Intelligent forming system of enteric capsule based on computer real-time monitoring control

Abstract

The present application relates to the general control system technical field, specifically relates to the enteric capsule intelligent forming system based on computer real-time monitoring control, the system carries out pellet detection and tracking in the video data in the diffusion chamber in each coating process, obtains pellet detection tracking result, wherein the coating thickness of the pellet is controllable, according to the pellet detection tracking result in real-time coating process, in combination with the temperature and humidity data in the fluidized bed diffusion chamber at each moment, the wind pressure of real-time coating process is determined to be high intensity, in combination with the difference between the pellet detection tracking result in real-time coating process and all historical coating processes, the wind pressure demand index of real-time coating process is determined, finally, in combination with the pressure data of the flow guide cylinder distribution plate in all coating processes at all times, the wind pressure control is carried out on the next coating process in the future.The present application improves the real-time monitoring control efficiency of computer on enteric capsule forming process.

Description

Technical Field

[0001] This invention relates to the field of general control system technology, and more specifically to an intelligent forming system for enteric-coated capsules based on real-time computer monitoring and control. Background Technology

[0002] Enteric-coated capsules are capsules manufactured using a special process. Their outer layer is coated with an enteric coating, ensuring that the drug does not dissolve in the acidic environment of the stomach but only disintegrates and releases its contents in the alkaline environment of the intestine, thus maximizing its efficacy. Enteric microcapsule coating technology is a crucial process in the preparation of enteric-coated capsules, and the precision of the coating process directly affects the drug's solubility.

[0003] Traditionally, fluidized bed bottom spray microcapsule coating technology is used for enteric microcapsule coating. In the microcapsule coating process, microcapsules made of pharmaceutical materials undergo reciprocating fluidized motion in a fluidized bed. During each fluidized motion, the microcapsules are mixed with the coating liquid sprayed from the bottom nozzle. Through multi-process fluidized motion, a coating film is formed on the surface of the enteric microcapsules, thus achieving microcapsule coating. The entire coating process is monitored and controlled by a computer in real time.

[0004] Existing problems: In traditional computer-monitored and controlled fluidized bed bottom spray microcapsule coating machines, the fluidized bed fan pressure is often set to a constant value. However, in actual scenarios, there are differences in microcapsule particle size and morphology, as well as variations in atomization pressure, temperature, and humidity within the fluidized bed diffusion chamber. This results in different fan pressure requirements for each fluidization process. If the fan pressure is too high, the microcapsules may fluidize too quickly, leading to a shorter contact time with the coating liquid and insufficient drug application. If the fan pressure is too low, it may cause increased adhesion between microcapsules and uneven coating. Therefore, the traditional coating method using constant fan pressure is difficult to adapt to the coating requirements of enteric-coated microcapsules, thereby reducing the efficiency of computer monitoring and control of the intelligent molding process of enteric capsules. Summary of the Invention

[0005] This invention provides an intelligent molding system for enteric-coated capsules based on real-time computer monitoring and control to solve existing problems.

[0006] The intelligent enteric-coated capsule forming system based on real-time computer monitoring and control of the present invention adopts the following technical solution:

[0007] One embodiment of the present invention provides an intelligent forming system for enteric-coated capsules based on real-time computer monitoring and control. The system includes the following modules:

[0008] Enteric-coated microsphere coating data acquisition module: used to acquire video data in the diffusion chamber during each coating process of fluidized bed enteric-coated microsphere coating, as well as temperature and humidity data in the fluidized bed diffusion chamber and pressure data of the guide tube distribution plate at each moment; the coating process includes one real-time coating process and several historical coating processes; microsphere detection and tracking are performed on the video data in the diffusion chamber during each coating process to obtain microsphere detection and tracking results;

[0009] Wind pressure rise intensity analysis module: used to determine the wind pressure rise intensity of the real-time coating process based on the microparticle detection and tracking results during the real-time coating process, combined with the temperature and humidity data in the fluidized bed diffusion chamber at each moment;

[0010] Wind pressure demand index analysis module: used to determine the wind pressure demand index of the real-time coating process based on the difference between the real-time coating process and the microparticle detection and tracking results in all historical coating processes, combined with the wind pressure increasing intensity.

[0011] The air pressure control module is used to control the air pressure of the next coating process based on the real-time air pressure demand index of the coating process and the pressure data of the guide tube distribution plate at all times in all coating processes.

[0012] Furthermore, the determination of the wind pressure increase intensity during the real-time coating process includes:

[0013] The microparticle detection and tracking results include: the maximum fluidization height, fluidization velocity, and the time of termination of complete fluidization for each intact fluidized microparticle, as well as several microparticles that exhibited adhesion during the fluidization process, selected from all intact fluidized microparticles.

[0014] Based on the maximum fluidization height, fluidization velocity, and complete fluidization termination time of each complete fluidized pellet in the real-time coating process, the coating uniformity degradation index of the real-time coating process is determined.

[0015] Based on the number of all intact fluidized pellets in the real-time coating process, the number of all pellets that are agglomerated during the fluidization process selected from all intact fluidized pellets, and the temperature and humidity data in the fluidized bed diffusion chamber at all times, the wind pressure adhesion influence of the real-time coating process is determined.

[0016] Based on the coating uniformity degradation index and the influence of wind pressure adhesion during the real-time coating process, the wind pressure trend intensity during the real-time coating process is determined.

[0017] Furthermore, the coating uniformity degradation index for determining the real-time coating process includes:

[0018] The average of the inversely proportional normalized value of the maximum fluidization height and the inversely proportional normalized value of the fluidization velocity of each complete fluidized microsphere during the real-time coating process is obtained and denoted as the particle size factor of each complete fluidized microsphere.

