Tobacco shred online filling value detection equipment and model

Abstract

The invention relates to a cut tobacco online filling value detection device and model, which comprises a supporting seat, a belt weighing system is fixedly mounted at the top of the supporting seat, two brackets are fixedly mounted at the top of the supporting seat and located at the top of the belt weighing system, fixed blocks are mounted on the surfaces of the two brackets, and the fixed blocks are connected with the belt weighing system. And mounting plates are fixedly mounted at the bottoms of the two fixing blocks correspondingly, a sensor and a 3D laser intelligent camera are fixedly mounted at the bottoms of the two mounting plates correspondingly, and protection assemblies are mounted in the two mounting plates correspondingly. According to the tobacco shred on-line filling value equipment and model, the tobacco shred on-line filling value equipment realizes real-time on-line detection of the tobacco shred filling value by means of components such as a 3D laser intelligent camera, a belt weighing system and a software analysis and calculation system; the 3D laser intelligent camera scans in real time to obtain volume information, the belt weighing system weighs in real time to obtain weight information, and the software analysis and calculation system obtains a filling value through calculation and proportion conversion.

Description

Technical Field

[0001] This invention relates to the field of tobacco shred detection technology, specifically to an online tobacco shred filling value detection device and model. Background Technology

[0002] The development of modern industry has promoted the rapid development of the tobacco industry, and the level of tobacco process testing has also developed rapidly. Tobacco process testing involves parameters such as tobacco moisture, tobacco size, and tobacco filling value. Among them, the tobacco filling value is related to the appearance, draw resistance, smoking experience, and cost of cigarettes. It is an economic indicator reflecting the production cost and raw material consumption of cigarettes, and an important physical indicator of tobacco used in cigarette processing.

[0003] Traditional methods for detecting tobacco filling value involve manual sampling. A certain weight of tobacco is weighed in a laboratory, placed in a container, and pressure is applied to the tobacco for a certain period of time. The volume of the tobacco is then measured, and the filling value is calculated from the weight and volume. This manual method is time-consuming, labor-intensive, and subject to human error. Therefore, an online tobacco filling value detection device and model are proposed to solve the aforementioned problems. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an online tobacco filling value detection device and model, which has the advantages of convenient operation and solves the problems of traditional methods for detecting tobacco filling value, which involve manually sampling samples, weighing a certain weight of tobacco in a laboratory and placing it in a container, applying a certain pressure to the tobacco and maintaining it for a certain time, measuring the volume of the tobacco, and calculating the tobacco filling value by converting weight and volume. This manual method is time-consuming, labor-intensive, and also affected by human factors.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an online tobacco filling value detection device, comprising a support base, a belt weighing system fixedly installed on the top of the support base, two brackets fixedly installed on the top of the support base above the belt weighing system, fixing blocks installed on the surface of each of the two brackets, mounting plates fixedly installed on the bottom of each of the two fixing blocks, a sensor and a 3D laser intelligent camera fixedly installed on the bottom of each of the two mounting plates, and protective components installed inside each of the two mounting plates;

[0006] The protective assembly includes two mounting plates, each with a mounting groove at its bottom. Each mounting plate has an internal movable cavity. A motor is fixedly mounted at one end of each mounting plate. Two threaded rods are rotatably mounted inside each of the two movable cavities. Threaded blocks are threaded onto the surfaces of each threaded rod. A sprocket is fixedly mounted on the outer surface of each threaded rod. The two sprockets are connected by a chain drive. A movable plate is fixedly mounted between every two threaded blocks. A wiping cloth is attached to one side of each movable plate.

[0007] As a preferred technical solution, the sensor and the 3D laser smart camera are respectively located inside the mounting slot.

[0008] As a preferred technical solution, a visible-near-infrared spectral camera is installed next to the 3D laser smart camera to simultaneously collect the spectral reflectance characteristics of tobacco.

[0009] As a preferred technical solution, the sensor, 3D laser intelligent camera, belt weighing system and visible-near infrared spectral camera are all electrically connected to the software analysis and calculation system.

