Sorting control method and system and rubber product sorting equipment

Abstract

The invention belongs to the technical field of intelligent sorting control, and discloses a sorting control method and system and rubber product sorting equipment. The method comprises the following steps: collecting sorting data, wherein the sorting data comprises acceleration and angular velocity of materials, speed of a conveyor belt and position coordinates of the materials; frequency domain analysis is carried out on the acceleration, the mapping relation between the conveyor belt speed and the vibration dominant frequency of the acceleration is obtained, and the frequency domain analysis is carried out according to the conveyor belt speed to obtain a linear coefficient of the mapping relation; constructing a dynamic weight compensation model according to the sorting data and the mapping relation; carrying out resource adaptive scheduling through a simplified dynamic weight compensation model in combination with the speed of the conveyor belt to obtain compensation weight, and carrying out sorting in combination with a grading rule according to the compensation weight; cooperative control is carried out through multiple execution mechanisms according to the sorting result; according to the method, the problems of low precision, poor efficiency and the like caused by single-factor modeling and fixed model lagging of a traditional sorting system are solved, and the sorting performance of the system under complex working conditions is remarkably enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent sorting control, more particularly, the present application relates to a sorting control method, system and rubber product sorting equipment. BACKGROUND

[0002] In the scene of industrial equipment balancing, mechanical system counterweight, etc., rubber counterweight is widely used due to its buffering and shock-absorbing characteristics, but the coupling effect of dynamic working conditions poses a severe challenge to sorting accuracy.

[0003] Chinese patent application with publication number CN119126665A discloses a high-precision metering weighing system based on PLC control, which comprises a pressure sensor, a weighing module, a PLC controller, and a set weight grading sorting module. The pressure sensor, weighing module, PLC controller, and set weight grading sorting module are connected in sequence. The pressure sensor converts the weight of the product to be measured into a voltage signal through vertical displacement and sends it to the weighing module. The weighing module amplifies and digitizes the received voltage signal to obtain a digital signal corresponding to the pressure signal and transmits it to the PLC controller. The PLC controller analyzes and processes the digital signal using a fuzzy PID algorithm to obtain weight information and transmits it to the corresponding set weight grading sorting module. The set weight grading sorting module receives the weight information and controls the corresponding electromagnetic valve to open and close to actuate the rotary air cylinder to move the product to be measured to the corresponding storage basket when the displacement signal meets the corresponding sorting level discharge port. This invention realizes accurate product weight measurement and sorting by combining pressure sensor, weighing module, and fuzzy PID control algorithm. The system can quickly convert the weight information of the product to be measured into a voltage signal, which is amplified and digitized to provide accurate digital signals for the PLC. The PLC uses a fuzzy PID algorithm for error analysis and parameter adaptive adjustment to improve control accuracy and response speed.

[0004] Although the above method can meet most scenarios, research and practical application of the above method and existing technology have found that the above method and existing technology at least have the following defects:

[0005] The above method only considers the weight deviation caused by a single factor (such as speed or vibration) under uniform speed conditions, resulting in a large actual weight measurement deviation, and uses a fixed model for weight compensation, affecting sorting efficiency.

[0006] In view of this, the present application provides a sorting control method, system and rubber product sorting equipment to solve the above problems. SUMMARY

[0007] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purpose, the present application provides the following technical solutions: a sorting control method, comprising the following steps:

[0008] Collect sorting data, the sorting data including acceleration and angular velocity of the material, the conveying belt speed and the position coordinates of the material;

[0009] Perform frequency domain analysis on the acceleration to obtain a mapping relationship between the conveying belt speed and the vibration main frequency of the acceleration, wherein a linear coefficient of the mapping relationship is obtained by performing frequency domain analysis on the conveying belt speed;

[0010] Construct a dynamic weight compensation model according to the sorting data and the mapping relationship;

[0011] Combine the conveying belt speed, simplify the dynamic weight compensation model to perform resource adaptive scheduling, obtain a compensation weight, and perform sorting according to the compensation weight and in combination with a grading rule;

[0012] Perform collaborative control according to the sorting result through multiple actuators.

[0013] Further, the method for analyzing and obtaining the mapping relationship between the conveying belt speed and the vibration main frequency of the acceleration comprises:

[0014] Obtain the conveying belt speed, perform Hanning window weighting FFT on vibration data of each uniform speed section, and extract a frequency peak with an amplitude greater than P% of total energy;

[0015] Remove the frequency peak exceeding a preset range from the extracted frequency peak to obtain a vibration main frequency corresponding to each speed v;

[0016] Build a linear model between the conveying belt speed and the main frequency, and solve the linear model by the least square method to obtain a linear coefficient of the linear model;

[0017] Verify the linear coefficient by a vibration source model;

[0018] Obtain the conveying belt speed, divide a variable speed section vibration signal into overlapping frames according to a preset window, add a Hanning window to each frame to obtain a framed signal, perform FFT on each framed signal to obtain a time-varying frequency spectrum, and extract a frequency with the largest amplitude from each time-varying frequency spectrum as a vibration main frequency;

[0019] Real-time calculate the vibration main frequency, calculate a speed differential by dividing a conveying belt speed difference in a preset time period by the preset time period, and compensate the linear coefficient in combination with the speed differential when the speed differential is greater than a preset differential threshold.

