Dynamic cut tobacco weight real-time detection system and method based on microwave technology

Abstract

The invention discloses a dynamic tobacco shred weight real-time detection system and method based on a microwave technology, and belongs to the technical field of mass flow measurement. Comprising the steps that a microwave resonant cavity assembly (1) is installed on a vertical section of a cut tobacco conveying pipeline (2), and the cut tobacco density is measured in real time through microwave signal frequency and amplitude variation; the wind speed sensor assembly (6) is installed on a straight pipe section of the negative pressure pipeline (7) and used for monitoring negative pressure wind speed and calculating the cut tobacco conveying speed in combination with a cut tobacco speed correction coefficient. An internal electrical assembly of the electric cabinet (5) carries out fusion operation based on the tobacco shred density, the conveying speed and the tobacco shred weight correction coefficient, meanwhile, an adaptive Kalman filtering algorithm and a multi-parameter dynamic compensation model are introduced, and it is guaranteed that the real-time weight of the tobacco shred is accurately output. A special detection system and method are designed, the density and speed of the tobacco shreds in the conveying pipeline are synchronously detected on line, the limitation of single-parameter detection is broken through, and the real-time consumption metering precision of the tobacco shreds of each tobacco machine and the production process control capability can be improved.

Description

Technical Field

[0001] This invention relates to the field of mass flow measurement technology, specifically to a real-time dynamic tobacco weight detection system and method based on microwave technology. It is applicable to real-time metering and process control of bulk materials in industries such as tobacco, grain processing, chemicals, and pharmaceuticals. Background Technology

[0002] Currently, the tobacco industry mainly relies on traditional weighing methods or offline detection technologies for measuring tobacco consumption, which presents the following prominent problems: Traditional methods suffer from low accuracy and significant time lag. Traditional manual sampling inspection methods generally have an error exceeding 10%, and the inspection time per batch is long, making it difficult to meet the needs of intelligent manufacturing for real-time control of the production process and optimization of raw material utilization.

[0003] Offline detection cannot be synchronized with production. Offline detection technology cannot operate synchronously with the production process, making it difficult to provide real-time feedback on tobacco consumption dynamics. This results in delayed adjustments to production parameters, a reduction in processing efficiency of over 15%, and a decrease in product quality uniformity, severely hindering the automation and intelligent development of tobacco consumption measurement.

[0004] Existing online detection technologies have significant limitations, and current online detection technologies (such as density measurement schemes based on capacitance or optical sensors) face multiple challenges in their widespread application: Poor environmental adaptability: As a loose, discontinuous, and poorly flowing bulk material, tobacco shreds have limited sensor stability in dusty and vibrating industrial environments, with data fluctuations reaching 8% to 12%. Significant impact of speed fluctuations: The maximum speed fluctuation during tobacco conveying exceeds 20%, and the uneven distribution within the pipeline will introduce significant measurement errors; The system is costly and complex to install: the existing equipment has a complex structure, with a single unit costing more than 100,000 yuan. Moreover, it focuses on single-parameter monitoring, requiring large-scale modification of the production line during installation. The modification cycle is long, compatibility is poor, and the promotion cost is high.

[0005] Therefore, existing detection technologies generally suffer from poor environmental adaptability, significant limitations in single-parameter detection, and insufficient anti-interference capabilities, making it impossible to accurately detect the real-time weight of tobacco shreds, resulting in increased costs and difficulties in quality control. There is an urgent need for a high-precision, high-stability, easily integrated, and suitable dynamic real-time tobacco shred weight detection system for complex industrial environments. Summary of the Invention

[0006] The purpose of this invention is to address the aforementioned problems by providing a real-time online tobacco weight detection system and method based on the dual-parameter fusion of a microwave resonant cavity and a wind speed sensor. The microwave resonant cavity component detects tobacco density in real time, the wind speed sensor monitors the negative pressure wind speed and calculates the tobacco conveying speed, and the speed correction coefficient is calibrated on-site. and weight correction factor Meanwhile, an adaptive Kalman filter algorithm and a dynamic compensation model are introduced, integrating intelligent self-calibration, adaptive working condition identification, remote diagnosis and data interaction functions to ensure that the real-time detection accuracy of tobacco consumption of each tobacco machine reaches ±3.5%, reducing tobacco raw material waste by 5% to 10% and improving production efficiency by more than 20%.

