Automatic conveying posture adjusting method and system of raw material batching device and storage medium

Abstract

The invention relates to the technical field of conveying control, and discloses an automatic conveying posture adjusting method and system of a raw material batching device and a storage medium. The method comprises the steps of obtaining a conveying line real-time image based on a raw material batching instruction, and extracting a conveying target and a conveying line; calculating real-time attitude data of a conveying target and a conveying line label of a conveying line through an image recognition algorithm; matching the attitude reference data and the conveying direction data from a preset database according to the conveying line mark, and setting a first controlled part and a second controlled part of a conveying target according to the attitude reference data and the conveying direction data; calculating attitude deviation data according to the real-time attitude data and the attitude reference data by using an attitude comparison algorithm; first force control data and second force control data are obtained through a deviation control algorithm and the like, the stress of the conveying line on the two controlled parts is adjusted respectively, and the control trends of the two controlled parts are opposite. According to the method, posture deflection caused by the roller body problem in the conveying process of the batching frame can be automatically corrected, stable raw material transfer is guaranteed, and the conveying efficiency and the batching accuracy are improved.

Description

Technical Field

[0001] The present application relates to the technical field of conveying control, and in particular to a method, system and storage medium for automatically adjusting the conveying posture of a raw material batching device. Background Art

[0002] In traditional Chinese medicine production lines, automation of ingredient delivery has become an industry trend. Traditional manual batching methods are inefficient, subject to large errors, and unable to meet the demands of large-scale production. Currently, most companies have introduced automated conveying equipment, which achieves continuous transfer of raw materials through mechanical transmission.

[0003] Traditional Chinese medicine dried materials are typically loaded into a batching basket in a fixed quantity by a loading mechanism. The filled basket is then placed on conveyor rollers. The friction generated by the rollers' rotation transports the basket along a pre-set path, ultimately arriving at the processing equipment. From there, the basket is transferred to the next process step by a robotic arm or manually. Maintaining the stable position of the batching basket is crucial for ensuring accurate material transfer, preventing spillage, and ensuring smooth transitions to subsequent processes.

[0004] The conveyor rollers are driven by multiple independent rollers linked by chains or gears. If a section of the rollers becomes inconsistent in height due to wear, material jamming, or installation errors, especially when dried Chinese medicinal materials are light and don't press the ingredient frame against the conveyor rollers, the driving force applied to different parts of the ingredient frame's bottom will be uneven, easily causing the ingredient frame to skew. Over time, problems such as roller bearing wear and chain loosening can further exacerbate the synchronization deviation between the rollers, causing the ingredient frame to skew more significantly, affecting not only conveying efficiency but also the accuracy of ingredient distribution. Summary of the Invention

[0005] In order to make the posture of the batching frame on the conveying roller more stable, the present application provides a method, system and storage medium for automatically adjusting the conveying posture of a raw material batching device.

[0006] In a first aspect, the present application provides a method for automatically adjusting the conveying posture of a raw material batching device, which adopts the following technical solution: A method for automatically adjusting the conveying posture of a raw material batching device comprises the following steps: Based on the raw material batching instructions, obtain real-time images of the conveyor line; Extract the conveying target and conveying line from the real-time image; Calculate the real-time posture data of the conveying target and the conveying line mark of the conveying line through the image recognition algorithm; Match the posture reference data from the preset posture database according to the conveyor line mark; Match the conveying direction data from the preset conveying database according to the conveying line mark; Setting the two sides of the conveying target along the conveying direction as a first controlled part and a second controlled part according to the conveying direction data; Calculate the posture deviation data through the posture comparison algorithm based on the real-time posture data and the posture reference data; The first force control data is calculated based on the posture deviation data through the deviation control algorithm, and the force applied by the conveyor line to the first controlled part is adjusted based on the first force control data; and / or, the second force control data is calculated based on the posture deviation data through the posture force algorithm, and the force applied by the conveyor line to the second controlled part is adjusted based on the second force control data, and the control trend of the first force control data is opposite to the control trend of the second force control data.

[0007] By adopting the above technical solution, by acquiring the conveyor line image in real time, extracting the conveying target and conveyor line information, combining the preset database to match the reference data, and calculating the posture deviation, the force of the conveyor line on both sides of the batching frame is adjusted in a targeted manner. This can automatically correct the posture skew caused by roller problems, ensure the stability of raw material transfer, and improve the conveying efficiency and batching accuracy.

[0008] Optionally, the image recognition algorithm includes: Extract the contour line of the part of the conveying target close to the conveying line; The intersection center and intersection angle are calculated based on the contour line, and the intersection center and intersection angle correspond one to one; Extract the conveying edge line of the conveying line; fit the conveying line mark according to the conveying edge line; Select a pair of intersection centers and intersection angles that are closest to the conveying edge line; Calculate the offset distance between the conveying edge line and the center of the intersection; Calculate the offset data based on the offset distance and the preset reference distance; Calculate the posture angle between the conveying edge line and the median line of the intersection angle; Calculating an angle difference based on the posture angle and a preset first reference angle; The real-time attitude data is formed based on the angle difference and offset data and output.