[0010] As a preferred technical solution, one end of each of the two threaded rods is fixedly connected to the output shaft of the motor, and the two wiping cloths are in contact with the sensor and the 3D laser smart camera, respectively.

[0011] As a preferred technical solution, both of the fixed blocks are installed with the bracket by four fastening bolts. The two mounting plates are provided with sliding grooves inside and on both sides of the mounting groove, and the two ends of the two movable plates are slidably connected to the sliding grooves.

[0012] An online tobacco filling value detection model includes a data processing algorithm module, an intelligent learning model, and an interface, wherein...

[0013] Data processing algorithm module: This module is used to construct data processing algorithms from the weight data collected by the belt weighing system, the volume data acquired by the 3D laser smart camera, and the environmental data sensed by the sensors.

[0014] Intelligent learning model: used to adaptively adjust calculation parameters and optimize the calculation results of tobacco filling value to improve the accuracy and stability of measurement;

[0015] Interface: Enables real-time data interaction and sharing.

[0016] As a preferred technical solution, the data processing algorithm module uses a multi-source data fusion algorithm to fuse the weight data collected by the belt weighing system, the volume data acquired by the 3D laser smart camera, and the environmental data sensed by the sensor.

[0017] Let W be the weight data collected by the belt weighing system, V be the volume data acquired by the 3D laser intelligent camera, and E be the environmental data (such as temperature and humidity) sensed by the sensors. n The formula for calculating the integrated data D after data fusion is as follows:

[0018]

[0019] Where, ω W ω V , The weights are respectively for weight data, volume data, and various environmental data, and satisfy the following conditions: The initial weights can be set based on experience, and then optimized through training with historical data and error backpropagation algorithm to improve the accuracy of tobacco filling value calculation.

[0020] As a preferred technical solution, the intelligent learning model is constructed based on a deep learning algorithm and is trained using a long short-term memory network on a large amount of historical tobacco filling value data;

[0021] Let the input sequence be x t The hidden layer state is h t The cell state is c t The forget gate, input gate, and output gate are f, respectively. t i t ,

[0022] o t The calculation process of the LSTM network is as follows:

[0023] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0024] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0025]

[0026] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0027] h t =o t ⊙tanh(c t)

[0028] Among them, W f W i W c W o Let b be the weight matrix. f b i b c b o Let σ be the bias vector, σ be the sigmoid activation function, and tanh be the hyperbolic tangent activation function. This represents element-wise multiplication. By training an LSTM network, it is possible to effectively capture the time-series features in the data, thereby more accurately and adaptively adjusting the calculation parameters and optimizing the calculation results of the tobacco filling value.

[0029] As a preferred technical solution, the interface supports multiple data transmission protocols, including but not limited to HTTP, TCP / IP, etc., and can perform stable and efficient real-time data interaction and sharing with the software analysis and computing system (9), while also having data encryption function;

[0030] The transmitted data is encrypted using AES (Advanced Encryption Standard). Let the plaintext data be P, the key be K, and the encrypted ciphertext be...

[0031] The formula for calculating C is as follows:

[0032] C = AES K (P)

[0033] At the receiving end, the ciphertext data C is decrypted using the same key K to recover the plaintext data P.

[0034]

[0035] Among them, AES -1 This is an AES decryption function that uses the AES encryption algorithm to ensure the security of data transmission.

[0036] Compared with the prior art, the technical solution of this application has the following beneficial effects:

[0037] 1. This online tobacco filling value device and model, by setting up the online tobacco filling value device with components such as a 3D laser intelligent camera, a belt weighing system, and a software analysis and calculation system, realizes real-time online detection of tobacco filling value; the 3D laser intelligent camera scans the volume information in real time, the belt weighing system weighs the weight information in real time, and the software analysis and calculation system calculates the filling value through calculation and proportional conversion. It can also obtain empirical coefficients to optimize the data through offline data analysis, which greatly simplifies the detection process and provides accurate and effective tobacco filling value data for tobacco processing departments, helping to control production processes and product quality.

