chamber, and a hot air module, and the device includes: Atomization image acquisition module, which is used to remotely acquire atomization images of the drug after atomization based on an atomization monitor arranged on the atomizer, and identify and obtain droplet characteristic distribution; A feature array acquisition module, the feature array acquisition module is used to collect a hot air feature array based on a hot air monitor arranged in the hot air module, and to collect a temperature array and a humidity array in the drying chamber; A drug atomization drying prediction module, the drug atomization drying prediction module is used to predict drug atomization drying based on the droplet characteristic distribution, hot air characteristic array, temperature array and humidity array, and obtain drug drying distribution and drug deterioration distribution; a first drug drying quality information acquisition module, the first drug drying quality information acquisition module being used to analyze and obtain first drug drying quality information according to the drug drying distribution and the drug deterioration distribution; A cumulative iterative prediction module, wherein the cumulative iterative prediction module is used to predict the change of the humidity array and the temperature array according to the drug drying distribution when the first drug drying quality information meets the requirements, obtain the predicted humidity array and the predicted temperature array, and perform cumulative iterative prediction until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative prediction time length is obtained; An alarm module, the alarm module is used to determine whether the cumulative predicted time length is greater than the adjusted inertia time length, if so, no alarm is issued, if not, a first alarm message is generated according to the cumulative predicted time length and the adjusted inertia time length, and the first alarm message is corrected according to the cumulative predicted drug drying mass information to obtain a second alarm message and issue an alarm; The inertia adjustment time length includes the reaction time length for maintenance adjustment of the drying equipment; Determine whether the accumulated predicted time length is greater than the adjusted inertia time length, and if so, do not issue an alarm; If not, then calculating the ratio of the accumulated predicted time length to the adjusted inertia time length, and classifying to obtain first alarm information, wherein the first alarm information includes an alarm level; According to the deviation between the accumulated predicted drug drying quality information and the drug drying quality level threshold, an alarm correction coefficient is set, wherein the magnitude of the alarm correction coefficient is positively correlated with the magnitude of the deviation; The alarm correction coefficient is used to perform correction calculation on the first alarm information to obtain second alarm information.

2. The remote intelligent monitoring and alarm device for drying equipment according to claim 1 is characterized in that: The atomization image acquisition module is used for: Based on the atomization monitor arranged on the atomizer, an atomization image of the drug after atomization is collected, wherein the atomization image includes images of multiple atomized drug droplets; Collecting a set of sample atomized images, identifying and marking the size of atomized drug droplets in each sample atomized image, and obtaining a set of sample droplet characteristic distributions; Using the sample atomization image set and the sample droplet feature distribution set to train an atomization droplet feature identifier; The atomized droplet feature identifier is used to identify the atomized image to obtain droplet feature distribution.

3. The remote intelligent monitoring and alarm device for drying equipment according to claim 1 is characterized in that: The feature array acquisition module is used for: Based on a hot air monitor arranged on the hot air module, hot air characteristic parameters of the hot air module in multiple directions in the drying chamber are collected to obtain a hot air characteristic array, wherein each hot air characteristic parameter includes wind speed and wind temperature; The temperature and humidity of multiple areas in the drying chamber are collected to obtain a temperature array and a humidity array.

4. The remote intelligent monitoring and alarm device for drying equipment according to claim 1 is characterized in that: The drug atomization drying prediction module is used for: According to the historical data of drug drying by the drying device, a sample droplet feature distribution set, a sample hot air feature array set, a sample temperature array set and a sample humidity array set are collected, and according to the evaporation drying ratio of atomized drug droplets of different sizes and the ratio of atomized drug droplets of different sizes deteriorating after drying, a sample drug drying distribution set and a sample drug deterioration distribution set are marked and obtained; The sample droplet feature distribution set, the sample hot air feature array set, the sample temperature array set and the sample humidity array set are used as inputs, and the sample drug drying distribution set and the sample drug deterioration distribution set are used as outputs to train a drug drying predictor; The drug drying predictor is used to predict drug atomization drying based on the droplet characteristic distribution, hot air characteristic array, temperature array and humidity array to obtain drug drying distribution and drug deterioration distribution.

5. The remote intelligent monitoring and alarm device for drying equipment according to claim 1 is characterized in that: The first drug drying quality information acquisition module is used for: According to the drug drying experiment data records, a sample drug drying distribution set and a sample drug deterioration distribution set are collected, and a sample drug drying quality information set is collected, wherein each sample drug drying quality information includes a drug drying quality grade; Using the sample drug drying distribution set, the sample drug deterioration distribution set and the sample drug drying quality information set, based on a decision tree, a drug drying quality classifier is constructed; The drug drying quality classifier is used to classify the drug drying distribution and the drug deterioration distribution to obtain first drug drying quality information.

6. The remote intelligent monitoring and alarm device for drying equipment according to claim 1 is characterized in that: The cumulative iterative prediction module is used for: According to the historical data of drug drying performed by the drying device, a sample initial temperature array set, a sample initial humidity array set, and a sample drug drying distribution set are collected, and a sample change humidity array set and a sample change temperature array set after the influence of the sample drug drying distribution are obtained; Using the sample initial temperature array set, the sample initial humidity array set, the sample drug drying distribution set, and the sample change humidity array set and the sample change temperature array set, to train a temperature and humidity influence predictor; Determining whether the first drug drying quality information meets the drug drying quality level threshold, and if not, generating the highest alarm information; If yes, the temperature and humidity impact predictor is used to predict the changes of the drug drying distribution, humidity array and temperature array to obtain a predicted humidity array and a predicted temperature array; Based on the predicted humidity array and the predicted temperature array, in combination with the droplet characteristic distribution and the hot air characteristic array, iteratively performs drug atomization drying prediction and temperature and humidity change prediction until cumulative predicted drug drying quality information that does not meet the drug drying quality level threshold is obtained; The cumulative prediction time length is calculated based on the cumulative iterative prediction times of the cumulative iterative prediction and the unit time of each atomized drug drying.

7. A remote intelligent monitoring and alarm method for drying equipment, characterized in that: The method is implemented by the remote intelligent monitoring and alarm device for drying equipment according to any one of claims 1 to 6, and the method comprises: Based on the atomization monitor installed in the nebulizer, the atomization image of the drug after atomization is remotely collected to identify and obtain the characteristic distribution of droplets; Based on the hot air monitor arranged in the hot air module, the hot air feature array is collected, as well as the temperature array and humidity array in the drying chamber; Based on the droplet characteristic distribution, hot air characteristic array, temperature array and humidity array, a drug atomization drying prediction is performed to obtain drug drying distribution and drug deterioration distribution; Analyze and obtain first drug drying quality information according to the drug drying distribution and drug deterioration distribution; When the first drug drying quality information meets the requirements, according to the drug drying distribution, the humidity array and the temperature array change prediction are performed to obtain the predicted humidity array and the predicted temperature array, and cumulative iterative prediction is performed until the cumulative predicted drug drying quality information that does not meet the requirements is obtained, and the cumulative prediction time length is obtained; Determine whether the accumulated predicted time length is greater than the adjusted inertia time length. If so, no alarm is issued. If not, generate a first alarm message based on the accumulated predicted time length and the adjusted inertia time length, and modify the first alarm message based on the accumulated predicted drug drying mass information to obtain a second alarm message and issue an alarm.

Images