Simulation method and system for particle blanking of medicine bin

Abstract

The invention discloses a simulation method and system for particle blanking of a medicine bin, relates to the technical field of simulation methods for particle blanking, and determines a falling track of each medicine relative to the medicine bin based on detection of a falling event of each medicine particle. And the simulation picture of medicine particle blanking is determined according to the falling track of each medicine relative to the medicine bin and falling simulation of the monitoring space, so that the accuracy of the simulation picture of medicine particle blanking is ensured. Therefore, a plurality of flow velocity layers are formed based on division of the simulation picture of medicine particle blanking, the residual particle amount of the plurality of flow velocity layers is determined according to the speed of the plurality of flow velocity layers and the medicine particle proportion of the plurality of flow velocity layers, and posture adjustment of the medicine bin is triggered, so that the material accumulation condition of the medicine bin after each time of discharging is avoided; the accuracy of the residual particle amount of the multiple flow velocity layers is guaranteed, posture adjustment of the medicine bin is triggered so as to optimize the residual particle amount of the multiple flow velocity layers, and the material accumulation condition of the medicine bin after each time of discharging is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of a simulation method for particle dropping, and in particular to a simulation method and system for particle dropping in a medicine bin. Background Art

[0002] With the development of science and technology, medicine silos are gradually used in people's lives and store multiple drug particles. The multiple drug particles are all in the same medicine silo and discharged downward from the discharge port of the medicine silo. In the existing technology, the multiple drug particles are discharged downward from the discharge port of the medicine silo. However, in each discharge of the medicine silo, there is a large amount of material accumulation inside the medicine silo, which affects the subsequent use of the medicine silo. The existing technology cannot measure the multiple flow rate layers of the drug particles, and thus cannot optimize the residual particle amount in the multiple flow rate layers. Summary of the Invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art, and the present invention provides a simulation method and system for particle dropping from a medicine bin.

[0004] An embodiment of the present invention provides a simulation method for particle dropping from a medicine bin, comprising: determining a monitoring space based on the current position of the medicine bin and the previous falling trajectories of the medicine particles; in the dynamic detection of the monitoring space, determining the falling events of each medicine particle in the medicine bin according to the position nodes, time and specifications of the medicine particles during the falling process; determining the falling trajectories of each medicine relative to the medicine bin based on the detection of the falling events of each medicine particle, and determining the simulation picture of the medicine particle dropping based on the falling trajectories of each medicine relative to the medicine bin and the falling simulation of the monitoring space; forming a plurality of flow velocity layers based on the division of the simulation picture of the medicine particle dropping, the plurality of flow velocity layers being a central fast layer, an inner wall slow layer and a transition speed layer; determining the amount of residual particles in the plurality of flow velocity layers according to the speeds of the plurality of flow velocity layers and the proportions of the medicine particles in the plurality of flow velocity layers, and triggering the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge.

[0005] An embodiment of the present invention provides a simulation system for particle dropping from a medicine bin. The simulation system for particle dropping from a medicine bin is applied to the above-mentioned simulation method for particle dropping from a medicine bin. The simulation system for particle dropping from a medicine bin includes: The monitoring space module is used to determine the monitoring space based on the current position of the medicine warehouse and the previous falling trajectory of the medicine particles; A falling event module is used to determine the falling events of each drug particle in the drug bin according to the position node, time and specifications of each drug particle during the falling process in the dynamic detection of the monitoring space; a simulation screen module for determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining a simulation screen of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space; A velocity layer module is used to form multiple velocity layers based on the division of the simulation screen of drug particle falling, and the multiple velocity layers are respectively a central fast layer, an inner wall slow layer and a transition velocity layer; The posture adjustment module is used to determine the amount of residual particles in multiple flow rate layers according to the speed of multiple flow rate layers and the proportion of drug particles in multiple flow rate layers, and trigger the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge.

[0006] Compared with the prior art, the present invention has the following beneficial effects: In an embodiment of the present invention, through the method in the embodiment of the present invention, the falling trajectory of each drug relative to the medicine warehouse is determined based on the detection of the falling events of each drug particle, and the simulation picture of the drug particle falling is determined according to the falling trajectory of each drug relative to the medicine warehouse and the falling simulation of the monitoring space, which is compatible with the falling trajectory of each drug and the overall consideration of the detection space, and ensures the accuracy of the simulation picture of the drug particle falling.

[0007] Therefore, based on the division of the simulation screen of drug particle dropping, multiple flow velocity layers are formed, and the multiple flow velocity layers are respectively a central fast layer, an inner wall slow layer and a transition velocity layer; the residual particle amount of the multiple flow velocity layers is determined according to the speed of the multiple flow velocity layers and the proportion of drug particles in the multiple flow velocity layers, and the posture adjustment of the medicine warehouse is triggered to avoid material accumulation in the medicine warehouse after each discharge. It is compatible with the overall consideration of the speed of multiple flow velocity layers and the proportion of drug particles in multiple flow velocity layers, ensures the accuracy of the residual particle amount of the multiple flow velocity layers, and triggers the posture adjustment of the medicine warehouse to optimize the residual particle amount of the multiple flow velocity layers, and avoids material accumulation in the medicine warehouse after each discharge. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 1 is a flow chart of a simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 2 1 is a flow chart of step S11 in the simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 3 1 is a flow chart of step S12 in the simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 4 1 is a flow chart of step S13 in the simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 51 is a flow chart of step S14 in the simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 6 1 is a flow chart of step S15 in the simulation method for particle dropping from a medicine bin in an embodiment of the present invention; Figure 7 It is a schematic diagram of the structural composition of the simulation system for particle dropping from the medicine bin in an embodiment of the present invention. DETAILED DESCRIPTION

[0009] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0010] See also Figures 1 to 7 A simulation method for particle dropping in a medicine bin includes: Step S11: determining the monitoring space according to the current position of the medicine bin and the previous falling trajectory of the medicine particles; Step S12: in the dynamic detection of the monitoring space, determining the falling events of each drug particle in the drug bin according to the position nodes, time and specifications of each drug particle during the falling process; Step S13: determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining a simulation image of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space; Step S14: forming a plurality of flow velocity layers based on the division of the simulation screen of the drug particle falling, wherein the plurality of flow velocity layers are respectively a central fast layer, an inner wall slow layer and a transitional velocity layer; Step S15: determining the amount of residual particles in the multiple flow rate layers according to the speeds of the multiple flow rate layers and the proportions of the drug particles in the multiple flow rate layers, and triggering the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge; refer to Figure 2 ,In step S11, the monitoring space is determined according to the current position of the medicine bin and the previous falling trajectory of the medicine particles; In the specific implementation process of the present invention, the specific steps are: S111: In the drug testing room, positioning detection is performed on the drug warehouse based on the camera of the drug testing room, and a current image of the drug warehouse is collected. The current position of the drug warehouse is determined based on the matching of the current image of the drug warehouse and the indoor distribution map of the drug testing room; S112: collecting the corresponding drug particle carrying platform based on the peripheral detection of the current position of the drug warehouse, and determining the relative space between the drug warehouse and the drug particle carrying platform based on the spatial comparison between the drug warehouse and the drug particle carrying platform; S113: Determine the previous falling trajectory of the drug particles based on the current position of the drug bin and the drug bin database, and determine the monitoring space according to the previous falling trajectory of the drug particles and the relative space between the drug bin and the drug particle carrying platform.

[0011] In an embodiment of the present application, in a drug testing room, the medicine warehouse is positioned and detected based on the camera of the drug testing room, and the current image of the medicine warehouse is collected. The current position of the medicine warehouse is determined based on the matching of the current image of the medicine warehouse and the indoor distribution map of the drug testing room, thereby ensuring the accuracy of the current position of the medicine warehouse.

[0012] At this time, in order to capture the location and status of the medicine warehouse in real time, high-resolution cameras need to be arranged in the drug testing room; at the same time, the cameras should be installed in a position that can fully cover the activity area of ​​the medicine warehouse to ensure that key views such as the front, side or top of the medicine warehouse can be clearly captured; the number, position and angle of the cameras should be reasonably planned according to the size and layout of the drug testing room.

[0013] The images captured by the camera are used to perform real-time positioning detection of the medicine warehouse. At this time, the images captured by the camera are transmitted to the computer vision system, which uses image recognition technology (such as edge detection, feature matching, etc.) to identify the medicine warehouse in the image. By comparing the image features of the medicine warehouse with the template features pre-stored in the database, the position and direction of the medicine warehouse in the image are determined.