[0020] Further, the method for constructing the dynamic weight compensation model comprises:

[0021] Calculate a coupling term according to a coupling coefficient, a conveying belt speed and a material acceleration;

[0022] Calculate a jerk inertia term according to a mass inertia coefficient, a speed differential and a material mass;

[0023] a sine quantity of a product of the vibration main frequency and the motion time is calculated, and a speed vibration compensation term is calculated according to the vibration amplitude gain and the sine quantity;

[0024] an angular velocity term is calculated according to the centrifugal compensation coefficient, the material angular velocity and the distance from the material centroid to the center of the conveyor belt;

[0025] a dynamic weight compensation model is obtained by superimposing the speed vibration compensation term and the angular velocity term on the basis of the original measured weight, and deducting the coupling term and the jerk inertia term;

[0026] The method for obtaining the coupling coefficient comprises:

[0027] a standard block with a known mass is used;

[0028] the conveyor belt is controlled to run at a constant acceleration;

[0029] the output force of the load cell when the conveyor belt is empty is measured;

[0030] a product of the material acceleration and the conveyor belt speed is calculated, a ratio of the output force to the product of the material acceleration and the conveyor belt speed is calculated, n sets of tests are performed, a mean value of the n sets of ratios is calculated, and the coupling coefficient is obtained.

[0031] Further, the method for obtaining the mass inertia coefficient comprises:

[0032] three standard blocks with different masses are selected, a preset acceleration working condition is set, and the deviation between the measured weight value and the true value is measured; the mass inertia coefficient is obtained by solving the jerk inertia term based on the least square method;

[0033] The method for obtaining the vibration amplitude gain comprises:

[0034] the vertical vibration speed of the conveyor belt is measured, and the output fluctuation value of the load cell is synchronously collected;

[0035] FFT is performed on the vertical vibration speed and the output fluctuation value, the amplitudes A1 and A2 at the vibration main frequency are extracted respectively, the ratio of A1 and A2 is calculated, and the vibration amplitude gain is obtained.

[0036] Further, the method for performing resource adaptive scheduling in combination with the conveyor belt speed comprises:

[0037] when the conveyor belt speed is lower than a first speed threshold, a precise mode is divided, and the compensation weight is obtained using the dynamic weight compensation model;

[0038] when the conveyor belt speed is not lower than the first speed threshold and is lower than a second speed threshold, a balance mode is divided, and the compensation weight is obtained using a simplified dynamic weight compensation model; wherein the simplified dynamic weight compensation model is obtained by removing the speed vibration compensation term in the dynamic weight compensation model;

[0039] When the conveyor speed is not lower than the second speed threshold, the classification is classified as a high-speed mode, and the compensated weight is obtained using a simplified dynamic weight compensation model; the simplified dynamic weight compensation model is obtained by removing the speed vibration compensation term, the angular velocity term, and the jerk inertia term in the dynamic weight compensation model.

[0040] Further, the method of sorting combined with the classification rule includes:

[0041] The deviation amount of the compensated weight from the standard material weight is calculated, and when the deviation amount is lower than the first weight threshold, the detected weight is marked as accurate;

[0042] When the compensated weight is lower than the standard material weight, and the deviation amount exceeds the second weight threshold, the detected weight is marked as underweight;

[0043] When the compensated weight is higher than the standard material weight, and the deviation amount exceeds the second weight threshold, the detected weight is marked as overweight.

[0044] Further, the method of performing cooperative control includes:

[0045] When it is detected that the conveyor speed is greater than the speed threshold and the detected weight is marked as underweight, the injection duration is calculated according to the absolute value of the deviation amount, the conveyor speed, and the air blowing calibration coefficient, and air blowing sorting is performed according to the injection duration, wherein the air blowing calibration coefficient is obtained by blowing calibration experiment combined with data fitting calculation;

[0046] When the detected weight is marked as overweight, the thrust is calculated according to the absolute value of the deviation amount and the thrust calibration coefficient, and the sorting is started according to the thrust, wherein the thrust calibration coefficient is obtained by thrust calibration experiment;

[0047] When the distance between adjacent materials is less than the distance threshold, the push rod is disabled, and cooperative control is performed by a double air blowing cooperative mode.

[0048] Further, the method of performing cooperative control by the double air blowing cooperative mode includes:

[0049] The trigger time of each air blowing device is calculated according to the time when the material reaches the reference of the injection area, the distance threshold, and the conveyor speed, and the injection duration of each air blowing device is calculated according to the absolute value of the deviation amount, the conveyor speed, and the air blowing calibration coefficient; wherein when the detected weight is marked as underweight, the air blowing calibration coefficient is obtained by blowing calibration experiment combined with data fitting calculation, and when the detected weight is marked as overweight, the air blowing calibration coefficient is obtained by thrust calibration experiment combined with data fitting calculation.

