Intelligent internet-of-things belt weigher self-calibration real-time monitoring system based on multi-mode AI

Abstract

The invention discloses an intelligent Internet of Things belt weigher self-calibration real-time monitoring system based on multi-mode AI, relates to the technical field of belt weighers, solves the problem that an original error adjustment mode is relatively delayed and cannot quickly confirm errors, and provides a self-calibration real-time monitoring system for an intelligent Internet of Things belt weigher through careful checking of pressure data of branch monitoring nodes on the two sides. The problem source can be accurately positioned; whether the main pressure sensor is abnormal or the position of the main monitoring node deviates, the system rapidly generates corresponding signals or moving direction information, clear guidance is provided for subsequent correction work, the efficiency and accuracy of fault diagnosis are greatly improved, and the maintenance cost and time loss are reduced; simulation movement verification is carried out by means of a preset two-dimensional model, and an accurate position reference can be provided for correction work by determining an associated included angle in real time and locking a verification base point; the virtual verification mode is not only efficient, but also can perform deep analysis and positioning on the problems of the belt weigher under the condition that the actual production is not influenced.

Description

Technical Field

[0001] The present invention relates to the technical field of belt scales, and specifically to an intelligent Internet-of-Things belt scale self-calibration real-time monitoring system based on multi-modal AI. Background Art

[0002] Belt scales mainly measure the weight of materials based on gravity induction. When the materials are carried on the belt and pass through the scale frame, the weighing bridge of the belt scale supports a section of the belt and the materials on the belt. The gravity of the materials and the belt is converted into an electrical signal by the weighing sensor. For example, in a common electronic belt scale, the sensor converts the gravity into a voltage or current signal proportional to the weight, and this signal is transmitted to the subsequent processing unit.

[0003] In order to calculate the material flow rate passing through the belt scale per unit time, it is also necessary to measure the belt running speed. Usually, a speed sensor, such as an optoelectronic or magnetoelectric speed sensor, is installed on the driving roller of the belt or other appropriate positions. The speed sensor generates a pulse signal corresponding to the belt speed, and its frequency is proportional to the belt speed. The control system performs a comprehensive operation on the weight signal output by the weighing sensor and the speed signal output by the speed sensor to obtain the instantaneous flow rate of the material (the weight of the material passing through per unit time). Integrating the instantaneous flow rate over a period of time can obtain the cumulative flow rate (the total weight of the material passing through over a period of time).

[0004] The application with the publication number CN114383699B discloses a belt scale weighing device and a metering method. The device includes a tension scale frame, a multi-roller weighing bridge, a speed sensor, a weighing display instrument, and an AI big data instrument. The weighing display instrument and the AI big data instrument are respectively used to receive the data signals sent by the connected sensors for processing and analysis, and establish communication. The method includes installing and calibrating the belt scale weighing device, recording its range coefficient, temperature, the initial values of each weighing sensor, and the average signal value output by each weighing sensor during no-load operation; dividing the weighing interval into S parts, and recording the average signal value of each weighing interval in each temperature region; establishing a standard ratio coefficient; recording the average signal value output by each weighing sensor in each weighing interval for a period of time; calculating the real-time ratio coefficient; and correcting the range coefficient of each weighing interval in each temperature region. This invention can detect the change of belt tension, and can correct the tension influence amount in real time, improving the material metering accuracy and stability of the belt scale.

[0005] During the real-time calibration process of its intelligent belt scale, generally only based on the comparison process of the pressure data and the actual pressure data during the corresponding monitoring process, to evaluate whether there is an accuracy error problem with the belt scale. And for such problems, the error adjustment is relatively delayed, and it cannot quickly confirm the error, and cannot achieve a more accurate and rapid error verification and confirmation effect. Summary of the Invention

[0006] In view of the deficiencies of the prior art, the present invention provides an intelligent Internet of Things belt scale self-calibration real-time monitoring system based on multimodal AI, which solves the problem that the original error adjustment method is relatively delayed and cannot quickly confirm errors.