[0007] The technical solution of the present invention is as follows: A real-time dynamic tobacco weight detection system based on microwave technology, characterized in that it includes: A microwave resonant cavity assembly is directly connected in series to the tobacco conveying path by replacing a portion of the vertical section of the tobacco conveying pipe. It is used to measure the density of the tobacco flowing through it in real time based on the changes in the microwave resonant frequency and amplitude within the cavity. The wind speed sensor assembly is installed through mounting holes on the straight section of the negative pressure pipeline, with the sensing diaphragm extending into the pipeline for real-time monitoring of negative pressure wind speed. The data processing and control unit is integrated within the electrical cabinet and is connected to the microwave resonant cavity assembly via a first coaxial cable and to the wind speed sensor assembly via a second signal line to receive density and wind speed signals. The data processing and control unit is configured to: based on the wind speed signal and a pre-calibrated speed correction coefficient... The tobacco shred velocity is calculated, and an adaptive Kalman filter algorithm is used to fuse the density signal with the calculated tobacco shred velocity for optimal estimation. Finally, a weight correction factor is incorporated. Based on the pipe cross-sectional area parameters, calculate and output the real-time mass flow rate and cumulative weight of the tobacco.

[0008] The aforementioned system establishes a core measurement architecture based on "dual-parameter fusion." By simultaneously and in real-time detecting tobacco density and conveying speed, it fundamentally solves the problem of inaccurate measurement in traditional single-parameter detection methods (such as measuring only density) under material flow and speed fluctuation scenarios, laying a system foundation for achieving high-precision dynamic mass flow rate detection.

[0009] Furthermore, the microwave resonant cavity assembly includes an internal coaxial through conduit. During installation, the two ends of the conduit are directly connected and fastened to the tobacco conveying pipes at both ends of the cut-off section through a variable diameter clamp device, and an O-ring is radially arranged on the docking surface to achieve sealing. The wind speed sensor assembly is fixed to the outer wall of the negative pressure pipe by a fixed bracket, and the thickness direction of its sensing diaphragm is parallel to the axis of the negative pressure pipe.

[0010] The above system clarifies the specific installation and sealing structure of the resonant cavity and wind speed sensor. A "replacement" series installation ensures the consistency of the flow channel in the density measurement section, and reliable sealing is achieved through O-rings and clamps. This design guarantees the long-term stability of the sensor in dusty environments while minimizing modifications to existing production lines, thus improving the system's engineering applicability and reliability.

[0011] Furthermore, the electrical cabinet contains a microwave source component, an isolator component, an amplifier component, an analog-to-digital converter module, and a microcontroller connected in sequence. The excitation signal generated by the microwave source component is transmitted to the microwave resonant cavity component via the isolator component and a coaxial cable. The response signal returned by the microwave resonant cavity component after sensing the tobacco returns along the original path. After being isolated by the same isolator component to prevent reflected interference, it is sent to the amplifier component for amplification. After being sampled by the analog-to-digital converter module, it is sent to the microcontroller for processing.

[0012] The aforementioned system defines a dedicated signal processing link within the electrical cabinet. Through a closed-loop design of "microwave source → isolator → resonant cavity → isolator → amplifier," mutual interference between the signal transmission and reception paths is effectively isolated, and the weak detection signal is amplified. This significantly improves the signal-to-noise ratio and anti-interference capability of the microwave density measurement signal, providing a clean and stable signal source for high-precision data processing at the back end.

[0013] Furthermore, the data processing and control unit is communicatively connected to an industrial control screen, which is integrated into the electrical cabinet and is used to display density, wind speed, weight data and system status in real time. It is also equipped with parameter calibration buttons, historical data query buttons and system reset buttons.

[0014] The system integrates a localized human-machine interface and display terminal (industrial control screen). This enables real-time visualization of detection data, convenient calibration of key parameters (such as correction coefficients), and direct monitoring of system status. This significantly lowers the technical barrier to system operation and maintenance, allowing on-site personnel to promptly grasp production consumption and intervene.

[0015] Furthermore, the data processing and control unit is configured to run an adaptive Kalman filter algorithm, which takes time-series data synchronously collected from the microwave resonant cavity assembly and the wind speed sensor assembly as input, and iteratively outputs the optimal estimates of tobacco density and velocity by dynamically updating the statistical characteristics of process noise and observation noise for subsequent weight calculation.

[0016] The aforementioned system incorporates an adaptive Kalman filter algorithm as its core data processing engine. This algorithm automatically learns and tracks the statistical characteristics of noise during tobacco conveying, dynamically optimizing the filtering parameters. This effectively suppresses random measurement noise caused by airflow pulsation and mechanical vibration, significantly improving the smoothness and accuracy of density and velocity estimates, and giving the system excellent dynamic adaptability.

[0017] This application also includes a method for real-time detection of dynamic tobacco weight based on microwave technology, which employs a real-time detection system for dynamic tobacco weight based on microwave technology, and includes the following steps: Signal acquisition: The simulated signal of tobacco density is acquired in real time through a microwave resonant cavity component connected in series to the tobacco conveying path; the simulated signal of negative pressure wind speed is acquired in real time through a wind speed sensor component installed in the negative pressure pipeline; Signal transmission and conversion: The density analog signal is transmitted to the electrical cabinet via a coaxial cable, and the wind speed analog signal is transmitted to the electrical cabinet via a signal line, and synchronous analog-to-digital conversion is performed inside the electrical cabinet; Data fusion and estimation: An adaptive Kalman filter algorithm is used to fuse the converted digital density signal and wind speed signal to output the optimal tobacco density estimate. With the optimal tobacco speed estimate ; Weight calculation and output: Based on the optimal estimate, combined with the pre-calibrated tobacco weight correction factor. and pipe cross-sectional area and time Calculate the real-time weight of the tobacco: ; The mean value is estimated for the tobacco density. The mean value is the estimated speed of the tobacco shreds; Calculate the total weight of the tobacco: ; It is displayed and output through an industrial control panel.