[0050] A sorting control system for implementing the sorting control method, comprising:

[0051] Data acquisition module: collect sorting data, including the acceleration and angular velocity of the material, the speed of the conveyor belt and the position coordinates of the material;

[0052] Mapping modeling module: frequency domain analysis is performed on the acceleration to obtain a mapping relationship between the speed of the conveyor belt and the vibration main frequency of the acceleration, wherein a linear coefficient of the mapping relationship is obtained by frequency domain analysis according to the speed of the conveyor belt;

[0053] Model building module: a dynamic weight compensation model is constructed according to the sorting data and the mapping relationship;

[0054] Adaptive scheduling module: in combination with the speed of the conveyor belt, resource adaptive scheduling is performed by simplifying the dynamic weight compensation model to obtain a compensation weight, and sorting is performed according to the compensation weight in combination with grading rules;

[0055] Cooperative control module: cooperative control is performed according to the sorting result by multiple actuators.

[0056] A rubber product sorting device is applied to the sorting control method.

[0057] The sorting control method, system and rubber product sorting device have the following technical effects and advantages:

[0058] The present application breaks through the limitation of traditional methods that only rely on uniform speed single factor modeling by collecting multi-source data of material acceleration, angular velocity, conveyor belt speed and material position coordinates, combining frequency domain analysis to distinguish uniform speed and variable speed working conditions, improving the mapping accuracy of speed and vibration main frequency, and providing a reliable frequency reference for dynamic weight compensation; a multi-physical quantity dynamic weight compensation model is constructed, which integrates coupling terms, jerk inertia terms, speed vibration compensation terms and angular velocity terms, and is combined with experimental calibration of parameters such as coupling coefficient and mass inertia coefficient to fully offset interference such as speed-acceleration coupling, vibration fluctuation and centrifugal deviation, reduce weight measurement deviation, and improve sorting accuracy; a resource adaptive scheduling mechanism based on the speed of the conveyor belt improves the accuracy of low speed working conditions and reduces the calculation delay of high speed working conditions, solving the efficiency imbalance problem of fixed models at different speeds; an actuator cooperative control strategy reduces the sorting conflict rate of adjacent materials, and improves the material sorting success rate by differentiating the calibration coefficient of air blowing through experiments; the present application effectively solves the problems of low precision and poor efficiency caused by single factor modeling and fixed model lag in traditional sorting systems, significantly enhances the sorting performance of the system under complex working conditions, and provides a high precision, high efficiency and high reliability solution for the industrial automatic sorting field. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 It is a flowchart of a sorting control method of the present application;

[0060] Figure 2A data flow diagram of the present application;

[0061] Figure 3 A dynamic weight compensation model construction diagram of the present application;

[0062] Figure 4 A sorting control system structure diagram of the present application. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0064] Embodiment 1:

[0065] Please refer to Figure 1 、 Figure 2 The present embodiment provides a sorting control method, comprising the following steps:

[0066] Collecting sorting data, the sorting data including acceleration and angular velocity of the material, conveying belt speed and position coordinates of the material;

[0067] Deploying sensors in the conveying belt weighing section to collect sorting data of the material in real time, including acceleration and angular velocity of the material, conveying belt speed and position coordinates of the material;

[0068] Collecting multi-dimensional sorting data including acceleration, angular velocity, conveying belt speed and position coordinates of the material, which breaks through the limitation of the existing method relying on only a single factor (such as speed or vibration), so that the system can comprehensively perceive the comprehensive motion characteristics of the material under dynamic conditions such as acceleration, deceleration and static working conditions;

[0069] Collecting multi-dimensional sorting data including acceleration, angular velocity, conveying belt speed and position coordinates of the material, which is convenient for frequency domain analysis of acceleration and distinguishes uniform speed and variable speed working conditions for separate processing, constructs a dynamic weight compensation model integrating coupling coefficient, mass inertia coefficient, vibration amplitude gain and angular velocity term and other multiple physical quantities, replaces the fixed model considering only uniform speed single factor, and simultaneously implements resource adaptive scheduling according to the conveying belt speed, solving the problem of large weight measurement deviation caused by single factor modeling in the existing method, and realizing accurate compensation of dynamic weight under multi-factor interference.

[0070] Frequency domain analysis is performed on the acceleration to obtain a mapping relationship of the vibration main frequency of the conveying belt speed and the acceleration, wherein a linear coefficient of the mapping relationship is obtained by frequency domain analysis according to the conveying belt speed;

[0071] The method for analyzing the mapping relationship between the speed of the conveyor belt and the vibration main frequency of the acceleration comprises the following steps:

[0072] Obtain the speed of the conveyor belt, and perform Hanning window weighting FFT on the vibration data of each uniform speed section to extract a frequency peak with an amplitude greater than P% of the total energy;

[0073] Remove the frequency peak that exceeds the preset range from the extracted frequency peak to obtain the vibration main frequency corresponding to each speed v;

[0074] Build a linear model between the speed of the conveyor belt and the main frequency, and solve the linear model by the least square method to obtain the linear coefficient of the linear model;

[0075] Verify the linear coefficient by a vibration source model, for example, whether the linear coefficient satisfies the eccentricity of the roller of the conveyor belt and the meshing of the gear of the motor;

[0076] Obtain the speed of the conveyor belt, divide the variable speed section vibration signal into overlapping frames according to a preset window, add a Hanning window to each frame to obtain a framed signal, perform FFT on each framed signal to obtain a time-varying frequency spectrum, and extract the frequency with the maximum amplitude from each time-varying frequency spectrum as the vibration main frequency;

[0077] Real-time calculate the vibration main frequency, calculate the speed differential of a preset time period and the ratio of the preset time period to obtain the speed differential; when the speed differential is greater than a preset differential threshold, compensate the linear coefficient in combination with the speed differential.

