Control method and system for injection ampoule bottle filling and sealing machine

Abstract

The invention provides a control method and system for an injection ampoule bottle filling and sealing machine, and relates to the technical field of automation control. Collecting bottle body data information of the injection ampoule bottle; obtaining filling and sealing data information of the injection ampoule bottle filled and sealed by the filling and sealing machine; the bottle body data information and the filling and sealing data information are input into a vacuum self-adaptive adjusting model embedded in a vacuum suction cup platform, first self-adaptive vacuum parameters are output, and the first self-adaptive vacuum parameters are used for controlling a vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle; after the injection ampoule bottle is fixed, a filling and sealing execution instruction is sent to the filling and sealing machine, and the filling and sealing machine carries out filling and sealing on the injection ampoule bottle according to the filling and sealing data information. The technical problem that an existing injection ampoule bottle filling and sealing machine is fixed in control parameter and cannot be adaptively adjusted according to specification parameters of an ampoule bottle, so that adhesion is prone to occurring in the filling and sealing process is solved, and the filling and sealing success rate and the filling and sealing quality are improved.

Description

Technical Field

[0001] The present application relates to the field of automation control technology, and in particular to a control method and system for an injection ampoule filling machine. Background Art

[0002] In the pharmaceutical industry, injection filling is a key production link that is directly related to the quality and safety of drugs. At present, the filling method of injection ampoules mainly relies on mechanical equipment, using vacuum suction cups and other technologies to fix the ampoules for liquid filling. These methods are usually operated with preset parameters and are suitable for bottles of specific specifications.

[0003] However, as the pharmaceutical industry's requirements for product diversity and production efficiency continue to increase, when the production line needs to handle ampoules of different shapes or specifications, operators often need to manually adjust equipment parameters, which not only increases the workload, but also easily leads to human errors, resulting in unstable filling quality. Secondly, the linkage control mechanism of traditional filling equipment is relatively simple, and there is a lack of effective coordination between the fixing and filling processes. Each link works independently. When faced with complex bottle features, it is difficult to accurately adjust according to the needs of different drugs, resulting in insufficient fixing force, resulting in adhesion between the ampoule and the filling head, causing damage to the ampoule or filling failure, seriously affecting production efficiency and product quality. Summary of the invention

[0004] The present application provides a control method and system for an injection ampoule filling machine, which solves the technical problem that the control parameters of the existing injection ampoule filling machine are fixed and cannot be adaptively adjusted according to the specification parameters of the ampoule, resulting in unstable fixation of the ampoule during the filling process and reduced filling quality, and achieves the technical effect of accurately fixing the ampoule to improve the filling success rate and the filling quality of the ampoule.

[0005] In view of the above problems, on the one hand, the present application provides a control method for an injection ampoule filling machine, the method comprising: setting a vacuum suction cup platform, the control terminal of the vacuum suction cup platform is communicatively connected to the control terminal of the filling machine, and vacuum suction cups are distributed on the vacuum suction cup platform; collecting bottle body data information of the injection ampoule; obtaining filling data information of the injection ampoule filled by the filling machine; inputting the bottle body data information and the filling data information into a vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and outputting a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule; when the injection ampoule is fixed, sending a filling execution instruction to the filling machine, and the filling machine fills the injection ampoule according to the filling data information.

[0006] On the other hand, the present application also provides a control system for an injection ampoule filling machine, the system comprising: a bottle fixing module, the bottle fixing module is used to set a vacuum suction cup platform, the control terminal of the vacuum suction cup platform is communicatively connected with the filling machine control terminal, and vacuum suction cups are distributed on the vacuum suction cup platform; a bottle data acquisition module, the bottle data acquisition module is used to collect bottle data information of the injection ampoule; a filling data acquisition module, the filling data acquisition module is used to obtain the filling data information of the injection ampoule filled by the filling machine; a vacuum parameter adjustment module, the The vacuum parameter adjustment module is used to input the bottle body data information and the filling data information into the vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and output the first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle; the filling execution module is used to send a filling execution instruction to the filling machine after the injection ampoule bottle is fixed, and the filling machine fills the injection ampoule bottle according to the filling data information.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] A vacuum suction cup platform is provided, and vacuum suction cups are distributed on the vacuum suction cup platform to provide stable support for the ampoule bottle; the control terminal of the vacuum suction cup platform is connected to the control terminal of the filling machine by communication, so as to ensure the information interaction of the filling process. The bottle body data information of the injection ampoule bottle is collected, and the filling data information of the injection ampoule bottle filled by the filling machine is obtained. The bottle body data information reflects the characteristics of the ampoule bottle itself, and the filling data information clarifies the requirements of the filling. These data will be used as the basis for the subsequent adjustment of the filling machine to adapt to different bottle bodies and filling conditions. The bottle body data information and the filling data information are input into the vacuum adaptive adjustment model embedded in the vacuum suction cup platform. The adaptive vacuum adjustment model automatically optimizes the fixed parameters according to the real-time data and outputs the first adaptive vacuum parameter, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle, so as to ensure the stability of the ampoule bottle during the filling process and avoid affecting the filling quality due to poor fixation. When the injection ampoule is fixed, a filling execution instruction is sent to the filling machine, and the filling machine fills the injection ampoule according to the filling data information, ensuring that the filling operation is accurately performed on the basis of the ampoule being stably fixed.