[0014] In order to subsequently match it with the indoor distribution map, it is necessary to collect the current high-definition image of the medicine warehouse; at this time, after the camera captures the image of the medicine warehouse, the computer will automatically save the image as the current image. This image should contain sufficient details for subsequent image matching and position determination.

[0015] The specific location of the medicine warehouse in the room is determined by matching the current image of the medicine warehouse with the indoor distribution map of the drug testing room. In this case, the indoor distribution map is a two-dimensional or three-dimensional image that contains information such as the room layout and equipment location. The computer vision system will compare the current image of the medicine warehouse with the indoor distribution map and determine the relative position of the medicine warehouse in the distribution map by finding common feature points or edge contours. This position is usually expressed in coordinate form, such as (x, y) or (x, y, z) coordinates.

[0016] Based on the results of image matching, the exact location of the drug warehouse in the drug testing room is determined. At the same time, once the relative position of the drug warehouse in the distribution map is determined, the actual position of the drug warehouse in the room is calculated based on the size and scale of the distribution map. This position is used in subsequent analysis and simulation processes.

[0017] Furthermore, the corresponding drug particle carrier is collected based on the peripheral detection of the current position of the drug warehouse, and the relative space between the drug warehouse and the drug particle carrier is determined according to the spatial comparison of the drug warehouse and the drug particle carrier, which is compatible with the overall consideration of the spatial comparison of the medicine warehouse and the drug particle carrier, and ensures the accuracy of the relative space between the medicine warehouse and the drug particle carrier.

[0018] At this point, after determining the current location of the medicine warehouse, it is necessary to detect the environment around the medicine warehouse to identify other equipment or structures related to the medicine warehouse, especially the drug particle carrier; at the same time, this usually involves using devices such as sensors or cameras to scan or photograph the space around the medicine warehouse; sensors detect physical parameters (such as distance, light, etc.), while cameras capture visual information, which is transmitted to the computer system for subsequent analysis and processing.

[0019] In the results of the peripheral detection, the information of the drug particle carrier directly related to the drug warehouse is identified and collected; at this time, the computer system will filter out the data related to the drug particle carrier from the results of the peripheral detection according to the preset rules or algorithms, which includes the position, size, shape and other information of the carrier. This information is crucial for the subsequent determination of the relative space between the drug warehouse and the carrier.

[0020] By comparing the spatial information of the drug chamber and the drug particle carrier, the relative position and spatial relationship between them are determined. At this time, this usually involves the use of technologies such as three-dimensional modeling or spatial geometry calculation. First, based on the collected information, a three-dimensional model or spatial coordinates of the drug chamber and the carrier are constructed. Then, by calculating the relative distance, direction and other parameters between these models or coordinates, the relative spatial relationship between them is determined. This relative spatial relationship is used in the subsequent particle falling simulation to simulate the process of particles falling from the drug chamber to the carrier.

[0021] Specifically, the medicine bin is mounted on a fixed bracket, and the drug particle carrier is located somewhere below the medicine bin; the current position of the medicine bin has been determined to be (x=500, y=600, z=1000) through step S111 (assuming this is a three-dimensional coordinate system); in step S112, the environment around the medicine bin is first scanned using a camera; the scanning results show that at a certain position directly below the medicine bin, there is a drug particle carrier that matches the preset rule; by further analyzing the scanning results, the position information of the carrier is collected as (x=500, y=600, z=900).

[0022] Next, 3D modeling technology was used to construct 3D models of the medicine bin and the carrier platform, and the relative spatial relationship between them was determined based on their positional information. In this example, the positions of the medicine bin and the carrier platform are the same in the x and y directions, but in the z direction, the carrier platform is 100 units lower than the medicine bin, which means that when the particles fall from the medicine bin, they will fall 100 units along the z-axis to reach the carrier platform. By determining this relative spatial relationship, accurate spatial parameters are provided for the subsequent particle dropping simulation to ensure the accuracy and practicality of the simulation. During the simulation process, the particles are simulated falling from different positions in the medicine bin, and their accumulation pattern and distribution on the carrier platform are observed.

[0023] Furthermore, the previous falling trajectory of the drug particles is determined based on the current position of the drug warehouse and the drug warehouse database, and the monitoring space is determined based on the previous falling trajectory of the drug particles and the relative space between the drug warehouse and the drug particle supporting platform. This is compatible with the overall consideration of the previous falling trajectory of the drug particles and the relative space between the drug warehouse and the drug particle supporting platform, ensuring the accuracy of the monitoring space.

[0024] At this time, the current location information of the medicine warehouse is used to retrieve relevant historical data from the medicine warehouse database, especially the previous falling trajectories of drug particles. At this time, the medicine warehouse database is a system that stores historical data about the medicine warehouse and its related operations (such as particle falling, discharge, etc.). After determining the current location of the medicine warehouse, the system will search for matching or similar records in the database based on this location information. These records contain trajectory data of particles falling from the medicine warehouse under different conditions (such as different particle types, different operating parameters, etc.).

[0025] From the retrieved historical data, the drug particle falling trajectories that are closest or similar to the current conditions are screened out; at this point, this usually involves comparing and analyzing the retrieved data; the system will evaluate the similarity of each record with the current conditions based on preset rules or algorithms. These rules or algorithms take into account the particle type, drug bin type, operating parameters (such as discharge speed, vibration frequency, etc.) and environmental factors (such as temperature, humidity, etc.); finally, the system will select the falling trajectory that is closest or similar to the current conditions as a reference.

[0026] Based on the previous falling trajectory of drug particles and the relative spatial relationship between the drug bin and the drug particle supporting platform, a monitoring space for monitoring particle falling is determined. At this time, first, the system will predict the area and range of particle falling under the current conditions based on the falling trajectory data. Then, combined with the relative spatial relationship between the drug bin and the supporting platform (such as distance, direction, etc.), a monitoring space containing this predicted area is determined. This monitoring space should be large enough to cover all particle landing points, but not too large to avoid unnecessary computational burden and false alarms.

[0027] Specifically, the current position of the medicine warehouse has been determined to be (x=500, y=600, z=1000) through steps S111 and S112, and the position information of the medicine particle carrier directly related to the medicine warehouse has been collected as (x=500, y=600, z=900); at the same time, a medicine warehouse database containing particle falling trajectory data under various conditions has also been established.

[0028] In step S113, first, based on the current position information of the medicine bin, relevant historical data is retrieved from the database; assuming that a record very close to the current conditions is found, which describes the falling trajectory of the particles when the same type of particles are used for discharge in the same type of medicine bin, this record shows that the particles start to fall from the outlet of the medicine bin, fall about 100 units along the z-axis direction (matching the height difference of the support platform), and form a specific accumulation shape on the support platform; next, combined with the relative spatial relationship between the medicine bin and the support platform, a monitoring space for monitoring particle falling is determined; in this example, the monitoring space is defined as a cylindrical area with a radius of R centered on the support platform, where R is determined based on the diffusion degree of the particle falling trajectory and the size of the support platform. This monitoring space will be used in subsequent particle falling simulation and real-time monitoring to ensure that the particles can accurately fall on the support platform and avoid accumulation or leakage.

[0029] In one embodiment of the present application, a falling trajectory parameter matching table is collected to record the records in the medicine warehouse database that are close to the current position and their related parameters; the falling trajectory parameter matching table is shown in Table 1: Table 1 Falling trajectory parameter matching table

[0030] refer to Figure 3 In step S12, in the dynamic detection of the monitoring space, the falling events of the respective drug particles in the drug bin are determined according to the position nodes, time and specifications of the drug particles during the falling process; In the specific implementation process of the present invention, the specific steps are: S121: Dynamically detecting the monitoring space, marking each drug particle discharged through the discharge port of the drug bin based on the dynamic detection of the monitoring space, determining a position node of each drug particle during its falling process based on the marked spatial position of each drug particle, and collecting the position node of each drug particle during its falling process; S122: For each drug particle, determining a predicted trajectory of the drug particle based on connections of multiple position nodes of the drug particle; S123: collecting the particle carrying capacity of the drug particle carrying platform, and determining the falling event of each drug particle based on the predicted trajectories of the multiple drug particles, the specifications of the multiple drug particles, and the particle carrying capacity of the drug particle carrying platform; In an embodiment of the present application, dynamic detection is performed on the monitoring space, and based on the dynamic detection of the monitoring space, each drug particle discharged through the discharge outlet of the drug bin is marked, and the position node of each drug particle during the falling process is determined according to the marked spatial position of each drug particle to collect the position node of each drug particle during the falling process.