[0007] To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent Internet of Things belt scale self-calibration real-time monitoring system based on multimodal AI, comprising:

[0008] A multi-terminal pressure data monitoring terminal that monitors the pressure data of the belt above the belt scale in real time and transmits the monitoring data associated with different monitoring nodes to the error verification analysis terminal in real time;

[0009] An error verification analysis terminal that verifies the monitoring data associated with different monitoring nodes in real time, locks the pressure data changes of the main monitoring node, and based on the data change characteristics, locks the verification time. Based on the pressure data change characteristics of the branch monitoring nodes at the verification time, it is confirmed whether there is a verification error in this belt scale. The specific method is as follows:

[0010] Based on the different pressure data associated with the main monitoring node at different times, generate its pressure data change curve, and select data peak points from the generated pressure data change curve, and record the time associated with the data peak points as the verification time;

[0011] Calibrate the pressure data associated with the verification time as Ys, and based on the current pressure data Ys, confirm the sinking distance JL of the belt position associated with the main monitoring node, where JL = Ys × C1, and C1 is a preset fixed coefficient factor;

[0012] Randomly confirm a set of pressure data associated with the branch monitoring nodes at the verification time, and calibrate the associated pressure data as YL;

[0013] Based on the horizontal distance L between the main monitoring node and the branch monitoring node, where L is a preset value, generate a set of right triangles based on the associated sinking distance JL and horizontal distance L. Take the side associated with the horizontal distance L as the base, the side associated with the sinking distance JL as the right angle side, and the other side as the hypotenuse, and confirm the associated included angle A between the right angle side and the hypotenuse. Use: Ys ÷ cosA = YX to confirm the stress data YX associated with the hypotenuse;

[0014] Then confirm whether the difference standard between the stress data YX and the pressure data YL meets: YX - YL ≥ Y1, where Y1 is a preset value. If not met, it means the error does not meet the standard, generate an error verification signal, and transmit the error verification signal to the feature analysis terminal; if met, it means the error meets the standard and no processing is required;

[0015] The feature analysis end, based on the generated error verification signal, checks the pressure data associated with the branch monitoring nodes on both sides of the verification moment, and generates a verification signal based on the verification result. The specific method is as follows:

[0016] Calibrate the two sets of pressure data associated with the branch monitoring nodes on both sides at the verification moment as YL1 and YL2 respectively, and evaluate the numerical difference between YL1 and YL2:

[0017] If YL1 = YL2, generate a main pressure sensor abnormal signal and directly display it;

[0018] If YL1 ≠ YL2, confirm the numerical magnitudes of the two. If YL1 > YL2, take the path direction from the branch monitoring node associated with YL2 to the branch monitoring node associated with YL1 as the moving direction, and transmit the confirmed moving direction into the virtual verification end. Conversely, confirm the opposite moving direction and transmit it;

[0019] The virtual verification end, based on the moving direction confirmed by the feature analysis end, performs a moving verification on the preset two-dimensional model, and based on the real-time moving verification process, determines its associated included angle in real time and locks the verification base point. The specific method is as follows:

[0020] Based on the preset two-dimensional model and the sinking distance JL, perform a sinking process on the position associated with the main monitoring node in the two-dimensional model, and the sinking distance value is the same as JL;

[0021] Perform a moving verification on the two-dimensional model that has completed the sinking process, confirm the specific position of the sinking point, and according to the confirmed moving direction, make the sinking point move horizontally according to the moving direction, and the sinking distance value remains unchanged, and confirm the associated included angles of the left branch monitoring node and the right branch monitoring node during the real-time moving process:

[0022] Taking the vertical direction of the sinking point as the reference direction, taking the line connecting the left monitoring node and the sinking point as the left side line, taking the line connecting the right monitoring node and the sinking point as the right side line, taking the included angle between the left side line and the reference direction as the included angle associated with the left monitoring node and denoted as J1, and taking the included angle between the right side line and the reference direction as the included angle associated with the right monitoring node and denoted as J2;

[0023] Based on the pressure data YL1 of the left monitoring node and the pressure data YL2 of the right monitoring node, based on the corresponding moving process, lock the moving process that satisfies: YL1 ÷ cosJ1 = YL2 ÷ cosJ2, then denote the sinking point associated with this moving process as the verification base point. If it has not been locked all the time, output an error signal for display, and transmit the confirmed verification base point into the correction data output end.