[0009] In summary, the present application realizes a high degree of automation and intelligence in the filling process of injection ampoules, can flexibly adapt to bottles of different specifications and filling requirements, and accurately fix ampoules, thereby significantly improving the filling success rate and ensuring the filling quality of ampoules.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic flow chart of a control method for an injection ampoule filling machine provided in an embodiment of the present application.

[0012] Figure 2 A schematic diagram of a flow chart of sending an anti-adhesion control instruction to a vacuum suction cup platform in a control method of an injection ampoule filling machine provided in an embodiment of the present application.

[0013] Figure 3 A schematic diagram of the process of training a vacuum adaptive adjustment model in a control method for an injection ampoule filling machine provided in an embodiment of the present application.

[0014] Figure 4 A schematic structural diagram of a control system for an injection ampoule filling machine provided in an embodiment of the present application.

[0015] Description of the accompanying drawings: bottle fixing module 10 , bottle data acquisition module 20 , filling data acquisition module 30 , vacuum parameter adjustment module 40 , filling execution module 50 . DETAILED DESCRIPTION

[0016] The embodiment of the present application provides a control method and system for an injection ampoule filling machine, which solves the technical problem that the control parameters of the existing injection ampoule filling machine are fixed and cannot be adaptively adjusted according to the specification parameters of the ampoule, resulting in unstable fixation of the ampoule during the filling process and reduced filling quality, and achieves the technical effect of accurately fixing the ampoule to improve the filling success rate and the filling quality of the ampoule.

[0017] Embodiment 1, as Figure 1 As shown, the embodiment of the present application provides a method for controlling an injection ampoule filling machine, the method comprising:

[0018] Step S1: Setting a vacuum suction cup platform, wherein a control terminal of the vacuum suction cup platform is communicatively connected with a control terminal of a filling and sealing machine, and vacuum suction cups are distributed on the vacuum suction cup platform.

[0019] Specifically, the vacuum suction cup platform is a platform for placing injection ampoules, on which multiple vacuum suction cups are distributed to fix the ampoules. The control terminal is the control center of the vacuum suction cup platform and the filling and sealing machine, and is used to send and receive instructions, data and other information.

[0020] First, a vacuum suction cup platform is set up on the production line, on which several vacuum suction cups are distributed. Then, the control terminal of the vacuum suction cup platform is connected to the control terminal of the filling and sealing machine through industrial Ethernet or serial communication, so that the two can transmit data and interact with each other.

[0021] Step S2: Collecting bottle body data information of the injection ampoule bottle.

[0022] Specifically, the bottle data information is data about the characteristics of the injection ampoule itself, including the shape, size, weight, material, center of gravity position and other information of the bottle. For example, the length, diameter and other dimensional data of ampoules of different specifications are part of the bottle data information. Connect the sensors and measuring equipment deployed on the filling and sealing production line to read the collected bottle data information of the injection ampoule. These sensors and measuring equipment can obtain information such as the size and shape of the bottle in real time, providing a data basis for subsequent adsorption. For example, a laser rangefinder is used to measure the height and diameter of the ampoule, a weight sensor is used to measure the weight of the bottle, and the shape of the bottle is analyzed through an image recognition system. These measurement data are summarized to form bottle data information to determine the specific characteristics of the ampoule, so as to facilitate the subsequent vacuum parameter adjustment for the specific parameters of different bottles.

[0023] Step S3: obtaining the filling data information of the injection ampoule bottle filled by the filling machine.

[0024] Specifically, the filling data information is data related to the filling machine filling injection ampoules, such as liquid injection volume, filling speed, filling force peak, etc. For example, a certain injection needs to be accurately injected with 5 ml of liquid, and this 5 ml is the liquid injection volume in the filling data information.

[0025] The control system of the filling machine (such as a PLC control system) has data recording and storage functions, and will record information such as the filling volume and filling speed of the filling process. For example, the operation interface of the filling machine can display data such as the filling speed and liquid injection volume. Communicate with the filling machine control system, read the filling data information of the filling machine filling injection ampoules through the corresponding interface (such as a USB interface, network interface, etc.), determine the specific filling requirements, and provide data reference for subsequent adjustments.

[0026] Step S4: input the bottle body data information and the filling data information into the vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and output a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle.

