Flexible array sensor counting method, system and storage medium adapted to irregularly shaped filter rods

Abstract

This invention relates to the field of intelligent cigarette manufacturing technology, specifically to a flexible array sensor counting method, system, and storage medium adapted to irregularly shaped filter rods. The method includes: using a flexible sensor array to acquire the original pressure signal of the filter rod in real time; constructing a counting model for irregularly shaped filter rods, including: constructing a pressure profile template for each type of filter rod; identifying the current filter rod type based on the pressure profile template and the original pressure signal to obtain a current counting threshold; constructing the counting model for irregularly shaped filter rods based on the current counting threshold; and counting the filter rods using the counting model. This invention adapts to the complex surface of irregularly shaped filter rods using a flexible sensor array, avoiding signal distortion; and combines a pressure profile template with a dynamic threshold model to achieve adaptive counting for multiple filter types, overcoming the shortcomings of fixed thresholds that are prone to misjudgment, and significantly improving counting accuracy under high-speed transmission.

Description

Technical Field

[0001] This invention relates to the field of intelligent cigarette manufacturing technology, specifically to a flexible array sensor counting method, system, and storage medium adapted to irregularly shaped filter rods. Background Technology

[0002] In the cigarette manufacturing industry, accurate counting of filter rods is a key link in ensuring production efficiency and controlling product quality. Among them, irregularly shaped filter rods are the core auxiliary materials in cigarette production, and their counting accuracy directly affects the continuity of cigarette production and the product qualification rate.

[0003] Currently, the filter rod counting technology widely used in cigarette production lines is mainly designed for round filter rods. The closest existing technology is a filter rod counting method and device based on image acquisition and recognition (CN109741303B). This technology is the mainstream solution in the current filter rod counting field. Its technical structure mainly includes a movable box, a drawer-type camera fixing structure, a filter rod plastic grid support structure, an imaging supplementary light, a camera, and a laptop computer. A UPS is installed at the bottom of the box to power all components. The camera is fixed to the drawer-type structure and can be adjusted to the optimal imaging focal length according to the size of the filter rod plastic grid. The supplementary light is used to improve image clarity. The laptop computer communicates with the camera, and after receiving the filter rod images, it sequentially performs filtering and noise reduction, image binarization, morphological processing, and counting analysis to achieve the statistical output of the number of filter rods. This technology improves the counting accuracy of round filter rods to a certain extent by optimizing the image acquisition structure and processing algorithm, and is suitable for batch counting scenarios within filter rod plastic grids in cigarette production. In addition, related technologies include a machine vision-based online cigarette filter counting method (CN104537671A) and a deep learning-based video stream filter counting method and device (CN218332619U). CN104537671A uses a line-scan camera to acquire images of the filter packs and counts the filters using template matching and area segmentation, attempting to solve the recognition error problem caused by filter compression deformation. CN218332619U, on the other hand, utilizes a filter receiving drum, a sensor module, and a counting module. The sensor detects the filter within the drum groove and outputs pulse signals, while the counting module counts the number of pulses to achieve counting.

[0004] However, the aforementioned existing technologies have many intractable defects and shortcomings in practical applications, especially for counting irregularly shaped filter rods such as elliptical and triangular ones. They cannot meet the accurate counting requirements of industrial production lines for irregularly shaped filter rods. The specific problems are as follows: First, the adaptability is extremely poor. Existing technologies are all based on circular filter rods. The image recognition algorithms used in CN109741303B and CN104537671A rely on circular contour features, and the drum groove used in CN218332619U is a circular adaptation structure. When faced with irregularly shaped filter rods, they either cannot accurately identify their contours or cannot stably carry and transport them, resulting in high rates of missed counts and miscounts, and the counting accuracy is generally lower than 98%. Secondly, the operation is not stable enough. Existing technologies lack effective conveying and adjustment mechanisms. For example, CN109741303B relies on the regular placement of filter rods in plastic grids, and the drum groove of CN218332619U cannot adapt to the size of irregularly shaped filter rods. This makes it easy for filter rods to get stuck, shake or tilt during the conveying process, which further aggravates the counting error. Moreover, troubleshooting is complicated and affects the continuity of the production line.

[0005] In addition, existing technologies have limited versatility. Different types and sizes of irregularly shaped filter rods require the replacement of corresponding imaging templates, drums, or plastic grids, which is cumbersome and increases production input and maintenance costs. At the same time, some technologies, such as deep learning-based counting methods, have problems such as low algorithm efficiency and high dependence on ambient light, which cannot adapt to the real-time counting requirements of high-speed production lines (10~20 rods / second).