[0018] The above method defines a complete and coherent detection process from signal acquisition to weight output, and embeds core calculation formulas. Standardizing and proceduralizing the entire detection process ensures the clarity and repeatability of the calculation logic. The explicit formulas directly establish a mathematical model between the physical measurement values ​​and the final management target (weight), giving the detection results clear physical meaning and engineering value.

[0019] Furthermore, in the data fusion and estimation, the adaptive Kalman filter algorithm performs the following process: Establish a system model: construct the state equation with tobacco density and velocity as state variables, and construct the observation equation with sensor measurements as observations; Dynamic noise update: The process noise covariance matrix is ​​dynamically adjusted based on the residuals between the predicted state and the observed values. Covariance matrix of observation noise The updated formula is: , , in, To predict residuals, To observe the residuals, and A forgetting factor between 0 and 1, used to balance historical noise and current noise; Iterative optimal estimation: Calculate the Kalman gain using the updated noise matrix, iteratively execute the prediction and update steps, and output the optimal estimates of tobacco density and velocity at each time step.

[0020] The above method refines the execution steps and core update formula of the adaptive Kalman filter algorithm. Through a formulaic dynamic noise update mechanism (updating the Q and R matrices), the filtering process is no longer based on fixed parameters, but can self-adjust according to real-time operating conditions (such as signal abrupt changes caused by tobacco clumping). This ensures that the system maintains superior estimation accuracy and robustness even in complex industrial environments with non-steady-state, non-Gaussian noise.

[0021] Furthermore, before weight calculation and output, it also includes: On-site calibration: During initial system installation or changes in operating conditions, the control system is operated to collect 10 to 30 sets of data covering different operating conditions, including tobacco density, negative pressure wind speed, and corresponding standard weight. The least squares method is used for fitting to determine the tobacco speed correction coefficient under the current operating condition. Correction factor for tobacco weight .

[0022] Based on the above methods, a "small-sample on-site calibration" method is proposed. This overcomes the shortcomings of traditional calibration methods, which require a large number of samples, are time-consuming and labor-intensive, and are difficult to adapt to differences in different production lines. Only a small amount (10-30 sets) of targeted on-site data is needed to quickly and accurately fit correction coefficients suitable for the specific equipment and material characteristics, greatly improving system deployment efficiency and personalized accuracy in different application scenarios.

[0023] Furthermore, a multi-parameter dynamic compensation mechanism is introduced in the weight calculation and output: when the optimal estimate sequence output by the adaptive Kalman filter algorithm shows a specific abrupt change or continuous fluctuation pattern, it is determined to be a tobacco clumping or unstable airflow condition, and the pre-stored compensation coefficient is automatically called to correct the density or velocity value used for the final calculation in real time.

[0024] The above method incorporates a dynamic compensation mechanism based on operating condition identification. This enables the system to not only perform general filtering but also intelligently identify specific abnormal operating conditions (such as clumping and strong airflow disturbances). By invoking preset compensation rules, the algorithm output is specifically corrected, effectively mitigating special measurement errors caused by drastic changes in the physical state of materials and expanding the system's reliable measurement range under extreme conditions.

[0025] Furthermore, it also includes: Remote interaction and diagnostics: The calculated real-time weight, system status, and alarm information are uploaded to the factory management system through the communication interface integrated with the data processing and control unit; when data exceeds limits or equipment failure is detected, a local audible and visual alarm is triggered and remote alarm information is sent simultaneously.

[0026] The above methods improved data interaction and diagnostic functions. The local detection system was upgraded to a networked intelligent node. Remote centralized monitoring and production management informatization were achieved through data uploads; the combination of local alarms and remote diagnostics enabled rapid fault detection, location, and response, improving the overall operational efficiency and intelligence level of the production line.

[0027] Compared with existing technologies, the advantages of this invention are: 1. This application realizes real-time detection of tobacco density through a microwave resonant cavity, and combines it with a wind speed sensor to measure the tobacco conveying speed, thus constructing a dynamic measurement link of density-speed-weight, which can realize accurate online detection of real-time tobacco weight; 2. This application introduces an adaptive Kalman filter algorithm and a dynamic compensation model, which can automatically adjust the filter parameters according to non-uniform working conditions such as looseness, clumping, and speed fluctuations during the tobacco conveying process, thereby improving the stability and accuracy of density, speed, and weight calculations. 3. This application adopts a fully automated data acquisition and processing flow, which can complete signal filtering, compensation calculation, real-time display and weight statistics without manual intervention, greatly improving the detection efficiency and the automation level of the production process; 4. The system structure of this application is modular and easy to install. It can be directly applied to the renovation and upgrading of existing smoke machines, and has good on-site adaptability and engineering feasibility. 5. This application can operate stably for a long time under complex working conditions such as high dust, strong vibration, and significant wind speed fluctuations. It helps to reduce raw material waste, reduce energy consumption, improve raw material utilization and product consistency, and has significant industrial application value and broad promotion prospects. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the overall system structure of this application.