[0031] At this time, the changes in the monitoring space are monitored in real time, especially the movement of drug particles after being discharged from the discharge port of the drug warehouse. At the same time, this usually involves the use of high-speed cameras, laser scanners or other types of sensors to continuously capture or scan the monitoring space. These devices can record images or data in the space at a high frequency, thereby capturing the dynamic changes of drug particles.

[0032] Each drug particle discharged from the drug chamber outlet is uniquely identified for subsequent tracking and analysis. At this time, after capturing the image or data of the drug particles, each particle is marked using image processing technology (such as edge detection, object recognition, etc.) or sensor data analysis methods. This is achieved by assigning a unique ID to each particle or using certain characteristics of the particles (such as shape, size, color, etc.) to distinguish them.

[0033] The key positions of each drug particle during its fall are recorded. These position nodes will be used for subsequent trajectory prediction and analysis. At this time, based on the marked drug particles, the object tracking algorithm in image processing technology or the time series analysis of sensor data is used to determine the position of the particles at different time points. These position information are recorded as position nodes, each of which contains the spatial coordinates and timestamp of the particle.

[0034] Collect the location node information of all drug particles for subsequent processing and analysis; at this time, store the determined location node information in a database or data file. This information usually includes the particle ID, spatial coordinates of each location node, timestamp, etc., so that the data can be accessed at any time for further analysis or visualization when needed.

[0035] Furthermore, in each drug particle, the predicted trajectory of the drug particle is determined based on the connection of multiple position nodes of the drug particle, which is compatible with the overall consideration of the connection of multiple position nodes of the drug particle and ensures the accuracy of the predicted trajectory of the drug particle.

[0036] At this point, multiple location node information of each drug particle is obtained from the previous step (such as S121). This information is the basis for determining the predicted trajectory of the particle. At this point, the location node information usually includes the particle ID, the spatial coordinates of each node (such as x, y, z), and the corresponding timestamp. This information is stored in a database or data structure for subsequent processing.

[0037] Select an appropriate trajectory prediction algorithm based on the motion characteristics and accuracy requirements of the drug particles. In this case, trajectory prediction algorithms include but are not limited to linear interpolation, polynomial fitting, Bezier curve fitting, Kalman filtering, or machine learning algorithms (such as time series analysis and neural networks). The choice of algorithm should be based on factors such as the smoothness of particle motion, velocity changes, acceleration, and the required prediction accuracy.

[0038] Configure the necessary parameters for the selected trajectory prediction algorithm and initialize it. At this point, this involves setting the number of interpolation points, the order of the polynomial, the noise parameters of the Kalman filter, etc. In addition, it is also necessary to initialize the data structures required by the algorithm, such as matrices and vectors.

[0039] The collected position node information is used to calculate the predicted trajectory of the particle through the trajectory prediction algorithm. At this time, the position node information is input into the algorithm, and the algorithm calculates the predicted position of the particle at different time points based on the spatial coordinates and timestamps of the node. These predicted positions are connected to form the predicted trajectory of the particle.

[0040] The accuracy of the predicted trajectory is evaluated and adjusted as needed. This is achieved by comparing it with the actual observed particle motion trajectory. If there is a significant difference between the predicted and actual trajectories, the algorithm parameters need to be adjusted or a more appropriate algorithm needs to be selected. In addition, statistical methods (such as mean square error, correlation coefficient, etc.) are used to quantify the accuracy of the predicted trajectory.

[0041] Specifically, the medicine chamber continuously discharges small round particles; in step S121, multiple position node information of these particles has been collected through high-speed cameras and image processing technology; now, in step S122, the predicted trajectory of the particles will be determined based on this information; first, the position node information of each particle is extracted from the database; then, considering that the movement of particles is usually smooth and the speed does not change much, quadratic polynomial fitting is selected as the trajectory prediction algorithm.

[0042] Next, configure the parameters of the polynomial fit, such as the order of fit (quadratic is chosen here) and the number of interpolation points (determined according to the density and accuracy requirements of the position nodes); then, initialize the data structures required by the algorithm, such as the matrix for storing the fitting coefficients; now, execute the polynomial fitting algorithm; for each particle, input the spatial coordinates and timestamp of the position node into the algorithm, and the algorithm calculates the predicted position of the particle at different time points. These predicted positions are connected to form the predicted trajectory of the particle; finally, verify the accuracy of the predicted trajectory; by comparing with the actual observed particle motion trajectory, it is found that the predicted trajectory is very close to the actual trajectory, which shows that the algorithm selection and parameter configuration are reasonable.

[0043] Therefore, the particle carrying capacity of the drug particle carrier is collected, and the falling event of each drug particle is determined based on the predicted trajectories of multiple drug particles, the specifications of multiple drug particles, and the particle carrying capacity of the drug particle carrier. This is compatible with the overall consideration of the predicted trajectories of multiple drug particles, the specifications of multiple drug particles, and the particle carrying capacity of the drug particle carrier, ensuring the accuracy of the falling event of each drug particle.

[0044] At this time, the number or mass of particles currently piled on the drug particle support platform is obtained in real time or periodically to understand the load condition of the support platform. At this time, this is achieved in a variety of ways, such as using a weight sensor to directly measure the total weight on the support platform, or calculating the coverage area or number of particles on the support platform through image processing technology. If the particles are evenly distributed and the mass or volume of each particle is known, then the total weight or coverage area is converted into the number or mass of the particles.

[0045] The predicted trajectory of each drug particle calculated in the previous step (such as S122) is used to subsequently analyze the falling condition of the particles. At this time, these predicted trajectories are usually stored in the form of time series, which contain the spatial position information of the particles at different time points. In order to analyze whether the particles have successfully landed on the support platform, these predicted trajectories need to be compared with the position information of the support platform.

[0046] Consider the particle's shape, size, mass, and other specifications, as this information affects the particle's falling behavior and accumulation pattern. These specifications are typically acquired during production or testing. These factors must be considered when analyzing particle drop events, as they affect the particle's resistance in the air, its falling velocity, and its accumulation density.

[0047] Based on the particle carrying capacity of the support platform, the predicted trajectory of the particles, and the specification information of the particles, it is judged whether each particle has successfully landed on the support platform and their stacking method; at this time, this is achieved by comparing the predicted trajectory with the position information of the support platform; if the end point of the predicted trajectory falls within the range of the support platform and the particle carrying capacity of the support platform does not reach the maximum value (or a preset threshold), the particle is considered to have successfully fallen; in addition, the particle specification information and stacking method are also used to judge whether the particles are stacked evenly, whether there are overlaps or gaps, etc.

[0048] Specifically, the medicine chamber continuously discharges small round particles, which are collected on a medicine particle carrying platform; in step S122, the predicted trajectory of each particle has been calculated; now, in step S123, the particle carrying capacity on the carrying platform is first collected by the weight sensor; it is assumed that the maximum carrying capacity of the carrying platform is 1000 grams, and the current carrying capacity is 800 grams; then, the predicted trajectory of each particle is obtained and compared with the position information of the carrying platform; for example, it is found that the end point of the predicted trajectory of a particle falls within the range of the carrying platform, and its falling time does not conflict with other particles (that is, no other particles fall at the same time and block its path).

[0049] Then, the specification information of the particles was considered; in this example, all particles are round particles of the same size and mass, so their falling behavior and accumulation methods are considered to be similar; finally, the falling event of the particle was determined; because the end point of the predicted trajectory falls within the range of the carrier platform and the particle carrying capacity of the carrier platform has not reached the maximum value, the particle is judged to have fallen successfully and is included in the current carrying capacity of the carrier platform (that is, the carrying capacity is updated to 800 grams + the mass of the particle); through this example, it can be seen that each sub-step in step S123 is interrelated and interdependent; together, they constitute a complete process for determining the falling event of drug particles based on the predicted particle trajectory, particle specifications and carrier platform carrying capacity.

[0050] In one embodiment of the present application, a drop event matching table is collected, and the drop event matching table is shown in Table 2: Table 2. Matching table of drop events

[0051] refer to Figure 4 In step S13, based on the detection of the falling events of the respective drug particles, the falling trajectories of the respective drugs relative to the drug bin are determined, and according to the falling trajectories of the respective drugs relative to the drug bin and the falling simulation of the monitoring space, a simulation image of the drug particles falling is determined; In the specific implementation process of the present invention, the specific steps are: S131: collecting the falling events of each drug particle, and determining the falling trajectory of each drug particle relative to the drug bin according to the falling events of each drug particle and the specifications of each drug particle; S132: In the monitoring space, the falling trajectory of each drug particle relative to the drug bin is presented as a trajectory line of different colors; S133: Determine the hierarchical image corresponding to each time node based on the multiple trajectory lines and the time nodes of the monitoring space, and determine the simulation picture of the drug particle falling according to each time node, the multiple hierarchical images and the corresponding trajectory lines.