[0006] The core reason for the aforementioned problems lies in the fact that existing technologies have not been specifically designed for the irregular cross-sectional characteristics of irregularly shaped filter rods. They all follow the counting approach used for circular filter rods, failing to address the core pain points of incomplete contour acquisition, unstable conveying, and incompatible recognition algorithms for irregularly shaped filter rods. Currently, the industry is attempting to adapt to irregularly shaped filter rods by optimizing image recognition algorithms or adjusting mechanical structures, but both have limitations: simply optimizing the algorithm cannot solve the shaking problem during the conveying of irregularly shaped filter rods, and adjusting the mechanical structure requires frequent replacement of parts, making it impossible to achieve universal adaptation for multiple types of irregularly shaped filter rods. It is also difficult to balance counting accuracy, versatility, and operational stability. Therefore, developing a counting technology that can adapt to various irregularly shaped filter rods, achieve accurate counting, and operate stably has become an urgent technical challenge to be solved in the current industrial production field. Summary of the Invention

[0007] The purpose of this invention is to provide a flexible array sensor counting method, system, and storage medium adapted to irregularly shaped filter rods, so as to solve the technical problems of existing filter rod counting technology being only compatible with circular filter rods and having low counting accuracy and poor versatility for irregularly shaped filter rods such as elliptical and triangular ones.

[0008] To achieve the above objectives, embodiments of the present invention provide a flexible array sensor counting method, system, and storage medium adapted to irregularly shaped filter rods, including: A flexible sensor array is used to collect the original pressure signal of the filter rod in real time; Constructing a counting model for irregularly shaped filter rods includes: constructing a pressure profile template for each type of filter rod; identifying the current filter rod type based on the pressure profile template and the original pressure signal to obtain a current counting threshold; and constructing a counting model for irregularly shaped filter rods based on the current counting threshold. The filter rods are counted using the aforementioned irregularly shaped filter rod counting model.

[0009] Optionally, constructing the pressure profile template for each filter rod includes: Collect sample data and preprocess it; The pre-processed samples were averaged to obtain a standard pressure profile template for each filter rod. Core features are extracted from each standard pressure profile template and used as the basis for filter rod identification; the core features include edge pressure peak, center pressure peak, length-to-short axis ratio, and profile similarity index.

[0010] Optionally, based on the pressure profile template and the original pressure signal, identifying the current filter rod type to obtain the current counting threshold includes: Calculate the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template; The filter rod type is determined based on the matching degree. Adaptive adjustment of the counting threshold based on filter rod type.

[0011] Optionally, calculating the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template includes: The matching degree is calculated according to formula (1). (1) in, For real-time pressure profile and the first The degree of matching of the template, For the real-time acquisition and preprocessing of the pressure matrix, This represents the average pressure value of the real-time pressure matrix. For the first Pressure matrix of the template For the first The average pressure value of the template.

[0012] Optionally, adaptive adjustment of the counting threshold based on the filter rod type includes: The adjusted counting threshold is obtained according to formula (2). (2) in, The adaptively adjusted counting threshold, The default counting threshold is used. To adjust the coefficient, This represents the maximum matching degree.

[0013] Optionally, constructing a counting model for irregularly shaped filter rods based on the current counting threshold includes: The counting model is obtained according to formula (3). (3) in, The cumulative count of filter rods. This represents the total number of sampling times. For indicator functions, for The maximum pressure value of the real-time pressure matrix. This is the adaptively adjusted counting threshold.

[0014] Optionally, the method further includes adaptively adjusting the transmission mechanism, including: Real-time acquisition of filter rod edge pressure; Adjusting the guide plate spacing based on the filter rod edge pressure includes: obtaining the adjusted guide plate spacing according to formula (4). (4) in, The adjusted guide plate spacing, This is the current guide plate spacing. This is the proportional adjustment coefficient. For the preset target edge pressure, This refers to the real-time collected pressure value at the edge of the filter rod.

[0015] On the other hand, the present invention also provides a flexible array sensing and counting system adapted to irregularly shaped filter rods, the system comprising: The data acquisition module includes a flexible sensor array unit for real-time acquisition of two-dimensional pressure distribution signals on the surface of the filter rod. The data processing module is used to preprocess the data collected by the data acquisition module; The irregularly shaped filter rod adaptation algorithm module is used to identify the filter rod type and adaptively adjust the counting threshold to accurately count the number of filter rods; An adaptive transmission mechanism is used to adjust the spacing between guide plates in real time to ensure the smooth transport of different irregularly shaped filter rods; A processor, connected to the data acquisition module, data processing module, irregular filter rod adaptation algorithm module, and adaptive transmission mechanism, is configured to perform any of the methods described above.

[0016] Optionally, the flexible sensing array unit adopts a high-density piezoresistive design, consisting of multiple independent piezoresistive sensing units arranged along the filter rod conveying direction and the channel width direction; the array unit uses a flexible polymer film as a substrate, on which a conductive piezoresistive material is coated to form a piezoresistive layer for detecting the filter rod contact pressure; the array unit adopts a partitioned wiring design, dividing the sensing unit into multiple regions, each region being independently connected to the data acquisition module to obtain a two-dimensional pressure matrix; the sampling rate of the array unit is adaptively adjusted according to the transmission speed to capture the pressure peak value when each filter rod passes through.

[0017] In another aspect, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement any of the methods described above.