[0029] Figure 2This is a schematic diagram of the microwave resonant cavity assembly structure.

[0030] Figure 3 A fixed cross-sectional view of the resonant component and the tobacco duct.

[0031] Figure 4 for Figure 3 A magnified view of a portion of point A in the middle.

[0032] Figure 5 This is a schematic diagram of the wind speed sensor assembly.

[0033] Figure 6 for Figure 5 A magnified view of a section at point B.

[0034] Figure 7 This is a schematic diagram of the sensor diaphragm installation.

[0035] Figure 8 This is a schematic diagram of the external structure of the electrical cabinet.

[0036] Figure 9 This is a schematic diagram of the front structure of the electrical cabinet.

[0037] Figure 10 This is a schematic diagram of the connection structure between the industrial control panel and the electrical cabinet.

[0038] Figure 11 This is a schematic diagram of the external structure of an industrial control panel.

[0039] Figure 12 This is a schematic diagram of the electrical components.

[0040] Figure reference numerals: 1-Microwave resonant cavity assembly, 2-Tobacco conveying pipe, 3-Coaxial cable, 4-Tobacco machine, 5-Electrical cabinet, 6-Wind speed sensor assembly, 7-Negative pressure pipe, 8-Cavity, 9-Cavity top cover, 10-Top cover fixing bolt, 11-Conduit, 12-Variable diameter clamp device, 13-Resonant cavity assembly wiring terminal, 14-Tobacco flow direction, 15-O-ring, 16-Sealant, 17-Electrode, 18-Fixing bracket, 19-Induction diaphragm, 20-Mounting hole, 21-Cooling fan, 22-Electrical cabinet wiring area, 23- - Signal line interface, 24- Tactile switch, 25- Interface mounting plate, 26- Industrial control panel, 27- Industrial control panel power switch, 28- External power supply conduit, 29- USB interface, 30- Emergency stop button, 31- Industrial control panel housing, 32- Industrial control panel screen, 33- Query historical data button, 34- Parameter calibration button, 35- System reset button, 36- Terminal block, 37- Cable tray, 38- Microwave source assembly, 39- Amplifier assembly, 40- Isolator assembly, 41- Temperature control assembly, 42- Main power supply for electrical cabinet. Detailed Implementation

[0041] It should be noted that relational terms such as "first" and "second" are used merely 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.

[0042] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0043] Please see Figure 1-12 A real-time dynamic tobacco weight detection system based on microwave technology, such as Figure 1 As shown, the system includes a microwave resonant cavity assembly 1, a tobacco conveying pipe 2, an electrical cabinet 5, a wind speed sensor assembly 6, and a negative pressure pipe 7. The microwave resonant cavity assembly 1 is installed in the vertical section of the tobacco conveying pipe 2, sealed with an O-ring 15, and fixed with a variable diameter clamp device 12. It measures the tobacco density in real time by measuring the changes in the frequency and amplitude of the microwave signal; the greater the density, the greater the change. The wind speed sensor assembly 6 is installed in the straight section of the negative pressure pipe 7 to monitor the negative pressure wind speed and calculate the tobacco conveying speed. The electrical components inside the electrical cabinet 5 perform fusion calculations on the tobacco density and flow rate to calculate and output the real-time weight of the tobacco.

[0044] In an embodiment, Figure 2 This is a schematic diagram of the microwave resonant cavity assembly. The microwave resonant cavity assembly 1 includes a cavity 8, which is made of brass to improve electromagnetic shielding performance and signal stability. The central conduit 11 of the cavity 8 is made of polytetrafluoroethylene (PTFE), a special engineering plastic with a friction coefficient of 0.04–0.1. This material is wear-resistant and does not easily stick to tobacco, ensuring smooth tobacco delivery. The inner diameter of the conduit 11 is consistent with the inner diameter of the tobacco delivery pipe 2. The diameter is 119mm to prevent tobacco accumulation or speed fluctuations caused by changes in the pipe's inner diameter. A microwave coupling structure is integrated below the top cover 9 of the cavity, and the cavity 8 is installed together with the top cover 9 using the top cover fixing bolts 10. The resonant cavity adopts a semi-open structure design, with the optimized resonant mode being the TE101 mode. It is highly sensitive to changes in tobacco density and has an electromagnetic shielding effectiveness ≥100dB, effectively isolating electromagnetic interference from the industrial environment. The inner wall of the cavity 8 is coated with a fluorocarbon coating to prevent dust and condensation, avoiding interference from dust adhesion and water vapor condensation on the microwave signal.