[0052] In an embodiment of the present application, the falling events of each drug particle are collected, and the falling trajectory of each drug particle relative to the medicine bin is determined based on the falling events of each drug particle and the specifications of each drug particle. This takes into account the overall consideration of the falling events of each drug particle and the specifications of each drug particle, and ensures the accuracy of the falling trajectory of each drug particle relative to the medicine bin.

[0053] At this point, the entire process of each drug particle being discharged from the drug bin and falling onto the carrier (or other target location) is recorded, including information such as whether it has successfully fallen, the time and location of the fall. At this point, this usually relies on high-precision monitoring equipment such as sensors or cameras, which can monitor and record the movement state of the particles in real time. For example, a high-speed camera is used to capture the falling process of the particles, or an infrared sensor is used to detect the passing of the particles. The collected falling event data needs to be accurately recorded and stored in a computer system for subsequent analysis and processing.

[0054] Based on the collected drop event data and particle size information, the complete motion trajectory of each particle from discharge from the drug chamber to final fall is calculated. At the same time, particle size includes physical properties such as particle shape, size, and density, which affect the particle's motion state in the air (such as speed and direction). Trajectory estimation: Use physical models (such as Newton's second law, air resistance model, etc.) or machine learning algorithms (such as neural networks, support vector machines, etc.) to infer the trajectory of particles. These models or algorithms require as input the initial state of the particle (such as position, velocity), specification information, and external environmental parameters (such as gravitational acceleration, air density, etc.), and then output the position information of the particle at each point in time; the inferred trajectory is usually represented by a series of spatial coordinate points, which are connected to form the falling trajectory of the particle.

[0055] Furthermore, in the monitoring space, the falling trajectory of each drug particle relative to the drug bin is presented as a trajectory line of different colors.

[0056] At this point, the presentation range of the particle falling trajectory should be clarified, which is usually a three-dimensional or two-dimensional spatial area, including the drug bin, the particle falling path and the landing point (such as the supporting platform); at this time, the size and shape of the monitoring space should be determined according to the actual layout of the production line to ensure that the falling trajectory of all particles can be fully captured.

[0057] The calculated trajectory data for each drug particle relative to the drug chamber is obtained from the previous step (e.g., S131). This trajectory data is typically a series of spatial coordinate points representing the particle's position at different points in time. To distinguish different particle trajectories, each particle is assigned a unique color. The color scheme is a gradient based on a rainbow color band, which is a fixed set of colors that are used in a cycle. The colors should have sufficient contrast to visually distinguish different trajectories.

[0058] In the monitoring space, the corresponding trajectory lines are drawn according to the particle falling trajectory data. At this time, graphics drawing software or programming libraries (such as MATLAB, Python's matplotlib or Plotly, etc.) are used to draw the trajectory lines. When drawing, each coordinate point in the trajectory data needs to be connected to form a smooth curve or broken line. According to the color scheme selected in sub-step 3, the corresponding color is assigned to the trajectory line of each particle.

[0059] The drawn trajectory lines are presented in the monitoring space for visual analysis and observation; the presentation method: static images, as well as dynamic animations; animations can more intuitively show the falling process of particles; as needed, interactive functions can be added, such as hovering the mouse to display trajectory information, zooming in and out to view details, etc.

[0060] Specifically, the medicine bin continuously discharges medicine particles of different specifications; the falling trajectory data of each particle relative to the medicine bin has been calculated through the previous step (such as S131); In step S132, the monitoring space is first determined, which is a three-dimensional area that includes the drug chamber, the particle falling path, and the supporting platform; then, the particle falling trajectory data is obtained from step S131, which contains the spatial coordinates of each particle at different time points; next, a gradient color scheme based on a rainbow band is selected to assign a unique color to each particle, so that even if the trajectories of multiple particles overlap in space, they can be distinguished by color.

[0061] Then, Python's matplotlib library is used to draw the trajectory lines. Each coordinate point in the trajectory data is connected to form a smooth curve. At the same time, each trajectory line is assigned a corresponding color according to the color scheme. Finally, the drawn trajectory lines are presented in the form of a three-dimensional animation. In the animation, trajectory lines of different colors are seen starting from the medicine chamber mouth and falling along different paths to the support platform. By observing the animation, the falling process of each particle and the relative position relationship between them can be intuitively understood.

[0062] Therefore, based on multiple trajectory lines and time nodes of the monitoring space, the hierarchical images corresponding to each time node are determined, and the simulation picture of the drug particles falling is determined according to each time node, multiple hierarchical images and corresponding trajectory lines. It is compatible with the overall consideration of each time node, multiple hierarchical images and corresponding trajectory lines, ensuring the accuracy of the simulation picture of the drug particles falling. At the same time, it is compatible with the overall consideration of the falling trajectory of each drug and the detection space, ensuring the accuracy of the simulation picture of the drug particles falling.

[0063] At this time, in order to generate continuous simulation images, it is necessary to determine a series of time nodes, which will be used to extract the corresponding position information from the trajectory lines. At this time, the selection of time nodes should be determined according to the falling speed of the particles and the required simulation accuracy. For example, if the falling speed of the particles is fast and a high-precision simulation image is required, a shorter time interval should be selected. On the contrary, if the falling speed of the particles is slow and the simulation accuracy requirement is not high, a longer time interval should be selected.

[0064] At each time point, a two-dimensional or three-dimensional image representing the current position of the particles is generated based on the trajectory lines of all particles. This image is called a hierarchical image. At this time, the generation of the hierarchical image usually involves converting the position information of the trajectory lines at each time point into pixel points on the image. This is achieved through graphics drawing software or programming libraries, such as Python's matplotlib, Plotly, or three-dimensional graphics libraries (such as VTK, OpenGL, etc.). When generating the image, it is necessary to consider the particle's size, shape, color and other properties to ensure that the image can accurately reflect the actual state of the particles.

[0065] Combine hierarchical images from multiple time nodes in chronological order to form a dynamic simulation picture. At this point, this usually involves converting the hierarchical images into video frames and combining them into a coherent video using video editing software or programming libraries (such as FFmpeg, OpenCV, etc.). During the combination process, it is necessary to ensure that the transition between frames is smooth to avoid jumps or freezes.

[0066] In the simulation screen, in addition to the current position of the particles represented by the hierarchical image, trajectory lines need to be added to represent the movement path of the particles. At this time, the addition of trajectory lines is achieved by drawing lines on the video frame. This requires calculating the starting and end points of the lines based on the particle trajectory data at each time node and connecting them using appropriate drawing functions. In order to maintain the continuity of the trajectory, interpolation algorithms (such as linear interpolation, Bezier curves, etc.) are used to smoothly connect the position points on adjacent time nodes.

[0067] Optimize the simulation screen to improve its visual quality and readability, and output the final simulation results. Optimization includes adjusting parameters such as image brightness, contrast, and color saturation, as well as adding auxiliary information such as annotations, labels, or scales. The output format is a video file, GIF animation, or a series of static images, depending on application requirements and user preferences.

[0068] Specifically, the medicine chamber continuously discharges drug particles of different specifications and colors; the falling trajectory data and color information of the particles have been obtained through previous steps (such as S131 and S132); in step S133, the time node is first determined, and a sampling rate of 10 frames per second is selected to capture the falling process of the particles, which means that at each time node, the position information of the particles will be extracted from the trajectory line.

[0069] Then, hierarchical images were generated for each time node. These images are two-dimensional, in which each pixel represents the position of the particle at that time node; different colors are used to represent particles of different specifications to ensure that the image can clearly reflect the actual state of the particles; next, these hierarchical images are combined in chronological order to form a dynamic simulation picture; during the combination process, smooth transitions between frames are ensured to avoid jumping or freezing; in order to more intuitively display the movement path of the particles, trajectory lines are added to the simulation picture. These lines connect the position points on adjacent time nodes and are represented by the same color as the particles; in order to maintain the continuity of the trajectory, a linear interpolation algorithm is used to smoothly connect these position points; finally, the simulation picture is optimized, and parameters such as image brightness, contrast and color saturation are adjusted, and auxiliary information such as scale and annotations are added; finally, the simulation results are output as a video file for subsequent analysis and presentation.