[0027] Specifically, the vacuum adaptive adjustment model is a model specifically used to output suitable adaptive vacuum parameters according to the input bottle data information and potting data information. The model is constructed based on a machine learning algorithm and can learn the optimal vacuum parameter settings under different data combinations. The bottle data information and potting data information obtained in steps S2 and S3 are input into the vacuum adaptive adjustment model. The model outputs a first adaptive vacuum parameter based on a machine learning algorithm and through analysis of historical data. The first adaptive vacuum parameter is sent to the control terminal of the vacuum suction cup platform, and the control terminal controls the vacuum suction cup to fix the injection ampoule bottle according to the first adaptive vacuum parameter to ensure that the ampoule bottle does not move during the canning process. The first adaptive vacuum parameter includes the distribution position of the vacuum suction cup, the adsorption strength, the adsorption rate, and the adsorption holding time. Exemplarily, for a heavier ampoule bottle, a larger adsorption strength and a longer adsorption holding time are required as the first adaptive vacuum parameter.

[0028] Step S5: After the injection ampoule is fixed, a filling execution instruction is sent to the filling machine, and the filling machine fills the injection ampoule according to the filling data information.

[0029] Specifically, after the injection ampoule is fixed, the control terminal of the vacuum suction cup platform sends a filling execution instruction to the control terminal of the filling machine. After receiving the instruction, the filling machine starts to fill the ampoule according to the filling data information obtained in step S3. For example, the filling machine controls the injection pump to accurately inject the liquid medicine into the ampoule according to the liquid injection volume data, and controls the filling rhythm according to the filling speed data to ensure that each bottle can be accurately filled.

[0030] Furthermore, after step S5, the embodiment of the present application further includes:

[0031] Step S6: After the filling of the injection ampoule is completed, the adhesion detection device is used to determine whether the injection ampoule is adhered to the filling head of the filling machine. If the injection ampoule is adhered to the filling head of the filling machine, an anti-adhesion control instruction is sent to the vacuum suction cup platform.

[0032] Step S7: providing feedback to the vacuum adaptive adjustment model according to the anti-adhesion control instruction.

[0033] Specifically, the filling head is a component on the filling machine used to inject the drug solution into the ampoule and perform sealing operations. It is in direct contact with the ampoule. During the filling process, if the filling conditions are not suitable (such as drug solution residue, temperature changes, etc.), the filling head may stick to the ampoule.

[0034] After the filling of the injection ampoule is completed, the adhesion detection device is started to detect whether the injection ampoule is adhered to the filling head of the filling machine. When the adhesion detection device determines that adhesion occurs, the device sends an anti-adhesion control instruction to the vacuum suction cup platform. This instruction can be transmitted through the communication line connecting the two (such as a data line, etc.), triggering the vacuum adaptive adjustment model to provide feedback and adjust the vacuum suction cup control parameters to deal with the adhesion state. For example, adjust the parameters of the vacuum suction cup, control the vibration of the vacuum suction cup platform, etc. to separate the filling head from the injection ampoule.

[0035] The adhesion detection device can detect adhesion in time and send feedback information to the vacuum suction cup platform to adjust the vacuum suction cup state, thereby reducing or removing adhesion, ensuring the smooth progress of the potting process, and improving the stability and reliability of the overall potting production line.

[0036] Furthermore, the adhesion detection device described in step S6 of the embodiment of the present application includes a tension sensor and a resistance sensor, the tension sensor is arranged on the vacuum suction cup platform, and the resistance sensor is arranged on the filling head of the filling machine; wherein, the tension sensor is used to detect the tension of the injection ampoule bottle, and the resistance sensor is used to detect the recovery resistance of the filling head of the filling machine, and when the tension or the recovery resistance exceeds a preset threshold value, it is determined that the injection ampoule bottle is adhered to the filling head of the filling machine.

[0037] Specifically, the adhesion detection device is composed of components such as a tension sensor and a resistance sensor. The tension sensor is set on the vacuum suction cup platform to detect the tension applied to the injection ampoule. When the ampoule is pulled up (normal detachment or adhesion), the tension sensor will output an electrical signal corresponding to the magnitude of the tension. The resistance sensor is set on the filling head of the filling machine to detect the recovery resistance of the filling head. When the filling head completes the filling operation and starts to recover, the resistance sensor will detect the resistance encountered by the filling head in this process and output a corresponding electrical signal.

[0038] The adhesion detection device receives the electrical signals output by the tension sensor and the resistance sensor, and converts them into actual tension values ​​and recovery resistance values. These values ​​are then compared with a preset threshold value that has been set in advance. The preset threshold value is the upper limit of the tension and recovery resistance when the ampoule bottle is normally separated from the filling head. When the tension or recovery resistance exceeds the preset threshold value, the adhesion detection device determines that the ampoule bottle and the filling head are adhered. Exemplarily, the preset threshold value of the tension is 3N, and the preset threshold value of the recovery resistance is 2N. When the tension sensor detects that the tension is 4N, or when the resistance sensor detects that the recovery resistance is 3N, the adhesion detection device determines that adhesion has occurred. This detection method utilizes the characteristics of the tension sensor and the resistance sensor. By comparing with the preset threshold value, it can accurately determine whether there is adhesion, thereby providing a basis for subsequent anti-adhesion operations.