[0009] By adopting the above technical solution, by extracting the key features of the conveying target and the conveying line, calculating the offset distance and posture angle to generate real-time posture data, the position offset and angle deviation of the conveying target can be accurately captured, thereby improving the accuracy of posture recognition.

[0010] Optionally, the image recognition algorithm includes: Extract the conveying edge line of the conveying line; Extract the conveying center line according to the conveying edge line; Extract the target contour of the conveying target on the conveying line; According to the relative position of the conveying center line and the target contour, the projection line of the conveying center line on the target contour surface is fitted; Calculating the volumes of the target contour on both sides of the projection line as the first volume and the second volume; calculating a volume difference between the first volume and the second volume; Calculating offset data based on the volume difference and a preset volume reference value; Extract the center line of the target contour, which is parallel to the conveying center line when the conveying target is in a standard posture; Calculate the posture angle between the conveying center line and the contour center line; Calculating an angle difference based on the posture angle and a preset second reference angle; The real-time attitude data is formed based on the angle difference and offset data and output.

[0011] By adopting the above technical solution, by analyzing the relative position relationship between the conveying centerline and the target contour, and combining the volume difference and posture angle to calculate real-time posture data, the posture deviation of the conveying target can be accurately identified from the perspective of overall shape and center of gravity distribution, thereby improving the accuracy of posture recognition.

[0012] Optionally, the pose comparison algorithm includes: Extract angle difference and offset data from real-time posture data; Extracting angle reference difference and offset reference data from attitude reference data; Calculating first deviation data according to the angle difference and the angle reference difference; Calculating second deviation data based on the offset data and the offset reference data; The posture deviation data is calculated according to the first deviation value and the second deviation value.

[0013] By adopting the above technical solution, the angle difference and offset data in the real-time posture data are compared and calculated with the reference data respectively, and then the comprehensive posture deviation data is obtained by weighted average. This can comprehensively consider the deviations in both angle and position, making the final deviation assessment more comprehensive and accurate.

[0014] Optionally, the deviation control algorithm includes: The difference between the calculated attitude deviation data and the preset deviation control data is the control difference; A control average value is obtained based on multiple control differences using a preset filtering algorithm; The control data is calculated based on the control average value and the preset correction reference value; If the control data is less than the preset first reference data, and the control data is used to adjust the first force-bearing part, the control data is used as the first force control data; If the control data is greater than the preset second reference data, and the control data is used to adjust the second force-bearing part, the control data is used as the second force control data; The second reference data is greater than the first reference data.

[0015] By adopting the above technical solution, the control difference is calculated and the control average value is obtained through filtering. The control data is determined in combination with the preset reference value, and differentiated thresholds are set according to different force-bearing parts. This can effectively filter out interference, smooth the control process, and achieve precise force adjustment for different parts.

[0016] Optionally, adjusting the force applied by the conveyor line to the first controlled portion and / or the second controlled portion comprises the following steps: A first controlled conveying roller is provided at a set position on the conveying line. The first controlled conveying roller includes two sets of active small rollers. The two sets of active small rollers are independently provided and are respectively connected to a driving assembly with independently controllable rotational speed. Initialize the speed of the two sets of active small rollers to the preset conveying speed of the conveyor line; When the force exerted by the conveyor line on the first controlled part is adjusted according to the first force control data, the rotation speed increment of the active small roller corresponding to the first controlled part is adjusted in a positive correlation with the first force control data to adjust the driving force exerted on the first controlled part; When the force exerted by the conveyor line on the second controlled part is adjusted according to the second force control data, the rotation speed increment of the active small roller corresponding to the second controlled part is adjusted in a positive correlation with the second force control data to adjust the driving force exerted on the second controlled part; The rotation speed adjustment trends of the active small roller corresponding to the first controlled part and the active small roller corresponding to the second controlled part are opposite.

[0017] By adopting the above technical solution, by setting up a first control conveying roller including two groups of active small rollers with independently controllable speeds, the conveying speed is used as a reference, and the speed increment of the small rollers at the corresponding position is adjusted in a positive correlation according to the force control data, and the adjustment trends on both sides are opposite. The driving force on both sides can be changed accurately and quickly through the speed difference, thereby realizing real-time and precise correction of the conveying target posture and ensuring conveying stability.