[0038] 2. This online tobacco filling value device and model, through the setting of protective components consisting of a motor, threaded rod, sprocket, chain, moving plate and wiping cloth, can effectively remove dust from the surface of the sensor and 3D laser intelligent camera. When dust adheres during monitoring, the motor drives the wiping cloth to move and wipe, ensuring that the device accurately collects tobacco weight, volume and environmental data, ensuring accurate calculation of tobacco filling value, providing a reliable basis for production process, while reducing equipment maintenance frequency and extending equipment life. Attached Figure Description

[0039] Figure 1 This is a three-dimensional view of the structure of the present invention;

[0040] Figure 2 This is a bottom view of the mounting plate structure of the present invention;

[0041] Figure 3 This is a cross-sectional view of the mounting plate structure of the present invention;

[0042] Figure 4 This is a structural connection diagram of the movable plate and wiping cloth of the present invention;

[0043] Figure 5 This is a framework diagram for calculating fill values ​​using the 3D laser intelligent camera of the present invention;

[0044] Figure 6 This is a flowchart illustrating the software calculation framework for filling values ​​in this invention.

[0045] Figure 7 This is a graph showing the online fill value detection data of the present invention.

[0046] In the diagram: 1. Support base; 2. Belt weighing system; 3. Bracket; 4. Fixing block; 5. Mounting plate; 6. Sensor; 7. 3D laser intelligent camera; 8. Protective components; 801. Mounting slot; 802. Movable cavity; 803. Motor; 804. Threaded rod; 805. Threaded block; 806. Sprocket; 807. Chain; 808. Moving plate; 809. Wiping cloth; 9. Software analysis and calculation system. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Example 1

[0049] Please see Figure 1-7The online tobacco filling value detection device in this embodiment includes a support base 1. A belt weighing system 2 is fixedly installed on the top of the support base 1. Two brackets 3 are fixedly installed on the top of the support base 1 and the top of the belt weighing system 2. Fixing blocks 4 are installed on the surface of the two brackets 3. Mounting plates 5 are fixedly installed on the bottom of the two fixing blocks 4. Sensors 6 and 3D laser smart cameras 7 are fixedly installed on the bottom of the two mounting plates 5 respectively. Protective components 8 are installed inside the two mounting plates 5.

[0050] It should be noted that the development of modern industry has promoted the rapid development of the tobacco industry, and the level of tobacco processing testing has also developed rapidly. Tobacco processing testing involves parameters such as tobacco moisture, tobacco size, and tobacco filling value. Among them, the tobacco filling value is related to the appearance, draw resistance, smoking experience, and cost of cigarettes. It is an economic indicator reflecting the production cost and raw material consumption of cigarettes, and an important physical indicator of tobacco processed in cigarettes. Therefore, online tobacco filling value equipment is used to test tobacco.

[0051] In this embodiment, the online tobacco filling value calculation system mainly consists of a 3D laser intelligent camera 7, a belt weighing system 2, and a software analysis and calculation system 9. The 3D laser intelligent camera 7 is installed on the bracket 3 of the belt weighing system 2. The belt weighing system 2 weighs and outputs the weight in real time. The 3D laser intelligent camera 7 scans and obtains the volume information in real time. The software analysis and calculation system 9 calculates the initial value of tobacco filling with a certain weight and volume in real time, and then performs proportional conversion to obtain the tobacco filling value under a certain pressure.

[0052] Furthermore, not shown in the diagram, a visible-near-infrared (VIS-NIR) spectral camera (such as the HySpex Mjolnir series) can be installed next to the 3D laser intelligent camera. This camera boasts a resolution of up to 2048×2048 pixels and over 200 spectral channels, used to simultaneously acquire the spectral reflectance characteristics of tobacco shreds. This allows for the identification of properties such as tobacco moisture content and stem ratio. These characteristics can then be integrated into the detection model.

[0053] The sensors, 3D laser smart camera, belt weighing system, and visible-near infrared spectroscopy camera are all electrically connected to the software analysis and calculation system.