[0039] Further, such as Figure 2 As shown, in step S6 of the embodiment of the present application, sending an anti-adhesion control instruction to the vacuum suction cup platform also includes:

[0040] Step S61: Obtain adhesion detection data according to the adhesion detection device, wherein the adhesion detection data includes tension sensing data and resistance sensing data.

[0041] Step S62: Setting a vibration module, wherein the vibration module is embedded with a trained adhesion-vibration control model, wherein the adhesion-vibration control model is obtained through training with adhesion detection samples, vibration control samples, and identification information identifying the degree of adhesion detachment.

[0042] Step S63: inputting the tension sensing data and the resistance sensing data into the adhesion-vibration control model, and outputting vibration control parameters, including vibration frequency, vibration amplitude and vibration duration.

[0043] Step S64: the vibration module controls the vacuum suction cup platform to vibrate according to the vibration control parameters.

[0044] Specifically, the adhesion detection data is data obtained by the adhesion detection device, including tension sensing data and resistance sensing data. These data are the basis for judging the adhesion situation and determining the subsequent vibration control parameters. The vibration control parameters include vibration frequency, vibration amplitude, and vibration duration, etc., which are parameters used to control the vibration characteristics of the vibration module. The vibration frequency indicates the number of vibrations per unit time, the vibration amplitude indicates the strength of the vibration, and the vibration duration indicates the duration of the vibration. These parameters work together to enable the vibration module to vibrate in an appropriate manner, thereby solving the adhesion problem between the ampoule bottle and the filling head.

[0045] A vibration module is configured for the vacuum suction cup platform, and a trained adhesion-vibration control model is pre-embedded in the vibration module. The vibration module is a component capable of generating vibrations, which is arranged near the vacuum suction cup platform or as a part of the vacuum suction cup platform, and is used to generate vibrations according to the vibration control parameters output by the adhesion-vibration control model to solve the adhesion problem between the ampoule bottle and the filling head. The adhesion-vibration control model is a trained model, which is used to output appropriate vibration control parameters according to the input adhesion detection data (tension sensor data and resistance sensor data) to achieve the best adhesion release effect.

[0046] When training the adhesion-vibration control model, a large number of adhesion detection samples (such as tension and resistance data under different production environments, ampoule types, etc.), vibration control samples (such as testing vibrations of different frequencies, amplitudes, and durations in the laboratory, and recording the corresponding adhesion resolution effect data) and identification information that identifies the degree of adhesion detachment (such as determining whether the ampoule is completely detached or partially detached through manual observation or other detection methods) are collected. Then, a machine learning algorithm, such as a neural network or a support vector machine, is used to train the model based on these sample data so that the model can accurately output vibration control parameters based on the input adhesion detection data.

[0047] The interactive adhesion detection device reads the adhesion detection data, including tension sensor data and resistance sensor data, and inputs the adhesion detection data into the adhesion-vibration control model. This model processes the input data based on the previous training algorithm, calculates and outputs the corresponding vibration control parameters under the current adhesion situation, including vibration frequency, vibration amplitude and vibration duration.

[0048] After receiving the vibration control parameters output from the adhesion-vibration control model, the vibration module performs vibration operations according to these parameters. For example, if the vibration control parameters include a vibration frequency of 50 Hz, a vibration amplitude of 2 mm, and a vibration duration of 10 seconds, then the vibration generator inside the vibration module (which may be an electromagnetic vibrator or other component) will vibrate continuously for 10 seconds at a frequency of 50 Hz and an amplitude of 2 mm, causing the vacuum suction cup platform to vibrate, thereby solving the adhesion problem between the ampoule bottle and the filling head.

[0049] The adhesion-vibration control model outputs the corresponding vibration control parameters according to the adhesion detection data to control the vibration module to accurately apply vibration operation, thereby effectively eliminating the adhesion problem between the bottle body and the filling head, reducing potential material loss, and enhancing the stability and reliability of the entire filling process.

[0050] Furthermore, in step S4 of the embodiment of the present application, outputting the first adaptive vacuum parameter according to the vacuum adaptive adjustment model also includes:

[0051] Step S41: inputting the bottle body data information and the filling data information into the vacuum adaptive adjustment model to obtain the center of gravity position of the bottle, the liquid injection volume and the filling force peak value.

[0052] Step S42: using the center of gravity position of the bottle, the liquid injection volume and the peak value of the filling force as input features, outputting the first adaptive vacuum parameters, wherein the first adaptive vacuum parameters include the distribution position of the vacuum suction cup, the adsorption force strength, the adsorption rate and the adsorption holding time.