[0019] In the real-time coating process, complete fluidized microspheres with a particle size factor greater than or equal to the preset particle size threshold are denoted as large-size microspheres.

[0020] The real-time coating process is divided into a first time period and a second time period. The number of large-diameter microparticles that are in the first time period when the fluidization process ends is counted and recorded as the first quantity value. The number of large-diameter microparticles that are in the second time period when the fluidization process ends is counted and recorded as the second quantity value. The ratio of the second quantity value to the first quantity value is recorded as the coating uniformity degradation index of the real-time coating process.

[0021] Furthermore, the determination of the wind pressure adhesion influence degree of the real-time coating process includes:

[0022] The ratio of the number of all micro-particles that are agglomerated during fluidization to the total number of all intact fluidized micro-particles in the real-time coating process is recorded as the material adhesion rate of the real-time coating process.

[0023] In the real-time coating process, the average temperature data in the fluidized bed diffuser at all times is obtained and recorded as the temperature characteristic value. The average humidity data in the fluidized bed diffuser at all times is obtained and recorded as the humidity characteristic value. The average of the inversely normalized values ​​of the temperature characteristic value and the inversely normalized values ​​of the humidity characteristic value is obtained and recorded as the influence weight. The product of the influence weight and the material adhesion rate in the real-time coating process is recorded as the wind pressure adhesion influence degree of the real-time coating process.

[0024] Furthermore, determining the wind pressure increase intensity of the real-time coating process based on the coating uniformity degradation index and wind pressure adhesion influence degree includes:

[0025] The inversely proportional normalized value of the product of the coating uniformity deterioration index and the wind pressure adhesion influence degree in the real-time coating process is denoted as the wind pressure rise intensity in the real-time coating process.

[0026] Furthermore, the wind pressure requirement indicators for determining the real-time coating process include:

[0027] The microparticle detection and tracking results include: the microparticle region of each complete fluidized microparticle in consecutive video frames, and the motion trajectory line of each complete fluidized microparticle;

[0028] The coating solution impregnation effect of each coating process is determined based on the color change of the microparticle region of each complete fluidized microparticle in consecutive video frames during each coating process.

[0029] The mean value of the coating solution immersion effect of all historical coating processes is obtained and recorded as the historical coating solution immersion effect. The normalized value of the difference between the coating solution immersion effect of the real-time coating process and the historical coating solution immersion effect is recorded as the microcapsule loading quality of the real-time coating process.

[0030] The atomization matching degree of the real-time coating process is determined based on the curvature of the trajectory points on the motion trajectory line of each complete fluidized microparticle in each coating process.

[0031] The wind pressure strength coefficient of the real-time coating process is determined based on the drug placement advantage and atomization matching degree of the microcapsule in the real-time coating process.

[0032] The normalized value of the average of the inverse proportional value of the wind pressure intensity and the inverse proportional value of the wind pressure strength coefficient during the real-time coating process, and the sum of the average value and the preset constant, are denoted as the wind pressure demand index for the real-time coating process.

[0033] Furthermore, determining the coating solution impregnation effect for each coating process includes:

[0034] Construct the LAB color space. In each coating process, obtain the average value of all pixels in the LAB color channel of the microsphere region in each frame of a continuous video frame for each complete fluidized microsphere. Determine the microsphere point position in the LAB color space of the microsphere region in each frame of a continuous video frame for each complete fluidized microsphere. Obtain the Euclidean distance between the microsphere point positions in the LAB color space of the microsphere region in the first and last frames of a continuous video frame for each complete fluidized microsphere. Record this distance as the chromaticity difference before and after fluidization for each complete fluidized microsphere.

[0035] The average color difference before and after fluidization of all intact fluidized pellets in each coating process is obtained as the coating solution impregnation effect of each coating process.

[0036] Furthermore, determining the atomization matching degree of the real-time coating process includes:

[0037] In each coating process, the mean curvature of all trajectory points on the motion trajectory line of each complete fluidized microsphere is obtained as the coating liquid encapsulation effect of each complete fluidized microsphere.

[0038] Get the maximum value of the coating liquid encapsulation effect of all complete fluidized microparticles in all coating processes, and record it as the maximum coating liquid encapsulation effect.

[0039] The ratio of the average coating liquid encapsulation effect of all complete fluidized microparticles in the real-time historical coating process to the maximum coating liquid encapsulation effect is recorded as the atomization matching degree of the real-time coating process.

[0040] Furthermore, the wind pressure strength coefficient for determining the real-time coating process includes:

[0041] The average of the inverse proportionality between the drug loading superiority and the atomization matching degree of the real-time coating process is obtained and denoted as the wind pressure strength coefficient of the real-time coating process.

[0042] Furthermore, the wind pressure control for the next coating process includes:

[0043] The default proportional gain coefficient of the PID controller is obtained using the tuning method;

[0044] The product of the wind pressure demand index of the real-time coating process and the default proportional gain coefficient is recorded as the adjusted proportional gain value.

[0045] Based on the adjusted proportional gain value, the pressure data of the guide tube distribution plate at all times during all coating processes are input to the PID controller, which outputs the wind pressure control signal for the next coating process.