[0070] In one embodiment of the present application, first, a series of time nodes need to be determined, and corresponding hierarchical images need to be generated for each time node. These hierarchical images reflect the positions of all drug particles in the monitoring space at that time point. To simplify the description, a time node matching table is used to record this information. The time node matching table is shown in Table 3: Table 3 Time node matching table

[0071] In order to more accurately describe the importance of the state and trajectory of particles at each time node, the concepts of weight and score are introduced; the weight is determined by the particle's size, speed, color and other attributes, while the score is calculated based on the weight and the particle's position at each time node; For each particle at each time node, a score is calculated based on its weight and position; for example, if the particle is located at a key position of the trajectory (such as the starting point, turning point, or end point), the score is higher; the score is calculated by the following formula: score = weight × position factor (the position factor is determined according to the position of the particle in the trajectory, for example, the factor is higher for key positions).

[0072] With information on time nodes, hierarchical images, weights, and scores, we begin to determine the simulation screen of drug particle dropping. Based on the matching table, we draw the corresponding hierarchical image for each time node. On the hierarchical image, we draw lines based on the particle trajectory data, and use different colors or thicknesses to represent different weights or scores. The hierarchical images and trajectory lines at multiple time nodes are combined to form a dynamic simulation screen. This is achieved through video editing software or programming libraries (such as FFmpeg, OpenCV, etc.).

[0073] Specifically, the final simulation image is a video file that contains the entire process from the initial state to the final accumulation form; at each time node, the trajectory lines of the particles and their position in space can be seen; by adjusting the distribution of weights and scores, the process of falling and accumulation of particles of different specifications and types can be more accurately simulated.

[0074] refer to Figure 5 In step S14, a plurality of flow velocity layers are formed based on the division of the simulation screen of the drug particle falling, and the plurality of flow velocity layers are respectively a central fast layer, an inner wall slow layer and a transitional velocity layer; In the specific implementation process of the present invention, the specific steps are: S141: collecting a simulation image of the drug particles falling, and marking the speed of each drug particle at different height positions based on the simulation image of the drug particles falling; S142: forming a plurality of speed ranges according to the classification of the speeds of the respective drug particles at different height positions, wherein the speed ranges are related to the specifications and height positions of the drug particles; S143: Based on the multiple speed ranges and the corresponding height positions, multiple velocity layers are formed, wherein the multiple velocity layers are a central fast layer, an inner wall slow layer, and a transition velocity layer. The speeds of the central fast layer, the inner wall slow layer, and the transition velocity layer are different and are targeted at different height positions.

[0075] In an embodiment of the present application, a simulation screen of drug particle falling is collected, and the speed of each drug particle at different height positions is marked based on the simulation screen of drug particle falling, thereby introducing the speed of each drug particle at different height positions.

[0076] At this point, obtain simulation images of the dynamic process of drug particle falling. These images should be able to clearly show the movement state of particles at different height positions. At this time, use professional simulation software (such as CFD simulation software, discrete element method (DEM) simulation software, etc.) to simulate the drug particle falling process. After the simulation is completed, export images or video files containing particle motion information from the software. These files are usually stored in image sequence or video format. Ensure that the exported images are of high enough quality so that the particles can be accurately identified and tracked later.

[0077] In the captured simulation image, the speed of each drug particle at different height positions is marked; at this time, image processing techniques (such as edge detection, threshold segmentation, template matching, etc.) are used to identify the particles in the image; the identified particles are tracked, which usually involves matching the positions of the particles between consecutive image frames to form a complete motion trajectory; on the particle trajectory, different height positions are selected as key points, and the position changes and time intervals between adjacent key points are used to calculate the instantaneous speed of the particles when they pass through these key points; the calculated speed data is associated with the corresponding height position and particle identification to form a data set, which usually involves superimposing speed markers (such as arrows, digital labels, etc.) on the simulation image, or using an external data table to record this information.

[0078] Specifically, assume that the dropping process of round drug particles in a cylindrical medicine bin is being analyzed; the diameter of the particles ranges from 2 mm to 5 mm, and the height of the medicine bin is 1 m; DEM simulation software is used to simulate this process, and a series of simulation images containing particle motion information are exported; in step S141, these simulation images are first imported into the image processing software; then, edge detection and threshold segmentation techniques are used to identify the particles in the image; then, the identified particles are tracked, and a complete motion trajectory is formed by matching the positions of the particles in consecutive image frames; based on the trajectory tracking, different height positions in the medicine bin (such as 0.2 m, 0.5 m, 0.8 m) are selected as key points; the instantaneous velocity of each particle when passing through these key points is calculated using the position change and time interval between adjacent key points; finally, the calculated velocity data is associated with the corresponding height position and particle identification, and a velocity arrow is superimposed on the simulation image to indicate the velocity direction and size of the particle.

[0079] Furthermore, a plurality of speed ranges are formed according to the classification of the speeds of the respective drug particles at different height positions. The speed ranges are related to the specifications and height positions of the drug particles, and a plurality of speed ranges are introduced.

[0080] At this point, the velocity data of each drug particle at different height positions is collected from the previous step (e.g., S141). At this point, the data integrity and accuracy are ensured, including the particle specifications (e.g., diameter, shape, etc.), height position, and corresponding velocity values. These data are usually stored in the form of a data set or table, with each row representing the velocity information of a particle at a certain height position.

[0081] The collected speed data is preprocessed to eliminate outliers, fill missing values, and ensure data quality. Statistical methods (such as box plots and Z scores) are used to detect and process abnormal speed values, which are caused by measurement errors or unstable factors in the simulation. Missing speed data are filled by interpolation based on the trends of adjacent data points, or other reasonable estimation methods are used.

[0082] The velocity data is divided into multiple velocity ranges based on the specifications and height positions of the drug particles. At this point, a preliminary classification is performed based on the particle specifications (such as diameter size), because particles of different specifications have different velocity characteristics under the same conditions. Within each specification category, the velocity is further subdivided based on the height position, because particles are affected by multiple factors such as gravity, air resistance, and inter-particle interactions during the falling process, causing the velocity to vary with height. Statistical methods (such as cluster analysis, histogram analysis, etc.) or experience-based rules are used to determine the velocity range. Cluster analysis divides the velocity data into naturally formed groups, while histogram analysis helps identify the distribution pattern of the velocity.

[0083] The classified speed ranges are associated with the specifications and height positions of the drug particles to form a complete speed classification system. At this point, a table or database is created in which each row represents a speed range, including the upper and lower limits of the speed range, the corresponding particle specifications, and the height position information. By comparing with actual observation data or experimental results, the accuracy and rationality of the speed classification system are verified, and adjustments are made as needed.

[0084] Specifically, assume that the dropping process of drug particles containing two specifications (large particles and small particles) in a cylindrical medicine bin is being analyzed; a large amount of velocity data of particles at different heights has been collected from step S141; in step S142, the data is first preprocessed, abnormal velocity values ​​caused by measurement errors are deleted, and a small amount of missing velocity data is filled using the interpolation method; next, a preliminary classification is performed according to the specifications of the particles; for large particles and small particles, the average velocity at different heights (such as 0.2m, 0.5m, and 0.8m) is calculated respectively, and the trend of velocity change with height is observed.

[0085] Then, the speed data was further subdivided using the cluster analysis method; for each specification of particles, they were divided into several speed ranges, such as "low speed", "medium speed" and "high speed" according to the natural distribution of speed values, and the upper and lower limits of these speed ranges were determined based on the clustering results; finally, an association table was established between speed ranges and particle specifications and height positions; for example, for large particles, at a height of 0.2m, the speed ranges were "low speed" (0-1 m / s), "medium speed" (1-2 m / s) and "high speed" (>2 m / s); at a height of 0.8m, the speed ranges would be different due to the influence of air resistance and inter-particle interaction; through this example, we can see that each sub-step in step S142 is interrelated, and together they constitute a complete process for classifying speeds according to the specifications and height positions of drug particles and forming multiple speed ranges. This information provides an important basis for subsequent drug particle blanking analysis and optimization.

[0086] Therefore, multiple velocity layers are formed based on multiple velocity ranges and corresponding height positions. The multiple velocity layers are respectively a central fast layer, an inner wall slow layer and a transition velocity layer. There are differences in the velocities of the central fast layer, the inner wall slow layer and the transition velocity layer. The central fast layer, the inner wall slow layer and the transition velocity layer are introduced for different height positions.