[0018] Optionally, adjusting the force applied by the conveyor line to the first controlled portion and / or the second controlled portion comprises the following steps: A second control conveying roller is provided at a set position on the conveying line. The second control conveying roller includes two sets of driven small rollers. The two sets of driven small rollers are independently provided and are respectively connected to an adjustment component that can be independently controlled to rise and fall. Initialize the height of the two sets of driven small rollers to be lower than the height of other conveyor rollers on the conveyor line; When the force exerted by the conveyor line on the first controlled part is adjusted according to the first force control data, the time for the driven small roller corresponding to the first controlled part to be raised to the set height is adjusted in a positive correlation with the first force control data to adjust the friction time on the first controlled part; When the force exerted by the conveyor line on the second controlled part is adjusted according to the second force control data, the time for the driven small roller corresponding to the second controlled part to be raised to the set height is adjusted in a positive correlation with the second force control data to adjust the friction time on the second controlled part; The rising time adjustment trends of the driven small roller corresponding to the first controlled portion and the driven small roller corresponding to the second controlled portion are opposite.

[0019] By adopting the above technical solution, a second controlled conveying roller is set up, which includes two groups of independently liftable driven small rollers. The initial height is lower than that of other rollers. The rising time of the small rollers at the corresponding positions is adjusted in a positive correlation according to the force control data, and the adjustment trends on both sides are opposite. By controlling the difference in friction action time, the force on both sides can be changed, thereby realizing flexible adjustment of the conveying target posture.

[0020] In a second aspect, the present application provides a system for automatically adjusting the conveying posture of a raw material batching device, which adopts the following technical solutions: A system for automatically adjusting the conveying posture of a raw material batching device comprises a processor, wherein the processor executes the steps of any one of the above-mentioned methods for automatically adjusting the conveying posture of a raw material batching device.

[0021] In a third aspect, the present application provides a storage medium that adopts the following technical solution: A storage medium stores a program, which, when executed by a processor, implements the steps of the method for automatically adjusting the conveying posture of a raw material batching device described above.

[0022] In summary, this application includes at least one of the following beneficial technical effects: By designing two image recognition algorithms, the system not only focuses on the critical contact points between the conveying target and the conveyor line, accurately capturing detailed features of positional offset and angular deviation, but also analyzes the center of gravity distribution and overall deflection trends based on the overall shape, combining volume differences and centerline angles. This dual recognition logic covers both local details and global shape, effectively improving the accuracy of posture data and providing a reliable basis for subsequent adjustments.

[0023] The posture comparison algorithm compares real-time angle differences and offset data with reference data, and then calculates a weighted average to generate comprehensive deviation data. This algorithm eliminates interference from different physical dimensions and, through weighted distribution, reflects the differential impact of angle and position deviations on the overall posture. This quantitative approach makes deviation assessment more aligned with actual production needs and avoids the bias of single-dimensional judgments.

[0024] The deviation control algorithm smooths transient interference through filtering and, combined with differentiated threshold judgment, enables targeted adjustment of different controlled parts. This not only filters out malfunctions caused by noise signals, but also ensures the accuracy of adjustment direction through directional differentiation. Simultaneously, two adjustment methods achieve force control through the speed difference of the active roller and the friction time difference of the driven roller, respectively. This balances rapid response and flexible adaptation. The appropriate method can be selected based on the characteristics of the conveying target, ensuring efficient and stable posture correction. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The present invention is a step diagram of a method for automatically adjusting the conveying posture of a raw material batching device.

[0026] Figure 2 This is a diagram of the steps of the first image recognition algorithm.

[0027] Figure 3 This is a step diagram of the second image recognition algorithm. DETAILED DESCRIPTION

[0028] Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings.

[0029] Throughout this specification, reference to the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0030] The present application discloses a method for automatically adjusting the conveying posture of a raw material batching device, referring to Figure 1 , including the following steps: Based on raw material batching instructions, real-time images of the conveyor line are captured. Upon receiving these instructions, image acquisition devices deployed at key locations along the conveyor line are triggered to activate. These devices can utilize industrial-grade high-definition cameras, clearly capturing the conveyor target: the overall shape of the batching racks containing the dried Chinese medicinal materials, as well as the surrounding conveyor line structure.

[0031] The captured images are preprocessed, including grayscale conversion, denoising filtering, and contrast enhancement, to eliminate image interference caused by factors such as ambient light changes and equipment vibration. After preprocessing, a deep learning-based image segmentation algorithm or a combination of traditional edge detection and morphological processing is used to extract the pixel regions of the conveying target and conveyor line. For the conveying target, the complete ingredient frame region is separated from the image background by identifying the characteristic contours of the ingredient frame (such as a rectangular border) and color identification (such as a preset distinguishing color). For the conveyor line, the entire conveyor line region composed of multiple rollers is extracted based on the arrangement pattern of the conveyor rollers (such as parallel distribution characteristics) and the reflective properties of the material.

