Self-test control method for working quality of cotton cleaner based on multi-device cooperation

Abstract

The application provides a cotton cleaning machine working quality self-test control method based on multi-device cooperation, relates to the technical field of quality detection, and comprises the following steps: sequentially arranging a laser sensor, a CCD sensor and a photoelectric sensor; configuring an initial control parameter set, performing raw cotton motion control, and recording the raw cotton motion speed; collecting the laser sensor acquisition result, the CCD sensor acquisition result and the photoelectric sensor acquisition result, performing quality detection, and generating a quality detection result; generating a preset quality standard; when the quality detection result does not satisfy the preset quality standard, performing optimization adjustment, generating an optimized control parameter set and an optimized raw cotton motion speed; and adjusting and controlling the cotton cleaning machine. The application solves the technical problems that the prior art cannot comprehensively cover the key quality indexes in the raw cotton processing process, lacks real-time monitoring and regulation and control of the raw cotton quality, and results in poor production efficiency and product quality stability.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of quality detection, in particular to a cotton cleaning machine working quality self-test control method based on multi-device cooperation. BACKGROUND

[0002] The cotton cleaning machine is a key raw cotton processing device in the textile industry, and its main function is to clean, separate impurities and screen fibers of raw cotton to ensure the quality of the final product. With the improvement of industrial automation level, the traditional cotton cleaning machine faces some challenges in quality control. On the one hand, the traditional cotton cleaning machine usually uses a single sensor or a small number of sensors for monitoring, and the monitoring devices are monitored separately, lacking cooperation and being unable to comprehensively cover the key quality indicators in the raw cotton processing process. On the other hand, the quality control of the traditional cotton cleaning machine in the process of processing raw cotton mainly relies on post-manual inspection or offline detection, which cannot respond and adjust the quality problems in the production process in real time, resulting in large quality fluctuations and affecting the production efficiency and stability of product quality. SUMMARY

[0003] The application provides a cotton cleaning machine working quality self-test control method based on multi-device cooperation, aiming to solve the technical problems that the existing technology usually uses a single sensor or a small number of sensors for separate monitoring, cannot comprehensively cover the key quality indicators in the raw cotton processing process, and lacks real-time monitoring and regulation of the quality of raw cotton, resulting in poor production efficiency and stability of product quality.

[0004] The cotton cleaning machine working quality self-test control method based on multi-device cooperation disclosed in the application is applied to a cotton cleaning machine, the cotton cleaning machine comprises a feeding mechanism, an opening mechanism, an impurity separation mechanism and a fiber sorting mechanism, and the method comprises the following steps: sequentially arranging a laser sensor at the outlet of the opening mechanism, arranging a CCD sensor at the outlet of the impurity separation mechanism, and arranging a photoelectric sensor at the outlet of the fiber sorting mechanism; configuring an initial control parameter set of the opening mechanism, the impurity separation mechanism and the fiber sorting mechanism, and performing raw cotton motion control of the feeding mechanism to record the raw cotton motion speed; collecting laser sensor collection results, CCD sensor collection results and photoelectric sensor collection results through a data acquisition system, performing quality detection, and generating quality detection results; interacting with a production task, generating a preset quality standard based on the production task; when the quality detection results do not meet the preset quality standard, optimizing and adjusting the initial control parameter set and the raw cotton motion speed to generate an optimized control parameter set and an optimized raw cotton motion speed; and based on the optimized control parameter set and the optimized raw cotton motion speed, adjusting and controlling the cotton cleaning machine.

[0005] One or more technical solutions provided in the application have at least the following technical effects or advantages:

[0006] By arranging laser sensors, CCD sensors and photoelectric sensors at the outlets of the opening mechanism, impurity separation mechanism and fiber sorting mechanism, real-time monitoring of each key link in the raw cotton processing is realized. The data acquisition system collects and integrates the collection results of multiple sensors in real time, realizes comprehensive monitoring of the quality indicators of raw cotton, including fiber density, impurity content and short fiber ratio, etc. Based on the real-time data acquisition system, quality detection can be quickly performed and quality detection results can be generated. Combined with the preset quality standards, when the quality detection results do not meet the preset standards, the initial control parameter set and the raw cotton movement speed can be automatically triggered to optimize the adjustment, and the optimized control parameter set and the raw cotton movement speed are generated to ensure that the quality is within the standard range. Through the optimization of the control parameter set and the raw cotton movement speed, precise adjustment and control of the scutcher are realized, and the efficiency and stability of the product quality in the production process are improved. In summary, the method effectively solves the limitations of the traditional scutcher in quality control by integrating multi-sensor data, real-time monitoring and adjustment, and significantly improves the quality management level and operation efficiency in the production process.