[0018] The beneficial effects of this invention are: This invention utilizes a flexible sensor array to acquire filter rod pressure signals in real time. This allows for comprehensive adaptation to the complex surface structures of irregularly shaped filter rods, preventing signal loss or distortion due to poor contact and improving the completeness and accuracy of signal acquisition from the source. Furthermore, by constructing a counting model that includes a pressure profile template, filter rod type identification, and dynamic threshold acquisition, it achieves adaptive matching of counting thresholds based on different filter rod cross-sectional characteristics. This overcomes the shortcomings of traditional fixed threshold methods, such as poor adaptability and susceptibility to misjudgments in multi-product mixed-line production. Overall, this invention combines flexible sensing with dynamic modeling, significantly improving the counting accuracy of irregularly shaped filter rods under high-speed transmission conditions and effectively reducing missed and over-counting phenomena.

[0019] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0020] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings: Figure 1 A flowchart of a flexible array sensor counting method for adapting irregularly shaped filter rods according to an embodiment of the present invention; Figure 2 A flowchart illustrating a method for constructing a counting model of irregularly shaped filter rods according to an embodiment of the present invention; Figure 3 A flowchart illustrating a method for constructing a pressure profile template for each filter rod according to an embodiment of the present invention; Figure 4This is a flowchart illustrating a method for identifying the current filter rod type to obtain the current counting threshold according to an embodiment of the present invention. Detailed Implementation

[0021] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0022] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with relevant laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0023] like Figure 1 The diagram shows a flowchart of a flexible array sensing counting method for adapting irregularly shaped filter rods according to an embodiment of the present invention. Figure 1 In this context, the counting method may include the following steps: In step S10, a flexible sensor array is used to collect the original pressure signal of the filter rod in real time; In step S11, a counting model for irregularly shaped filter rods is constructed, including: constructing a pressure profile template for each type of filter rod; identifying the current filter rod type based on the pressure profile template and the original pressure signal to obtain the current counting threshold; and constructing a counting model for irregularly shaped filter rods based on the current counting threshold. In step S12, the filter rods are counted using an irregularly shaped filter rod counting model.

[0024] In such Figure 1 In the flexible array sensing counting method for adapting to irregularly shaped filter rods, step S10 is used to acquire the original pressure signal of the filter rod in real time using a flexible sensor array. The flexible sensor array is the foundation for accurate acquisition of pressure signals from irregularly shaped filter rods, and its structural design directly determines the accuracy and adaptability of signal acquisition. In this embodiment, the array adopts a high-density piezoresistive design, balancing flexibility, wear resistance, and signal acquisition accuracy. Specific parameters and design details are as follows, and all parameters are set based on existing industrial production conditions: The array is designed to consist of M×N independent piezoresistive sensing units, with M=200 units along the filter rod conveying direction and N=50 units along the width of the filter rod transmission channel. The area of ​​a single sensing unit is 1mm×1mm. The high-density design can accurately capture the pressure distribution differences on the surface of irregularly shaped filter rods and avoid signal loss due to excessive unit spacing.

[0025] Substrate selection: Polyimide (PI) film with a thickness of 50μm is used as the array substrate. This material has excellent flexibility, wear resistance and high temperature resistance, can withstand long-term friction during filter rod transmission, has a service life of ≥10,000 hours, and can be slightly bent to adapt to the curvature of the inner wall of the transmission channel.

[0026] Piezoresistive layer design: A conductive piezoresistive material (preferably graphene piezoresistive material) is coated on the surface of a polyimide film. The piezoresistive layer has a thickness of 10μm, high pressure response sensitivity, and a pressure detection range of 50kPa, which can accurately capture minute pressure changes generated by the contact between the filter rod and the array.

[0027] Signal acquisition method: A partitioned wiring design is adopted, dividing the 200×50 sensing units into 10 areas. Each area is independently wired and connected to the data acquisition module to avoid signal interference caused by messy wiring, realize the real-time acquisition of the two-dimensional pressure matrix, and ensure the timeliness of signal transmission.

[0028] Sampling rate parameter: The sampling rate can be adaptively adjusted within a range of 15kHz. In the case of transmission of regular irregularly shaped filter rods, a sampling rate of 3kHz is used. In the case of high-speed transmission (20 rods / second), it can be adjusted to 5kHz to ensure that the pressure peak of each filter rod passing through can be accurately captured and to avoid counting omissions due to insufficient sampling rate.

[0029] Because the raw pressure signals acquired by the flexible sensor array suffer from noise interference and signal offset, they cannot be directly used for filter rod identification and counting. Therefore, preprocessing operations such as normalization and Gaussian filtering are required to improve the accuracy of the pressure profile. In this example, the specific preprocessing method is as follows: The purpose of normalization is to eliminate the inherent errors of different sensing units and the signal shift caused by changes in ambient temperature, normalizing the original pressure signal to the [0,1] interval to facilitate subsequent template matching and feature extraction. The normalization formula is as follows: ; in, For the first line, number The raw pressure signal (unit: kPa) collected by the column sensing unit. The minimum pressure value collected by all sensing units (default 0 kPa, i.e., pressure without contact). This is the maximum detection pressure value of the sensing unit (e.g., 50 kPa). This is the normalized pressure value.