[0032] The real-time posture data of the conveying target and the conveyor line's conveyor line mark are calculated using an image recognition algorithm. For the conveyor line, a straight-line fitting algorithm, such as the Hough transform, is used to process the extracted conveyor line area, fitting a reference line that represents the overall extension direction of the conveyor line, namely the conveyor line mark. This reference line can be quantitatively represented in the image coordinate system through its slope and intercept. For the conveying target, real-time posture data is generated by calculating parameters such as the geometric center coordinates and long-side angle of the batching frame: the geometric center coordinates are obtained by calculating the pixel center of gravity of the batching frame's outline area, and the long-side angle is calculated by fitting the slopes of the straight lines along the two long sides of the batching frame. These parameters are converted to actual physical parameters using a preset conversion relationship between pixel size and actual physical size (e.g., 1 pixel corresponds to 0.5mm) to accurately reflect the spatial posture of the batching frame.

[0033] The system matches posture reference data from a preset posture database based on the conveyor line mark. This database pre-stores standard posture parameters corresponding to different conveyor line marks. These parameters are based on the posture characteristics of the batching frame in an ideal conveying state (i.e., no skew and center aligned with the conveyor line). These parameters include a reference center offset (usually 0) and a reference long side angle (parallel to the conveyor line mark, with an angle difference of 0). By comparing the currently calculated conveyor line mark with the conveyor line mark characteristics stored in the database, the system automatically matches the corresponding posture reference data, which serves as a benchmark for determining whether the batching frame posture is abnormal.

[0034] The conveyor database records the actual conveying direction information corresponding to each conveyor line mark, including parameters such as the horizontal angle and tilt direction. Tilt directions, such as left and right, are determined by the slope of the conveyor line mark and preset direction definition rules, such as a positive slope corresponding to an upward and right tilt. By matching the conveying direction data, the system can clearly determine the forward path of the batching frame, providing a spatial orientation reference for the subsequent division of controlled areas.

[0035] For example, if the conveying direction is horizontal and to the right, the left edge of the batching frame is defined as the first controlled area, and the right edge is defined as the second controlled area. If the conveying direction is inclined upward, the corresponding areas on either side of the batching frame are defined as two controlled areas, based on the perpendicular line of the conveying direction. The specific range of the controlled areas is determined by a preset proportional parameter (such as 1 / 3 of the batching frame width) to ensure that it accurately reflects the force state on both sides of the batching frame.

[0036] The posture deviation data is calculated using a posture comparison algorithm based on the real-time posture data and the posture reference data. The posture comparison algorithm includes extracting the center offset and angle deviation values ​​from the real-time posture data, as well as the baseline offset and baseline angle values ​​from the posture reference data; calculating the difference between the two to obtain the actual offset deviation and actual angle deviation. The actual offset deviation is the difference between the real-time center offset and the baseline offset; the actual angle deviation is the difference between the real-time long side angle and the baseline angle. A preset weighted calculation model is used to combine these two deviation values ​​to obtain quantified posture deviation data. The magnitude of this data directly reflects the severity of the skew of the batching frame. The weighted calculation model may have an offset deviation weight of 0.4 and an angle deviation weight of 0.6.

[0037] Based on the posture deviation data, a deviation control algorithm calculates first force control data, and the force applied by the conveyor line to the first controlled portion is adjusted based on the first force control data. And / or, based on the posture deviation data, a posture force algorithm calculates second force control data, and the force applied by the conveyor line to the second controlled portion is adjusted based on the second force control data. The control trend of the first force control data is opposite to the control trend of the second force control data. Both the deviation control algorithm and the posture force algorithm are constructed based on a preset control model, such as a PID control model. The first force control data and the posture deviation data are positively correlated; that is, the greater the deviation, the greater the control data. The second force control data and the posture deviation data are negatively correlated; that is, the greater the deviation, the smaller the control data, or vice versa. For example, when the posture deviation data indicates that the batching frame is tilted to the left, the first force control data drives the conveyor line to increase the driving force on the first controlled portion (left side), while the second force control data drives the conveyor line to reduce the driving force on the second controlled portion (right side). The force difference between the two sides generates a corrective torque, gradually restoring the batching frame to its standard posture. When the batching frame tilts to the left, regardless of whether the initial friction forces on both sides are balanced, a rightward correcting torque can be formed by making the force on the left side greater than the right side or the force on the right side less than the left side; the increase in the force on the left side can push the left side forward faster, and the reduction in the force on the right side can slow down the forward speed of the right side relatively. The combination of the two causes the batching frame to rotate around the center of gravity and gradually return to the right side.

[0038] Through the above steps, the method can monitor the posture changes of the ingredient frame in real time, accurately identify skew deviations, and realize automatic deviation correction by differentially adjusting the forces on both sides, effectively solving the posture instability problem caused by conveying roller wear, installation errors, etc., and ensuring the stability of Chinese medicinal dry materials during the transportation process.