[0046] The beneficial effects of the technical solution of this invention are as follows: First, the coating process is divided, and then the particle size factor of each microsphere is obtained based on the maximum fluidization height and fluidization velocity of the microspheres in the real-time coating process. Then, the coating uniformity degradation index of the real-time coating process is evaluated based on the distribution performance of microspheres with different sizes. Then, the wind pressure intensity of the real-time process is obtained by combining the material adhesion performance. Then, based on the wind pressure intensity, the drug delivery effect of the microspheres and the matching degree between the fluidization and atomization pressure of the microspheres are analyzed to obtain the wind pressure strength coefficient. And the wind pressure demand index is obtained by combining the microsphere input performance of the guide tube. Then, the gain parameter of PID control is adjusted according to the real-time wind pressure demand index. Compared with the traditional method of using constant fan wind pressure throughout the coating process, this invention can obtain more accurate coating fluidized bed fan wind pressure control results by combining the microsphere particle size distribution performance, microsphere drug delivery performance, and matching performance with atomization pressure in the actual scenario, thereby improving the efficiency of real-time monitoring and control of the enteric capsule forming process by the computer. Attached Figure Description

[0047] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart of the module of the intelligent forming system for enteric-coated capsules based on real-time computer monitoring and control according to the present invention.

[0049] Figure 2 This is a schematic diagram of a fluidized bed bottom spray coating machine. Detailed Implementation

[0050] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the intelligent enteric-coated capsule forming system based on real-time computer monitoring and control proposed in this invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0051] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0052] The specific solution of the intelligent molding system for enteric-coated capsules based on real-time computer monitoring and control provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0053] Please see Figure 1 The diagram illustrates a module flowchart of an intelligent enteric-coated capsule forming system based on real-time computer monitoring and control, according to an embodiment of the present invention. The system includes the following modules:

[0054] Module 101: Enteric-coated microsphere coating data acquisition module.

[0055] This module is used to acquire video data in the diffusion chamber during each coating process of fluidized bed enteric microsphere coating, as well as temperature and humidity data in the fluidized bed diffusion chamber and pressure data of the guide tube distribution plate at each moment; the coating process includes a real-time coating process and several historical coating processes; microsphere detection and tracking are performed on the video data in the diffusion chamber during each coating process to obtain microsphere detection and tracking results.

[0056] It should be noted that enteric-coated microcapsule coating technology is an important forming process in the preparation of enteric-coated capsules. This technology is often achieved using a fluidized bed bottom-spray microcapsule coating machine. The entire process is monitored and controlled in real time by a computer. A coating film is applied to the outside of the microcapsules to ensure that the drug does not dissolve in the stomach but fully dissolves in the intestines, thereby maximizing its efficacy. The coating thickness of the microcapsules is controllable. A schematic diagram of a fluidized bed bottom-spray coating machine is shown below. Figure 2 As shown. Figure 2The fluidized bed bottom spray coating machine includes a guide tube, a diffusion chamber, a material container, an airflow distribution plate, and a spray gun, with air intake from the bottom. The process for coating enteric-coated microcapsules is as follows: First, all the enteric-coated microcapsules to be coated are poured into the material container of the coating machine. The coating machine is turned on, and the airflow distribution plate applies fan pressure to blow the microcapsules into a "fountain-like" fluidized state (rising inside the guide tube and falling outside the tube). The spray gun sprays coating liquid droplets. When the microcapsules rise inside the guide tube, they come into contact with the droplets, forming a thin film on their surface. Subsequently, the microcapsules fall to the external drying zone, completing one "spray to dry" cycle. This process is repeated until the microcapsule group reaches the target coating time.

[0057] It should be further noted that the constant fan pressure used in the traditional micro-pellet coating process is difficult to adapt to the coating requirements of micro-pellets. Therefore, in this embodiment, the computer monitoring and control results of the fan pressure of the coating fluidized bed are obtained by combining the micro-pellet particle size distribution, micro-pellet drug loading, and matching performance with atomization pressure in the actual scenario.

[0058] The preset single-coating process duration is 3 seconds, and this will be used as an example for explanation.

[0059] During the fluidized bed enteric pellet coating process, video data of the diffusion chamber during each coating process was acquired every 3 seconds, along with temperature and humidity data of the fluidized bed diffusion chamber and pressure data of the guide tube distribution plate at each moment.

[0060] In chronological order, the last coating process is designated as the real-time coating process, and the other coating processes are designated as historical coating processes.

[0061] It should be noted that in this embodiment, the fan pressure remains constant within a single coating process. The fan pressure is adaptively adjusted during the transition between adjacent coating processes, i.e., automatic adjustment of the fan pressure occurs every 3 seconds. Therefore, there are no cases where the coating process does not meet the 3-second interval requirement during the adjustment analysis. Specifically, real-time temperature and humidity data within the fluidized bed diffusion chamber are read by a temperature and humidity monitoring module, and real-time pressure data of the guide tube distribution plate is read by a pressure sensing module on the side of the guide tube distribution plate. The temperature, humidity, and pressure data are collected at a frequency of 12 times per second. Video data of the frontal view within the diffusion chamber is then collected using an industrial camera at a frequency of 24 frames per second. All the read data undergoes data cleaning and preprocessing, and finally, the preprocessed data is uploaded to the computer monitoring module for subsequent analysis.

[0062] It should be further explained that the purpose of this embodiment is to monitor and control the fan pressure at the bottom of the fluidized bed during the enteric microcapsule coating process. Therefore, it is necessary to judge the air pressure requirements of each coating process. First, the particle size of each microcapsule in the real-time coating process is evaluated. Then, the coating uniformity degradation index of the real-time process is obtained by the distribution performance of microcapsules of different sizes. Then, the air pressure intensity is obtained by combining the adhesion of microcapsules in the fluidization process. Based on the air pressure intensity, the drug delivery effect of microcapsules and the matching degree of microcapsule fluidization and atomization pressure are analyzed to obtain the air pressure strength coefficient. Then, the real-time air pressure requirement index is obtained by combining the microcapsule input performance of the guide tube. The air pressure requirement index is input into the PID control module of the computer for gain parameter adjustment to control the fan pressure during the microcapsule coating process.