[0007] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented in accordance with the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0008] Figure 1 A scutcher working quality self-test control method flowchart based on multi-device cooperation is provided for the embodiments of the present application.

[0009] Figure 2 An execution quality detection flowchart in a scutcher working quality self-test control method based on multi-device cooperation is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0010] The embodiments of the present application provide a scutcher working quality self-test control method based on multi-device cooperation, which solves the technical problem that the prior art usually uses a single sensor or a small number of sensors for separate monitoring of the scutcher, which cannot comprehensively cover the key quality indicators in the raw cotton processing process, and lacks real-time monitoring and regulation of the quality of raw cotton, resulting in poor production efficiency and product quality stability.

[0011] After introducing the basic principles of the present application, the various non-limiting embodiments of the present application will be specifically introduced in combination with the drawings of the specification.

[0012] As Figure 1As shown, the embodiment of the present application provides a cotton cleaning machine working quality self-test control method based on multi-device cooperation, which is applied to a cotton cleaning machine. The cotton cleaning machine includes a feeding mechanism, an opening mechanism, an impurity separation mechanism, and a fiber sorting mechanism. The method includes:

[0013] A laser sensor is arranged at the outlet of the opening mechanism, a CCD sensor is arranged at the outlet of the impurity separation mechanism, and a photoelectric sensor is arranged at the outlet of the fiber sorting mechanism.

[0014] The cotton cleaning machine working quality self-test control method based on multi-device cooperation is applied to a cotton cleaning machine. The cotton cleaning machine includes a feeding mechanism, an opening mechanism, an impurity separation mechanism, and a fiber sorting mechanism. The feeding mechanism is a device for conveying raw cotton to the opening mechanism. The opening mechanism is a mechanical device for preliminary opening and carding of raw cotton. Through mechanical action, the fibers in the raw cotton are separated, part of the impurities are removed, and the arrangement state of the fibers is preliminarily arranged. The impurity separation mechanism is a device for separating impurities from raw cotton. Through physical methods such as airflow and vibration, impurities such as sand and seed shells in raw cotton are separated out, improving the purity of raw cotton. The fiber sorting mechanism is a device for further separating and arranging fibers. According to the length, density, and other characteristics of the fibers, the fibers are sorted and arranged to ensure the uniformity and quality of the fibers.

[0015] A laser sensor is installed at the outlet of the opening mechanism to accurately scan the fiber bundle passing through the outlet. The laser sensor measures the density change of the fiber bundle by emitting and receiving a laser beam. A CCD (Charge-Coupled Device) sensor is installed at the outlet of the impurity separation mechanism. The CCD sensor detects the remaining impurities in the raw cotton by taking pictures and analyzing the images. It can identify impurities of different sizes and shapes and generate corresponding data. A photoelectric sensor is installed at the outlet of the fiber sorting mechanism. The photoelectric sensor measures and analyzes the length distribution of the fibers by detecting different lengths of fibers in the passing fiber bundle using the principle of photoelectric conversion. By arranging the above sensors, real-time monitoring and self-test control of the working quality of the cotton cleaning machine can be realized to ensure the quality stability of the fiber processing process.

[0016] The initial control parameter set of the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism is configured, and the raw cotton motion control of the feeding mechanism is performed. The raw cotton motion speed is recorded.

[0017] The initial control parameters are configured for the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism to ensure their normal operation. The opening mechanism configuration parameters include the speed of the opening roller, the gap of the opening clothing, the beating frequency, etc. The impurity separation mechanism configuration parameters include the speed of the separation roller, the air flow intensity, the aperture of the separation screen, etc. The fiber sorting mechanism configuration parameters include the speed of the sorting roller, the width of the sorting channel, the suction intensity, etc.

[0018] The raw cotton motion control of the feeding mechanism is performed to control the motion speed of the raw cotton in the feeding mechanism, ensuring that the raw cotton can enter the opening mechanism stably and uniformly, and the motion speed of the raw cotton is recorded in real time by using the speed sensor.

[0019] The laser sensor acquisition results, the CCD sensor acquisition results, and the photoelectric sensor acquisition results are collected by the data acquisition system, and the quality detection is performed to generate the quality detection results.