[0053] Specifically, the center of gravity of the bottle refers to the center of the overall mass of the injection ampoule, which depends on factors such as the shape of the bottle, the material distribution, and the distribution of the internal liquid. It is used to analyze the possibility of tilting during the filling process. The liquid injection volume is the volume of liquid injected into the injection ampoule by the filling machine. In pharmaceutical production, different drug dosage forms and specifications require different liquid injection volumes. Different liquid injection volumes will affect the stability of the ampoule during the filling process, and the liquid injection volume can be used to evaluate the difficulty of adsorption of the bottle. The filling force is the force applied by the filling machine when filling the ampoule, and the filling force peak is the maximum value reached by this force during the filling process. The filling force peak can be used to determine the fixing force required for the bottle and the possible tilt during the filling process.

[0054] After obtaining the bottle data information and the filling data information, the bottle data information and the filling data information are input into the vacuum adaptive adjustment model. The model extracts the bottle body weight center position information from the bottle body data information, and the liquid injection volume, filling force peak value and other information from the filling data information. Exemplarily, the bottle body data information may be a complex data structure obtained by multiple methods such as image recognition and weight sensor. The model obtains the bottle body weight center position and other information through data analysis and feature extraction algorithms; the filling data information is obtained from the filling machine control system, and the model directly reads the liquid injection volume and filling force peak value and other data therein.

[0055] The model then inputs the acquired information such as the center of gravity position of the bottle, the amount of liquid injected, the peak value of the filling force, etc. as input features into the output calculation part of the vacuum adaptive adjustment model. This model may be built based on a machine learning algorithm, such as a neural network algorithm. The model calculates and outputs the first adaptive vacuum parameter based on the pre-trained weights and algorithm logic, taking into account these input features. These parameters include the distribution position, adsorption strength, adsorption rate, and adsorption holding time of the vacuum suction cup. For example, if the center of gravity position of the bottle is high, the amount of liquid injected is large, and the peak value of the filling force is large, the model may output that the vacuum suction cup is distributed in the middle and lower part of the bottle body, adopting a higher adsorption strength, a faster adsorption rate, and a longer adsorption holding time.

[0056] The vacuum adaptive adjustment model can realize the automatic vacuum suction cup adjustment parameters according to different bottles and filling conditions to achieve the best adsorption effect, thereby improving the stability of the filling process.

[0057] Further, such as Figure 3 As shown, the method described in the embodiment of the present application also includes training a vacuum adaptive adjustment model, and the specific steps are as follows:

[0058] Step S-01: According to the historical filling process, historical bottle data samples, filling data samples and vacuum adjustment data samples are obtained.

[0059] Step S-02: Obtain a hybrid network model, the hybrid network model comprising a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to perform feature extraction on the historical bottle body data samples, the filling data samples, and the vacuum adjustment data samples, and the long short-term memory network is used to perform time series synchronization processing on the historical bottle body data samples, the filling data samples, and the vacuum adjustment data samples.

[0060] Step S-03: Minimizing the error between the adaptive vacuum parameters output by the hybrid network model and the vacuum regulation data samples so as to bring the hybrid network model into a convergence state, thereby generating the vacuum adaptive regulation model.

[0061] Step S-04: embed the vacuum adaptive adjustment model into the vacuum suction cup platform for calling.

[0062] Specifically, historical bottle data samples are relevant data samples about injection ampoule bottles collected during the past filling process. Including data such as the shape, size, weight, and material of ampoules of different batches and specifications. Filling data samples are data record samples of previous filling machines during filling operations. Contains information such as the liquid injection volume, filling speed, and filling force peak of different drugs, reflecting the working parameters of the filling machine in the past. Vacuum adjustment data samples are vacuum adjustment parameter data samples actually used for different bottle bodies and filling conditions during the past filling process. For example, data such as the distribution position, adsorption strength, adsorption rate, and adsorption retention time of the vacuum suction cups used previously.

[0063] Query and obtain data related to the historical filling process from the historical production record database. For historical bottle data samples, they can be collected from the historical records of the bottle inspection equipment, the filling data samples can be obtained from the historical operation records of the filling machine, and the vacuum adjustment data samples can be obtained from the vacuum suction cup platform control system.

[0064] A hybrid network model is constructed by combining convolutional neural networks and long short-term memory networks using a deep learning framework (such as TensorFlow or PyTorch). The convolutional neural network is used to extract features from historical bottle data samples, filling data samples, and vacuum adjustment data samples, thereby identifying the key factors that affect the filling effect. The long short-term memory network is used to process the time series characteristics of these data in order to capture the relationship between data at different time points. The two networks are combined according to the designed structure, so that the data can be sequentially processed through the convolutional neural network for feature extraction, and then through the long short-term memory network for time series synchronization processing.

[0065] Input historical bottle data samples, filling data samples, and vacuum adjustment data samples into the hybrid network model. The model will first extract features from the input data through a convolutional neural network, then perform time series synchronization on the data through a long short-term memory network, and finally output the adaptive vacuum parameters. Calculate the error between the output adaptive vacuum parameters and the vacuum adjustment data samples. This error can be calculated using an error function such as the mean square error (MSE). Then adjust the network weights through an optimization algorithm (such as a gradient descent algorithm) to reduce the error, and continuously minimize the error until the model is in a converged state, that is, the error no longer decreases significantly. When the model converges, the final vacuum adaptive adjustment model is generated.