[0087] At this time, the relationship between the multiple velocity ranges and the corresponding height positions determined in step S142 is analyzed in depth, which is the basis for forming the velocity layer; at this time, the velocity classification data is reviewed, and attention is paid to the velocity change trends of particles of different specifications at different heights, and how these trends change with changes in particle specifications and height positions.

[0088] Based on the correlation between velocity range and height position, three main velocity layers are defined: central fast layer, inner wall slow layer and transition velocity layer; at this time, central fast layer: usually located in the center of the flow area, composed of faster particles, these particles have larger specifications or higher falling potential energy; inner wall slow layer: close to the inner wall of the flow area, composed of slower particles, these particles are affected by friction resistance, inter-particle interaction or boundary effect; transition velocity layer: located between the central fast layer and the inner wall slow layer, the particle velocity is between the two, and it is the area with the most significant velocity gradient change.

[0089] The boundaries of each velocity layer are clearly defined so that they can be accurately identified and distinguished in simulation images or actual experiments. At this time, a clear velocity threshold is set for each velocity layer based on the velocity range. Particles with velocity exceeding a certain threshold belong to the central fast layer, particles below another threshold belong to the inner wall slow layer, and particles in between belong to the transition velocity layer. Considering the change in velocity with height, the boundaries of the velocity layer are not fixed, but vary with the height position. Therefore, the boundaries of the velocity layer need to be determined separately for each height position.

[0090] Verify the rationality of the defined velocity layers by comparing them with actual observation data or simulation results. Overlay velocity layer markers on the simulation screen to see if the actual particle movement matches the defined velocity layers. Perform statistical analysis on the particle velocities within the velocity layers to check if the velocity distribution meets expectations.

[0091] Specifically, assume that the dropping process of drug particles containing two specifications (large particles and small particles) in a cylindrical drug bin is being analyzed; in step S142, multiple speed ranges have been determined, and their relationship with particle specifications and height positions has been understood; in step S143, the changing trends of these speed ranges with height are first deeply analyzed; it is found that large particles generally have a faster speed in the central area, forming a central fast layer; while small particles, due to the influence of boundary effects, have a slower speed near the inner wall, forming an inner wall slow layer; particles between the two form a transition speed layer.

[0092] Next, clear velocity layer boundaries were defined for each height position. For example, at a height of 0.2 m, particles with velocities greater than 1.5 m / s were assigned to the central rapid layer, particles with velocities less than 0.5 m / s to the inner wall slow layer, and particles with velocities between 0.5 m / s and 1.5 m / s to the transition velocity layer. These thresholds were determined based on a comprehensive consideration of the velocity range and particle size. Velocity layer markers were then superimposed on the simulation screen, and the rationality of the velocity layers was verified through visualization. It was found that the actual motion trajectories of the particles were highly consistent with the defined velocity layers: particles in the central rapid layer fell rapidly, particles in the inner wall slow layer had slower velocities, and particles in the transition velocity layer had velocities between the two. Finally, a statistical analysis of the particle velocities within each velocity layer was performed, further verifying the rationality of the velocity layers. The statistical results showed that the particle velocity distribution in the central rapid layer was more concentrated, with higher velocity values; the particle velocity distribution in the inner wall slow layer was more dispersed, with lower velocity values; and the particle velocity in the transition velocity layer showed a clear gradient.

[0093] In one embodiment of the present application, a velocity layer matching table is collected, which intuitively displays the correspondence between velocity range, height position and velocity layer; the velocity layer matching table is shown in Table 4: Table 4 Velocity layer matching table Height position (m) Speed ​​range (m / s) Velocity layer 0.2 >1.5 Central Rapid Layer 0.2 0.5-1.5 Transition Speed ​​Layer 0.2 <0.5 Inner wall slow layer 0.5 >1.2 Central Rapid Layer 0.5 0.4-1.2 Transition Speed ​​Layer 0.5 <0.4 Inner wall slow layer ... ... ... 0.8 >1.0 Central Rapid Layer 0.8 0.3-1.0 Transition Speed ​​Layer 0.8 <0.3 Inner wall slow layer In this velocity layer matching table, different velocity ranges are listed for each height position, and these ranges are matched with the corresponding velocity layers; for example, at a height of 0.2 m, particles with a velocity greater than 1.5 m / s are classified as the central fast layer, particles with a velocity between 0.5-1.5 m / s are classified as the transition velocity layer, and particles with a velocity less than 0.5 m / s are classified as the inner wall slow layer.

[0094] refer to Figure 6 In step S15, the residual particle amounts of the multiple flow rate layers are determined according to the speeds of the multiple flow rate layers and the proportions of the drug particles in the multiple flow rate layers, and the posture adjustment of the medicine bin is triggered to avoid material accumulation in the medicine bin after each discharge; In the specific implementation process of the present invention, the specific steps are: S151: In each flow rate layer, a simulation image of the flow rate layer at a preset time point is collected, and the total amount of drug particles is determined based on the traversal of the simulation image, and the distribution quantity of each type of drug particles is determined based on the total amount of drug particles and the specifications of the drug particles; S152: Determine the proportion of drug particles in the velocity layer based on a comparison of the distribution quantities of each type of drug particles; determine the corresponding residual layer based on a comparison of the flow velocity layers and the area of ​​the drug bin, determine a drug particle residual image based on tracing the residual layer, and determine the amount of residual particles in the velocity layer based on the detection of the drug particle residual image; S153: Determine the residual layer with the maximum residual particle amount based on the residual particle amount of each flow rate layer and the height position of each flow rate layer, and determine the residual position of the drug particles based on the detection of the residual layer with the maximum residual particle amount, and trigger the posture adjustment of the drug bin according to the flow rate of the flow rate layer corresponding to the residual position of the drug particles and the current posture of the drug bin, so that the current posture of the drug bin changes, and gradually regulates the accumulation of drug particles at the residual position of the drug particles.

[0095] In an embodiment of the present application, in each flow rate layer, a simulation image of the flow rate layer at a preset time point is collected, and the total amount of drug particles is determined based on the traversal of the simulation image, and the distribution quantity of each type of drug particles is determined based on the total amount of drug particles and the specifications of the drug particles, which is compatible with the overall consideration of the total amount of drug particles and the specifications of the drug particles, and ensures the accuracy of the distribution quantity of each type of drug particles.

[0096] At this point, obtain simulation images at preset time points in each velocity layer for subsequent analysis; at this point, use simulation software or simulation tools to capture images of the velocity layer at preset time points (such as a period of time after the particle flow stabilizes); ensure that the images are clear and can accurately reflect the distribution and number of particles; the images should contain sufficient details to facilitate subsequent identification and counting of particles.

[0097] By traversing the simulation screen, the total amount of drug particles in the flow layer is calculated; at this time, image processing techniques such as edge detection and binarization are used to separate the particles in the screen from the background; the processed screen is traversed, each particle is identified and counted; the number of all identified particles is accumulated to obtain the total amount of particles in the flow layer.

[0098] Based on the total amount and specifications of the particles, the distribution number of each type of drug particles is determined; at this time, the identified particles are classified and divided into different categories according to their specifications (such as size and shape); the particles in each category are counted to obtain the distribution number of each type of particles; the spatial distribution of each type of particles in the flow layer is further analyzed, such as uniformity, aggregation, etc.

[0099] Specifically, assume that the flow of a medicine bin containing two sizes of particles (large particles and small particles) is being analyzed; some time after the flow in the medicine bin stabilizes, simulation software is used to capture images of the central fast layer, transition speed layer, and inner wall slow layer at preset time points; the images are clear and can accurately reflect the distribution and number of particles; the captured images are processed, and edge detection and binarization techniques are used to separate the particles from the background; the processed images are traversed to identify and count each particle; for example, in the central fast layer, 1,000 large particles and 800 small particles are identified and counted, so the total number of particles in this layer is 1,800.

[0100] The identified particles are classified into two categories: large particles and small particles; the particles in each category are counted to obtain the distribution number of each type of particles; for example, in the central rapid layer, the distribution number of large particles is 1,000, and the distribution number of small particles is 800; further analysis shows that large particles are more evenly distributed in the central rapid layer, while small particles have a certain degree of aggregation. This information is of great significance for the subsequent drug warehouse design and optimization; through the above examples, we can see that each sub-step in step S151 is interrelated, and they together constitute a complete process for determining the total amount and distribution number of drug particles in each flow rate layer. This information provides important basic data for the subsequent drug warehouse flow analysis and optimization.