[0024] Preferably, pressure sensors associated therewith are provided at different monitoring nodes, and the monitoring nodes include a set of main monitoring nodes and two sets of branch nodes. The main monitoring nodes are arranged at the middle position at the lower end of the belt of the belt scale, and the branch nodes are arranged on both sides of the main monitoring nodes and symmetrically.

[0025] Preferably, the calibration data output end generates and outputs a calibration data packet based on the determined sinking point and the calibration reference point, with the direction from the calibration reference point to the sinking point as the calibration direction and the distance between the calibration reference point and the sinking point as the calibration distance.

[0026] The present invention provides an intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI. Compared with the prior art, it has the following beneficial effects:

[0027] By carefully checking the pressure data of the branch monitoring nodes on both sides, the present invention can accurately locate the root cause of the problem. Whether the main pressure sensor is abnormal or the position of the main monitoring node is shifted, the system can quickly generate corresponding signals or movement direction information, providing clear guidance for subsequent calibration work, greatly improving the efficiency and accuracy of fault diagnosis, and reducing the maintenance cost and time loss.

[0028] By means of a preset two-dimensional model for simulated movement calibration, in a complex working condition environment, by determining the associated angle in real time and locking the calibration reference point, it can provide an accurate position reference for the calibration work; this virtual calibration method is not only efficient, but also can deeply analyze and locate potential problems of the belt scale without affecting actual production, effectively improving the reliability and stability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 is a schematic diagram of the principle framework of the present invention;

[0030] Figure 2 is a schematic diagram of the two-dimensional model of the belt scale of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0032] The First Embodiment

[0033] Please refer to Figure 1, this application provides an intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI, including a multi-terminal pressure data monitoring terminal, an error verification and analysis terminal, a feature analysis terminal, a virtual verification terminal, and a calibration data output terminal. Among them, the multi-terminal pressure data monitoring terminal, the error verification and analysis terminal, the feature analysis terminal, the virtual verification terminal, and the calibration data output terminal are electrically connected in sequence from the output node to the input node;

[0034] Among them, the multi-terminal pressure data monitoring terminal monitors the pressure data of the belt above the belt scale in real time, and transmits the monitoring data associated with different monitoring nodes to the error verification and analysis terminal in real time. Pressure sensors are set at different monitoring nodes, and the monitoring nodes include a group of main monitoring nodes and two groups of branch nodes. The main monitoring node is set at the middle position at the lower end of the belt of the belt scale, and the branch nodes are set on both sides of the main monitoring node and symmetrically. When setting the pressure sensors, they are all preset by the operator at the designated positions below the belt scale. Generally, there is a main pressure sensor, which is set at the middle position below the belt scale, and two branch pressure sensors are set at the middle positions on both sides below the belt scale, generally symmetrically, for specific verification of the accuracy of the pressure value and evaluation of whether there are associated differences in the pressure data;

[0035] Among them, the error verification and analysis terminal verifies the monitoring data associated with different monitoring nodes in real time, locks the change of the pressure data of the main monitoring node, and based on the data change characteristics, locks the verification moment. Based on the pressure data change characteristics of the branch monitoring nodes at the verification moment, it is confirmed whether there is a verification error in this belt scale. Among them, the specific method for confirming the verification error is:

[0036] Based on the different pressure data associated with the main monitoring node at different times, generate its pressure data change curve, and select a data peak point from the generated pressure data change curve (the data peak point is the point of the maximum pressure value corresponding to the main monitoring node, that is, the specific situation where the above goods are located in the middle position of the belt scale), and record the moment associated with the data peak point as the verification moment. Calibrate the pressure data associated with the verification moment as Ys. Based on the current pressure data Ys, confirm the sinking distance JL of the belt position associated with the main monitoring node, where JL = Ys × C1, and C1 is a preset fixed coefficient factor, and its specific value is determined by the operator according to experience. The pressure data and the sinking distance are corresponding. The larger the sinking distance, the larger the associated pressure data, and the smaller the sinking distance, the smaller the associated pressure data. The sinking distance and the pressure data show a linear relationship;

[0037] Randomly confirm the pressure data associated with a group of branch monitoring nodes at the verification moment, and calibrate the associated pressure data as YL;

[0038] Based on the horizontal distance L associated with the main monitoring node and the branch monitoring node, this L is a preset value, which has been confirmed and preset in advance in the system according to the set positions of the main monitoring node and the branch monitoring node. Based on the associated subsidence distance JL and the horizontal distance L, a set of right triangles are generated. The side associated with the horizontal distance L is used as the base, the side associated with the subsidence distance JL is used as the right-angle side, and the other set of sides is used as the hypotenuse. The associated included angle A between the right-angle side and the hypotenuse is confirmed, and the following formula is used: Ys÷cosA = YX to confirm the stress data YX associated with the hypotenuse;

[0039] Then, it is confirmed whether the difference standard between the stress data YX and the pressure data YL meets: YX - YL ≥ Y1, where Y1 is a preset value determined in advance by the operator according to experience. If it is satisfied, it means the error meets the standard and no processing is required. If it is not satisfied, it means the error does not meet the standard, an error calibration signal is generated, and the error calibration signal is transmitted to the feature analysis end;

[0040] Specifically, when the error calibration signal exists, generally, the position of the monitoring sensor set corresponding to the main monitoring node is unreasonable, or the corresponding sensor has a calibration error. Under normal circumstances, the force data between the main monitoring node and the branch node should meet certain standards. When exceeding such standards, it means an error exists and calibration is required to ensure the real-time accuracy of the belt scale and avoid problems related to costs and material control caused by excessive errors;

[0041] Among them, the feature analysis end, based on the generated error calibration signal, checks the pressure data associated with the branch monitoring nodes on both sides at the calibration moment, and generates a check signal based on the check result. The specific method of checking is as follows:

[0042] The two sets of pressure data associated with the branch monitoring nodes on both sides at the calibration moment are respectively calibrated as YL1 and YL2, and the numerical difference between YL1 and YL2 is evaluated:

[0043] If YL1 = YL2, a main pressure sensor abnormal signal is generated and directly displayed. When the pressure data on both sides are the same, but there is a large error in the pressure data monitored by the main pressure monitoring node, then there is a corresponding pressure monitoring problem with the main pressure monitoring node, and it is necessary to display the signal in time and relevant personnel should take corresponding measures in time;

[0044] If YL1≠YL2, the numerical values ​​of the two are confirmed. If YL1>YL2, the path direction from the branch monitoring node associated with YL2 to the branch monitoring node associated with YL1 is the moving direction, and the confirmed moving direction is transmitted to the virtual verification end. Otherwise, the opposite moving direction is confirmed and transmitted. For example: if the branch monitoring node associated with YL1 is the left monitoring node, the branch monitoring node associated with YL2 is the right monitoring node. When YL1>YL2, it means that the gravity data associated with the left monitoring node is higher than that of the right monitoring node. Monitoring node, so the corresponding main monitoring node may be relatively to the left. In order to confirm the main position of its main monitoring node, it is necessary to move in the left direction and perform virtual verification to determine the corresponding specific position and perform real-time virtual verification. When the main monitoring node is to the left, then during actual monitoring, its left branch monitoring node is relatively close to the corresponding weighing object, then the pressure data generated will be larger. Therefore, in the subsequent verification process, the virtual verification adjustment method can be used to determine the actual offset distance and perform associated calibration.