[0020] The laser sensor scans the fiber bundle passing through the outlet of the opening mechanism in real time, records the density change, and transmits the data to the data acquisition system. The CCD sensor takes an image of the raw cotton passing through the outlet of the impurity separation mechanism, identifies and records the impurity content and type by using the image processing technology, and transmits the data to the data acquisition system. The photoelectric sensor detects the fiber length passing through the outlet of the fiber sorting mechanism, records the short fiber proportion and other data, and transmits the data to the data acquisition system. Based on the data acquisition results, the quality detection is performed, including calculating the fiber uniformity coefficient by analyzing the fiber density distribution, calculating the impurity content proportion coefficient by analyzing the impurity content, and calculating the short fiber proportion coefficient by analyzing the fiber length distribution. The analysis results are integrated to generate the overall quality detection results, and the working quality of the scutcher is evaluated.

[0021] The production tasks are interacted, the preset quality standards are generated based on the production tasks, and when the quality detection results do not meet the preset quality standards, the initial control parameter set and the raw cotton motion speed are optimized and adjusted to generate an optimized control parameter set and an optimized raw cotton motion speed.

[0022] The specific requirements of the current production task are read from the production planning system or the management system, including the fiber quality requirements, etc. The target values of various quality indicators are determined according to the specific requirements of the fiber quality in the production task, such as fiber uniformity, impurity content, and fiber length, and the preset quality standards are generated.

[0023] The fiber uniformity coefficient, impurity content proportion coefficient, and short fiber proportion coefficient in the quality detection result are compared with the preset quality standard. If the detection result is lower than the preset standard, optimization adjustment is needed. Through optimization adjustment, the working quality of the scutcher is improved to meet the preset quality standard. Through the iterative optimization process, an optimized control parameter set is obtained, including the optimized control parameters of the opening mechanism, impurity separation mechanism, and fiber sorting mechanism, and the optimized raw cotton movement speed.

[0024] The scutcher is adjusted and controlled based on the optimized control parameter set and the optimized raw cotton movement speed.

[0025] The optimized parameters and speed are applied to the scutcher for adjustment and control of each mechanism. Specifically, the speed of the opening roller, the gap of the opening clothing, and the vibration frequency of the opening mechanism are adjusted according to the optimized control parameters. The speed of the separation roller, the airflow intensity, and the aperture of the separation screen of the impurity separation mechanism are adjusted according to the optimized control parameters. The speed of the sorting roller, the width of the sorting channel, and the air suction intensity of the fiber sorting mechanism are adjusted according to the optimized control parameters. The speed of the feeding roller is adjusted according to the optimized raw cotton movement speed to ensure that the raw cotton is fed at the optimized speed. Through parameter adjustment, the scutcher reaches the best working state.

[0026] Further, as shown in Figure 2 The method comprises:

[0027] According to the laser sensor acquisition result, fiber density distribution analysis is performed to generate a fiber uniformity coefficient. According to the CCD sensor acquisition result, impurity detection is performed to generate an impurity content proportion coefficient. According to the photoelectric sensor acquisition result, fiber length distribution analysis is performed to generate a short fiber proportion coefficient. The fiber uniformity coefficient, the impurity content proportion coefficient, and the short fiber proportion coefficient are integrated to generate the quality detection result.

[0028] The laser sensor scans the fibers at the outlet of the opening mechanism, records the density data of the fiber bundle, performs data preprocessing such as removing outliers, and performs data smoothing processing to reduce the influence of noise. Based on the preprocessed data, the fiber density values of each scanning point are calculated to generate a density distribution curve. According to the density distribution curve, the fiber uniformity coefficient is calculated. For example, a statistical method such as the coefficient of variation is used to represent the uniformity of the fibers. The coefficient of variation is defined as the ratio of the standard deviation to the mean. The obtained fiber uniformity coefficient is used to measure the uniformity of the fiber density distribution. The smaller the fiber uniformity coefficient, the more uniform the fiber distribution.

[0029] The CCD sensor takes an image of the raw cotton at the outlet of the impurity separation mechanism, performs grayscale and binaryzation processing on the image to enhance the contrast of the impurities, uses image processing algorithms such as edge detection and morphological processing to identify the impurity regions in the image, extracts the area, shape, and position of the impurities, calculates the total area of the impurity regions in the image, and calculates the impurity content ratio by dividing the total area of the impurities by the total area of the raw cotton in the image. The smaller the impurity content ratio coefficient, the less impurities there are in the raw cotton, and the higher the quality.

[0030] The photoelectric sensor measures the fibers at the outlet of the fiber sorting mechanism, records the length of each fiber, counts the length of each fiber, generates a length distribution graph, evaluates the length distribution of the fibers, and generates a short fiber ratio coefficient. Specifically, a threshold length of short fibers is set, such as fibers less than 25 mm, which are defined as short fibers, and the proportion of short fibers in the total number of fibers is calculated to obtain the short fiber ratio coefficient.