[0030] Gaussian filtering is used to remove random noise from the original signal, such as electromagnetic interference during transmission and pressure fluctuations caused by tiny impurities on the filter rod surface, smoothing the pressure profile and improving the accuracy of subsequent filter rod identification. The filtering formula is as follows: ; in, This is the smoothed pressure matrix after filtering, where * denotes convolution operation. Gaussian filter kernel, standard deviation The preferred value is (Take 1.0 in normal scenarios), the filter kernel size is 3×3, which can effectively filter noise while retaining the core features of the filter rod pressure profile and avoiding profile distortion.

[0031] Preprocessed pressure matrix It can accurately reflect the surface pressure distribution of irregularly shaped filter rods, providing high-quality data support for subsequent irregularly shaped filter rod adaptation algorithms.

[0032] Step S11 is used to construct the irregularly shaped filter rod counting model. In this example, the specific method for constructing the irregularly shaped filter rod counting model in step S11 may include, for example: Figure 2 The steps shown are described. Figure 2 In this context, step S11 may include: In step S20, a pressure profile template for each type of filter rod is constructed; In step S21, the current filter rod type is identified based on the pressure profile template and the original pressure signal to obtain the current counting threshold; In step S22, a counting model for irregularly shaped filter rods is constructed based on the current counting threshold.

[0033] In such Figure 2 In the method shown, step S20 is used to construct a pressure profile template for each type of filter rod. Template construction is the foundation for filter rod type identification. By collecting pressure profile samples of various typical irregularly shaped filter rods, a standard template set is established to ensure the accuracy and universality of identification. In this example, the specific method for constructing the pressure profile template for each type of filter rod can be of various forms known to those skilled in the art. In one example of the present invention, this step may include, for example... Figure 3 The steps shown are described. Figure 3 In this step, constructing the pressure profile template may include: In step S30, sample data is collected and preprocessed; In step S31, the pre-processed samples are averaged to obtain a standard pressure profile template for each filter rod. In step S32, core features are extracted from each standard pressure profile template as the basis for filter rod identification; the core features include edge pressure peak, center pressure peak, length-to-short axis ratio and profile similarity index.

[0034] In such Figure 3In the method shown, step S30 is used to collect sample data and perform preprocessing. Specifically, for three typical irregularly shaped filter rods—circular, elliptical, and triangular (covering existing mainstream irregularly shaped filter rod types)—≥20 sets of pressure profile samples are collected for each type. Each set of samples is collected at different conveying speeds (20 rods / second) and different contact pressures (5-20 kPa) to ensure the comprehensiveness of the samples. Next, sample preprocessing is performed, including normalization and Gaussian filtering of the collected raw samples to remove abnormal samples (such as samples with profile distortion caused by filter rod tilting or material jamming) to ensure the accuracy of the template.

[0035] Step S31 is used to average the pre-processed samples to obtain a standard pressure profile template for each filter rod. Specifically, the processed valid samples are averaged to obtain a standard pressure profile template for each filter rod, forming a template set, as shown in the following expression: ; in, For the first Standard pressure profile template for filter rods, ~ This is the effective sample pressure matrix for this type of filter rod. For the number of valid samples, This represents the total number of filter rod types. In this example, Corresponding to circles, Corresponding to ellipse, Corresponding triangle; number of valid samples ; It can be expanded according to actual needs.

[0036] Step S32 extracts core features from each standard pressure profile template as the basis for filter rod identification. In this example, four core features are extracted from each standard template to ensure the accuracy of filter rod type identification. The specific features are as follows: Edge pressure peak The maximum pressure value generated when the filter rod edge comes into contact with the sensor array reflects the edge profile of the filter rod cross-section; Peak central pressure The maximum pressure value generated when the central region of the filter rod comes into contact with the sensor array reflects the central profile of the filter rod cross-section; Major-minor axis ratio The major-minor axis ratio of the filter rod cross-section is obtained by fitting the pressure profile. The major-minor axis ratio of the circular filter rod is 1, the major-minor axis ratio of the elliptical filter rod is 1.5, and the major-minor axis ratio of the triangular filter rod is 1.8. Contour similarity index: The overall distribution characteristics of the template pressure contour, used for subsequent matching with the pressure contour acquired in real time.

[0037] Step S21 is used to identify the current filter rod type based on the pressure profile template and the original pressure signal, in order to obtain the current counting threshold. In this embodiment, the specific method for identifying the current filter rod type and obtaining the current counting threshold in step S21 can be of various forms known to those skilled in the art. In one example of the present invention, step S21 may include, for example... Figure 4 The steps shown are described. Figure 4 In this context, step S21 may include: In step S40, the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template is calculated. In step S41, the filter rod type is determined based on the matching degree; In step S42, the counting threshold is adaptively adjusted based on the filter rod type.