[0045] like Figure 3 and Figure 4 As shown, when installing the microwave resonant cavity assembly 1, the vertical section of the tobacco conveying pipe 2 must first be cut to a suitable length at the corresponding position. Then, the two ends of the conduit 11 are aligned with the cut ends of the tobacco conveying pipe 2, while ensuring that the axial direction of the conduit 11 coincides with the axial direction of the cavity 8. An O-ring 15 is radially fitted onto the mating surface of both to achieve a seal, preventing dust from entering the cavity 8 and interfering with the microwave signal. The connection between the microwave resonant cavity assembly 1 and the tobacco conveying pipe 2 is then tightened and fixed using a variable-diameter clamping device 12. When the tobacco flows along the tobacco flow direction 14 through the conduit 11, the dielectric properties of the tobacco will change the resonant state of the electromagnetic field inside the cavity. The density can be measured by the change in frequency and amplitude of the microwave signal relative to the cavity state of the tobacco conveying pipe 2. .

[0046] like Figure 5 , Figure 6 and Figure 7 As shown, the wind speed sensor assembly 6 of this application includes a wind speed sensor electrode 17. During installation, an installation hole 20 is opened in the straight section of the negative pressure pipe 7. The sensor end extends into the interior of the negative pressure pipe 7 through the installation hole 20. The thickness direction of the sensing diaphragm 19 must be parallel to the axial direction of the negative pressure pipe 7 to ensure that the airflow acts uniformly on the sensing diaphragm 19 and ensure accurate wind speed measurement. The wind speed sensor assembly 6 is fixed to the outer wall of the negative pressure pipe 7 by a fixing bracket 18. The gap between the fixing bracket 18 and the outer wall of the negative pressure pipe 7 is filled with sealant 16 to achieve dustproof sealing and airflow isolation. The tobacco conveying pipe 2 is connected to the negative pressure pipe 7 to form a closed space with equal pressure. At this time, it can be approximately considered that the tobacco flow velocity is equal to the negative pressure wind speed of the negative pressure pipe 7. The wind speed sensor assembly 6 senses the change in airflow velocity through the sensing diaphragm 19, converts the velocity signal into an collectable electrical signal, and after built-in temperature compensation, combines it with the tobacco velocity correction coefficient calibrated on site. This allows for the accurate calculation of the real-time tobacco conveying speed. .

[0047] To adapt to different tobacco types, environmental changes, and speed differences between different tobacco machines under negative pressure conditions, this application constructs a parameter adaptive correction method based on small-sample calibration. During the installation and commissioning phase or initial production run, a small number (typically 10-30 sets) of microwave density signals, negative pressure wind speed signals, and corresponding standard weight data are collected. A small-sample regression model is then used to adjust the tobacco speed correction coefficient. With weight correction factor Automatic fitting and optimization are performed. The small-sample calibration method does not rely on large-scale training data, but rather on a limited number of samples collected under actual field conditions. It establishes the mapping relationship between density, wind speed, and weight through least-squares estimation or a lightweight regression model, and outputs the optimal result. , The coefficient is used to improve the accuracy of subsequent real-time detection.

[0048] The small-sample calibration mechanism in this application can continue to calibrate based on real-time residuals during system operation. , Fine-tuning is performed. When a persistent deviation is detected between the estimated density or wind speed and the actual weight reference value, the system automatically initiates an incremental update strategy to iterate the calibration model in small increments, ensuring that the parameters remain optimal as operating conditions change. This method effectively solves the problems of traditional calibration methods, such as reliance on a large number of samples, long calibration cycles, and high levels of human intervention, significantly improving the system's adaptability to different tobacco varieties, different ambient temperatures and humidity levels, and different negative pressure wind velocities.

[0049] like Figure 8 and Figure 9 As shown, the electrical cabinet 5 of this application includes electrical components and an industrial control panel 26. A cooling fan 21 is installed on the side of the electrical cabinet 5 for heat dissipation. The front of the electrical cabinet 5 is the electrical wiring area 22, which includes three signal line interfaces 23 and two tactile switches 24. The three signal line interfaces 23 are used for the input and output of microwave signals, fan speed signals, and alarm signals, respectively, and are fixed to the front of the electrical cabinet 5 by interface fixing plates 25. The electrical cabinet 5 can be installed and fixed on the top of the tobacco machine 4 near the tobacco conveying pipe 2.