[0045] Among them, the virtual verification end performs mobile verification on the preset two-dimensional model based on the moving direction confirmed by the feature analysis end, and determines its associated angle in real time based on the real-time mobile verification process, and locks the verification base point. The specific method of locking the verification base point is:

[0046] Based on the preset two-dimensional model and the sinking distance JL, the main monitoring node is sunk at the position associated with the two-dimensional model, and the sinking distance value is consistent with JL;

[0047] Perform mobile verification on the two-dimensional model that has completed the sinking process to confirm the specific location of the sinking point, and according to the confirmed moving direction, make the sinking point move horizontally according to the moving direction, and its sinking distance value remains unchanged. In addition, confirm the associated angles associated with the left branch monitoring node and the right branch monitoring node during the real-time movement:

[0048] The vertical direction of the sinking point is taken as the reference direction, the line between the left monitoring node and the sinking point is taken as the left line, the line between the right monitoring node and the sinking point is taken as the right line, the angle between the left line and the reference direction is taken as the angle associated with the left monitoring node and is recorded as J1, and the angle between the right line and the reference direction is taken as the angle associated with the right monitoring node and is recorded as J2;

[0049] Based on the pressure data YL1 of the left monitoring node and the pressure data YL2 of the right monitoring node, and based on the corresponding movement process, the movement process that satisfies: YL1÷cosJ1=YL2÷cosJ2 is locked, and the sinking point associated with this movement process is recorded as the verification base point. If it has not been locked, an error signal is output for display, and the confirmed verification base point is transmitted to the correction data output terminal.

[0050] Specifically, in combination with Figure 2 , inside the belt scale, the main monitoring node is set at the middle position of the belt scale, and its branch monitoring nodes are respectively set on both sides of the main monitoring node, used to monitor the pressure data on both sides, and based on the monitoring process, confirm the corresponding sinking distance.

[0051] Among them, the calibration data output end, based on the determined sinking point and the calibration reference point, takes the direction from the calibration reference point to the sinking point as the calibration direction, takes the distance between the calibration reference point and the sinking point as the calibration distance, generates a calibration data packet and outputs it. Subsequently, relevant external operators, according to this calibration data packet, perform relevant calibration on the main monitoring sensor in the actual application process.

[0052] Some of the data in the above formula are numerically calculated after removing their dimensions, and the content not described in detail in this specification belongs to the prior art well-known to those skilled in the art.

[0053] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent IoT belt scale self - calibration real - time monitoring system based on multi - modal AI, characterized in that, include: The multi-terminal pressure data monitoring terminal monitors the pressure data of the belt above the belt scale in real time, and transmits the monitoring data associated with different monitoring nodes to the error verification analysis terminal in real time; The error verification analysis end verifies the monitoring data associated with different monitoring nodes in real time, locks the pressure data changes of the main monitoring node, and locks the verification time based on the data change characteristics. Based on the pressure data change characteristics of the branch monitoring node at the verification time, it confirms whether there is a verification error for this belt scale; The characteristic analysis end verifies the pressure data associated with the branch monitoring nodes on both sides at the verification time based on the generated error verification signal, and generates a verification signal based on the verification result; The virtual verification end performs mobile verification on the preset two-dimensional model based on the moving direction confirmed by the feature analysis end, and determines its associated angle in real time and locks the verification base point based on the real-time mobile verification process.

2. The intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI according to claim 1, wherein, Associated pressure sensors are provided at different monitoring nodes, and the monitoring nodes include a group of main monitoring nodes and two groups of branch nodes. The main monitoring node is arranged at the middle position of the lower end of the belt of the belt scale, and the branch nodes are arranged on both sides of the main monitoring node and are symmetrically arranged.

3. The intelligent IoT belt scale self - calibration real - time monitoring system based on multi - modal AI according to claim 1, characterized in that, The specific method of locking the verification time at the error verification analysis end is: Based on the different pressure data associated with the main monitoring node at different times, a pressure data change curve is generated, and a data peak point is selected from the generated pressure data change curve, and the time associated with the data peak point is recorded as the verification time.