[0078] The mapping relationship between the speed of the conveyor belt and the vibration main frequency is constructed by performing frequency domain analysis on the acceleration signal. For the uniform speed condition, the frequency peak with an amplitude greater than P% of the total energy is extracted by using the Hanning window weighting FFT, and the abnormal frequency peak is removed to accurately obtain the vibration main frequency corresponding to each speed v. A linear model is built by using the least square method, and the physical rationality of the linear coefficient is verified by a vibration source model (such as the eccentricity of the roller and the meshing of the gear) to avoid overfitting caused by single data driving and solve the problem of distortion of the frequency-speed relationship caused by the existing method which only relies on the uniform speed single factor modeling. For the variable speed condition, the vibration signal is divided into frames and windowed by using the STFT, the maximum amplitude frequency of the time-varying frequency spectrum is extracted as the real-time vibration main frequency, the speed differential is calculated, and the linear coefficient is dynamically compensated when the speed differential exceeds the differential threshold to effectively cope with the deviation of the main frequency extraction caused by the spectrum widening during the variable speed, make up for the defect that the traditional fixed model cannot adapt to the speed change, and finally realize the accurate modeling of the mapping relationship between the speed of the conveyor belt and the vibration main frequency under all working conditions, provide reliable frequency parameters for the dynamic weight compensation, and improve the accuracy of the weight measurement and the adaptability of the sorting system under multiple working conditions.

[0079] Build a dynamic weight compensation model according to the sorting data and the mapping relationship;

[0080] Please refer to Figure 3As shown, the method for constructing a dynamic weight compensation model comprises:

[0081] According to the coupling coefficient, the conveyor belt speed and the material acceleration, a coupling term is calculated and obtained;

[0082] According to the mass inertia coefficient, the speed differential and the material mass, a jerk inertia term is calculated and obtained;

[0083] A sine quantity of the product of the vibration main frequency and the motion time is calculated, and according to the vibration amplitude gain and the sine quantity, a speed vibration compensation term is calculated and obtained;

[0084] According to the centrifugal compensation coefficient, the material angular velocity and the distance from the material centroid to the center of the conveyor belt, an angular velocity term is calculated and obtained, wherein the distance from the material centroid to the center of the conveyor belt is calculated by the Euclidean distance;

[0085] On the basis of the original measured weight, the acceleration vibration compensation term and the angular velocity term are superimposed, and the coupling term and the jerk inertia term are deducted, to obtain a dynamic weight compensation model.

[0086] The method for constructing a dynamic weight compensation model compensates the dynamic weight by constructing a dynamic weight compensation model that integrates multiple physical quantities. The coupling term is calculated by introducing the coupling coefficient, the conveyor belt speed and the material acceleration, which compensates for the interference of the coupling of the conveyor belt speed and acceleration on the weight measurement, solving the defect that the traditional method does not consider the mutual influence of the two under variable speed conditions. The jerk inertia term is calculated by using the mass inertia coefficient, the speed differential (jerk) and the material mass, which offsets the inertia deviation caused by the speed change rate and makes up for the neglect of the dynamic inertia force by the fixed model. The speed vibration compensation term is calculated by the vibration main frequency, the motion time and the vibration amplitude gain, which converts the periodic vibration interference into a sine compensation quantity, effectively suppressing the modulation error of the conveyor belt vibration on the weight signal. The angular velocity term is calculated based on the centrifugal compensation coefficient, the angular velocity and the Euclidean distance from the material centroid to the center of the conveyor belt, which quantifies the influence of the centrifugal force generated by the eccentric motion of the material, filling the gap of the traditional model that does not cover the rotating centrifugal interference. Finally, the multi-dimensional compensation mechanism of superimposing the acceleration vibration compensation term and the angular velocity term on the original weight and deducting the coupling term and the jerk inertia term is constructed, which covers the dynamic model of multiple physical factors such as speed-acceleration coupling, jerk inertia, vibration fluctuation and centrifugal deviation, breaks through the limitations of the existing method that relies only on a single factor and a fixed model, realizes the precise correction of the weight measurement deviation under complex conditions, provides reliable weight data support for sorting decisions, and significantly improves the adaptability and sorting precision of the sorting system in multiple interference scenarios.

[0087] The method for obtaining the coupling coefficient comprises:

[0088] A standard block with a known mass is used for testing;

[0089] The conveyor belt is controlled to run at a constant acceleration;

[0090] measuring the output force of the load cell when unloaded;

[0091] calculating the product of the material acceleration and the conveying belt speed, calculating the ratio of the output force and the product of the material acceleration and the conveying belt speed, performing n sets of tests, calculating the average of the n sets of ratios, and obtaining the coupling coefficient.