[0038] In such Figure 4 In the method shown, step S40 is used to calculate the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template. Specifically, in this example, the correlation coefficient method is used to calculate the matching degree between the real-time pressure profile and each standard template. The larger the correlation coefficient, the higher the matching degree between the two. The calculation formula is as follows: (1) in, For real-time pressure profile and the first The degree of matching of the template, For the real-time acquisition and preprocessing of the pressure matrix, This represents the average pressure value of the real-time pressure matrix. For the first Pressure matrix of the template For the first The average pressure value of the template.

[0039] Step S41 is used to determine the filter type based on the matching degree. In this example, the filter type corresponding to the template with the highest matching degree is selected as the type of filter that passes through in real time, and the expression is as follows: ; in, To determine the type of filter rod, Indicates matching degree The maximum time corresponding Value, when the maximum matching degree When the filter bar is identified as an unknown type, the default threshold is used for counting.

[0040] Step S42 is used for adaptive adjustment of the counting threshold based on the filter rod type. Since the pressure distribution of different types of filter rods varies significantly, a fixed threshold will lead to counting errors; therefore, the threshold is adjusted based on the maximum matching degree. The counting threshold is adaptively adjusted to ensure counting accuracy. The adjustment formula is as follows: (2) in, The adaptively adjusted counting threshold (unit: V). The default counting threshold is used. To adjust the coefficients, the higher the matching degree, the smaller the threshold adjustment range, ensuring the stability of the count. For maximum matching degree, when hour, This means using the default threshold. In this example, the default counting threshold is... Fixed at 0.2V, based on the conventional filter rod pressure signal setting; adjustment coefficient The range of values ​​is (0.4 is preferred).

[0041] By calculating the matching degree between the real-time collected pressure profile and the standard template, the filter rod cross-section type is determined, and the counting threshold is adaptively adjusted according to the matching degree to avoid counting errors caused by different filter rod types.

[0042] Step S22 is used to construct a counting model for irregularly shaped filter rods based on the current counting threshold. Based on the adjusted counting threshold... By combining the peak value changes of the real-time pressure profile, a peak value accumulation counting model is established to achieve accurate counting of irregularly shaped filter rods and avoid omissions and miscounts. Specifically, the counting model expression is as follows: (3) in, The cumulative count of filter rods. Total number of sampling times (unit: ms). This is an indicator function; when the condition inside the parentheses is true, ,otherwise ; for The maximum pressure value of the real-time pressure matrix corresponds to the peak pressure when the filter rod passes through.

[0043] By establishing pressure profile templates, optimizing recognition algorithms, and adaptively adjusting thresholds, the system achieves accurate identification and counting of various irregularly shaped filter rods, addressing the pain point that existing technologies cannot adapt to irregularly shaped filter rods.

[0044] In step S12, the irregularly shaped filter rod counting model is used to count the filter rods. The counting logic is as follows: when the filter rod enters the transmission channel and comes into contact with the flexible sensing array, the real-time pressure peak value is... It will gradually rise, when Exceeding the adjusted counting threshold At that time, indicator function The count increments by 1; when the filter rod has completely passed through the array and the pressure peak drops below the threshold, the indicator function... The counting stops accumulating, thus achieving accurate counting of each filter rod and effectively avoiding problems such as missed counting (e.g., peak values ​​not being captured) and miscounting (e.g., noise interference).

[0045] In one example of this invention, the flexible array sensor counting method adapted to irregularly shaped filter rods further includes adaptive adjustment of the transmission mechanism. The adaptive transmission mechanism is key to achieving stable transport of irregularly shaped filter rods. It can be linked with the flexible sensor array and algorithm module to form a closed-loop adjustment, ensuring stable passage of filter rods with different cross-sections and sizes, avoiding counting errors caused by jamming, shaking, or tilting, and improving counting accuracy. The transmission mechanism structure includes a guide plate and a pneumatic adjustment unit. The flexible guide plate adopts a deformable design, and its spacing can be adjusted in real time according to the filter rod width and pressure signal to achieve closed-loop control. The specific design and adjustment logic are as follows: Guide plate spacing The adjustment range covers the width range of existing mainstream irregular-shaped filter rods, ensuring that filter rods of all specifications can pass through smoothly. The adjustment range is: ,in, The minimum width of the irregularly shaped filter rod is preferred (3mm). The maximum width of the irregularly shaped filter rod is preferably 8mm. This is a buffer gap (which can be 0.5mm). Therefore, in this example, the guide plate spacing... The adjustment range is The buffer gap setting can prevent the filter rod from deforming due to excessive compression between the guide plate and the filter rod, while ensuring that the filter rod and the flexible sensing array are in full contact.

[0046] The guide plate spacing is adjusted in a closed loop based on real-time pressure signal feedback to ensure that the spacing is always at the optimal state. The adjustment formula is as follows: (4) in, The adjusted guide plate spacing, This is the current guide plate spacing. This is the proportional adjustment coefficient. For the preset target edge pressure, This refers to the real-time collected pressure value at the edge of the filter rod. In this example, the proportional adjustment coefficient... The range of values ​​is 0.2 is preferred; the preset target edge pressure The pressure is 10 kPa to ensure stable contact of the filter rods and prevent them from being squeezed.