[0066] The trained vacuum adaptive adjustment model is embedded into the control system of the vacuum suction cup platform, so that in the actual filling operation, the vacuum suction cup platform calls this model and outputs the first adaptive vacuum parameter according to the input bottle data information and filling data information, thus enhancing the adaptability to different bottles and filling conditions.

[0067] Furthermore, the vacuum suction cup on the vacuum suction cup platform described in the embodiment of the present application is a vacuum suction cup with multiple suction holes, and the opening or closing of each suction hole is controlled by a solenoid valve; when the filling of the injection liquid ampoule is completed, the current suction hole is closed by the solenoid valve to allow the injection liquid ampoule to detach from the vacuum suction cup platform.

[0068] Specifically, the vacuum suction cup on the vacuum suction cup platform has multiple adsorption holes. The opening and closing of each hole is controlled by a solenoid valve. The multiple adsorption holes work together to create a vacuum environment to adsorb the ampoule bottle, ensuring the stability of the ampoule bottle during the filling process.

[0069] During the filling process of the injection ampoule, the vacuum suction cup platform activates the solenoid valve, opens the multiple suction holes on the vacuum suction cup, and generates vacuum suction force to fix the ampoule. When the filling operation is completed, in order to make the ampoule detach from the vacuum suction cup platform, a closing command is issued to the solenoid valve. When the solenoid valve receives the closing command, it cuts off the connection between the suction hole and the vacuum source, connects the suction hole to the external environment, and the suction force disappears, so that the injection ampoule can be easily detached from the suction cup without causing damage or leakage.

[0070] Each adsorption hole is independently controlled by a solenoid valve, so that the adsorption method can be flexibly adjusted when facing bottles of different shapes and sizes, so that the ampoule can be accurately and quickly separated, improving production efficiency and stability. For example, if the bottle is heavy, more adsorption holes can be opened to increase the adsorption force; for light bottles, the number of adsorption holes can be appropriately reduced to save energy.

[0071] In summary, the control method of an injection ampoule filling machine provided in the embodiment of the present application has the following technical effects:

[0072] A vacuum suction cup platform is provided, and vacuum suction cups are distributed on the vacuum suction cup platform to provide stable support for the ampoule bottle; the control terminal of the vacuum suction cup platform is connected to the control terminal of the filling machine by communication, so as to ensure the information interaction of the filling process. The bottle body data information of the injection ampoule bottle is collected, and the filling data information of the injection ampoule bottle filled by the filling machine is obtained. The bottle body data information reflects the characteristics of the ampoule bottle itself, and the filling data information clarifies the requirements of the filling. These data provide data basis for the subsequent filling machine adjustment to adapt to different bottle bodies and filling conditions. The bottle body data information and the filling data information are input into the vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and the first adaptive vacuum parameter is output according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle, so as to ensure the stability of the ampoule bottle during the filling process, and avoid affecting the filling quality due to poor fixation. When the injection ampoule is fixed, a filling execution instruction is sent to the filling machine, and the filling machine fills the injection ampoule according to the filling data information, ensuring that the filling operation is accurately performed on the basis of the stable fixation of the ampoule. After the filling is completed, the adhesion detection device determines whether the ampoule is adhered to the filling head. If adhesion occurs, an anti-adhesion control instruction is sent to the vacuum suction cup platform, and the vibration control parameters are output through the adhesion-vibration control model. The vibration module vibrates according to the parameters to solve the adhesion problem, avoiding damage to the ampoule and filling failure caused by adhesion, and improving the success rate of production. After the filling is completed, the opening and closing of the adsorption hole on the vacuum suction cup is controlled by the solenoid valve, so that the ampoule is separated from the vacuum suction cup platform and prepared for the next round of filling.

[0073] In general, the embodiment of the present application adopts a vacuum adaptive adjustment model, which intelligently adjusts the vacuum parameters according to the real-time input of the bottle and the filling data to ensure the stable fixation of bottles of different specifications. This adaptive control not only improves the fixing accuracy, but also reduces the need for human intervention and improves the operational safety. In addition, the method also integrates an adhesion detection device and a vibration module, which can send anti-adhesion control instructions in time when adhesion is detected, start the vibration program, reduce the risk of bottle adhesion, and ensure the smooth progress of the filling process. This series of improvements significantly improves the success rate and filling quality of injection ampoule filling, reduces product loss during the filling process, and enhances the adaptability to different bottles and filling requirements, meeting the dual requirements of high efficiency and high quality in the modern pharmaceutical industry.