[0063] In each coating process, microparticles are detected and tracked in the video data within the diffusion chamber to obtain the microparticle detection and tracking results. The microparticle detection and tracking results include: the microparticle region of each complete fluidized microparticle in consecutive video frames, as well as the motion trajectory line, maximum fluidization height, fluidization motion speed, and complete fluidization termination time of each complete fluidized microparticle. Several microparticles that exhibit adhesion during the fluidization process are selected from all complete fluidized microparticles.

[0064] It should be noted that in this embodiment, YOLOv5 and DeepSORT are used respectively to perform micro-pellet detection and tracking in the video data diffused during each coating process. YOLOv5 is a target detection algorithm based on the YOLO (You Only Look Once) series, while DeepSORT (Deep Learning based SORT) is a target tracking algorithm combining deep learning. Both are well-known technologies, and their specific methods are not described here. YOLOv5 is used to identify and locate multiple targets (micro-pellets) in the video, and then combined with the Mask R-CNN algorithm to obtain the micro-pellet regions in the video frames. Mask R-CNN is a classic instance segmentation algorithm used to obtain the connected components of the targets (micro-pellets), which is a well-known technology, and its specific methods are not described here. DeepSORT is used to track the motion trajectory of the micro-pellets in the video. In this embodiment, the process of ejecting microparticles from the guide tube to the highest point and then dropping them into the outer material container from the highest point is recorded as a single complete fluidization cycle. DeepSORT can be used to extract all complete fluidized microparticles that have completed a single complete fluidization cycle during the coating process. Furthermore, DeepSORT can be used to determine whether there is adhesion between complete fluidized microparticles during the fluidization process. The condition for determining adhesion is that the motion trajectory lines of different complete fluidized microparticles merge at the same time, and the trajectory changes tend to be consistent in the remaining fluidization motion after merging.

[0065] Further explanation is needed: In this embodiment, a planar coordinate system is constructed with the lower left corner of the video frame as the origin, the horizontal direction to the right as the positive horizontal and vertical directions, and the vertical direction upward as the positive vertical axis. The center point of the microsphere region in each complete fluidized microsphere within consecutive video frames is taken as the trajectory point, and the time of the last trajectory point in the video frame is recorded as the completion fluidization termination time for each complete fluidized microsphere. On the planar coordinate system, the least squares method is used to perform curve fitting on the position coordinates of all trajectory points of each complete fluidized microsphere in consecutive video frames to obtain the motion trajectory line of each complete fluidized microsphere. Since this embodiment uses an industrial camera to collect video data from the frontal viewing angle surface inside the diffusion chamber, the vertical axis direction on the planar coordinate system is the height direction of the microsphere's motion trajectory; that is, the maximum vertical axis coordinate on the motion trajectory line of each complete fluidized microsphere is the maximum fluidization height of each complete fluidized microsphere. On the trajectory of each complete fluidized pellet, the ratio of the Euclidean distance between trajectory points corresponding to adjacent video frames to the time interval between adjacent video frames is taken as the pellet's motion velocity between adjacent video frames. The average of the pellet's motion velocities between all adjacent video frames is taken as the fluidization velocity of each complete fluidized pellet. The least squares method is a well-known technique, and its specific method will not be described here.

[0066] Module 102: Wind Pressure Trend Intensity Analysis Module.

[0067] This module is used to determine the wind pressure rise intensity of the real-time coating process based on the microparticle detection and tracking results during the real-time coating process, combined with the temperature and humidity data in the fluidized bed diffusion chamber at each moment.

[0068] It should be noted that during the fluidized bed bottom spray micro-pellet coating process, the particle size of each micro-pellet may vary due to differences in manufacturing process precision. Different particle sizes will exhibit different effects under the same air pressure. Therefore, it is necessary to evaluate the particle size distribution to distinguish between large and small particle sizes. When the air pressure is too low, large-diameter micro-pellets have a shorter fluidization path, while small-diameter micro-pellets diffuse further, leading to uneven drug coating. Therefore, an analysis of the distribution of micro-pellets of different sizes is used to obtain an index of coating uniformity degradation during the coating process. For a single micro-pellet, under the same air pressure, a higher maximum fluidization height and faster fluidization velocity indicate a higher probability of it being a small-diameter micro-pellet, and vice versa.

[0069] Preferably, in one embodiment of the present invention, the method for obtaining the wind pressure intensity of the real-time coating process includes:

[0070] During the real-time coating process, the maximum fluidization height of each complete fluidized pellet is obtained. The inversely proportional normalized value and fluidization velocity The mean of the inverse proportional normalized values ​​is denoted as the particle size factor for each intact fluidized pellet.

[0071] It should be noted that in this embodiment, the following is used: and These are respectively used as the maximum fluidization height With fluidization velocity The inverse proportional normalized value, where, This is a linear normalization function used to normalize data values ​​to a range of 0 to 1. In other words, the higher the maximum fluidization height of the microsphere and the faster the fluidization velocity (i.e., the smaller the particle size factor), the higher the probability that the microsphere is a small-sized microsphere; conversely, the lower the probability that the microsphere is a large-sized microsphere.

[0072] The preset particle size threshold is 0.52, and this will be used as an example for explanation.

[0073] In the real-time coating process, complete fluidized pellets with a particle size factor greater than or equal to a preset particle size threshold are denoted as large-size pellets, and complete fluidized pellets with a particle size factor less than a preset particle size threshold are denoted as small-size pellets.