4. The intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI according to claim 3, characterized in that, The specific method of performing error checking at the error checking analysis end is as follows: The pressure data associated with the verification time is calibrated as Ys. Based on the current pressure data Ys, the sinking distance JL of the belt position associated with the main monitoring node is determined, where JL = Ys × C1, where C1 is a preset fixed coefficient factor; Randomly confirm the pressure data associated with a group of branch monitoring nodes at the verification time, and mark the associated pressure data as YL; Based on the horizontal distance L associated with the main monitoring node and the branch monitoring node, which L is a preset value, a set of right triangles is generated based on the associated sinking distance JL and the horizontal distance L, and the side associated with the horizontal distance L is used as the base, the side associated with the sinking distance JL is used as the right-angle side, and the other set of sides is used as the hypotenuse. The associated angle A between the right-angle side and the hypotenuse is confirmed, and the stress data YX associated with the hypotenuse is confirmed using: Ys÷cosA=YX; Then confirm whether the difference standard between the stress data YX and the pressure data YL meets the following conditions: YX-YL≥Y1, where Y1 is a preset value. If not, it means that the error does not meet the standard, and an error check signal is generated and transmitted to the feature analysis end.

5. The intelligent IoT belt scale self-calibration real-time monitoring system based on multimodal AI according to claim 4, characterized in that, If the stress data YX and the pressure data YL satisfy YX-YL≥Y1, it means that the error meets the standard and no processing is required.

6. The intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI according to claim 1, wherein, The characteristic analysis end verifies the pressure data associated with the branch monitoring nodes on both sides in the following specific manner: The two sets of pressure data associated with the branch monitoring nodes on both sides at the time of verification are calibrated as YL1 and YL2, and the numerical difference between YL1 and YL2 is evaluated: If YL1 = YL2, an abnormal signal of the main pressure sensor is generated and directly displayed; When YL1 ≠ YL2, the numerical magnitudes of the two are confirmed. If YL1 > YL2, the moving direction is the path direction from the branch monitoring node associated with YL2 to the branch monitoring node associated with YL1, and the confirmed moving direction is transmitted into the virtual verification end. Otherwise, the opposite moving direction is confirmed and transmitted.

7. The intelligent IoT belt scale self-calibration real-time monitoring system based on multimodal AI according to claim 6, characterized in that, The specific method for the virtual verification end to lock the verification base point is as follows: Based on the preset two-dimensional model and the sinking distance JL, the main monitoring node is subjected to a sinking process at the associated position within the two-dimensional model, and the sinking distance value is the same as JL; Perform a movement verification on the two-dimensional model after the sinking process, confirm the specific position of the sinking point, and according to the confirmed moving direction, make the sinking point move horizontally in the moving direction, with the sinking distance value remaining unchanged, and confirm the associated included angles of the left branch monitoring node and the right branch monitoring node during the real-time movement process: Taking the vertical direction of the sinking point as the reference direction, the connection line between the left monitoring node and the sinking point is used as the left side line, the connection line between the right monitoring node and the sinking point is used as the right side line, the included angle between the left side line and the reference direction is used as the included angle associated with the left monitoring node and denoted as J1, and the included angle between the right side line and the reference direction is used as the included angle associated with the right monitoring node and denoted as J2; Based on the pressure data YL1 of the left monitoring node and the pressure data YL2 of the right monitoring node, and based on the corresponding movement process, lock the movement process that satisfies YL1 ÷ cosJ1 = YL2 ÷ cosJ2. Then, the sinking point associated with this movement process is denoted as the verification base point. If it has not been locked all the time, an error signal is output for display, and the confirmed verification base point is transmitted into the correction data output end.

8. The intelligent IoT belt scale self-calibration real-time monitoring system based on multi-modal AI according to claim 7, characterized in that, The correction data output end, based on the determined sinking point and the verification base point, takes the direction from the verification base point to the sinking point as the correction direction, takes the distance between the verification base point and the sinking point as the correction distance, generates a correction data packet and outputs it.

Images