[0039] In this embodiment, the image recognition algorithm includes two implementation methods: Reference Figure 2 , the first image recognition algorithm: The contour of the target near the conveyor line is extracted. For the segmented target area in the image (i.e., the ingredient frame), the Canny edge detection algorithm is used to extract the contour of its bottom edge, where it contacts the conveyor line, to obtain a continuous set of edge pixels. To ensure the integrity of the contour, the extracted edge pixels are fitted and completed to eliminate contour breaks caused by image noise or partial occlusion, ultimately forming a closed contour line that accurately reflects the bottom shape of the target.

[0040] The intersection center and intersection angle are calculated based on the contour line, and the intersection center and intersection angle correspond one to one; the bottom contour line of the conveying target is usually a rectangle, corresponding to the bottom border of the ingredient box, and its contact area with the conveying line will form four corner points, that is, the turning points of the contour line. These four corner points are identified through corner detection algorithms, such as Harris corner detection, and the line segment between each two adjacent corner points is defined as a contour edge. The intersection of two adjacent contour edges is the corner point, that is, the intersection point. For each intersection point, its pixel coordinates in the image coordinate system are calculated as the intersection center, and the angle of the intersection point, that is, the intersection angle, is calculated by the difference in the slopes of the two adjacent contour edges. For example, the intersection angle of a rectangular border is theoretically 90°, but in reality, deviations may occur due to skewed posture.

[0041] Extract the conveyor edge lines of the conveyor line; fit the conveyor line mark based on the conveyor edge lines. For the conveyor line area segmented in the image, identify the edge contours on both sides, namely the two boundary lines formed by the conveyor roller arrangement. Fit the pixel points of the edge contours on both sides using a line fitting algorithm, such as the least squares method, to obtain two parallel or nearly parallel lines, namely the conveyor edge lines. Select one of the conveyor edge lines, usually the side closest to the center of the conveyor target, as the reference, or perform midline fitting on the two conveyor edge lines to obtain a reference line that can represent the overall extension direction of the conveyor line, which is the conveyor line mark.

[0042] Select a pair of intersection centers and intersection angles that are closest to the conveying edge line; calculate the vertical distances from the centers of all intersections to the conveying edge line, screen out the intersection with the smallest distance, and use the center coordinates of the intersection and its corresponding intersection angle as key analysis objects to focus on the direct contact area between the conveying target and the conveying line, thereby improving the pertinence of posture detection.

[0043] Calculate the offset distance between the conveyor edge and the intersection center. Based on the selected intersection center, calculate the perpendicular distance to the conveyor edge using the point-to-line distance formula (in pixels). Then, use a preset pixel-to-physical size conversion factor (e.g., 1 pixel = 0.3 mm) to convert this distance to the actual physical distance in mm, i.e., the offset distance. This distance directly reflects the degree of positional deviation of the conveyed target perpendicular to the conveying direction.

[0044] The offset data is calculated based on the offset distance and the preset reference distance. The preset reference distance is the theoretical distance from the center of the intersection to the conveyor edge when the conveying target is in a standard posture. For example, it is preset to 30mm. The offset data is the difference between the actual offset distance and the reference distance. Offset data = actual offset distance - reference distance. A positive offset data indicates that the conveying target is offsetting away from the center of the conveyor line; a negative offset data indicates that the target is offsetting towards the center of the conveyor line. The absolute value of the offset data reflects the degree of offset.

[0045] Calculate the posture angle between the conveying edge line and the median of the intersection angle. For the selected intersection angle, calculate its angle bisector, i.e., the direction vector of the median. This direction vector is synthesized by the direction vectors of the two intersecting contour edges. Then, use the vector angle formula to calculate the angle between this median and the conveying edge line, i.e., the posture angle. For example, in the standard posture, the median of the corner of the batching frame should be perpendicular to the conveying edge line. The theoretical value of the posture angle is 90°. This angle will deviate when the posture is skewed.

[0046] The angle difference is calculated based on the posture angle and a preset first reference angle. The preset first reference angle is the theoretical value of the posture angle in a standard posture, usually 90°. The angle difference is the difference between the actual posture angle and the first reference angle: angle difference = actual posture angle - first reference angle. A positive difference indicates that the conveyed object is deflected in a certain direction; a negative difference indicates deflection in the opposite direction. The absolute value of the difference reflects the severity of the angular deviation.

[0047] The real-time posture data is generated based on the angle difference and offset data and output. The angle difference (unit: degree) and offset data (unit: mm) are combined into real-time posture data and output. This data set fully reflects the posture characteristics of the conveyed target in terms of both angle and position.

[0048] The first image recognition algorithm focuses on the key contact points between the conveying target and the conveyor line, accurately extracts position offset and angle deviation features, and effectively improves the targeting and accuracy of posture recognition.