[0101] Furthermore, the proportion of drug particles in the flow rate layer is determined based on the comparison of the distribution numbers of various types of drug particles; the corresponding residual layer is determined based on the regional comparison of each flow rate layer and the medicine warehouse, and the residual image of the drug particles is determined based on the tracing of the residual layer. The amount of residual particles in the flow rate layer is determined based on the detection of the residual image of the drug particles, thereby ensuring the accuracy of the amount of residual particles in the flow rate layer.

[0102] At this point, by comparing the distribution number of each type of drug particles, the proportion of each type of particles in a certain flow rate layer is determined; at this point, the distribution number data of each type of drug particles in each flow rate layer are collected; the data are normalized, that is, the proportion of the number of each type of particles to the total number of particles in the flow rate layer is calculated; Analyze the proportion data to understand the distribution of particle types in different flow velocity layers.

[0103] By comparing each flow rate layer and the overall area of ​​the medicine chamber, the flow rate layer with the most significant particle accumulation, namely the residual layer, is identified; at this time, the particle flow conditions of each flow rate layer are analyzed, especially the accumulation and retention of particles; combined with the structural characteristics of the medicine chamber, the area where particles are likely to accumulate is identified; based on the above information, the location of the residual layer is determined.

[0104] The historical simulation images of the residual layer are traced back to extract images of drug particle residues. At this point, images of the residual layer at key time points are searched and extracted in the simulation software. The extracted images are pre-processed, such as denoising and contrast enhancement, to more clearly display the particle residue situation. The processed image is saved as a drug particle residue image.

[0105] Based on the residual image of drug particles, the particle amount in the residual layer is detected and determined. At this time, image processing technology (such as particle recognition algorithm) is used to process the residual image to identify and count the residual particles. Based on the particle recognition results, the total number or total area of ​​the residual particles is calculated (depending on whether the particle size is uniform). The changing trend of the residual particle amount is analyzed to provide a basis for subsequent adjustments and optimization.

[0106] Specifically, assume that the flow of a medicine bin containing three sizes of particles (large, medium, and small) is being analyzed, and the data collection in step S151 has been completed; the distribution quantity data of various types of particles in the central fast layer, transition velocity layer, and inner wall slow velocity layer have been collected; for example, in the central fast layer, large particles account for 30%, medium particles account for 40%, and small particles account for 30%. By comparing these data, the distribution of particle types in different flow velocity layers can be understood.

[0107] When analyzing the particle flow, it was found that particle accumulation was most significant in the slow layer on the inner wall, especially in certain corners and bends of the medicine warehouse. Combined with the structural characteristics of the medicine warehouse, the slow layer on the inner wall was determined to be a residual layer. Images of the slow layer on the inner wall at key time points were searched and extracted in the simulation software. The extracted images were preprocessed to enhance the contrast to more clearly display the particle residue. The processed images were saved as drug particle residue images.

[0108] The residual image was processed using a particle recognition algorithm to identify and count the residual particles. For example, 500 large particles, 800 medium particles, and 300 small particles were identified in the residual image. Based on the recognition results, the total number of residual particles (1,600 particles) and the proportion of each type of particle were calculated. By analyzing the changing trend of the residual particle amount, the main areas and trends of particle accumulation were discovered, providing a basis for subsequent adjustments and optimizations. For example, it is necessary to adjust the structure of the medicine bin or change the way the particles are delivered to reduce the accumulation of particles in the slow layer on the inner wall. From the above examples, we can see that each sub-step in step S152 is interrelated, and together they constitute a complete process for determining the position of the residual layer, extracting the residual image, and detecting the amount of residual particles. This information provides important basic data for subsequent medicine bin flow analysis and optimization.

[0109] Therefore, the residual layer with the maximum residual particle amount is determined according to the residual particle amount of each flow rate layer and the height position of each flow rate layer, and the residual position of the drug particles is determined based on the detection of the residual layer with the maximum residual particle amount, and the posture adjustment of the medicine warehouse is triggered according to the residual position of the drug particles and the flow velocity of the flow rate layer corresponding to the current posture of the medicine warehouse, so that the current posture of the medicine warehouse is changed, and the accumulation of drug particles in the residual position of the drug particles is gradually regulated, which is compatible with the overall consideration of the speed of multiple flow rate layers and the proportion of drug particles in multiple flow rate layers, ensures the accuracy of the residual particle amount of multiple flow rate layers, and triggers the posture adjustment of the medicine warehouse to optimize the residual particle amount of multiple flow rate layers, and avoid material accumulation in the medicine warehouse after each discharge.

[0110] At this time, by comparing the residual particle amounts of each flow rate layer, find out the flow rate layer with the largest residual particle amount; at this time, summarize the residual particle amount data of each flow rate layer; compare the data and find out the flow rate layer with the largest residual particle amount. This layer is the residual layer that needs to be paid special attention to.

[0111] In the velocity layer with the maximum amount of residual particles, the specific residual location of the drug particles is determined; at this time, the residual particle image or simulation data of the velocity layer is analyzed to identify the area where the particles are aggregated or accumulated; combined with the structural characteristics of the drug warehouse, the specific location of the residual particles is determined.

[0112] According to the residual position of the drug particles and the flow velocity of the flow layer corresponding to the current posture of the medicine warehouse, the posture of the medicine warehouse is adjusted to change the flow path of the particles; at this time, the flow velocity distribution of each flow velocity layer under the current posture of the medicine warehouse is analyzed; according to the residual position and flow velocity distribution, the direction and degree of adjustment of the medicine warehouse posture, such as the tilt angle, rotation angle, etc., are determined; the posture adjustment mechanism of the medicine warehouse is triggered to change the posture of the medicine warehouse.

[0113] By continuously observing the flow of particles after the medicine chamber posture is adjusted, the medicine chamber posture is gradually adjusted to reduce or eliminate the accumulation of drug particles. At this time, after the medicine chamber posture is adjusted, the flow of particles is continuously observed. Based on the observation results, the medicine chamber posture is gradually fine-tuned until the ideal particle distribution state is achieved. The key parameters and effects during the adjustment process are recorded to provide a reference for subsequent medicine chamber design and optimization.

[0114] Specifically, it is assumed that in step S152, the slow layer on the inner wall is determined to be the flow velocity layer with the largest amount of residual particles, and the residual particle image is extracted; the residual particle amount data of each flow velocity layer are summarized, and it is found that the residual particle amount of the slow layer on the inner wall is the largest, reaching 1,200 particles; therefore, the slow layer on the inner wall is determined to be the residual layer that needs special attention.

[0115] Analysis of the residual particle images in the slow inner layer revealed that the particles primarily accumulated in the bottom corners and bends of the chamber. Combining the chamber's structural characteristics, the specific locations of the residual particles were determined. Analysis of the velocity distribution of each velocity layer in the chamber's current posture revealed that the slow inner layer exhibited a low velocity, making particle accumulation more likely. To alter the particle flow path, the inclination of the chamber was adjusted to increase the velocity in the slow inner layer. This triggered the chamber's posture adjustment mechanism, tilting it to one side at a specific angle.

[0116] After the medicine bin posture was adjusted, the flow of particles was continuously observed. It was found that after the tilt angle was adjusted, the flow speed of particles in the slow layer on the inner wall increased, but some particles still gathered in the bottom corners. Therefore, the tilt angle and rotation angle of the medicine bin were gradually fine-tuned until the distribution of particles in the medicine bin became more uniform. The key parameters and effects of the adjustment process were recorded, and it was found that when the tilt angle of the medicine bin was 15 degrees and the rotation angle was 30 degrees, the accumulation of particles was significantly improved. Through the above examples, it can be seen that each sub-step in step S153 is interrelated, and they together constitute a complete process for determining the residual layer, residual position, triggering the adjustment of the medicine bin posture, and gradually regulating the accumulation of particles. This information provides an important reference for the subsequent medicine bin design and optimization.

[0117] In one embodiment of the present application, it is assumed that there are three flow velocity layers: A (low altitude), B (medium altitude), and C (high altitude), and the residual particle amount of each flow velocity layer has been obtained through the previous steps; a residual particle amount matching table is created, and the residual particle amount matching table is shown in Table 5: Table 5 Residual particle quantity matching table

[0118] The residual particle count matching table shows that velocity layer B has the largest residual particle count, making it the residual layer of interest. Assume that residual particle image analysis reveals that particles in layer B are primarily concentrated in a specific area of ​​the chamber (e.g., the bottom corner). This area represents the residual location. Based on the residual location and the velocity of the velocity layer corresponding to the chamber's current posture, adjustments to the chamber's posture are made. For example, if the chamber is currently horizontal, a slight tilt may be made to increase the flow velocity in layer B and aid particle flow. After adjusting the chamber's posture, the particle flow is continuously observed, and the chamber's posture is gradually fine-tuned as needed until particle accumulation improves.