[0031] The obtained fiber uniformity coefficient, impurity content ratio coefficient, and short fiber ratio coefficient are integrated, such as setting different weight values according to the importance of each coefficient, performing weighted comprehensive calculation, and generating a comprehensive score, i.e. the quality detection result, which is used to represent the working quality of the scutcher. The setting of the weight value is determined according to production requirements and experience.

[0032] Further, the impurity detection based on the CCD sensor acquisition result and the generation of the impurity content ratio coefficient include:

[0033] According to the laser sensor acquisition result, the emission light signal node and the reflected light signal node of the laser sensor are obtained, light signal analysis is performed, the impurity size set and the impurity coordinate set are determined, and the impurity size set and the impurity coordinate set have a mapping relationship; based on the impurity size set and the impurity coordinate set, an initial pre-aiming frame is set; an impurity feature recognition library of raw cotton is established; the initial pre-aiming frame is used to frame the CCD sensor acquisition result, the impurity feature recognition library is called based on the framing result, impurity feature recognition is performed, and the impurity content ratio coefficient is generated according to the impurity feature recognition result.

[0034] The laser sensor emits a laser beam when the raw cotton passes, records the emission light signal node, records the light signal node after the laser beam is reflected by the impurities, and generates reflected light signal data. The emission light signal node and the reflected light signal node are matched, and the specific position of the impurities is calculated through the signal intensity and the time difference. Specifically, according to the diffusion angle of the laser beam and the signal time difference, the size of the impurities is calculated; according to the installation position of the laser sensor and the position of the impurities on the conveying belt, the coordinates of the impurities are calculated. The size and coordinate data of all impurities are summarized to form an impurity size set and an impurity coordinate set, wherein there is a one-to-one mapping relationship between the impurity size set and the impurity coordinate set.

[0035] According to the size and coordinates of the impurities, a pre-aiming frame is set in the image to cover the entire area of the impurities. According to the size and shape of different impurities, the size and position of the pre-aiming frame are adjusted to ensure that the pre-aiming frame can effectively cover the impurities.

[0036] Impurity samples are extracted from different batches of raw cotton, including grass seeds, leaves, soil, insect pests, etc. High-definition images of each impurity sample are taken, and feature information of the impurity samples is extracted, including color features, shape features, texture features, size features, etc. An impurity feature recognition library is established.

[0037] The image collected by the CCD sensor is aligned with the initial pre-aiming frame, the area in the frame is cropped, and the image part of the impurities of interest is extracted. The impurity image part selected by the frame is subjected to feature extraction, including color, shape, texture, size, etc. The extracted features are matched with the features in the impurity feature recognition library, the similarity is calculated, and the type and quantity of the impurities are determined according to the matching result. According to the impurity feature recognition result, an impurity content proportion coefficient is generated. Specifically, the number of each type of impurities identified is counted, the proportions of various impurities are summarized, and a comprehensive impurity content proportion coefficient is generated.

[0038] Further, the method further comprises:

[0039] The abnormal impurity size set is obtained, wherein the abnormal impurity size set is part of the impurity size set that exceeds the preset impurity size interval; based on the abnormal impurity size set, a continuous tracking instruction of the CCD sensor is triggered; according to the continuous tracking instruction, the CCD sensor is controlled to perform continuous tracking image acquisition of an abnormal impurity coordinate set, and a continuous tracking image acquisition control parameter is recorded, wherein the continuous tracking image acquisition control parameter includes a lens focal length and an acquisition size; based on the continuous tracking image acquisition control parameter, a zoom ratio analysis of a continuous tracking image acquisition result is performed, and a zoom ratio evaluation coefficient set is established; based on the zoom ratio evaluation coefficient set, the initial pre-aiming frame is adjusted, and a pyramid aiming frame is established; the CCD sensor acquisition result is frame-selected step by step with the pyramid aiming frame, and the frame-selected result is supplemented to the frame selection result.

[0040] According to the preset impurity size standard, the allowed size range of each impurity is determined, and according to the data collected by the laser sensor, an abnormal impurity size set is established by screening those impurity sizes that exceed the preset size interval from the impurity size set.

[0041] For abnormal impurities, their positions and states need to be continuously tracked for further processing. Specifically, a continuous tracking instruction of the CCD sensor is triggered, which requires the sensor to continuously perform corresponding operations such as image acquisition when a specific event or condition is detected. The instruction is sent to the CCD sensor, which is required to continuously perform image acquisition when an abnormal impurity is detected.