[0092] The method for obtaining the coupling coefficient controls the variable through a standard block experiment, maintains a constant acceleration, eliminates the interference of the material weight itself in the case of a known standard block mass, and accurately quantifies the influence of the coupling effect of speed and acceleration on the output of the load cell; the superposition effect of the coupling interference and the material self-weight is separated by unloaded measurement to ensure that the experimental data only reflect the contribution of the coupling term; the random error is reduced by statistical average of n sets of ratios to obtain a stable coupling coefficient, which provides a reliable quantitative parameter for the coupling term in the dynamic weight compensation model; the coupling coefficient can accurately deduct the weight deviation caused by the coupling of speed and acceleration in the compensation process, solving the weight measurement error problem caused by not considering the coupling of the two in the traditional method, so that the dynamic weight compensation model can more accurately correct the deviation caused by the coupling of multiple physical quantities, and the weight measurement accuracy and sorting reliability of the sorting system under variable speed conditions are improved.

[0093] The method for obtaining the mass inertia coefficient comprises:

[0094] selecting three mass standard blocks, presetting an acceleration working condition, and measuring the deviation of the weighing value and the true value; solving the mass inertia coefficient based on the least square method according to the jerk inertia term.

[0095] The method for obtaining the mass inertia coefficient covers multiple mass scenarios using different mass standard blocks, simulates the influence of jerk (speed differential) in the actual variable speed process through a preset acceleration working condition, and establishes a quantitative correlation between the weighing deviation and the jerk inertia term; the mass inertia coefficient can be accurately extracted by fitting multiple sets of experimental data with the help of the least square method, and the coefficient can represent the contribution of the jerk inertia term in the dynamic weight compensation; by introducing the coefficient into the model, the interference of the inertial force caused by the speed change rate on the weighing result when the conveying belt accelerates or decelerates is effectively offset, solving the weight measurement deviation problem caused by not considering the jerk inertia effect in the traditional method, so that the dynamic weight compensation model can more comprehensively cover the multiple physical quantity interference under variable speed conditions, improving the accuracy of weight compensation and the adaptability of the sorting system to dynamic conditions, and ensuring the sorting accuracy under complex motion states.

[0096] The method for obtaining the vibration amplitude gain comprises:

[0097] measuring the vertical vibration speed of the conveying belt, and synchronously collecting the output fluctuation value of the load cell;

[0098] The vertical vibration speed and the output fluctuation value are subjected to FFT, the amplitudes A1 and A2 at the vibration main frequency are extracted respectively, the ratio of A1 and A2 is calculated, and the vibration amplitude gain is obtained.

[0099] The method utilizes the frequency domain analysis technology to quantitatively correlate the periodic interference of the conveyor belt vibration and the weighing fluctuation signal at the vibration main frequency, accurately represents the modulation strength of the vibration speed on the weight measurement through the amplitude ratio, and provides a key parameter for the speed vibration compensation term in the dynamic weight compensation model; the coefficient can convert the energy of the vibration speed into a corresponding weight fluctuation compensation amount, effectively suppresses the periodic interference of the conveyor belt vibration on the weighing signal, solves the weight measurement deviation problem caused by the non-separation of the vibration interference in the traditional method, makes the dynamic weight compensation model be able to accurately offset the amplitude influence of the vibration main frequency, and improves the stability of the weight detection and the anti-vibration performance of the sorting system in a multi-interference environment.

[0100] In combination with the conveyor belt speed, resource adaptive scheduling is performed through simplification of the dynamic weight compensation model to obtain a compensation weight, and the compensation weight is used in combination with a grading rule to sort;

[0101] The method for resource adaptive scheduling according to the conveyor belt speed comprises the following steps:

[0102] When the conveyor belt speed is lower than a first speed threshold, the accurate mode is divided, and the dynamic weight compensation model is used to obtain a compensation weight;

[0103] When the conveyor belt speed is not lower than the first speed threshold and is lower than a second speed threshold, the balanced mode is divided, and the simplified dynamic weight compensation model is used to obtain a compensation weight; wherein the simplified dynamic weight compensation model is obtained by removing the speed vibration compensation term in the dynamic weight compensation model;

[0104] When the conveyor belt speed is not lower than the second speed threshold, the high-speed mode is divided, and the minimal dynamic weight compensation model is used to obtain a compensation weight; the minimal dynamic weight compensation model is obtained by removing the speed vibration compensation term, the angular velocity term and the jerk inertia term in the dynamic weight compensation model.

[0105] The method of resource self-adaptive scheduling according to the conveying belt speed adopts a full-amount dynamic weight compensation model for a low-speed scene (the conveying belt speed is lower than a first threshold value), that is, all interference terms are included to ensure high precision of weight compensation when the calculation resource is sufficient, solving the problem of insufficient precision caused by redundant calculation or simplified model of the fixed model under low-speed working conditions; a simplified dynamic weight compensation model is used for a medium-speed scene (the conveying belt speed is between the first threshold value and a second threshold value), that is, the speed vibration compensation term is removed, the calculation amount is reduced while the core compensation precision is ensured, the balance between resource consumption and sorting efficiency is achieved, and the waste of computing power of the full-amount model is avoided; a minimalist model is adopted for a high-speed scene (the conveying belt speed is not lower than the second threshold value), the speed vibration compensation term, the angular velocity term and the jerk inertia term are removed, the processing speed is greatly improved through simplifying the calculation logic, the real-time performance of weight compensation under high-speed motion is ensured, and the problem of sorting decision lag caused by calculation delay of the fixed model is solved; the scheduling mechanism dynamically matches the model complexity with the conveying belt speed, realizes efficient use of calculation resources and optimized balance of sorting precision under different working conditions, and significantly improves the adaptability and overall sorting efficiency of the sorting system in the full-speed range.