[0054] The software analysis and calculation system 9 receives information from the belt scale in real time, including the weighing signal. The 3D laser intelligent camera 7 acquires the belt speed of the belt scale in real time and outputs volume information in real time. The software analysis and calculation system 9 combines the belt scale information, volume information, and conversion ratio relationship to obtain tobacco filling value data in real time and store it in the database. After a large amount of statistical calculation, the system obtains the effective filling value for the tobacco processing department to view.

[0055] In this embodiment, the 3D laser smart camera 7 can detect 3D data in real time and obtain volume information in real time. In addition, with the weighing belt data, a large amount of raw data of tobacco filling value can be obtained in real time. After offline data analysis and comparison with online data, empirical coefficients are obtained and converted into effective data of tobacco filling value for use by the process department, which greatly simplifies the detection process of tobacco filling value.

[0056] In this embodiment, sensors 6, such as temperature, humidity, and tobacco flow rate, are added to explore the potential relationships between data and provide more basis for process optimization. Temperature and humidity affect the physical properties of tobacco, thereby changing the filling value. For example, as temperature rises, tobacco expands, and the filling value changes; as humidity increases, tobacco softens, and the filling value is also affected. By monitoring temperature and humidity in real time through sensors 6 and combining the tobacco filling value data, a more accurate correction model can be established, eliminating the interference of environmental factors, making the filling value measurement more accurate, and providing reliable indicators for tobacco production.

[0057] Example 2

[0058] Please see Figure 1-7 This embodiment discloses an online tobacco filling value detection device, which includes a protective component 8. The bottom of each of the two mounting plates 5 has a mounting groove 801. Each of the two mounting plates 5 has a movable cavity 802. A motor 803 is fixedly mounted at one end of each of the two mounting plates 5. Two threaded rods 804 are rotatably mounted inside each of the two movable cavities 802. Threaded blocks 805 are threadedly connected to the surfaces of the two threaded rods 804. A sprocket 806 is fixedly mounted on the outer surface of each of the two threaded rods 804. The two sprockets 806 are connected by a chain 807. A movable plate 808 is fixedly mounted between every two threaded blocks 805. A wiping cloth 809 is attached to one side of each of the two movable plates 808.

[0059] In this embodiment, since a large amount of dust will adhere to the surface of the sensor 6 and the 3D laser smart camera 7 during the monitoring process, the motor 803 drives the transmission between the sprocket 806 and the chain 807, which in turn drives the two threaded rods 804 to rotate. The rotation of the two threaded rods 804 drives the moving plate 808 and the wiping cloth 809 to move through the threaded block 805, thereby wiping the surface of the sensor 6 and the 3D laser smart camera 7. The timely wiping by the protective component 8 can ensure that the equipment accurately collects the weight, volume and environmental data of the tobacco, ensures the accurate calculation of the tobacco filling value, and provides a reliable basis for the production process.

[0060] Example 3

[0061] This embodiment discloses an online tobacco filling value detection model, including a data processing algorithm module, an intelligent learning model, and an interface, wherein...

[0062] Data processing algorithm module: This module is used to construct a data processing algorithm based on the weight data collected by the belt weighing system 2, the volume data acquired by the 3D laser smart camera 7, and the environmental data sensed by the sensor 6.

[0063] Intelligent learning model: used to adaptively adjust calculation parameters and optimize the calculation results of tobacco filling value to improve the accuracy and stability of measurement;

[0064] Interface: Enables real-time data interaction and sharing.

[0065] In real-world applications, the production process in tobacco processing workshops is complex and dynamically changing. Different batches of tobacco exhibit variations in raw material characteristics and production process parameters, placing higher demands on the accuracy and adaptability of the online tobacco filling value model. This model can be deployed within the production monitoring system of tobacco companies to process data generated on the production line in real time.

[0066] The data processing algorithm module uses a multi-source data fusion algorithm to fuse the weight data collected by the belt weighing system 2, the volume data acquired by the 3D laser smart camera 7, and the environmental data sensed by the sensor 6.