[0119] See also Figure 7 , Figure 7 : is a schematic diagram of the structural composition of a simulation system for particle dropping from a medicine bin in an embodiment of the present invention; the simulation system for particle dropping from a medicine bin comprises: The monitoring space module 21 is used to determine the monitoring space based on the current position of the medicine bin and the previous falling trajectory of the medicine particles; A falling event module 22 is used to determine the falling event of each drug particle in the drug bin according to the position node, time and specifications of each drug particle during the falling process during dynamic detection in the monitoring space; a simulation image module 23 for determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining a simulation image of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space; A velocity layer module 24 is configured to form a plurality of velocity layers based on the division of the simulation screen of the drug particle falling, wherein the plurality of velocity layers are a central fast layer, an inner wall slow layer, and a transition velocity layer; The posture adjustment module 25 is used to determine the amount of residual particles in multiple flow rate layers according to the speed of multiple flow rate layers and the proportion of drug particles in multiple flow rate layers, and trigger the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge.

[0120] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A simulation method for particle dropping in a medicine bin, characterized in that: include: Determine the monitoring space based on the current position of the drug warehouse and the previous falling trajectory of the drug particles; In the dynamic detection of the monitoring space, the falling events of each drug particle in the drug bin are determined according to the position node, time and specifications of each drug particle during the falling process; Determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining a simulated image of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space; Based on the division of the simulation screen of drug particle falling, multiple flow velocity layers are formed, including a central fast layer, an inner wall slow layer and a transition velocity layer; The amount of residual particles in multiple flow rate layers is determined according to the speed of the multiple flow rate layers and the proportion of drug particles in the multiple flow rate layers, and the posture adjustment of the medicine bin is triggered to avoid material accumulation in the medicine bin after each discharge.

2. The simulation method for particle dropping of a medicine bin according to claim 1, characterized in that: The method of determining the monitoring space based on the current position of the medicine bin and the previous falling trajectory of the medicine particles includes: In the drug testing room, the drug warehouse is positioned and detected based on the camera in the drug testing room, and the current image of the drug warehouse is collected. The current position of the drug warehouse is determined based on the matching of the current image of the drug warehouse and the indoor distribution map of the drug testing room; Based on the peripheral detection of the current position of the medicine warehouse, the corresponding medicine particle carrying platform is collected, and the relative space between the medicine warehouse and the medicine particle carrying platform is determined based on the spatial comparison between the medicine warehouse and the medicine particle carrying platform; The previous falling trajectory of the drug particles is determined based on the current position of the drug bin and the drug bin database, and the monitoring space is determined according to the previous falling trajectory of the drug particles and the relative space between the drug bin and the drug particle carrying platform.

3. The simulation method for particle dropping of a medicine bin according to claim 1, characterized in that: In the dynamic detection of the monitoring space, the falling events of the respective drug particles in the drug bin are determined according to the position nodes, time and specifications of the drug particles during the falling process, including: Dynamic detection is performed on the monitoring space, and based on the dynamic detection of the monitoring space, each drug particle discharged through the discharge port of the drug bin is marked, and a position node of each drug particle during the falling process is determined according to the spatial position of the mark of each drug particle, so as to collect the position node of each drug particle during the falling process; In each drug particle, a predicted trajectory of the drug particle is determined based on connections of a plurality of position nodes of the drug particle; The particle carrying capacity of the drug particle carrying platform is collected, and the falling event of each drug particle is determined based on the predicted trajectories of the multiple drug particles, the specifications of the multiple drug particles, and the particle carrying capacity of the drug particle carrying platform.

4. The method for simulating particle dropping from a medicine bin according to claim 1, characterized in that: The method of determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining the simulation screen of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space, includes: The falling events of each drug particle are collected, and the falling trajectory of each drug particle relative to the drug bin is determined according to the falling events of each drug particle and the specifications of each drug particle.

5. The method for simulating particle dropping from a medicine bin according to claim 4, characterized in that: Based on the detection of the falling events of each drug particle, the falling trajectory of each drug relative to the drug bin is determined, and according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space, a simulation picture of the drug particle falling is determined, which also includes: In this monitoring space, the falling trajectory of each drug particle relative to the drug bin is presented as trajectory lines of different colors; Based on multiple trajectory lines and time nodes of the monitoring space, the hierarchical image corresponding to each time node is determined, and the simulation picture of the drug particle falling is determined according to each time node, the multiple hierarchical images and the corresponding trajectory lines.

6. The method for simulating particle dropping from a medicine bin according to claim 1, characterized in that: The division of the simulation screen based on the drug particle falling is used to form multiple flow velocity layers, and the multiple flow velocity layers are respectively a central fast layer, an inner wall slow layer and a transition velocity layer, including: A simulation screen of drug particle falling is collected, and the speed of each drug particle at different height positions is marked based on the simulation screen of drug particle falling.

7. The method for simulating particle dropping from a medicine bin according to claim 6, characterized in that: The method further comprises: forming a plurality of flow velocity layers based on the division of the simulation screen of the drug particle falling, wherein the plurality of flow velocity layers are respectively a central fast layer, an inner wall slow layer and a transitional velocity layer; A plurality of speed ranges are formed according to the classification of the speeds of the individual drug particles at different height positions, wherein the speed ranges are related to the specifications and height positions of the drug particles; Based on multiple speed ranges and corresponding height positions, multiple velocity layers are formed. The multiple velocity layers are a central fast layer, an inner wall slow layer, and a transition velocity layer. The speeds of the central fast layer, the inner wall slow layer, and the transition velocity layer are different and target different height positions.

8. The method for simulating particle dropping from a medicine bin according to claim 1, characterized in that: The method of determining the residual particle amounts of the multiple flow rate layers according to the speeds of the multiple flow rate layers and the proportions of the drug particles in the multiple flow rate layers, and triggering the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge, includes: In each flow rate layer, a simulation image of the flow rate layer at a preset time point is collected, and the total amount of drug particles is determined based on the traversal of the simulation image, and the distribution quantity of each type of drug particles is determined based on the total amount of drug particles and the specifications of the drug particles; The proportion of drug particles in the flow rate layer is determined based on the comparison of the distribution quantity of each type of drug particles; the corresponding residual layer is determined based on the regional comparison of each flow rate layer and the medicine warehouse, and the residual image of the drug particles is determined based on the tracing of the residual layer, and the amount of residual particles in the flow rate layer is determined based on the detection of the residual image of the drug particles.

9. The method for simulating particle dropping from a medicine bin according to claim 8, characterized in that: The method further includes determining the amount of residual particles in the plurality of flow rate layers according to the speeds of the plurality of flow rate layers and the proportions of the drug particles in the plurality of flow rate layers, and triggering the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge. The residual layer with the maximum residual particle amount is determined based on the residual particle amount of each flow rate layer and the height position of each flow rate layer, and the residual position of the drug particles is determined based on the detection of the residual layer with the maximum residual particle amount. The posture adjustment of the drug bin is triggered according to the flow rate of the flow rate layer corresponding to the residual position of the drug particles and the current posture of the drug bin, so that the current posture of the drug bin changes, and the accumulation of drug particles at the residual position of the drug particles is gradually regulated.

10. A simulation system for particle dropping in a medicine bin, characterized in that: The simulation system for particle dropping from a medicine bin is applied to the simulation method for particle dropping from a medicine bin as claimed in any one of claims 1 to 9, and the simulation system for particle dropping from a medicine bin comprises: The monitoring space module is used to determine the monitoring space based on the current position of the medicine warehouse and the previous falling trajectory of the medicine particles; A falling event module is used to determine the falling events of each drug particle in the drug bin according to the position node, time and specifications of each drug particle during the falling process during dynamic detection in the monitoring space; a simulation screen module for determining the falling trajectory of each drug relative to the drug bin based on the detection of the falling events of each drug particle, and determining a simulation screen of the drug particle falling according to the falling trajectory of each drug relative to the drug bin and the falling simulation of the monitoring space; A velocity layer module is used to form multiple velocity layers based on the division of the simulation screen of drug particle falling, and the multiple velocity layers are respectively a central fast layer, an inner wall slow layer and a transition velocity layer; The posture adjustment module is used to determine the amount of residual particles in multiple flow rate layers according to the speed of multiple flow rate layers and the proportion of drug particles in multiple flow rate layers, and trigger the posture adjustment of the medicine bin to avoid material accumulation in the medicine bin after each discharge.

Images