[0047] when If this occurs, it indicates insufficient contact between the filter rods and the array, and the spacing between the guide plates needs to be reduced. );when If this occurs, it indicates that the filter rod is being excessively compressed, and the spacing between the guide plates needs to be increased. );when At the same time, the current spacing remains unchanged, achieving real-time adaptive adjustment of the spacing.

[0048] The guide plate spacing is adjusted by a pneumatic adjustment unit, using readily available and mature industrial pneumatic components to ensure rapid and stable adjustment. In this example, the driving component uses a cylinder or pneumatic actuator (preferably an SC63×50 cylinder) with a rated working pressure of 0.4~0.6MPa and a response time ≤50ms, enabling rapid response to spacing adjustment commands. It can accommodate filter rod conveying speeds of 10~20 rods / second, meeting the needs of high-speed production lines. The adjustment process is smooth and shock-free, avoiding filter rod swaying caused by excessive adjustment speed. Control is achieved by an MCU / FPGA outputting control signals, which control the extension and retraction of the pneumatic actuator via a solenoid valve, realizing precise adjustment of the guide plate spacing and forming a closed-loop linkage with the algorithm module and flexible sensor array.

[0049] On the other hand, the present invention also provides a flexible array sensor counting system adapted to irregularly shaped filter rods. The system includes a data acquisition module, a data processing module, an irregularly shaped filter rod adaptation algorithm module, an adaptive transmission mechanism, and a processor. Each module is linked together through circuits, signal transmission lines, and control programs, and is integrated as a whole into the conveying channel of the filter rod conveying production line.

[0050] The data acquisition module includes a flexible sensor array unit for real-time acquisition of two-dimensional pressure distribution signals on the filter rod surface. In this example, the flexible sensor array unit is fixed to the core area of ​​the inner wall of the filter rod conveying channel. It is a piezoresistive flexible array structure and is electrically connected to the data acquisition and processing module through a dedicated signal transmission line. Its function is to convert the mechanical pressure signals generated by the filter rod contacting the array during transmission into electrical signals, acquire and output two-dimensional pressure matrix signals in real time, and provide raw data for subsequent counting. This module uses a polyimide film as the substrate and graphene piezoresistive material as the piezoresistive layer. It consists of 200×50 1mm×1mm piezoresistive sensor units. Interference-free signal transmission is achieved through partitioned wiring. The flexible material allows it to adapt to the surface contours of filter rods with different cross-sections, ensuring the integrity of pressure signal acquisition.

[0051] The data processing module is used to preprocess the data acquired by the data acquisition module. In this example, the module is installed in the control box next to the conveying channel. The input end is electrically connected to the flexible sensor array module, and the output end is signal connected to the irregular filter rod adaptation algorithm module. It has a built-in high-precision ADC with a resolution of ≥12 bits and a sampling error of ≤±0.5%, an STM32F4 series MCU, or a CycloneIV series FPGA. Its core function is to receive the raw pressure electrical signal transmitted by the flexible sensor array module, and complete the preprocessing operations such as signal normalization, Gaussian filtering, and pressure profile extraction through hardware circuits and built-in programs. It filters noise interference, corrects signal deviation, and outputs a smooth pressure profile matrix signal, providing high-quality processed data for the irregular filter rod adaptation algorithm module.

[0052] The irregularly shaped filter rod adaptation algorithm module is used to identify the filter rod type and adaptively adjust the counting threshold for accurate filter rod counting. In this example, this module is integrated into the MCU / FPGA of the data acquisition and processing module. It is a software algorithm module with bidirectional signal connection to the data acquisition and processing module and the adaptive transmission mechanism. It internally stores a set of standard pressure profile templates for irregularly shaped filter rods such as circles, ellipses, and triangles, as well as the core algorithm program. It can receive the smooth pressure profile matrix transmitted by the data acquisition and processing module, identify the filter rod cross-section type through template matching and correlation coefficient calculation, and then adaptively adjust the counting threshold according to the identification result. Finally, the filter rod is counted by peak accumulation. At the same time, the pressure signal at the edge of the filter rod is transmitted to the adaptive transmission mechanism to provide a basis for transmission adjustment.

[0053] The adaptive transmission mechanism is used to adjust the guide plate spacing in real time to ensure smooth transport of different irregularly shaped filter rods. In this example, the module is set at the entrance and core transmission section of the filter rod transport channel, and is signal-connected to the irregularly shaped filter rod adaptation algorithm module. It consists of flexible guide plates (made of silicone, 2~3mm thick, capable of slight deformation), pneumatic actuators (cylinder model preferably SC63×50), and a closed-loop control device linked with the flexible sensor array. The flexible guide plates are symmetrically arranged on both sides of the transport channel and fixedly connected to the telescopic ends of the pneumatic actuators. The pneumatic actuators are electrically connected to the closed-loop control device. The closed-loop control device receives the edge pressure signal transmitted by the irregularly shaped filter rod adaptation algorithm module, and controls the telescopic extension of the pneumatic actuators through a preset program to realize the real-time adjustment of the guide plate spacing, ensuring smooth transport of filter rods, avoiding jamming, shaking or tilting, and ensuring the stability of pressure signal acquisition.