[0050] In an embodiment, such as Figure 10 , Figure 11 and Figure 12 As shown, the electrical cabinet 5 includes an industrial control panel 26 and electrical components. It adopts a functional partition design, with each partition electrically connected to the terminal block 36 via prefabricated cable trays 37, forming a closed-loop system of signal input-processing-computation-output. Specifically, it encompasses a microwave source component 38, a data acquisition and processing unit, the industrial control panel 26, and a temperature control component 41. These units work together to complete the data processing and weight display process. After the main power supply 42 of the electrical cabinet supplies power to the microwave source component 38, the microwave source component 38 uses a voltage-controlled oscillator to provide a stable microwave signal. This signal is first filtered for noise by the isolator component 40, and then transmitted to the microwave resonant cavity component 1 via the coaxial cable 3. When the tobacco shreds cause a frequency shift and amplitude attenuation in the microwave signal after passing through the resonant cavity conduit 11, the attenuated signal returns to the electrical cabinet 5 along the original path. The isolator component 40 can block the reflected signal transmitted in the opposite direction, preventing it from interfering with the microwave source component 38. Subsequently, the weak attenuated signal enters the amplifier assembly 39, which first amplifies the signal with low loss, and then the power amplifier further enhances the signal strength, ultimately amplifying the signal amplitude to a range that the data acquisition module can recognize.

[0051] The data acquisition and processing unit establishes data connections with the microwave source assembly 38 and the wind speed sensor assembly 6 via terminal block 36, respectively, and constructs a state equation based on the instantaneous density measurement value of the microwave resonant cavity and the instantaneous wind speed value of the wind speed sensor: , in, A two-dimensional state vector containing density and wind speed; This is the state transition matrix; The noise is the process noise, and the noise covariance is... .

[0052] The analog-to-digital converter (ADC) receives microwave signals from amplifier assembly 39 and current signals from wind speed sensor assembly 6 via signal lines. The ADC synchronously converts these two analog signals into digital signals. The observation equation is as follows: , in, This is actual sensor measurement data; The observation matrix; To observe the noise, the noise covariance is: .

[0053] During the conversion process, a built-in adaptive Kalman filter algorithm is used to improve data accuracy. After the digital signal is transmitted to the microcontroller, the arithmetic module first starts a preprocessing program to perform a moving average filter on the instantaneous density and instantaneous wind speed signals. By selecting the mean of multiple consecutive sampling points, signal fluctuations caused by random interference factors such as tobacco clumping and airflow pulsation are reduced. After preprocessing, the data fusion algorithm is started, dynamically adjusting the weights of the density and wind speed signals, and dynamically adjusting the noise covariance matrix based on the variance or moving average of the residuals. , , in and The forgetting factor is between 0 and 1, used to balance historical noise and current noise.

[0054] The Kalman gain is calculated by predicting and updating the data, and using the updated noise covariance matrix. And update the state estimate: , This yields optimal estimates of density and wind speed. These are then combined with the previously calibrated speed correction factor. Weight correction factor The cross-sectional area parameters of the tobacco conveying pipe 2 are used to calculate the real-time tobacco flow rate using the mass flow rate formula, and then the total weight is obtained by accumulating the time.

[0055] The real-time tobacco flow rate is calculated as follows: , The actual weight of the tobacco is .

[0056] After receiving data, the internal drive module of the industrial control screen 26 first converts the data format before transmitting it to the industrial control screen 32. Finally, it displays the tobacco density, average wind speed, real-time weight, cumulative weight, and system operating status in real time, combining numerical values ​​and curves. Operators can interact with the industrial control screen 26 using the function buttons: pressing the parameter calibration button 34 sends a calibration command to the microcontroller, which then calls the calibration program in the on-chip memory to guide the system in completing the correction coefficient calibration; pressing the query historical data button 33 reads past detection data stored in the on-chip memory and displays it in chronological order, facilitating historical data tracing; pressing the system reset button 35 sends a reset signal to the microcontroller, which restarts the data processing program and clears abnormal data.

[0057] The data communication interface connects to the factory management system via an industrial standard communication protocol. The wireless communication module uses the SIM8200EA-M2 module, supporting 5G and WiFi communication protocols, enabling data interaction with the factory's MES system and remote monitoring platform. The microcontroller packages real-time weight data in a preset format and uploads it to the management system through the data communication interface for remote monitoring. The temperature control component 41 and the main power supply 42 within the electrical cabinet 5 provide a stable environment for the data processing flow: the main power supply 42 first converts the external AC voltage to DC voltage, then supplies power to each unit through the power distribution module. When the current in a unit exceeds a threshold or the voltage is abnormal, the protection circuit automatically cuts off the power supply to that unit to prevent the fault from spreading. The temperature control component 41 collects the internal temperature of the electrical cabinet 5 in real time. When the temperature exceeds the preset upper limit, the temperature control component 41 activates the cooling fan 21 to ensure that precision components such as the microcontroller and analog-to-digital converter always operate within a suitable temperature range, preventing temperature and humidity changes from affecting data processing accuracy.

[0058] To address the random fluctuations and measurement deviations in microwave resonant density and negative pressure wind speed signals caused by factors such as clumping, wind speed fluctuations, and dust disturbances during tobacco conveying, this application proposes a joint estimation method for tobacco density and wind speed based on adaptive Kalman filtering. This method is not an improvement on the Kalman filtering algorithm itself, but rather an adaptive design of the noise parameter update strategy in the Kalman filter based on the operating conditions of tobacco conveying, enabling it to achieve real-time stable estimation under dynamic conditions.