[0049] Reference Figure 3 , the second image recognition algorithm: Extract the conveyor line's conveyor edge lines. For the segmented conveyor line area in the image, a combination of edge detection and line fitting is used to identify the boundary contours on both sides of the conveyor line. These are the two parallel edge lines formed by the arrangement of multiple rollers, known as the conveyor edge lines. Least squares fitting is performed on the edge pixels to obtain two precise line parameters, such as slope and intercept, ensuring they accurately reflect the lateral boundary of the conveyor line.

[0050] The conveyor centerline is extracted from the conveyor edge lines. Using two conveyor edge lines as a reference, the geometric midline is calculated as the conveyor centerline. By solving the perpendicular distance between the two edge lines, the midpoint trajectory forms a straight line parallel to the edge lines. This line is the conveyor centerline, and its parameters are expressed as a straight line equation in the image coordinate system. As the reference line for the conveying direction, the conveyor centerline accurately reflects the extension direction and center position of the conveyor line.

[0051] Extract the target's outline on the conveyor line. For the identified target area in the image, use a contour extraction algorithm, such as a contour tree-based extraction method, to obtain its complete outer contour boundary, including the top, bottom, and side edges. Use morphological processing to remove noise points and redundant line segments from the outline to ensure that the target outline fully and clearly reflects the overall shape of the batching frame, providing a foundation for subsequent volume calculation and centerline extraction.

[0052] Based on the relative position of the conveyor centerline and the target contour, a projection line of the conveyor centerline on the target contour surface is fitted. The conveyor centerline is projected perpendicularly to the conveyor centerline onto the plane containing the target contour, covering the entire longitudinal length of the target contour, that is, its length along the conveying direction. By calculating the intersection of the conveyor centerline and the edge of the target contour, a straight line parallel to the conveyor centerline is fitted on the target contour surface, which is the projection line. This projection line serves as a reference for dividing the areas on both sides of the target contour. Its position relative to the conveyor centerline can reflect the overall deviation trend of the conveyed target.

[0053] The volumes of the target contour on either side of the projection line are calculated as the first and second volumes. Based on the 2D image features of the target contour and a preset conveying target height parameter, such as the standard height of the batching frame, the volumes of the areas on either side of the projection line are calculated using a three-dimensional geometric conversion algorithm. The 2D area of ​​the target contour on either side of the projection line is multiplied by the height parameter to obtain the first and second volumes. The first volume is the volume on one side of the projection line, and the second volume is the volume on the other side. During the volume calculation process, the image area is converted to actual area (unit: mm²) using a pixel-to-physical size conversion factor (e.g., 1 pixel corresponds to 0.5 mm) to ensure the physical accuracy of the volume parameters.

[0054] Calculate the volume difference between the first volume and the second volume; the volume difference is the difference between the first volume and the second volume, volume difference = first volume - second volume, and its positive or negative value reflects the offset direction of the center of gravity of the transport target relative to the projection line: a positive value indicates that the center of gravity is biased to one side of the projection line, and a negative value indicates that the center of gravity is biased to the other side. The absolute value reflects the degree of center of gravity offset.

[0055] The offset data is calculated based on the volume difference and the preset volume reference value; the preset volume reference value is the theoretical difference in volume on both sides of the projection line when the transport target is in a standard posture. The theoretical difference is usually 0, that is, the volumes on both sides are equal; the offset data is the ratio of the actual volume difference to the volume reference value, offset data = volume difference / volume reference value. Through normalization processing, the volume difference is converted into a quantitative indicator that can be directly used for posture analysis. The numerical value intuitively reflects the degree of position deviation of the transport target caused by the center of gravity offset.

[0056] Extract the target's contour centerline. By calculating the geometric center trajectory of the target's contour perpendicular to the conveying direction, a straight line is fitted through the target's contour, known as the contour centerline. When the target is in a standard posture, this contour centerline is parallel to the conveying centerline, and the two coincide horizontally or maintain a fixed distance.

[0057] Calculate the attitude angle between the conveying centerline and the contour centerline. Using the slope parameters of the two lines, calculate the acute angle between the conveying centerline and the contour centerline using the vector angle formula. This angle directly reflects the overall deflection of the conveying target: the angle is 0° in a standard attitude and increases as the attitude becomes skewed, with a maximum value of no more than 90°.

[0058] The angle difference is calculated based on the posture angle and the preset second reference angle; the preset second reference angle is the theoretical angle between the two center lines in the standard posture, and the theoretical angle is usually 0°. The angle difference is the difference between the actual posture angle and the second reference angle. The angle difference = actual posture angle - second reference angle, and its value is positively correlated with the degree of angular deviation of the conveying target.