[0106] The method of sorting according to the grading rule comprises:

[0107] The deviation amount of the compensation weight from the standard material weight is calculated, when the deviation amount is lower than a first weight threshold value, the detected weight is marked as accurate;

[0108] When the compensation weight is lower than the standard material weight and the deviation amount exceeds a second weight threshold value, the detected weight is marked as underweight;

[0109] When the compensation weight is higher than the standard material weight and the deviation amount exceeds the second weight threshold value, the detected weight is marked as overweight.

[0110] The method of sorting according to the grading rule quantifies the weight state of the material by quantifying the deviation amount, including accurate, underweight and overweight, provides accurate sorting decision basis for the actuator, solves the problem of sorting confusion caused by lack of clear classification standard in the traditional method; the first weight threshold value filters the small deviation (marked as accurate when lower than the threshold value), avoids mis-sorting caused by measurement error or normal fluctuation, and improves the sorting reliability; the second weight threshold value defines the significant deviation range (marked as underweight or overweight when exceeding the threshold value), ensures that only the material that needs intervention triggers the actuation action, reduces redundant operation, and improves the sorting efficiency; at the same time, the direction difference between underweight and overweight is distinguished, providing state identification for subsequent targeted control of the actuator such as air blowing and push rod, avoiding misaction of the actuator, and finally realizing standardization and intelligentization of the sorting process, improving the overall sorting quality and efficiency.

[0111] The multiple actuators are cooperatively controlled according to the sorting result;

[0112] The method of cooperative control comprises:

[0113] When the conveyor belt speed is greater than the speed threshold and the weight mark is detected as being too light, the jet duration is calculated according to the absolute value of the deviation, the conveyor belt speed and the air blowing calibration coefficient, and the air blowing sorting is performed according to the jet duration, wherein the air blowing calibration coefficient is obtained through air blowing calibration experiments combined with data fitting calculation;

[0114] When the weight mark is detected as being too heavy, the thrust is calculated according to the absolute value of the deviation and the thrust calibration coefficient, and the push rod is started to perform sorting according to the thrust, wherein the thrust calibration coefficient is obtained through thrust calibration experiments.

[0115] By cooperatively controlling the air blowing and the push rod actuator according to the material weight state (too light / too heavy) and the conveyor belt speed, the jet duration can be dynamically calculated based on the absolute value of the deviation, the conveyor belt speed and the air blowing calibration coefficient obtained through air blowing calibration experiments for the high-speed too light scenario, replacing the traditional fixed duration or empirical parameter control, ensuring that the air blowing intensity is accurately matched with the material lightness and movement speed, solving the problems of missed sorting due to insufficient air blowing duration and energy waste due to excessive duration for too light materials at high speed; for the too heavy scenario, the push rod thrust is calculated according to the absolute value of the deviation and the thrust calibration coefficient obtained through thrust calibration experiments, making the thrust positively related to the heavy degree, avoiding the mechanical damage caused by traditional single thrust setting for too heavy materials or excessive pushing, and improving the relevance and effectiveness of the actuator action; the cooperative control mechanism breaks through the limitations of single control logic and experience-dependent parameters in traditional methods by dynamically mapping the weight state-speed-actuator strategy and introducing experimentally calibrated coefficients, achieving the precision and intelligence of sorting actions, and significantly improving the sorting success rate and equipment operation reliability under different weight states and speed conditions.

[0116] When the distance between adjacent materials is less than the distance threshold, the push rod is disabled, and the double air blowing cooperative mode is switched to perform cooperative control, and the jet angles of the air blowing devices differ by 15°, which can cover the adjacent area.

[0117] The trigger time of each air blowing device is calculated according to the reference time when the material reaches the jet area, the distance threshold and the conveyor belt speed, and the jet duration of each air blowing device is calculated according to the absolute value of the deviation, the conveyor belt speed and the air blowing calibration coefficient; wherein when the weight mark is detected as being too light, the air blowing calibration coefficient is obtained through air blowing calibration experiments (the jet duration of too light materials is related to the upward air flow momentum, mainly used to resist gravity and inertia) combined with data fitting calculation, and when the weight mark is detected as being too heavy, the air blowing calibration coefficient is obtained through thrust calibration experiments (the jet duration of too heavy materials is related to the forward air flow momentum, mainly used to resist greater inertia) combined with data fitting calculation.

[0118] By disabling the push rod and switching to the double air blowing cooperative mode (15° difference in jet angle) when the distance between adjacent materials is less than the distance threshold, the triggering time is calculated based on the arrival time of the materials, the distance threshold and the conveyor speed, and different calibration coefficients are called according to the light or heavy state of the materials to calculate the jet duration. The effect is that: the mechanical collision risk of the traditional push rod when the distance is insufficient is effectively avoided, the non-contact sorting is realized by covering the adjacent area through the angle difference of double air blowing, and the sorting conflict problem of densely arranged materials is solved; the dynamic calculation of the triggering time ensures the precise synchronization of the air blowing action and the material movement, avoiding the failure of blowing off due to time deviation; the differential air blowing calibration coefficient for light or heavy materials respectively adopts the difference in gravity inertia (blowing off experiment) and forward inertia (pushing force experiment), so that the air flow momentum matches the weight state of the materials, breaking through the limitation of traditional single air blowing parameter that cannot consider different weight inertia; the cooperative control mechanism systematically solves the actuator conflict, sorting time error and weight state adaptability problem in the small distance scene, improves the robustness and sorting success rate of the sorting system in complex arrangement conditions, and ensures the accurate separation of light or heavy products in high-density material flow without damaging the equipment.