[0059] This application constructs a state equation based on microwave resonant cavity signals and wind speed sensor signals, and dynamically updates the process noise covariance matrix according to the measurement residuals. Covariance matrix of observation noise This allows for adaptive adjustment of filtering parameters to production conditions. The method effectively suppresses instantaneous errors caused by wind speed fluctuations, reduces the impact of density jumps caused by tobacco agglomeration, and improves the stability and accuracy of density, wind speed, and real-time weight calculations.

[0060] The computation module incorporates a Kalman filter algorithm and a dynamic compensation model to ensure both rapid response and high accuracy in the detection results. The Kalman filter algorithm, tailored to the dynamic transport characteristics of tobacco, constructs a dimensional state equation and an observation equation. Tobacco density and transport speed are used as core state variables, while environmental temperature and humidity changes, airflow disturbances, and other factors are included in the process noise matrix. Sensor noise is incorporated into the observation noise matrix. During computation, the density and velocity states at the current moment are predicted based on the optimal estimate and state transition matrix from the previous moment. The weights of the predicted and observed values ​​are dynamically adjusted by combining the current sensor observations. When sensor signal fluctuations are small, the weight of the observed values ​​is increased; conversely, the weight of the predicted values ​​is increased. This rapidly eliminates random noise, achieving a balance between dynamic response and detection accuracy.

[0061] The dynamic compensation model addresses complex operating conditions such as fluctuating tobacco density (sometimes loose, sometimes clumped, with fluctuating speed) by establishing a multi-parameter coupled correction mechanism. Based on extensive experimental data, the model constructs a mapping relationship between correction coefficients and tobacco state parameters: when the system detects a sudden change in tobacco density, it identifies a clumping condition and automatically applies the clumping compensation coefficient to correct the density measurement deviation caused by clumping; when wind speed signals fluctuate continuously, the speed correction coefficient is adjusted using the speed fluctuation coefficient to offset the impact of uneven airflow on speed measurement. The model possesses self-learning capabilities; during long-term operation, it records detection errors under different operating conditions and continuously optimizes compensation parameters using a gradient descent algorithm, gradually improving the system's adaptability to complex operating conditions.

[0062] In one embodiment, if an abnormal data is detected, the industrial control screen 26 is immediately triggered with an audible and visual alarm, and the fault information is recorded. At the same time, a fault diagnosis report is sent to a remote terminal. Staff can query the fault details through the industrial control screen 26 or receive the report through the remote terminal to troubleshoot and repair the fault in a timely manner.

[0063] In another embodiment, the system uses a small-sample calibration method to obtain the tobacco speed correction coefficient. and weight correction factor A regression model was constructed by collecting 10 to 30 sets of data on tobacco density, wind speed and standard weight, and correction coefficients were obtained using least squares or lightweight fitting algorithms to improve the accuracy of real-time weight calculation.

[0064] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A real-time dynamic tobacco weight detection system based on microwave technology, characterized in that, include: The microwave resonant cavity assembly (1) is directly connected in series to the tobacco conveying path by replacing a part of the vertical section of the tobacco conveying pipe (2), and is used to measure the density of the tobacco flowing through it in real time based on the change of the microwave resonant frequency and amplitude in the cavity. The wind speed sensor assembly (6) is installed through the mounting hole (20) on the straight section of the negative pressure pipe (7) and the sensing diaphragm (19) extends into the pipe for real-time monitoring of negative pressure wind speed. The data processing and control unit is integrated in the electrical cabinet (5) and is connected to the microwave resonant cavity assembly (1) via a first coaxial cable and to the wind speed sensor assembly (6) via a second signal line to receive density and wind speed signals. The data processing and control unit is configured to: receive density and wind speed signals based on the wind speed signal and a pre-calibrated speed correction coefficient. The tobacco shred velocity is calculated, and an adaptive Kalman filter algorithm is used to fuse the density signal with the calculated tobacco shred velocity for optimal estimation. Finally, a weight correction factor is incorporated. Based on the pipe cross-sectional area parameters, calculate and output the real-time mass flow rate and cumulative weight of the tobacco.

2. The real-time dynamic tobacco weight detection system based on microwave technology according to claim 1, characterized in that, The microwave resonant cavity assembly (1) includes an internal coaxial through conduit (11). During installation, the two ends of the conduit (11) are directly connected and fastened to the tobacco conveying pipes (2) at both ends of the cut-off section through a variable diameter clamp device (12), and an O-ring (15) is radially arranged on the docking surface to achieve sealing. The wind speed sensor assembly (6) is fixed to the outer wall of the negative pressure pipe (7) through a fixed bracket (18), and the thickness direction of its sensing diaphragm (19) is parallel to the axis of the negative pressure pipe (7).