[0042] According to the abnormal impurity coordinate set, the CCD sensor continuously tracks the position of the abnormal impurity according to the instruction, and records the image parameters at each acquisition, including the lens focal length, the acquisition size, which can affect the accuracy and range of image acquisition. The lens focal length refers to the focal length of the lens, which affects the field of view size and the ability to capture image details. The acquisition size refers to the resolution and the number of pixels of image acquisition, which determines the clarity and detail display of the image.

[0043] Using the collected continuous tracking image acquisition control parameters, the zoom ratio of the image under different settings is analyzed. The image zoom ratio refers to the proportional relationship between the image display and the actual physical size, which affects the display size and resolution of the abnormal impurities in the image. According to the analysis result, a zoom ratio evaluation coefficient set is established.

[0044] According to the evaluation coefficient set, the size of the initial pre-aiming frame is adjusted to ensure that the abnormal impurities are effectively captured in the image. This step can be completed by an automatic algorithm to ensure that the pre-aiming frame setting can cover the abnormal impurities as comprehensively as possible. By adjusting the pyramid aiming frame, the pyramid aiming frame is a multi-level frame selection strategy for gradually refining the identification range of abnormal impurities. It can include different levels of pre-aiming frames, each level corresponding to different image scaling factors and resolutions, from rough to fine step-by-step framing of abnormal impurities.

[0045] Using the established pyramid aiming frame, the image results collected by the CCD sensor are gradually framed. Each framing level is based on different pre-aiming frame settings to gradually refine the detection range of abnormal impurities. Gradual framing refers to the use of multiple levels of pre-aiming frames or framers for hierarchical abnormal impurity detection. Each level provides different precision and detail of abnormal impurity detection results. Further, the framing results of each level can complement each other to improve the comprehensive coverage and accurate detection of abnormal impurities. The supplementary framing results can further optimize the recognition and positioning of abnormal impurity features, ensuring the accuracy and efficiency of the self-test of the cleaning machine working quality.

[0046] Further, the method for performing impurity feature recognition includes:

[0047] Obtain the first framing image of the first framing result. Obtain the impurity feature image set of the impurity feature recognition library. Input the first framing image and the impurity feature image set into the impurity feature comparison analysis channel, and output the impurity feature similarity coefficient set. If the maximum similarity coefficient in the impurity feature similarity coefficient set meets the preset similarity coefficient, the impurity feature corresponding to the maximum similarity coefficient is taken as the impurity feature recognition result. Otherwise, output the impurity feature recognition result as empty.

[0048] Select any framing result from the image data collected by the CCD sensor as the first framing result, and obtain the first framing image corresponding to the first framing result. Obtain the impurity feature image set in the impurity feature recognition library, which includes various types and shapes of impurity feature images. These images are pre-determined and recorded.

[0049] The first framing image and each image in the impurity feature image library are input into the impurity feature comparison analysis channel for feature extraction and similarity calculation, generating an impurity feature similarity coefficient set. This set contains the similarity scores of each compared image and the first framing image.

[0050] The similarity values in the similarity coefficient set of the impurity characteristics are compared, the maximum similarity value is found, if the maximum similarity coefficient meets the preset similarity coefficient threshold, the impurity characteristic corresponding to the maximum similarity coefficient is taken as the final impurity characteristic recognition result and output. If the maximum similarity coefficient does not reach the preset similarity coefficient threshold, the specific impurity characteristic cannot be determined, and therefore the recognition result is empty.

[0051] Further, the impurity characteristic comparison analysis channel is constructed, and the method comprises:

[0052] A first impurity characteristic image is acquired, a sample first impurity image set of the same type of impurity in different historical records of the first impurity characteristic image is acquired, a sample first impurity image in the sample first impurity image set is randomly selected, image similarity analysis is performed on the sample first impurity image and the first impurity characteristic image, a sample first impurity characteristic similarity coefficient set is obtained, a twin network is used to construct an impurity characteristic comparison analysis channel, wherein the impurity characteristic comparison analysis channel comprises a first impurity characteristic comparison analysis path and a second impurity characteristic comparison analysis path, and the first impurity characteristic comparison analysis path and the second impurity characteristic comparison analysis path are of the same structure and share the same weight value. The first impurity characteristic image, the sample first impurity image set and the sample first impurity characteristic similarity coefficient set are used to train the impurity characteristic comparison analysis channel until convergence.

[0053] One of the impurity characteristic image sets is randomly extracted as a first impurity characteristic image, image data of other impurities similar to the first impurity characteristic stored in the historical record is collected and integrated into a sample first impurity image set, and these images are used for comparison analysis.