[0059] Real-time posture data is generated and output based on the angle difference and offset data. The angle difference (unit: degrees) and offset data (dimensionless normalized values) are combined into real-time posture data and output. This data set comprehensively reflects the posture characteristics of the transported target from two dimensions: center of gravity distribution (offset data derived from volume difference) and overall deflection (angle difference derived from posture angle), providing a multi-dimensional quantitative basis for subsequent posture adjustments.

[0060] The second image recognition algorithm analyzes the overall shape and center of gravity distribution of the transported target and combines it with the centerline angle relationship to achieve posture quantification. It can effectively capture posture deviations caused by uneven loading or unbalanced force, significantly improving the comprehensiveness and accuracy of posture recognition, and providing reliable support for posture adjustment under complex working conditions.

[0061] The pose comparison algorithm includes: The angle difference and offset data are extracted from the real-time posture data. Real-time posture data contains two core parameters output by the image recognition algorithm: the angle difference and the offset data. The angle difference reflects the angular deviation of the transported object relative to the standard posture, expressed in degrees; the offset data reflects the positional offset of the transported object, expressed in millimeters or normalized values. The algorithm uses a data parsing module to separate these two parameters and use them as quantitative indicators of real-time posture characteristics. The positive or negative sign of the angle difference indicates the direction of deviation, while the positive or negative sign of the offset data indicates the direction of offset.

[0062] The angle reference difference and offset reference data are extracted from the posture reference data. The posture reference data are pre-stored standard posture parameters. The angle reference difference is the maximum allowable angular deviation threshold for the conveying target in the ideal posture, such as ±1°. Any deviation outside this range is considered an angular anomaly. The offset reference is the maximum allowable position deviation threshold under the ideal posture, such as ±5mm. Any deviation outside this range is considered a position anomaly. These two reference parameters are pre-set based on the accuracy requirements of the conveyor line and the dimensional characteristics of the batching frame and stored in the posture database. The algorithm calls the corresponding reference data by matching the current conveyor line mark.

[0063] The first deviation data is calculated based on the angle difference and the angle reference difference; the calculation formula is: First Deviation Data = Angle Difference ÷ Angle Reference Difference. This calculation converts the angle deviation into a proportional value relative to the allowable threshold. For example, when the real-time angle difference is 2° and the angle reference difference is 1°, the first deviation data is 2.0, indicating that the angle deviation has exceeded the allowable threshold by twice. If the angle difference is -0.5°, the first deviation data is -0.5, indicating that the angle deviation is within the allowable range and is biased in the negative direction. Through the division operation, the first deviation data achieves normalization of the angle deviation, facilitating horizontal comparison of deviations of different magnitudes.

[0064] The second deviation data is calculated based on the offset data and the offset reference data. The calculation formula is: Second Deviation Data = Offset Data ÷ Offset Reference Data. Consistent with the calculation logic for angular deviation, this formula converts the position deviation into a proportional value relative to the allowable threshold. For example, if the offset data is 8mm and the offset reference data is 5mm, the second deviation data is 1.6, indicating that the position deviation exceeds 60% of the allowable threshold. If the offset data is -3mm, the second deviation data is -0.6, indicating that the position deviation is within the allowable range and is biased in the negative direction. The division operation also normalizes the position deviation, giving the angle and position deviations a unified quantitative scale.

[0065] The posture deviation data is calculated based on the first and second deviation values. The specific steps are as follows: First, the first and second deviation values ​​are assigned preset weights, such as a weight of 0.5 for angle deviation and 0.5 for position deviation, or adjusted based on their impact on posture stability, such as a weight of 0.6 for angle deviation and 0.4 for position deviation. Then, the first deviation value is multiplied by the corresponding weight, and the second deviation value is multiplied by the corresponding weight. The two products are then summed to obtain the final posture deviation data: posture deviation data = first deviation value × weight 1 + second deviation value × weight 2. During the weighted averaging process, the positive and negative signs of the deviation values ​​are retained to reflect the direction of the deviation. For example, a positive deviation value indicates that the transport target is deflected or offset to one side, while a negative deviation indicates that it is offset in the opposite direction.

[0066] Through the above process, the posture comparison algorithm first normalizes the angle and position deviations to ratios relative to an allowable threshold through division, eliminating the influence of different physical dimensions. It then uses weighted averaging to achieve a comprehensive assessment of multi-dimensional deviations. This approach preserves the directional characteristics and relative severity of each deviation while also reflecting the varying impact of different deviations on the overall posture through weighted distribution.