[0054] The processor is connected to a data acquisition module, a data processing module, an irregularly shaped filter rod adaptation algorithm module, and an adaptive transmission mechanism. The processor is configured to perform any of the methods described in the flexible array sensor counting method for adapting irregularly shaped filter rods.

[0055] The above modules form a complete closed loop of physical connection and signal transmission. The flexible sensor array module is the data input end, the data acquisition and processing module is the data processing center, the irregular filter rod adaptation algorithm module is the core control center, and the adaptive transmission mechanism is the execution and adjustment unit. Each module performs its own function and works together to realize the full-process automation from pressure signal acquisition to filter rod counting and transmission adjustment.

[0056] In one example of the present invention, the operation flow of the flexible array sensing and counting system adapted to irregularly shaped filter rods is as follows: Step 100: System Initialization: Fit the flexible sensor array tightly against the inner wall of the filter rod transmission channel, ensuring the array is wrinkle-free and without offset; power on the system, load the preset set of three typical irregularly shaped filter rod pressure profile templates, and set the default counting threshold. =0.2V, initial spacing of guide plates =6mm, complete system parameter initialization, and ensure normal linkage of all modules; Step 200: Filter rod delivery: Start the adaptive transmission mechanism. The filter rods enter the transmission channel at a preset speed (10~20 rods / second) and move forward smoothly under the guidance of the guide plate to ensure that the filter rods are in full contact with the flexible sensor array.

[0057] Step 300: Filter rod identification: The flexible sensor array collects the filter rod pressure signal in real time and transmits it to the data acquisition and processing module. Normalization and Gaussian filtering preprocessing are performed to obtain a smooth pressure matrix. The algorithm module calls the template matching algorithm to calculate the matching degree between the real-time pressure matrix and each standard template to determine the filter rod cross-section type.

[0058] Step 400: Threshold Adjustment: The algorithm module adjusts the threshold based on the maximum matching degree. Update the counting threshold according to the adaptive threshold adjustment formula. To ensure counting accuracy.

[0059] Step 500: Filter rod counting: The system monitors the pressure peak at every moment in real time. ,when Exceeding the adjusted threshold When the count is incremented by 1, the filter rods are counted in real time.

[0060] Step 600: Transmission Adjustment: The adaptive transmission mechanism adjusts according to the real-time collected filter rod edge pressure. The guide plate spacing is adjusted by a closed-loop regulation formula to ensure that the filter rods are always conveyed smoothly and to avoid jamming and shaking.

[0061] Step 700: Data Recording: The system automatically records the core information of each filter rod, including filter rod type, counting time, transmission speed, guide plate spacing, etc., and stores it to the local storage module (such as SD card) or the cloud for subsequent production statistics, system optimization and troubleshooting.

[0062] The entire process requires no manual intervention, achieving fully automated operation and continuous operation, thus meeting the high-efficiency requirements of cigarette production lines.

[0063] In another aspect, the present invention also provides a computer-readable storage medium storing instructions that, when executed by a processor, implement any of the methods described in the flexible array sensor counting method adapted to irregularly shaped filter rods.

[0064] The beneficial effects of this invention are: This invention, through the collaborative design of a flexible sensing array, an irregularly shaped filter rod adaptation algorithm, and an adaptive transmission mechanism, overcomes the limitations of existing circular filter rod counting technology and has the following significant technical effects and advantages: (1) High counting accuracy: The high-density flexible sensor array is adopted, combined with Gaussian filtering preprocessing and adaptive threshold adjustment, which effectively avoids counting errors caused by noise interference and differences in filter rod type. The counting accuracy can reach more than 99.8%, which completely solves the problem of missing or miscounting of irregularly shaped filter rods, far exceeding the existing technology.

[0065] (2) High versatility: Through pressure profile template matching and adaptive adjustment, it can be compatible with various irregularly shaped filter rods such as round, elliptical, and triangular, as well as filter rods of different widths, without the need to change mechanical molds, thus solving the pain point that existing technologies can only adapt to a single round filter rod.

[0066] (3) High operational stability: The adaptive guide plate is linked with the closed-loop pressure feedback control, which can adjust the transmission distance in real time, effectively avoiding filter rod jamming, shaking or tilting. The filter rod conveying jamming rate and the continuous operation failure rate of the system are reduced to below 0.1%, the maintenance cost is low, and it is suitable for the long-term continuous production needs of industry.

[0067] (4) Strong industrial feasibility: The polyimide film, graphene piezoresistive material, high-precision ADC, cylinder and other components used are all commonly used products in the existing industrial field, which are easy to purchase and cost controllable; the algorithm model can be implemented by conventional MCU / FPGA programming, without special equipment and processes, and can be directly industrialized and promoted for application.