[0119] Embodiment 2:

[0120] The embodiment provides an optimization method of a dynamic weight compensation model, including the following steps:

[0121] Caputo fractional derivative is used to replace the integer order of velocity differential, wherein the fractional order is calibrated through rheological experiment.

[0122] For rubber materials, the traditional integer order jerk (dv / dt) assumes that the material is a rigid body, which cannot describe the history dependence, decay characteristics and other viscoelastic behaviors of rubber stress-strain, such as the hysteresis deformation and energy dissipation of rubber when it is suddenly stopped. After introducing Caputo fractional derivative, the fractional order (such as 0.83 when the rubber hardness is 60HA) is calibrated through rheological experiment, so that the mathematical expression of the jerk inertia term is more consistent with the physical nature of the rubber memory and nonlinear relaxation; the non-local characteristic of fractional derivative can accurately capture the inertia force decay law of rubber under dynamic conditions such as acceleration and sudden stop, such as dv / dt = -8 m / s 2 when suddenly stopped. The traditional model ignores the viscoelasticity, which leads to strong or weak inertia force calculation, reduces the error of the jerk inertia term, greatly improves the weight compensation accuracy of viscoelastic materials under dynamic conditions (sudden acceleration and sudden stop), and solves the weight measurement deviation problem caused by the insufficient modeling of material characteristics of the traditional integer order model, providing a more accurate physical model support for sorting in complex viscoelastic scenes.

[0123] Embodiment 3:

[0124] Please refer to Figure 4The embodiment provides a sorting control system, comprising:

[0125] A data acquisition module acquires sorting data, the sorting data including acceleration and angular velocity of the material, a conveying belt speed and position coordinates of the material;

[0126] A mapping modeling module performs frequency domain analysis on the acceleration to obtain a mapping relationship between the conveying belt speed and vibration main frequency of the acceleration, wherein a linear coefficient of the mapping relationship is obtained by performing frequency domain analysis on the conveying belt speed;

[0127] A model building module builds a dynamic weight compensation model according to the sorting data and the mapping relationship;

[0128] An adaptive scheduling module performs resource adaptive scheduling by simplifying the dynamic weight compensation model in combination with the conveying belt speed to obtain a compensation weight, and performs sorting in combination with grading rules according to the compensation weight;

[0129] A cooperative control module performs cooperative control according to sorting results through multiple actuators.

[0130] The above merely describes specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0131] Finally, the above merely describes preferred embodiments of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A sort control method characterized by, The method comprises the following steps: Collect sorting data, including acceleration and angular velocity of the material, conveying belt speed and position coordinates of the material; Perform frequency domain analysis on the acceleration to obtain a mapping relationship between the conveying belt speed and the vibration main frequency of the acceleration, wherein a linear coefficient of the mapping relationship is obtained by performing frequency domain analysis on the conveying belt speed; Construct a dynamic weight compensation model according to the sorting data and the mapping relationship; Combine the conveying belt speed, simplify the dynamic weight compensation model, and perform resource adaptive scheduling to obtain a compensation weight, and sort according to the compensation weight and a grading rule; Control multiple actuators according to the sorting result.

2. The sort control method of claim 1, wherein, The method for obtaining the mapping relationship between the conveying belt speed and the vibration main frequency of the acceleration comprises: Obtain the conveying belt speed, perform Hanning window weighting FFT on the vibration data of each uniform speed section, and extract the frequency peaks with an amplitude greater than P% of the total energy; Remove the frequency peaks exceeding the preset range from the extracted frequency peaks to obtain the vibration main frequency corresponding to each speed v; Build a linear model between the conveying belt speed and the dominant frequency, and solve the linear model by the least square method to obtain the linear coefficient of the linear model; Verify the linear coefficient by a vibration source model; Obtain the conveying belt speed, divide the variable speed section vibration signal into overlapping frames according to a preset window, add a Hanning window to each frame, obtain a framed signal, perform FFT on each framed signal to obtain a time-varying frequency spectrum, and extract the frequency with the maximum amplitude from each time-varying frequency spectrum as the vibration main frequency; Real-time calculate the vibration main frequency, calculate the ratio of the conveying belt speed difference in a preset time period to the preset time period to obtain a speed differential; when the speed differential is greater than a preset differential threshold, compensate the linear coefficient in combination with the speed differential.