[0074] It should be noted that: due to the subtle differences in the morphology of microparticles, and the difficulty in determining their morphological details in all directions from video data, this embodiment analyzes the trajectory of the microparticles after positioning to determine their particle size. When the air pressure is too low, large-diameter microparticles are difficult to be blown to both sides of the diffusion chamber by the high-intensity fluidization motion. This may result in large-diameter microparticles falling to the near end of the guide tube and being repeatedly fluidized for drug application, while small-diameter microparticles blown to both sides of the diffusion chamber are difficult to enter the guide tube for drug application, which may reduce the uniformity of drug application among the microparticles. Therefore, the real-time coating process can be divided into two time periods based on its duration. If the number of large-diameter microparticles is significantly greater in the later time period than in the earlier time period, it indicates that the air pressure in the real-time coating process may be too low, causing poor fluidization uniformity of microparticles of different sizes.

[0075] The real-time coating process is divided into a first time period and a second time period. The number of large-diameter microparticles that are in the first time period when the fluidization process ends is counted and recorded as the first quantity value. The number of large-diameter microparticles that are in the second time period when the fluidization process ends is counted and recorded as the second quantity value. The ratio of the second quantity value to the first quantity value is recorded as the coating uniformity degradation index of the real-time coating process.

[0076] It should be noted that: A higher number of large-diameter microspheres in a later time period compared to an earlier time period during the real-time coating process indicates a higher likelihood of repeated drug application at low pressure and high frequency, resulting in poorer coating uniformity. Furthermore, the coating solution primarily alters the weight of the microspheres; the resulting surface coating is relatively thin and has minimal impact on microsphere morphology. Therefore, the effect of drug application on microsphere particle size is not considered.

[0077] It should be further explained that: considering that when the wind pressure is too low, the vertical movement of the microparticles weakens, causing some microparticles to remain in the coating area for a long time, excessively absorbing the coating liquid, increasing surface wettability, and increasing the probability of adhesion, the wind pressure tends to be higher, taking into account the coating uniformity and the adhesion of the microparticles. At the same time, excessively high wind pressure will cause the trajectory of the microparticles to directly pass through the coating liquid atomization layer, which will shorten the mixing contact time between the microparticles and the coating liquid, thus reducing the drug delivery efficiency. Meanwhile, the wind pressure of the highly adaptable fan should ensure that the residence time of the microparticles on the guide tube distribution plate is short, so as to help them quickly enter the fluidization state. Therefore, based on the wind pressure tending to be higher, the wind pressure requirement index for the real-time coating process is obtained by analyzing the drug delivery effect, atomization matching performance, and guide tube adaptability.

[0078] In the real-time coating process, the ratio of the number of all microspheres that are agglomerated during the fluidization process to the total number of all intact fluidized microspheres is recorded as the material adhesion rate of the real-time coating process.

[0079] It should be noted that, in addition to the fan pressure, the temperature and humidity inside the fluidized bed diffusion chamber also affect the adhesion between the microparticles. If the temperature and humidity inside the chamber are high, the fan pressure has a lower weight in influencing the adhesion of the microparticles.

[0080] During the real-time coating process, the average temperature data within the fluidized bed diffuser at all time points is obtained and recorded as the temperature characteristic value. The average humidity data within the fluidized bed diffuser at all time points is also obtained and recorded as the humidity characteristic value. The temperature characteristic value is then calculated. Inversely proportional normalized value and humidity characteristic value The mean of the inversely proportional normalized values ​​is denoted as the influence weight. The product of the influence weight and the material adhesion rate in the real-time coating process is denoted as the wind pressure adhesion influence degree of the real-time coating process.

[0081] It should be noted that in this embodiment, the following is used: and As temperature characteristic values ​​respectively With humidity characteristic value The inversely proportional normalized value. During the real-time coating process, the lower the temperature and humidity of the fluidized bed diffusion chamber, the higher the weight of the fan pressure on the microparticle adhesion.

[0082] It should be further noted that: if the coating uniformity degradation index of the real-time coating process is smaller, it means that the wind pressure intensity is sufficient to ensure uniform drug loading of microparticles of different sizes. In other words, the higher the wind pressure intensity, the lower the wind pressure adhesion effect, which means that there is less adhesion caused by wind pressure. This further indicates that the wind pressure intensity is higher at this time.

[0083] The product of the coating uniformity degradation index and the wind pressure adhesion influence of the real-time coating process. The inversely proportional normalized value is denoted as the wind pressure trend intensity during the real-time coating process.

[0084] In this embodiment, As The inverse proportional normalized value.

[0085] Module 103: Wind pressure demand index analysis module.

[0086] This module is used to determine the wind pressure requirement index for the real-time coating process based on the difference between the microparticle detection and tracking results in the real-time coating process and all historical coating processes, combined with the wind pressure increase intensity.

[0087] It should be noted that during the microparticle coating process, the coating solution often exhibits different chromaticity than the microparticles. That is, as the microparticles circulate and fluidize, the surface chromaticity of the microparticles gradually becomes more consistent with the coating solution. For each coating process, existing positioning and tracking algorithms are used to locate each microparticle ejected by the air pressure in the video frame. Then, connected component extraction technology is used to obtain detailed regions of each microparticle. Therefore, the greater the difference in chromaticity between the microparticle region in the final frame and the initial frame for each complete fluidized microparticle during the real-time coating process, the better the microparticle coating effect under the air pressure during the real-time coating process.