3. The real-time dynamic tobacco weight detection system based on microwave technology according to claim 1, characterized in that, The electrical cabinet (5) contains a microwave source component (38), an isolator component (40), an amplifier component (39), an analog-to-digital converter module, and a microcontroller connected in sequence. The excitation signal generated by the microwave source component (38) is transmitted to the microwave resonant cavity component (1) via the isolator component (40) and the coaxial cable (3). The response signal returned by the microwave resonant cavity component (1) after sensing the tobacco returns along the original path. After being isolated by the same isolator component (40) to isolate the reflected interference, it is sent to the amplifier component (39) for amplification. After being sampled by the analog-to-digital converter module, it is sent to the microcontroller for processing.

4. A real-time dynamic tobacco weight detection system based on microwave technology according to claim 1 or 3, characterized in that, The data processing and control unit is connected to an industrial control panel (26) for communication. The industrial control panel (26) is integrated in the electrical cabinet (5) and is used to display density, wind speed, weight data and system status in real time. It is equipped with a parameter calibration button (34), a historical data query button (33) and a system reset button (35).

5. The real-time dynamic tobacco weight detection system based on microwave technology according to claim 1, characterized in that, The data processing and control unit is configured to run an adaptive Kalman filter algorithm, which takes time-series data synchronously collected from the microwave resonant cavity assembly (1) and the wind speed sensor assembly (6) as input, and iteratively outputs the optimal estimates of tobacco density and velocity by dynamically updating the statistical characteristics of process noise and observation noise for subsequent weight calculation.

6. A method for real-time detection of dynamic tobacco weight based on microwave technology, characterized in that, The real-time dynamic tobacco weight detection system based on microwave technology as described in any one of claims 1-5 includes the following steps: Signal acquisition: The simulated signal of tobacco density is acquired in real time by a microwave resonant cavity assembly (1) connected in series to the tobacco conveying path; the simulated signal of negative pressure wind speed is acquired in real time by a wind speed sensor assembly (6) installed in the negative pressure pipeline (7); Signal transmission and conversion: The density analog signal is transmitted to the electrical cabinet (5) through the coaxial cable (3), and the wind speed analog signal is transmitted to the electrical cabinet (5) through the signal line, and synchronous analog-to-digital conversion is performed in the electrical cabinet (5); Data fusion and estimation: An adaptive Kalman filter algorithm is used to fuse the converted digital density signal and wind speed signal to output the optimal tobacco density estimate. With the optimal tobacco speed estimate ; Weight calculation and output: Based on the optimal estimate, combined with the pre-calibrated tobacco weight correction factor. and pipe cross-sectional area and time Calculate the real-time weight of the tobacco: ; The mean value is estimated for the tobacco density. The mean value for the tobacco speed estimate; Calculate the total weight of the tobacco: ; And it is displayed and output through the industrial control screen (26).

7. The method for real-time detection of dynamic tobacco weight based on microwave technology according to claim 6, characterized in that, In the data fusion and estimation process, the adaptive Kalman filter algorithm executes the following steps: Establish a system model: construct the state equation with tobacco density and velocity as state variables, and construct the observation equation with sensor measurements as observations; Dynamic noise update: The process noise covariance matrix is ​​dynamically adjusted based on the residuals between the predicted state and the observed values. Covariance matrix of observation noise The updated formula is: ， ， in, To predict residuals, To observe the residuals, and A forgetting factor between 0 and 1, used to balance historical noise and current noise; Iterative optimal estimation: Calculate the Kalman gain using the updated noise matrix, iteratively execute the prediction and update steps, and output the optimal estimates of tobacco density and velocity at each time step.

8. The method for real-time detection of dynamic tobacco weight based on microwave technology according to claim 6, characterized in that, Before weight calculation and output, the following is also included: On-site calibration: During initial system installation or changes in operating conditions, the control system is operated to collect 10 to 30 sets of data covering different operating conditions, including tobacco density, negative pressure wind speed, and corresponding standard weight. The least squares method is used for fitting to determine the tobacco speed correction coefficient under the current operating condition. Correction factor for tobacco weight .

9. The method for real-time detection of dynamic tobacco weight based on microwave technology according to claim 6, characterized in that, In the weight calculation and output, a multi-parameter dynamic compensation mechanism is also introduced: when the optimal estimate sequence output by the adaptive Kalman filter algorithm shows a sudden change or continuous fluctuation mode, it is determined to be a tobacco clumping or unstable airflow condition, and the pre-stored compensation coefficient is automatically called to correct the density value or velocity value used for the final calculation in real time.

10. The method for real-time detection of dynamic tobacco weight based on microwave technology according to claim 6, characterized in that, Also includes: Remote interaction and diagnostics: The calculated real-time weight, system status, and alarm information are uploaded to the factory management system through the communication interface integrated with the data processing and control unit; When data exceeds limits or equipment malfunctions are detected, a local audible and visual alarm is triggered and a remote alarm message is sent simultaneously.

Images