[0054] A sample first impurity image is randomly selected from the sample first impurity image set, comparison analysis is performed on the sample first impurity image and the first impurity characteristic image to obtain the similarity score therebetween, for example, image features are extracted through image processing and computer vision technology, the image features are compared, similarity evaluation is performed based on a structural similarity index, a similarity coefficient is generated, the sample first impurity image set is traversed, and a sample first impurity characteristic similarity coefficient set is obtained.

[0055] A twin network is a special neural network structure, which consists of two identical sub-networks, each receiving different inputs but sharing the same weights. This structure is suitable for tasks that require similarity or matching comparison, where one twin network path is used to process the first impurity feature image and its associated sample first impurity image set, and the other twin network path, with the same structure as the first path, is used to process another type of impurity feature image comparison. The network structure of the two paths is exactly the same, and their weight parameters are shared, which ensures that the feature representations learned by the two paths are comparable.

[0056] The first impurity feature image is input as the input of one path, and the sample first impurity image set is input as the input of the other path, while the sample first impurity feature similarity coefficient set is input, and the network learns an effective feature space through training, in which the same category of impurity features can approach each other, while different categories of features are far apart. The parameters of the network are gradually adjusted to the optimal state during the training process until the performance of the model no longer improves significantly or reaches the preset convergence standard.

[0057] Further, the initial control parameter set, the impurity separation mechanism, and the fiber sorting mechanism are optimized and adjusted to generate an optimized control parameter set and an optimized cotton motion speed. The method comprises:

[0058] The model of the cotton cleaning machine is obtained, and the control parameter adjustment interval set of the initial control parameter set of the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism is obtained according to the model of the cotton cleaning machine. The cotton motion speed constraint interval is obtained. The control parameter optimization function is established with the preset quality standard as the target. The control parameter set and the cotton motion speed are randomly generated within the control parameter adjustment interval set and the cotton motion speed constraint interval, and production simulation is performed to obtain simulation quality detection results, wherein the simulation quality detection results include simulation fiber uniformity coefficient, simulation impurity content proportion coefficient, and simulation short fiber proportion coefficient. According to the simulation quality detection results, the control parameter optimization function is combined to calculate the control parameter optimization fitness. The control parameter optimization is continued to obtain the optimized control parameter set and the optimized cotton motion speed with the minimum control parameter optimization fitness.

[0059] According to the model of the cotton cleaning machine, the control parameter adjustment interval set of the initial control parameter set of the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism is determined. These intervals define the adjustment range that each parameter can accept, which is to ensure the rationality and actual feasibility of parameter adjustment during optimization. Similarly, according to the model of the cotton cleaning machine, the constraint interval of the cotton motion speed is determined, which specifies the minimum and maximum speed range of the cotton motion in the cotton cleaning machine to ensure the safety and processing efficiency of the machine.

[0060] The preset quality standards for the cotton cleaner are determined. These standards include preset fiber uniformity, preset impurity content ratio, preset short fiber ratio, and other quality indicators. With the preset quality standards as the target, a control parameter optimization function is established. This function takes into account various control parameters of the cotton cleaner and the speed of raw cotton movement. By adjusting these parameters, the quality performance during the cotton cleaning process is optimized.

[0061] Within the defined set of control parameter adjustment intervals and the constraint interval of raw cotton movement speed, a set of control parameters and raw cotton movement speed are randomly generated. Using the generated set of control parameters and raw cotton movement speed, production simulation is performed. The quality of cleaned cotton is evaluated by simulating the actual operation of the cotton cleaning machine. During the simulation, the simulation quality detection results are recorded, including indicators such as the simulation fiber uniformity coefficient, the simulation impurity content ratio coefficient, and the simulation short fiber ratio coefficient. The specific method for obtaining the simulation quality detection results is the same as the aforementioned method for obtaining quality detection results.

[0062] The simulation quality test results are substituted into the control parameter optimization function to calculate the fitness of the control parameters. Within the control parameter adjustment interval set and the raw cotton movement speed constraint interval, multiple sets of random control parameters and raw cotton movement speeds are generated through algorithms such as genetic algorithms and particle swarm optimization. Simulations are performed on each set of parameters, and their fitness is calculated. Based on the fitness score, the optimized control parameter set with the minimum fitness and the corresponding optimized raw cotton movement speed are selected.

[0063] Furthermore, the formula for the control parameter optimization function is as follows:

[0064] ;

[0065] in, To optimize fitness by controlling parameters, , , These are the simulated fiber uniformity coefficient, the simulated impurity content ratio coefficient, and the simulated short fiber ratio coefficient, respectively. , , These are the preset fiber uniformity coefficient, preset impurity content ratio coefficient, and preset short fiber ratio coefficient, respectively, in the preset quality standards. , , These are the weight values, and , , The sum is 1.