[0088] Preferably, in one embodiment of the present invention, the method for obtaining the wind pressure demand index of the real-time coating process includes:

[0089] Construct the LAB color space. In each coating process, obtain the average value of all pixels in the LAB color channel within the microsphere region of each complete fluidized microsphere in each frame of a continuous video frame, and determine the microsphere point position in the LAB color space for the microsphere region of each complete fluidized microsphere in each frame of a continuous video frame.

[0090] The construction of the LAB color space is a well-known technique, and the specific method will not be introduced here.

[0091] In each coating process, the Euclidean distance between the microparticle regions in the first and last frames of consecutive video frames for each complete fluidized microparticle in the LAB color space is obtained and recorded as the chromaticity difference before and after fluidization for each complete fluidized microparticle.

[0092] The average color difference before and after fluidization of all intact fluidized pellets in each coating process is obtained as the coating solution impregnation effect of each coating process.

[0093] Obtain the average coating solution immersion effect of all historical coating processes, and denote it as the historical coating solution immersion effect. Subtract the historical coating solution immersion effect from the coating solution immersion effect of the current coating process. The normalized value is denoted as the drug loading priority of the microcapsule in the real-time coating process.

[0094] It should be noted that in this embodiment, the following is used: As The normalized value indicates that the greater the coating liquid impregnation effect of the real-time coating process compared to the historical coating process, the better the drug impregnation effect of the microparticles under the action of air pressure in the real-time coating process. Conversely, it indicates that the air pressure is too high, the contact time between the coating liquid and the microparticles is too short, and the drug impregnation effect is not good.

[0095] It should be further explained that when the wind pressure is suitable, the microparticles move at a slower speed. When the microparticles come into contact with the coating liquid atomization layer, they are encapsulated by the coating liquid. The coating liquid atomization layer exerts a lateral force on the microparticles, changing their original trajectory so that they are ejected from the edge of the coating liquid atomization layer, and the curvature of the trajectory increases. During this process, the microparticles and coating liquid have sufficient contact, reflecting a high degree of wind pressure suitability. However, if the wind pressure is too high, the microparticles move too fast, and they may directly break through the coating liquid atomization layer and fly directly out from the center of the guide tube. In this case, the interaction time between the microparticles and coating liquid is short, and the coating efficiency is low. Therefore, by using existing positioning and tracking algorithms to locate the motion trajectory of each microparticle ejected from the guide tube in the video frame, if the curvature of the ejected microparticle's trajectory is small during the coating process, the microparticle is more likely to directly break through the coating liquid atomization layer. This phenomenon reflects excessively high wind pressure; conversely, it indicates that the wind pressure results in a better matching effect between the microparticles and the atomization layer.

[0096] In each coating process, the mean curvature of all trajectory points on the motion trajectory line of each complete fluidized microsphere is obtained as the coating liquid encapsulation effect of each complete fluidized microsphere.

[0097] Get the maximum value of the coating liquid encapsulation effect among all complete fluidized pellets in all coating processes, and record it as the maximum coating liquid encapsulation effect.

[0098] The ratio of the average coating liquid encapsulation effect of all complete fluidized microparticles in the real-time historical coating process to the maximum coating liquid encapsulation effect is recorded as the atomization matching degree of the real-time coating process.

[0099] It should be noted that: the smaller the curvature of the microparticle's trajectory during the real-time coating process, the more likely the microparticle is to break through the physicochemical layer of the coating liquid due to excessive wind pressure, indicating that the wind pressure is too high. Conversely, the higher the curvature, the higher the matching degree between the wind pressure and the atomization layer during the real-time coating process.

[0100] It should be further noted that: if the matching degree between the real-time coating process air pressure and the atomization layer is lower, it indicates that the real-time air pressure is too strong. At the same time, if the drug placement of the microparticles is lower, it further indicates that the fluidization speed of the microparticles is too fast, which means that the air pressure intensity is too strong at this time.

[0101] The average of the inverse proportionality between the drug loading superiority of the microcapsule in the real-time coating process and the inverse proportionality between the atomization matching degree of the real-time coating process is denoted as the wind pressure strength coefficient of the real-time coating process.

[0102] It should be noted that in this embodiment, the difference between 1 and the difference in drug loading superiority of the microparticles during the real-time coating process is used as the inverse proportional value of drug loading superiority of the microparticles during the real-time coating process. Similarly, the difference between 1 and the difference in atomization matching degree during the real-time coating process is used as the inverse proportional value of atomization matching degree during the real-time coating process.

[0103] It should be further noted that if the wind pressure during the real-time coating process tends to be higher and the intensity is lower, and the wind pressure strength coefficient is also lower, it reflects that the wind pressure demand is higher.

[0104] The default constant is 1, and this will be used as an example for explanation.

[0105] The average of the inverse proportional value of the wind pressure intensity during the real-time coating process and the inverse proportional value of the wind pressure strength coefficient during the real-time coating process is obtained. The sum of the normalized value and the preset constant is denoted as the wind pressure demand index for the real-time coating process.

[0106] It should be noted that in this embodiment, the difference between 1 and the wind pressure intensity during the real-time coating process is used as the inverse proportional value of the wind pressure intensity during the real-time coating process; similarly, the difference between 1 and the wind pressure strength coefficient during the real-time coating process is used as the inverse proportional value of the wind pressure strength coefficient during the real-time coating process. As The normalized value, where, This is a hyperbolic tangent function used to normalize data values ​​to a range of -1 to 1. Higher wind pressure tends to have a lower intensity and stronger wind pressure coefficient, indicating a higher wind pressure demand.

[0107] Module 104: Wind pressure control module.