[0067] Let W be the weight data collected by the belt weighing system, V be the volume data acquired by the 3D laser intelligent camera, and E be the environmental data (such as temperature and humidity) sensed by the sensors. n The formula for calculating the integrated data D after data fusion is as follows:

[0068]

[0069] Where, ω W ω V , The weights are respectively for weight data, volume data, and various environmental data, and satisfy the following conditions: The initial weights can be set based on experience, and then optimized through training with historical data and error backpropagation algorithm to improve the accuracy of tobacco filling value calculation.

[0070] In the specific optimization process, the mean square error (MSE) between the actual measured value and the model calculated value of the tobacco filling value is used.

[0071] As the loss function, the formula is:

[0072] Where m is the number of samples, y j This is the actual fill value. The filling values ​​are calculated for the model. The weights are continuously adjusted using the gradient descent algorithm to minimize the loss function, thereby improving the accuracy of the tobacco filling value calculation. During data fusion, a Kalman filter algorithm can also be introduced to preprocess the weight, volume, and environmental data, removing noise interference and further improving data quality.

[0073] The intelligent learning model is built based on deep learning algorithms and is trained using a long short-term memory network (LSTM) on a large amount of historical tobacco filling value data.

[0074] Let the input sequence be x t The hidden layer state is h t The cell state is c t The forget gate, input gate, and output gate are f, respectively. t i t ,

[0075] o t The calculation process of the LSTM network is as follows:

[0076] f t =σ(W f ·[h t-1 ,x t ]+b f )

[0077] i t =σ(W i ·[h t-1 ,x t ]+b i )

[0078]

[0079] o t =σ(W o ·[h t-1 ,x t ]+b o )

[0080] h t =o t ⊙tanh(c t )

[0081] Among them, W f W i W c W o Let b be the weight matrix. f b i b c b o Let σ be the bias vector, σ be the sigmoid activation function, and tanh be the hyperbolic tangent activation function. This represents element-wise multiplication. By training an LSTM network, it is possible to effectively capture the time-series features in the data, thereby more accurately and adaptively adjusting the calculation parameters and optimizing the calculation results of the tobacco filling value.

[0082] Meanwhile, to prevent model overfitting, regularization techniques, such as L1 and L2 regularization, are introduced by adding a regularization term to the loss function: L = MSE + λΣ k ||θ k || 2

[0083] Where λ is the regularization parameter, θ k These are the model parameters; by training the LSTM network, the time series features in the data can be effectively captured, thereby more accurately and adaptively adjusting the calculation parameters and optimizing the calculation results of the tobacco filling value; in addition, the model can also periodically use newly generated production data for incremental learning, enabling the model to adapt to subtle changes in the production process.

[0084] The interface supports multiple data transmission protocols, including but not limited to HTTP and TCP / IP, and can perform stable and efficient real-time data interaction and sharing with the software analysis and computing system 9, while also having data encryption capabilities.

[0085] The transmitted data is encrypted using AES (Advanced Encryption Standard). Let the plaintext data be P, the key be K, and the encrypted ciphertext be...

[0086] The formula for calculating C is as follows:

[0087] C = AES K (P)

[0088] At the receiving end, the ciphertext data C is decrypted using the same key K to recover the plaintext data P.

[0089]

[0090] Among them, AES -1 This is an AES decryption function that uses the AES encryption algorithm to ensure the security of data transmission.

[0091] To further ensure the security and stability of data transmission, the interface can also employ digital signature technology to verify the transmitted data, ensuring that the data has not been tampered with during transmission. During data transmission, flow control and congestion control mechanisms are introduced to prevent network congestion caused by excessive data transmission volume, ensuring that data can be transmitted to the software analysis and computing system in real time and accurately. Simultaneously, the interface has good scalability, enabling easy integration with other production management systems to achieve comprehensive data sharing and collaborative applications.