[0066] Specifically, the formula for the control parameter optimization function is as follows:

[0067] ;

[0068] Herein , , respectively represent the simulation fiber uniformity coefficient, the simulation impurity content ratio coefficient, and the simulation short fiber ratio coefficient obtained by simulation, and , , are preset target values. The weight values , , control the influence of each quality index in the fitness function, and can be adjusted according to actual conditions to balance the relationship between the importance of different quality indexes and the optimization target.

[0069] When the simulation result is close to the preset standard, the square of the deviation, such as is small, resulting in a smaller corresponding logarithmic term, and vice versa. When the simulation result deviates from the preset standard, the square of the deviation increases, resulting in an increase in the corresponding logarithmic term. In the optimization of the control parameter set and the movement speed of raw cotton, different parameter combinations are tried, and , , are obtained by simulation. The optimization target is to minimize the fitness function, that is, to make the simulation result as close to the preset quality standard as possible. The function design can effectively reflect the complex relationship between multiple quality indexes, and smoothly guide parameter adjustment in the optimization process to achieve the expected quality standard.

[0070] In summary, the cotton cleaning machine working quality self-test control method based on multi-device cooperation provided by the embodiments of the present application has the following technical effects:

[0071] By arranging laser sensors, CCD sensors and photoelectric sensors at the outlets of the opening mechanism, impurity separation mechanism and fiber sorting mechanism, real-time monitoring of each key link in the processing of raw cotton is realized. The data acquisition system collects and integrates the collection results of multiple sensors in real time, realizes comprehensive monitoring of the quality indicators of raw cotton, including fiber density, impurity content and short fiber ratio, etc. Based on the real-time data acquisition system, quality detection can be quickly performed and quality detection results can be generated. Combined with the preset quality standards, when the quality detection results do not meet the preset standards, the initial control parameter set and the raw cotton movement speed can be automatically triggered to optimize the adjustment, and the optimized control parameter set and the raw cotton movement speed are generated to ensure that the quality is within the standard range. Through the optimization of the control parameter set and the raw cotton movement speed, precise adjustment and control of the scutcher are realized, and the efficiency and stability of product quality in the production process are improved. In summary, the method effectively solves the limitations of traditional scutchers in quality control by integrating multi-sensor data, real-time monitoring and adjustment, and significantly improves the quality management level and operation efficiency in the production process.

[0072] The above description of disclosed embodiments enables one of ordinary skill in the art to make or use the application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application will not be limited to the embodiments shown herein, but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for self-testing control of working quality of a cotton cleaner based on multi-device cooperation, characterized in that, The method is applied to a cotton cleaner, the cotton cleaner comprising a feeding mechanism, an opening mechanism, an impurity separation mechanism, and a fiber sorting mechanism, and the method comprises: sequentially arranging a laser sensor at an outlet of the opening mechanism, arranging a CCD sensor at an outlet of the impurity separation mechanism, and arranging a photoelectric sensor at an outlet of the fiber sorting mechanism; configuring an initial control parameter set of the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism, and performing motion control of raw cotton of the feeding mechanism, and recording a raw cotton motion speed; collecting laser sensor collection results, CCD sensor collection results, and photoelectric sensor collection results through a data acquisition system, and performing quality detection to generate quality detection results; wherein the method of performing quality detection to generate quality detection results comprises: performing fiber density distribution analysis according to the laser sensor collection results to generate a fiber uniformity coefficient; performing impurity detection according to the CCD sensor collection results to generate an impurity content proportion coefficient; wherein the method of performing impurity detection according to the CCD sensor collection results to generate an impurity content proportion coefficient comprises: acquiring a laser sensor emission light signal node and a reflected light signal node according to laser sensor collection results, performing light signal analysis, determining an impurity size set and an impurity coordinate set, and wherein the impurity size set and the impurity coordinate set have a mapping relationship; setting an initial pre-aiming frame based on the impurity size set and the impurity coordinate set; establishing an impurity feature recognition library of raw cotton; framing the CCD sensor collection results with the initial pre-aiming frame, calling the impurity feature recognition library based on the framing results, performing impurity feature recognition, and generating the impurity content proportion coefficient according to impurity feature recognition results; performing fiber length distribution analysis according to the photoelectric sensor collection results to generate a short fiber proportion coefficient; integrating the fiber uniformity coefficient, the impurity content proportion coefficient, and the short fiber proportion coefficient to generate the quality detection results; interacting with a production task, and generating a preset quality standard based on the production task; when the quality detection results do not meet the preset quality standard, optimizing and adjusting the initial control parameter set and the raw cotton motion speed to generate an optimized control parameter set and an optimized raw cotton motion speed; wherein the method of optimizing and adjusting the initial control parameter set and the raw cotton motion speed to generate an optimized control parameter set and an optimized raw cotton motion speed comprises: acquiring a cotton cleaner model, acquiring a control parameter adjustment interval set of the initial control parameter set of the opening mechanism, the impurity separation mechanism, and the fiber sorting mechanism according to the cotton cleaner model, and acquiring a raw cotton motion speed constraint interval; establishing a control parameter optimization function with the preset quality standard as a target; and performing optimization and adjustment on the control parameter optimization function to generate the optimized control parameter set and the optimized raw cotton motion speed. In the control parameter adjustment interval set and the raw cotton movement speed constraint interval, a control parameter set and a raw cotton movement speed are randomly generated for production simulation to obtain simulation quality detection results, wherein the simulation quality detection results include simulation fiber uniformity coefficients, simulation impurity content proportion coefficients, and simulation short fiber proportion coefficients; According to the simulation quality detection results, an optimized control parameter set and an optimized raw cotton movement speed are calculated based on the control parameter optimization function. The control parameter optimization function formula is as follows: wherein, is a control parameter optimization fitness, are respectively a simulation fiber uniformity coefficient, a simulation impurity content proportion coefficient, and a simulation short fiber proportion coefficient, are respectively a preset fiber uniformity coefficient, a preset impurity content proportion coefficient, and a preset short fiber proportion coefficient in the preset quality standard, are respectively weight values, and and are 1. The control parameter optimization is continued to obtain the optimized control parameter set and the optimized raw cotton movement speed with the minimum control parameter optimization fitness; The scutcher is adjusted and controlled based on the optimized control parameter set and the optimized raw cotton movement speed.