[0108] This module is used to control the air pressure of the next coating process based on the real-time air pressure demand index of the coating process and the pressure data of the guide tube distribution plate at all times in all coating processes.

[0109] Preferably, in one embodiment of the present invention, the method for obtaining the wind pressure control signal for the next coating process includes:

[0110] The default proportional gain coefficient of a PID controller is obtained using a tuning method. The tuning method is a well-known technique for obtaining PID controller parameters, and its specific details will not be described here. PID controllers are a very common and well-known type of controller used to control industrial processes, mechanical systems, and various other systems. PID stands for Proportional, Integral, and Derivative, representing the three main components of the controller.

[0111] The product of the wind pressure demand index of the real-time coating process and the default proportional gain coefficient is recorded as the adjusted proportional gain value.

[0112] Based on the adjusted proportional gain value, the pressure data of the guide tube distribution plate at all times during all coating processes are input to the PID controller, which outputs the wind pressure control signal for the next coating process.

[0113] It should be noted that in this embodiment, following the above method, the real-time adjusted parameters are continuously input into the PID control chip, and the calculated control output is converted into an actual control signal for fan pressure control. This allows for dynamic updates of the PID controller parameters during continuous micro-pellet coating monitoring, ensuring the stability of the fan pressure control for the micro-pellet coating fluidized bed during processing. Finally, after the control cycle ends, the system's control effect is evaluated to check for overshoot or oscillation. Based on the actual feedback results, relevant parameters can be fine-tuned and optimized to improve efficiency and system performance in subsequent molding processes.

[0114] Further explanation is needed: The complete process of intelligent molding of enteric-coated capsules based on real-time computer monitoring and control is as follows: First, the coating treatment of the current batch of drug microspheres is completed under the action of the above-mentioned computer adaptive fan pressure monitoring and control results. The coated microspheres are sieved through a mold to remove particles that are too large, too small or do not meet the particle size requirements, ensuring that the particle size distribution of the microspheres meets the design requirements. The screened coated microspheres are subjected to quality inspection, including but not limited to: (1) Particle size distribution: the microspheres are tested by sieving or laser particle size analyzer to ensure that the particle size is uniform. (2) Moisture content: the moisture content is tested by Karl Fischer method to ensure that the moisture content meets the requirements and to avoid the capsules absorbing moisture and deteriorating. (3) Coating weight gain: the weight gain of the enteric layer is tested by gravimetric method to ensure that it meets the design requirements, so as to ensure the enteric effect. The qualified microspheres need to be filled into the capsule shell to form the final enteric capsule. In the intelligent production line, the filling process usually uses automated equipment (such as capsule filling machine) to ensure accurate filling and good capsule sealing. Finally, each batch of pharmaceutical pellets is processed in the manner described above, thereby realizing an intelligent molding method for enteric-coated capsules based on real-time computer monitoring and control.

[0115] This invention is now complete.

[0116] In summary, in this embodiment of the invention, microparticle detection and tracking are performed on the video data in the diffusion chamber during each coating process to obtain the microparticle detection and tracking results. Based on the microparticle detection and tracking results in the real-time coating process, combined with the temperature and humidity data in the fluidized bed diffusion chamber at each moment, the wind pressure intensity of the real-time coating process is determined. Furthermore, considering the differences between the microparticle detection and tracking results of the real-time coating process and all historical coating processes, the wind pressure requirement index for the real-time coating process is determined. Finally, combining the pressure data of the guide tube distribution plate at all moments in all coating processes, wind pressure control is performed for the next coating process. This invention improves the efficiency of real-time monitoring and control of the enteric capsule forming process by the computer.