[0092] The volume calculation includes the following:

[0093] Part 1: Acquisition of Volume Data (Based on 3D Laser Smart Camera 7)

[0094] Detection principle

[0095] The 3D laser smart camera 7 scans the tobacco pile on the conveyor belt by emitting a laser beam, and uses laser triangulation to capture the three-dimensional contour data (such as height, width, length, etc.) of the tobacco pile in real time. Combined with the width parameter of the conveyor belt, it calculates the real-time volume of the tobacco pile.

[0096] For example: if the width of the conveyor belt is W, and the camera scans to obtain the cross-sectional area S of the tobacco pile, then the volume V = S × W.

[0097] Meanwhile, the 3D laser smart camera 7 uses a dynamic scanning-real-time modeling method to obtain instantaneous volume:

[0098] A laser beam is emitted to scan the tobacco pile on the conveyor belt at a high frequency (e.g., 500 times / second). Using laser triangulation, a three-dimensional point cloud model of the tobacco pile is constructed based on the time and angle differences between laser emission and reception, and contour data such as height, width, and length are extracted. Combined with the conveyor belt width parameter W, the tobacco pile is sliced ​​into sections along the conveyor belt's running direction, with each section having an area S. n The area is calculated by fitting contour data (e.g., by using polygon approximation to fit an irregular tobacco pile cross-section into a regular geometric shape and then summing the volumes V of adjacent slices). 切片 =S n ×L 切片 The instantaneous volume V = ∑V is finally obtained by converting the conveyor belt speed and the scanning interval. 切片 This enables dynamic and continuous volume monitoring.

[0099] Dynamic Adaptability

[0100] When the transmission speed of the belt weighing system 2 changes, the 3D laser smart camera 7 can ensure the accuracy of volume data at different speeds through high-frequency sampling (such as hundreds of times per second), avoiding errors caused by changes in the density of tobacco shreds.

[0101] By combining sensor 6 (such as a laser alignment sensor) to monitor the positional deviation of the tobacco on the conveyor belt in real time, the geometric parameters in the volume calculation are corrected, further improving the accuracy.

[0102] II. Acquisition of Weight Data (Based on Belt Weighing System 2)

[0103] Detection principle

[0104] The belt weighing system 2 measures the linear density (weight of tobacco per unit length) of the tobacco on the conveyor belt in real time using a weighing sensor, and calculates the real-time weight flow rate by combining the operating speed of the conveyor belt.

[0105] Calculation formula: Weight flow rate = Linear density × Conveyor belt speed;

[0106] For example: if the linear density is q kg / m and the conveyor belt speed is v m / s, then the weight per unit time is q × v kg / s.

[0107] Dynamic correspondence between weight and volume

[0108] The belt weighing system 2 and the 3D laser intelligent camera 7 achieve correlation between instantaneous flow rate and instantaneous volume through time synchronization and spatial calibration.

[0109] Time synchronization: All devices are connected to a high-precision clock module to ensure that the weight acquisition (linear density q) of the belt weighing system 2, the volume scanning (section slice S_n) of the 3D laser intelligent camera 7, and the speed acquisition (speed v) of the conveyor belt speed encoder are marked on the same time axis to ensure the data correspondence.

[0110] Spatial calibration: Using a laser alignment sensor, the spatial overlap between the scanning area of ​​the 3D laser intelligent camera 7 and the weighing area of ​​the belt weighing system 2 is calibrated (i.e., the physical interval from when the tobacco enters the weighing area to when the volume scan is completed). Combined with the conveyor belt speed v, the running time t of the tobacco in this interval is calculated, ensuring that the weight data (q×v) and volume data (V=∑S_n×L_slice, and L_slice=v×Δt, Δt is the scanning interval) of the same batch of tobacco are accurately matched in the spatial-temporal dimension, thereby establishing a dynamic correspondence between instantaneous weight flow and instantaneous volume, providing an accurate real-time data basis for filling value calculation.

[0111] Speed ​​compensation mechanism

[0112] Fluctuations in conveyor belt speed directly affect the accuracy of weight data. Therefore, the system monitors the conveyor belt speed in real time through an encoder and synchronizes the speed signal to the software analysis and calculation system 9 to dynamically correct the weight data.