[0074] Embodiment 2, as Figure 4 As shown, the embodiment of the present application provides a control system for an injection ampoule filling machine, the system comprising:

[0075] The bottle fixing module 10 is used to set a vacuum suction cup platform. The control terminal of the vacuum suction cup platform is communicatively connected with the control terminal of the filling machine. Vacuum suction cups are distributed on the vacuum suction cup platform.

[0076] The bottle body data collection module 20 is used to collect bottle body data information of the injection ampoule bottle.

[0077] The filling data acquisition module 30 is used to obtain the filling data information of the injection ampoule bottle filled by the filling machine.

[0078] A vacuum parameter adjustment module 40 is used to input the bottle body data information and the filling data information into a vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and output a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle.

[0079] The filling execution module 50 is used to send a filling execution instruction to the filling machine after the injection ampoule is fixed, and the filling machine fills the injection ampoule according to the filling data information.

[0080] Furthermore, the system described in the embodiment of the present application is also used to perform the following steps:

[0081] When the filling of the injection ampoule is completed, the adhesion detection device is used to determine whether the injection ampoule is adhered to the filling head of the filling machine. If the injection ampoule is adhered to the filling head of the filling machine, an anti-adhesion control instruction is sent to the vacuum suction cup platform; and the vacuum adaptive adjustment model is fed back according to the anti-adhesion control instruction.

[0082] Furthermore, the adhesion detection device described in the embodiment of the present application includes a tension sensor and a resistance sensor, the tension sensor is arranged on the vacuum suction cup platform, and the resistance sensor is arranged on the filling head of the filling machine; wherein, the tension sensor is used to detect the tension of the injection ampoule bottle, and the resistance sensor is used to detect the recovery resistance of the filling head of the filling machine, and when the tension or the recovery resistance exceeds a preset threshold value, it is judged that the injection ampoule bottle is adhered to the filling head of the filling machine.

[0083] Furthermore, the system described in the embodiment of the present application is also used to perform the following steps:

[0084] According to the adhesion detection device, adhesion detection data is obtained, and the adhesion detection data includes tension sensor data and resistance sensor data; a vibration module is set, and the vibration module is embedded with a trained adhesion-vibration control model, wherein the adhesion-vibration control model is obtained through training with adhesion detection samples, vibration control samples, and identification information identifying the degree of adhesion detachment; the tension sensor data and resistance sensor data are input into the adhesion-vibration control model, and vibration control parameters are output, including vibration frequency, vibration amplitude, and vibration duration; the vibration module controls the vacuum suction cup platform to vibrate according to the vibration control parameters.

[0085] Furthermore, the vacuum parameter adjustment module 40 of the embodiment of the present application is also used to perform the following steps:

[0086] The bottle body data information and the potting data information are input into the vacuum adaptive adjustment model to obtain the center of gravity position of the bottle body, the liquid injection volume and the peak value of the potting force; the center of gravity position of the bottle body, the liquid injection volume and the peak value of the potting force are used as input features to output the first adaptive vacuum parameters, wherein the first adaptive vacuum parameters include the distribution position of the vacuum suction cup, the adsorption force strength, the adsorption rate and the adsorption holding time.

[0087] Furthermore, the system described in the embodiment of the present application is also used to perform the following steps:

[0088] According to the historical filling process, historical bottle body data samples, filling data samples and vacuum adjustment data samples are obtained; a hybrid network model is obtained, the hybrid network model includes a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to extract features from the historical bottle body data samples, the filling data samples and the vacuum adjustment data samples, and the long short-term memory network is used to perform time series synchronization processing on the historical bottle body data samples, the filling data samples and the vacuum adjustment data samples; the vacuum adaptive adjustment model is generated by minimizing the error between the adaptive vacuum parameters output by the hybrid network model and the vacuum adjustment data samples so that the hybrid network model is in a convergence state; and the vacuum adaptive adjustment model is embedded in the vacuum suction cup platform for calling.

[0089] Furthermore, the vacuum suction cup on the vacuum suction cup platform described in the embodiment of the present application is a vacuum suction cup with multiple suction holes, and the opening or closing of each suction hole is controlled by a solenoid valve; when the filling of the injection liquid ampoule is completed, the current suction hole is closed by the solenoid valve to allow the injection liquid ampoule to detach from the vacuum suction cup platform.

[0090] Through the above-mentioned detailed description of a control method for an injection ampoule filling machine in this specification, those skilled in the art can clearly understand the control system of an injection ampoule filling machine in this embodiment. For the system disclosed in Example 2, since it corresponds to the method disclosed in Example 1 and has corresponding functional modules and beneficial effects, the relevant parts can be referred to the description of the method part.

[0091] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A control method for an injection ampoule filling machine, characterized in that: The method comprises: A vacuum suction cup platform is provided, wherein a control terminal of the vacuum suction cup platform is communicatively connected with a control terminal of the filling and sealing machine, and vacuum suction cups are distributed on the vacuum suction cup platform; Collect bottle data information of injection ampoule bottle; Acquiring filling data information of the injection ampoule bottle filled by the filling machine; Input the bottle body data information and the potting data information into a vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and output a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle; When the injection ampoule is fixed, a filling execution instruction is sent to the filling machine, and the filling machine fills the injection ampoule according to the filling data information.