[0068] (5) Significant economic benefits: No parts need to be replaced, saving enterprises more than 80% of parts costs and maintenance costs annually; improved counting accuracy and optimized conveying stability reduce product rework and scrap; combined with high-speed adaptability, production line efficiency can be increased by 15%~20%, significantly improving enterprise production efficiency.

[0069] (6) Outstanding industrial application value: It fills the technical gap of high-precision counting of irregularly shaped filter rods, can be directly applied to the cigarette production industry, quickly realize industrial production and promotion, promote the upgrading and iteration of intelligent and automated counting technology of cigarettes, and has significant industrial application value and promotion prospects.

[0070] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0072] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0073] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0074] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0075] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0076] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0077] It should also be noted that 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 process, method, article, or apparatus. Unless otherwise specified, 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 that element.

[0078] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A flexible array sensor counting method adapted to irregularly shaped filter rods, characterized in that, The counting method includes: A flexible sensor array is used to collect the original pressure signal of the filter rod in real time; Constructing a counting model for irregularly shaped filter rods includes: constructing a pressure profile template for each type of filter rod; identifying the current filter rod type based on the pressure profile template and the original pressure signal to obtain a current counting threshold; and constructing a counting model for irregularly shaped filter rods based on the current counting threshold. The filter rods are counted using the aforementioned irregularly shaped filter rod counting model.

2. The counting method according to claim 1, characterized in that, The pressure profile template for each filter rod includes: Collect sample data and preprocess it; The pre-processed samples were averaged to obtain a standard pressure profile template for each filter rod. Core features are extracted from each standard pressure profile template and used as the basis for filter rod identification; the core features include edge pressure peak, center pressure peak, length-to-short axis ratio, and profile similarity index.

3. The counting method according to claim 1, characterized in that, Based on the pressure profile template and the original pressure signal, identifying the current filter rod type and obtaining the current counting threshold includes: Calculate the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template; The filter rod type is determined based on the matching degree. Adaptive adjustment of the counting threshold based on filter rod type.

4. The counting method according to claim 3, characterized in that, The calculation of the matching degree between the real-time pressure profile of the current filter rod and each pressure profile template includes: The matching degree is calculated according to formula (1). ，（1） in, For real-time pressure profile and the first The degree of matching of the template, For the real-time acquisition and preprocessing of the pressure matrix, This represents the average pressure value of the real-time pressure matrix. For the first Pressure matrix of the template For the first The average pressure value of the template.

5. The counting method according to claim 3, characterized in that, Adaptive adjustment of the counting threshold based on filter rod type includes: The adjusted counting threshold is obtained according to formula (2). ，（2） in, The adaptively adjusted counting threshold, The default counting threshold is used. To adjust the coefficient, This represents the maximum matching degree.

6. The counting method according to claim 1, characterized in that, Constructing a counting model for irregularly shaped filter rods based on the current counting threshold includes: The counting model is obtained according to formula (3). ，（3） in, The cumulative count of filter rods. This represents the total number of sampling times. For indicator functions, for The maximum pressure value of the real-time pressure matrix. This is the adaptively adjusted counting threshold.

7. The counting method according to claim 1, characterized in that, The method further includes adaptive adjustment of the transmission mechanism, including: Real-time acquisition of filter rod edge pressure; Adjusting the guide plate spacing based on the filter rod edge pressure includes: obtaining the adjusted guide plate spacing according to formula (4). ，（4） in, The adjusted guide plate spacing, This is the current guide plate spacing. This is the proportional adjustment coefficient. For the preset target edge pressure, This refers to the real-time collected pressure value at the edge of the filter rod.

8. A flexible array sensing and counting system adapted to irregularly shaped filter rods, characterized in that, The system includes: The data acquisition module includes a flexible sensor array unit for real-time acquisition of two-dimensional pressure distribution signals on the surface of the filter rod. The data processing module is used to preprocess the data collected by the data acquisition module; The irregularly shaped filter rod adaptation algorithm module is used to identify the filter rod type and adaptively adjust the counting threshold to accurately count the number of filter rods; An adaptive transmission mechanism is used to adjust the spacing between guide plates in real time to ensure the smooth transport of different irregularly shaped filter rods; A processor connected to the data acquisition module, the data processing module, the irregular filter rod adaptation algorithm module, and the adaptive transmission mechanism, the processor being configured to perform the method as described in any one of claims 1 to 7.

9. The counting system according to claim 8, characterized in that, The flexible sensor array unit adopts a high-density piezoresistive design, consisting of multiple independent piezoresistive sensor units arranged along the filter rod conveying direction and the channel width direction. The array unit uses a flexible polymer film as a substrate, on which a conductive piezoresistive material is coated to form a piezoresistive layer for detecting the contact pressure of the filter rod. The array unit adopts a partitioned wiring design, dividing the sensor unit into multiple regions, each region being independently connected to the data acquisition module to obtain a two-dimensional pressure matrix. The sampling rate of the array unit is adaptively adjusted according to the transmission speed to capture the pressure peak value when each filter rod passes through.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores instructions that, when executed by a processor, implement the method as described in any one of claims 1 to 7.

Images