[0113] For example, when the conveyor belt accelerates, the software analysis and calculation system 9 will predict the actual weight of the current tobacco pile based on historical data of speed changes, avoiding calculation errors caused by delays.

[0114] In this embodiment, the data processing algorithm module employs a multi-source data fusion algorithm and assigns weights to the data. This not only fully leverages the value of data from belt weighing, the 3D laser intelligent camera 7, and the sensor 6, but also comprehensively considers the influence of various factors to improve calculation accuracy and provide reliable indicators for tobacco production. Furthermore, it adapts to complex production environments, ensuring accurate calculation of fill values ​​under different conditions. The intelligent learning model, based on the LSTM deep learning algorithm, can accurately capture the time-series characteristics of historical data, adaptively adjust parameters to optimize calculation results, and continuously improve performance through training with a large amount of historical data. The interface supports multiple transmission protocols, ensuring efficient and stable interaction with different software analysis and calculation systems, achieving seamless integration. It also features data encryption to ensure the security and integrity of production data, safeguarding enterprise interests and production reliability.

[0115] By comprehensively applying these technologies, the online tobacco filling value model can effectively improve the quality control level in the tobacco production process, reduce production costs, and enhance the market competitiveness of enterprises.

[0116] In summary, this online tobacco filling value device and model, by setting up the online tobacco filling value device with the help of components such as a 3D laser intelligent camera 7, a belt weighing system 2, and a software analysis and calculation system 9, realizes real-time online detection of tobacco filling value. The 3D laser intelligent camera 7 scans the volume information in real time, the belt weighing system 2 weighs the weight information in real time, and the software analysis and calculation system 9 calculates the filling value through calculation and proportional conversion. It can also obtain empirical coefficients to optimize the data through offline data analysis, which greatly simplifies the detection process and provides accurate and effective tobacco filling value data for tobacco processing departments, helping to control production processes and product quality.

[0117] Furthermore, by setting up a protective component 8 consisting of a motor 803, a threaded rod 804, a sprocket 806, a chain 807, a moving plate 808, and a wiping cloth 809, dust can be effectively removed from the surfaces of the sensor 6 and the 3D laser intelligent camera 7. When dust adheres during monitoring, the motor 803 drives the wiping cloth 809 to move and wipe, ensuring that the equipment accurately collects data on the weight, volume, and environment of the tobacco, ensuring accurate calculation of the tobacco filling value, providing a reliable basis for the production process, reducing the frequency of equipment maintenance, and extending the service life of the equipment.

[0118] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0119] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An online tobacco filling value detection device, comprising a support base (1), characterized in that, A belt weighing system (2) is fixedly installed on the top of the support base (1). Two brackets (3) are fixedly installed on the top of the support base (1) and the belt weighing system (2). Fixing blocks (4) are installed on the surface of the two brackets (3). Mounting plates (5) are fixedly installed on the bottom of the two fixing blocks (4). Sensors (6) and 3D laser smart cameras (7) are fixedly installed on the bottom of the two mounting plates (5). Protective components (8) are installed inside the two mounting plates (5). The protective assembly (8) includes two mounting plates (5) with mounting grooves (801) on their bottoms. Each of the two mounting plates (5) has a movable cavity (802) inside. A motor (803) is fixedly mounted on one end of each of the two mounting plates (5). Two threaded rods (804) are rotatably mounted inside each of the two movable cavities (802). Threaded blocks (805) are threadedly connected to the surfaces of the two threaded rods (804). Sprockets (806) are fixedly mounted on the outer surfaces of the two threaded rods (804). The two sprockets (806) are connected by a chain (807). A movable plate (808) is fixedly mounted between each pair of threaded blocks (805). A wiping cloth (809) is pasted on one side of each of the two movable plates (808).

2. The online tobacco filling value detection device according to claim 1, characterized in that, The sensor (6) and the 3D laser smart camera (7) are located inside the mounting slot (801), respectively.