[0117] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A computer real-time monitoring control based enteric capsule intelligent forming system, characterized in that, The system comprises the following modules: An enteric micro-pellet coating data acquisition module: used for acquiring video data in a diffusion chamber in each coating process, and temperature and humidity data in the diffusion chamber of the fluidized bed and pressure data of the distribution plate of the draft tube at each time during the fluidized bed enteric micro-pellet coating process; the coating process includes a real-time coating process and a plurality of historical coating processes; micro-pellet detection and tracking are performed on the video data in the diffusion chamber in each coating process to obtain micro-pellet detection and tracking results; A wind pressure trend high intensity analysis module: used for determining the wind pressure trend high intensity of the real-time coating process according to the micro-pellet detection and tracking results in the real-time coating process and in combination with the temperature and humidity data in the diffusion chamber of the fluidized bed at each time; A wind pressure demand index analysis module: used for determining the wind pressure demand index of the real-time coating process according to the difference between the micro-pellet detection and tracking results in the real-time coating process and in all historical coating processes and in combination with the wind pressure trend high intensity; A wind pressure control module: used for performing wind pressure control on the next coating process in the future according to the wind pressure demand index of the real-time coating process and in combination with the pressure data of the distribution plate of the draft tube at all times in all coating processes; The determination of the wind pressure trend high intensity of the real-time coating process comprises: The micro-pellet detection and tracking results include: the maximum fluidization height, the fluidization movement speed, and the complete fluidization termination time of each complete fluidized micro-pellet, and a plurality of micro-pellets selected from all complete fluidized micro-pellets that exist in the fluidization process; Determination of a coating uniformity degradation index of the real-time coating process according to the maximum fluidization height, the fluidization movement speed, and the complete fluidization termination time of each complete fluidized micro-pellet in the real-time coating process; Determination of a wind pressure adhesion influence degree of the real-time coating process according to the number of all complete fluidized micro-pellets, the number of all micro-pellets selected from all complete fluidized micro-pellets that exist in the fluidization process, and in combination with the temperature and humidity data in the diffusion chamber of the fluidized bed at all times; Determination of the wind pressure trend high intensity of the real-time coating process according to the coating uniformity degradation index and the wind pressure adhesion influence degree of the real-time coating process; The determination of the coating uniformity degradation index of the real-time coating process comprises: An inverse proportional normalized value of the maximum fluidization height and an inverse proportional normalized value of the fluidization movement speed of each complete fluidized micro-pellet in the real-time coating process are obtained, and the average value is recorded as a particle size degree factor of each complete fluidized micro-pellet; In the real-time coating process, the complete fluidized micro-pellets with the particle size degree factor greater than or equal to a preset particle size threshold value are recorded as large-particle-size micro-pellets; The real-time coating process is equally divided into a front time period and a rear time period, the number of the large-particle-size micro-pellets with the complete fluidization termination time in the front time period is recorded as a first number value, the number of the large-particle-size micro-pellets with the complete fluidization termination time in the rear time period is recorded as a second number value, and the ratio of the second number value to the first number value is recorded as the coating uniformity degradation index of the real-time coating process; The determination of the wind pressure demand index of the real-time coating process comprises: The micro-pellet detection and tracking results include: a micro-pellet area of each complete fluidized micro-pellet in a continuous video frame, and a movement trajectory line of each complete fluidized micro-pellet; Determine the coating liquid dyeing effect of each coating process according to the color change of the pellet area of each complete fluidized pellet in the continuous video frame in each coating process; Obtain the mean value of the coating liquid dyeing effect of all historical coating processes, denoted as the historical coating liquid dyeing effect, and the normalized value of the difference between the coating liquid dyeing effect of the real-time coating process and the historical coating liquid dyeing effect, denoted as the pellet medicine placement optimization degree of the real-time coating process; Determine the atomization matching degree of the real-time coating process according to the curvature of the trajectory points on the motion trajectory line of each complete fluidized pellet in each coating process; Determine the wind pressure strong coefficient of the real-time coating process according to the pellet medicine placement optimization degree and the atomization matching degree of the real-time coating process; Obtain the normalized value of the sum of the inverse value of the wind pressure trend high intensity of the real-time coating process and the inverse value of the mean value of the wind pressure strong coefficient and a preset constant, denoted as the wind pressure demand index of the real-time coating process.

2. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, characterized in that, The determination of the wind pressure adhesion influence degree of the real-time coating process includes: Obtain the ratio of the number of all the pellets that exist adhesion in the fluidization process to the number of all the complete fluidized pellets in the real-time coating process, denoted as the material adhesion rate of the real-time coating process; In the real-time coating process, obtain the mean value of the temperature data in the fluidized bed diffusion chamber at all times, denoted as the temperature characteristic value, obtain the mean value of the humidity data in the fluidized bed diffusion chamber at all times, denoted as the humidity characteristic value, obtain the mean value of the inverse normalized value of the temperature characteristic value and the inverse normalized value of the humidity characteristic value, denoted as the influence weight, and obtain the product of the influence weight and the material adhesion rate of the real-time coating process, denoted as the wind pressure adhesion influence degree of the real-time coating process.

3. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, characterized in that, The determination of the wind pressure trend high intensity of the real-time coating process according to the coating uniformity degradation index and the wind pressure adhesion influence degree of the real-time coating process includes: Obtain the inverse normalized value of the product of the coating uniformity degradation index and the wind pressure adhesion influence degree of the real-time coating process, denoted as the wind pressure trend high intensity of the real-time coating process.

4. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, characterized in that, The determination of the coating liquid dyeing effect of each coating process includes: Construct a LAB color space, in each coating process, obtain the mean value of the numerical value of all pixel points in the pellet area of each complete fluidized pellet in each frame of the continuous video frame in the LAB color channel, determine the pellet point position of the pellet area of each complete fluidized pellet in each frame of the continuous video frame in the LAB color space, obtain the Euclidean distance between the pellet point positions of each complete fluidized pellet in the first frame and the last frame of the continuous video frame in the LAB color space, denoted as the color difference before and after fluidization of each complete fluidized pellet; Obtain the mean value of the color difference before and after fluidization of all complete fluidized pellets in each coating process as the coating liquid dyeing effect of each coating process.

5. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, wherein, The determination of the atomization matching degree of the real-time coating process includes: In each coating process, obtain the mean value of the curvature of all trajectory points on the motion trajectory line of each complete fluidized pellet as the coating liquid wrapping effect of each complete fluidized pellet; Obtain the maximum value of the coating liquid wrapping effect of all complete fluidized pellets in all coating processes, denoted as the maximum coating liquid wrapping effect; The ratio of the average coating liquid wrapping effect of all complete fluidized pellets in the real-time historical coating process to the maximum coating liquid wrapping effect is denoted as the atomization matching degree of the real-time coating process.

6. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, wherein, The determination of the wind pressure strength coefficient of the real-time coating process comprises: The average of the inverse value of the pellet medicine placement degree of the real-time coating process and the inverse value of the atomization matching degree is denoted as the wind pressure strength coefficient of the real-time coating process.

7. The computer real-time monitoring control based enteric capsule intelligent forming system according to claim 1, wherein, The wind pressure control of the future next coating process comprises: The default proportional gain coefficient of the PID controller is obtained by tuning method; The product of the wind pressure demand index of the real-time coating process and the default proportional gain coefficient is denoted as the adjusted proportional gain value; According to the adjusted proportional gain value, the pressure data of the flow guide cylinder distribution plate at all times in all coating processes are input into the PID controller, and the wind pressure control signal of the future next coating process is output.

Images