2. The method according to claim 1, characterized in that After the filling machine fills the injection ampoule according to the filling data information, the method further includes: When the injection ampoule is filled, the adhesion detection device is used to determine whether the injection ampoule is adhered to the filling head of the filling machine. If the injection ampoule is adhered to the filling head of the filling machine, an anti-adhesion control instruction is sent to the vacuum suction cup platform; Feedback is provided to the vacuum adaptive adjustment model according to the anti-adhesion control instruction.

3. The method according to claim 2, characterized in that The adhesion detection device includes a tension sensor and a resistance sensor, wherein the tension sensor is arranged on the vacuum suction cup platform, and the resistance sensor is arranged on the potting head of the potting machine; Among them, the tension sensor is used to detect the tension of the injection ampoule bottle, and the resistance sensor is used to detect the recovery resistance of the filling head of the filling machine. When the tension or the recovery resistance exceeds a preset threshold, it is determined that the injection ampoule bottle is adhered to the filling head of the filling machine.

4. The method according to claim 3, characterized in that Sending an anti-adhesion control instruction to the vacuum suction cup platform, the method also includes: Acquiring adhesion detection data according to the adhesion detection device, wherein the adhesion detection data includes tension sensing data and resistance sensing data; Setting a vibration module, wherein the vibration module is embedded with a trained adhesion-vibration control model, wherein the adhesion-vibration control model is obtained through training with adhesion detection samples, vibration control samples, and identification information identifying the degree of adhesion detachment; Inputting the tension sensor data and the resistance sensor data into the adhesion-vibration control model, and outputting vibration control parameters, including vibration frequency, vibration amplitude, and vibration duration; The vibration module controls the vacuum suction cup platform to vibrate according to the vibration control parameters.

5. The method according to claim 1, characterized in that Outputting a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, the method comprises: Inputting the bottle body data information and the filling data information into the vacuum adaptive adjustment model to obtain the center of gravity position of the bottle body, the liquid injection volume and the filling force peak value; The center of gravity position of the bottle, the liquid injection volume and the peak value of the sealing force are used as input features to output the first adaptive vacuum parameters, wherein the first adaptive vacuum parameters include the distribution position of the vacuum suction cup, the adsorption force strength, the adsorption rate and the adsorption holding time.

6. The method according to claim 5, characterized in that Training a vacuum adaptive regulation model, the method comprising: According to the historical filling process, obtain historical bottle data samples, filling data samples and vacuum adjustment data samples; Acquire a hybrid network model, the hybrid network model comprising a convolutional neural network and a long short-term memory network, wherein the convolutional neural network is used to perform feature extraction on the historical bottle body data samples, the filling data samples, and the vacuum adjustment data samples, and the long short-term memory network is used to perform time series synchronization processing on the historical bottle body data samples, the filling data samples, and the vacuum adjustment data samples; The vacuum adaptive regulation model is generated by minimizing the error between the adaptive vacuum parameter output by the hybrid network model and the vacuum regulation data sample so as to make the hybrid network model in a convergence state; The vacuum self-adaptive adjustment model is embedded in the vacuum suction cup platform for calling.

7. The method according to claim 6, characterized in that The vacuum suction cup on the vacuum suction cup platform is a vacuum suction cup with multiple suction holes, and the opening or closing of each suction hole is controlled by a solenoid valve; When the injection ampoule is filled and sealed, the current adsorption hole is closed through the electromagnetic valve to make the injection ampoule separate from the vacuum suction cup platform.

8. A control system for an injection ampoule filling machine, characterized in that: The system is used to execute the control method of an injection ampoule filling machine according to any one of claims 1 to 7, and the system comprises: A bottle fixing module, wherein the bottle fixing module is used to set a vacuum suction cup platform, a control terminal of the vacuum suction cup platform is communicatively connected with a control terminal of the filling and sealing machine, and vacuum suction cups are distributed on the vacuum suction cup platform; A bottle data collection module, which is used to collect bottle data information of injection ampoule bottles; A filling data acquisition module, the filling data acquisition module is used to obtain the filling data information of the injection ampoule bottle filled by the filling machine; A vacuum parameter adjustment module, the vacuum parameter adjustment module is used to input the bottle body data information and the potting data information into a vacuum adaptive adjustment model embedded in the vacuum suction cup platform, and output a first adaptive vacuum parameter according to the vacuum adaptive adjustment model, wherein the first adaptive vacuum parameter is used to control the vacuum suction cup on the vacuum suction cup platform to fix the injection ampoule bottle; A filling execution module is used to send a filling execution instruction to the filling machine after the injection ampoule is fixed, and the filling machine fills the injection ampoule according to the filling data information.

Images