3. The online tobacco filling value detection device according to claim 2, characterized in that, A visible-near-infrared spectral camera is installed next to the 3D laser smart camera (7) to simultaneously collect the spectral reflectance characteristics of tobacco.

4. The online tobacco filling value detection device according to claim 1, characterized in that, The sensor (6), 3D laser intelligent camera (7), belt weighing system (2) and visible-near infrared spectroscopy camera are all electrically connected to the software analysis and calculation system (9).

5. The online tobacco filling value detection device according to claim 1, characterized in that, One end of each of the two threaded rods (804) is fixedly connected to the output shaft of the motor (803), and the two wiping cloths (809) are in contact with the sensor (6) and the 3D laser smart camera (7) respectively.

6. The online tobacco filling value detection device according to claim 1, characterized in that, The two fixed blocks (4) are installed with the bracket (3) by four fastening bolts. The two mounting plates (5) are provided with sliding grooves on both sides of the mounting groove (801), and the two ends of the two moving plates (808) are slidably connected to the sliding grooves.

7. An online tobacco filling value detection model, comprising an online tobacco filling value detection device as described in any one of claims 1-5, characterized in that, It includes a data processing algorithm module, an intelligent learning model, and an interface, among which, Data processing algorithm module: a data processing algorithm module used to construct the weight data collected by the belt weighing system (2), the volume data obtained by the 3D laser smart camera (7) and the environmental data sensed by the sensor (6); Intelligent learning model: used to adaptively adjust calculation parameters and optimize the calculation results of tobacco filling value to improve the accuracy and stability of measurement; Interface: Enables real-time data interaction and sharing.

8. The online tobacco filling value detection model according to claim 7, characterized in that, The data processing algorithm module uses a multi-source data fusion algorithm to fuse the weight data collected by the belt weighing system (2), the volume data obtained by the 3D laser smart camera (7), and the environmental data sensed by the sensor (6); Let W be the weight data collected by the belt weighing system, V be the volume data acquired by the 3D laser intelligent camera, and E be the environmental data sensed by the sensor. n The formula for calculating the integrated data D after data fusion is as follows: Where, ω W ω V , The weights are respectively for weight data, volume data, and various environmental data, and satisfy the following conditions: The initial weights can be set based on experience, and then optimized through training with historical data and error backpropagation algorithm to improve the accuracy of tobacco filling value calculation.

9. The online tobacco filling value detection model according to claim 7, characterized in that, The intelligent learning model is built based on deep learning algorithms and is trained using a long short-term memory network (LSTM) on a large amount of historical tobacco filling value data. Let the input sequence be x t The hidden layer state is h t The cell state is c t The forget gate, input gate, and output gate are f, respectively. t i t , o t The calculation process of the LSTM network is as follows: f t =σ(W f ·[h t-1 ,x t ]+b f ) i t =σ(W i ·[h t-1 ,x t ]+b i ) the t =σ(W o ·[h t-1 ,x t ]+b o ) h t =o t ⊙tanh(c t ) Among them, W f W i W c W o Let b be the weight matrix. f b i b c b o Let σ be the bias vector, σ be the sigmoid activation function, and tanh be the hyperbolic tangent activation function. This represents element-wise multiplication; by training an LSTM network, it is possible to effectively capture the time-series features in the data, thereby more accurately and adaptively adjusting the calculation parameters and optimizing the calculation results of the tobacco filling value.

10. The online tobacco filling value detection model according to claim 7, characterized in that, The interface supports multiple data transmission protocols, including but not limited to HTTP and TCP / IP, and can perform stable and efficient real-time data interaction and sharing with the software analysis and computing system (9), while also having data encryption function; The transmitted data is encrypted using AES (Advanced Encryption Standard). Let the plaintext data be P, the key be K, and the encrypted ciphertext be... The formula for calculating C is as follows: C=AES K (P) At the receiving end, the ciphertext data C is decrypted using the same key K to recover the plaintext data P: Among them, AES -1 This is an AES decryption function that uses the AES encryption algorithm to ensure the security of data transmission.

Images