2. The method for self-testing control of working quality of a cotton cleaner based on multi-device cooperation according to claim 1, characterized in that, The method further includes: An abnormal impurity size set is obtained, wherein the abnormal impurity size set is part of the impurity size set that is outside a preset impurity size interval; Based on the abnormal impurity size set, a continuous tracking instruction of the CCD sensor is triggered; According to the continuous tracking instruction, the CCD sensor performs continuous tracking image acquisition of an abnormal impurity coordinate set, and records continuous tracking image acquisition control parameters, wherein the continuous tracking image acquisition control parameters include lens focal lengths, and acquisition sizes; Based on the continuous tracking image acquisition control parameters, a zoom ratio analysis of a continuous tracking image acquisition result is performed to establish a zoom ratio evaluation coefficient set; Based on the zoom ratio evaluation coefficient set, the initial pre-aiming frame is adjusted to establish a pyramid aiming frame; The CCD sensor acquisition result is frame-selected level by level with the pyramid aiming frame, and the frame-selected result is supplemented with the frame-selected results.

3. The method for self-testing control of working quality of a cotton cleaner based on multi-device cooperation according to claim 2, characterized in that, The method for performing impurity feature recognition includes: A first frame-selected image of a first frame-selected result is obtained; An impurity feature image set of the impurity feature recognition library is obtained; The first frame-selected image and the impurity feature image set are input into an impurity feature comparison analysis channel to output an impurity feature similarity coefficient set; If a maximum similarity coefficient in the impurity feature similarity coefficient set meets a preset similarity coefficient, the impurity feature corresponding to the maximum similarity coefficient is taken as an impurity feature recognition result; Otherwise, the impurity feature recognition result is empty.

4. The method for self-testing control of working quality of a cotton cleaner based on multi-device cooperation according to claim 3, characterized in that, The method for constructing the impurity feature comparison analysis channel includes: A first impurity feature image is obtained; A sample first impurity image set of the same type of impurity in different historical records of the first impurity feature image is obtained; A sample first impurity image in the sample first impurity image set is randomly selected for image similarity analysis with the first impurity feature image to obtain a sample first impurity feature similarity coefficient set; Based on a twin network, an impurity feature comparison analysis channel is constructed, wherein the impurity feature comparison analysis channel includes a first impurity feature comparison analysis path and a second impurity feature comparison analysis path, and the first impurity feature comparison analysis path and the second impurity feature comparison analysis path have the same structure and share the same weight value. The first impurity feature image, the sample first impurity image set and the sample first impurity feature similarity coefficient set are used to train the impurity feature comparison and analysis channel until convergence.

Images