3. A sort control method according to claim 2, wherein, The method for constructing the dynamic weight compensation model comprises: Calculate a coupling term according to the coupling coefficient, the conveying belt speed and the material acceleration; Calculate a jerk inertia term according to the mass inertia coefficient, the speed differential and the material mass; Calculate a sine quantity of the product of the vibration main frequency and the motion time, and calculate a speed vibration compensation term according to the vibration amplitude gain and the sine quantity; Calculate an angular velocity term according to the centrifugal compensation coefficient, the material angular velocity and the distance from the material centroid to the center of the conveying belt; On the basis of the original measured weight, superimpose the acceleration vibration compensation term and the angular velocity term, deduct the coupling term and the jerk inertia term to obtain the dynamic weight compensation model; The method for obtaining the coupling coefficient comprises: Use a standard block with a known mass; control the conveying belt to run at a constant acceleration; Measure the output force of the load cell when the load is empty; Calculate the product of the material acceleration and the conveying belt speed, calculate the ratio of the output force to the product of the material acceleration and the conveying belt speed, perform n sets of tests, calculate the average of the n sets of ratios, and obtain the coupling coefficient.

4. A sort control method according to claim 3, wherein, The method for obtaining the mass inertia coefficient comprises: Select three standard blocks with different masses, measure the deviation between the weighing value and the true value under a preset acceleration condition, and solve the mass inertia coefficient based on the least square method according to the jerk inertia term to obtain the mass inertia coefficient. The method for obtaining the vibration amplitude gain comprises: Measure the vertical vibration speed of the conveying belt, and synchronously collect the output fluctuation value of the load cell; Perform FFT on the vertical vibration velocity and output fluctuation value, extract the amplitudes A1 and A2 at the main vibration frequency, calculate the ratio of A1 to A2, and obtain the vibration amplitude gain.

5. The sort control method of claim 1, wherein, Methods for adaptive resource scheduling based on conveyor belt speed include: When the conveyor belt speed is lower than the first speed threshold, it is classified as precision mode, and the dynamic weight compensation model is used to obtain the compensation weight. When the conveyor belt speed is not lower than the first speed threshold and is lower than the second speed threshold, it is classified as a balanced mode, and the compensation weight is obtained by using a simplified dynamic weight compensation model; wherein, the simplified dynamic weight compensation model is obtained by removing the speed vibration compensation term in the dynamic weight compensation model; When the conveyor belt speed is not lower than the second speed threshold, it is classified as a high-speed mode, and the compensation weight is obtained by using a simplified dynamic weight compensation model. The simplified dynamic weight compensation model is obtained by removing the velocity vibration compensation term, angular velocity term and jerk inertia term from the dynamic weight compensation model.

6. The sort control method of claim 1, wherein, Methods for sorting based on grading rules include: Calculate the deviation between the compensated weight and the standard material weight. When the deviation is lower than the first weight threshold, the detected weight is marked as accurate. When the compensation weight is lower than the standard material weight and the deviation exceeds the second weight threshold, the detected weight is marked as too light. When the compensation weight is higher than the standard material weight and the deviation exceeds the second weight threshold, the detected weight is marked as excessive.

7. A sort control method according to claim 6, wherein, Methods for implementing coordinated control include: When the conveyor belt speed is detected to be greater than the speed threshold and the detected weight is marked as too light, the injection time is calculated based on the absolute value of the deviation, the conveyor belt speed and the air-blowing calibration coefficient. Air-blowing sorting is then performed based on the injection time. The air-blowing calibration coefficient is obtained by combining the blow-off calibration experiment with data fitting. When the detected weight is marked as unbalanced, the thrust is calculated based on the absolute value of the deviation and the thrust calibration coefficient. The push rod is activated based on the thrust to perform sorting. The thrust calibration coefficient is obtained through a thrust calibration experiment. When the distance between adjacent materials is less than the distance threshold, the push rod is disabled, and collaborative control is performed through the dual air-blowing collaborative mode.

8. A sort control method according to claim 7, wherein, Methods for coordinated control using a dual-air-blowing coordinated mode include: The trigger time of each air-blowing device is calculated based on the reference time when the material arrives at the spray zone, the spacing threshold, and the conveyor belt speed. The spraying duration of each air-blowing device is calculated based on the absolute value of the deviation, the conveyor belt speed, and the air-blowing calibration coefficient. When the detected weight is marked as too light, the air-blowing calibration coefficient is obtained by combining the blow-off calibration experiment with data fitting. When the detected weight is marked as too heavy, the air-blowing calibration coefficient is obtained by combining the thrust calibration experiment with data fitting.

9. A sorting control system implementing the sorting control method according to any one of claims 1 to 8, characterized in that, include: Data acquisition module: Collects sorting data, including the material's acceleration and angular velocity, conveyor belt speed, and the material's position coordinates; Mapping Modeling Module: Performs frequency domain analysis on acceleration to obtain the mapping relationship between the conveyor belt speed and the dominant vibration frequency of acceleration. The linear coefficients of the mapping relationship are obtained based on the frequency domain analysis of the conveyor belt speed. Model building module: Constructs a dynamic weight compensation model based on sorting data and mapping relationships; Adaptive scheduling module: Combining conveyor belt speed, it performs adaptive resource scheduling by simplifying the dynamic weight compensation model to obtain the compensation weight, and sorts according to the compensation weight and hierarchical rules. Collaborative control module: Collaborative control is achieved through multiple actuators based on the sorting results.

10. A rubber article sorting apparatus characterized by, Applied to a sorting control method as described in any one of claims 1-8.

Images