[0067] Deviation control algorithms include: The difference between the calculated attitude deviation data and the preset deviation control data is called the control difference. The attitude deviation data is the comprehensive deviation indicator output by the attitude comparison algorithm. The comprehensive deviation indicator is signed and reflects the direction of the deviation. The preset deviation control data is the maximum deviation threshold allowed by the system and is typically set to 0, indicating no deviation under ideal conditions. The control difference is calculated as: Control difference = attitude deviation data - deviation control data. For example, if the attitude deviation data is 1.2 (positive deviation), the control difference is 1.2 - 0 = 1.2; if the attitude deviation data is -0.8 (negative deviation), the control difference is -0.8 - 0 = -0.8. This difference directly reflects the degree and direction of the current attitude deviation from the ideal state, providing the original error signal for subsequent control.

[0068] A preset filtering algorithm is used to calculate the control average based on multiple control differences. To eliminate transient deviation interference caused by factors such as image noise and equipment vibration, the algorithm filters multiple continuously collected control differences, such as those from the most recent 5-10 sampling periods. The preset filtering algorithm can use either a sliding average or a weighted moving average. The sliding average filter calculates the arithmetic mean of multiple control differences to calculate the control average, which is suitable for scenarios with low interference frequency. The weighted moving average filter assigns higher weight to recent control differences and lower weight to more distant values, better reflecting the dynamic trend of deviation. For example, using a 5-point sliding average, the control average = (nth control difference + n-1th + ... + n-4th) ÷ 5. Through filtering, the control average effectively smooths transient fluctuations and prevents actuator malfunctions caused by high-frequency interference.

[0069] The control data is calculated based on the control average value and the preset correction reference value; the preset correction reference value is the adjustment coefficient corresponding to the unit deviation preset by the system, such as 1.5N / unit deviation, which means that a correction force of 1.5N is required for each unit deviation. Its value is pre-calibrated according to the physical properties of the conveying target, such as the weight and friction coefficient. The calculation formula for the control data is: control data = control average value × correction reference value. For example, when the control average value is 1.2 and the correction reference value is 2N / unit, the control data = 1.2×2=2.4N, indicating that a correction force of 2.4N needs to be applied; if the control average value is -0.8, then the control data = -0.8×2=-1.6N, where the negative sign indicates that the direction of the correction force is opposite to the positive direction. This step converts the normalized deviation signal into a force control parameter with physical meaning.

[0070] If the control data is less than a preset first reference data and is used to adjust the first force-bearing location, the control data is used as the first force control data. If the control data is greater than a preset second reference data and is used to adjust the second force-bearing location, the control data is used as the second force control data. The second reference data is greater than the first reference data.

[0071] The first and second force control data are determined by threshold judgment. The preset first and second reference data serve as thresholds for distinguishing the adjustment targets, and the second reference data is greater than the first reference data. For example, the first reference data is set to -0.5N and the second reference data is set to 0.5N. Their values ​​are set according to the minimum corrective force requirement of the conveying target. The specific judgment logic is as follows: When the control data is less than the first reference data, such as -0.6N < -0.5N, and the control data is used to adjust the first controlled part (such as the left side of the batching frame), it is directly used as the first force control data, driving the actuator to apply the corresponding adjustment force to the first controlled part, such as increasing the driving force on the left side. When the control data is greater than the second reference data, such as 0.7N > 0.5N, and the control data is used to adjust the second controlled part (such as the right side of the batching frame), it is directly used as the second force control data, driving the actuator to apply the corresponding adjustment force to the second controlled part, such as increasing the driving force on the right side. If the control data is between the first reference data and the second reference data, such as -0.3N, the deviation is determined to be within the allowable range and the control data is temporarily not output to avoid posture oscillation caused by frequent adjustments.

[0072] Through the above process, the deviation control algorithm first obtains the original error signal through control difference calculation, then obtains a smooth control average value through filtering. This is then converted into quantized control data in combination with the correction reference value. Finally, differentiated threshold judgment is used to achieve precise adjustment of different controlled parts. This design not only effectively filters out interference and ensures the stability of the control process, but also enables targeted adjustment through direction differentiation and threshold setting, ensuring the accuracy and efficiency of posture correction.

[0073] In this embodiment, the force applied by the conveyor line to the first controlled portion and / or the second controlled portion is adjusted by the following two implementation methods: The first adjustment method: A first control conveyor roller is set at a set position on the conveyor line. The first control conveyor roller includes two groups of active small rollers. The two groups of active small rollers are independently set and are respectively connected to a drive component that can control the speed independently. The two groups of active small rollers are arranged in parallel along the horizontal direction of the conveyor line. The axis of each group of small rollers is perpendicular to the conveying direction, and a preset spacing is maintained between the two groups of small rollers, such as to adapt to the width of the ingredient frame, to ensure that they can act on the bottom of both sides of the ingredient frame respectively. Each group of active small rollers is independently connected to a drive component, such as a servo motor and a reduction gear set. The drive component can receive a control signal and independently adjust the speed of the corresponding small roller. There is no mechanical linkage between the two groups of drive components to ensure the independence of the speed adjustment.

[0074] Initialize the speed of the two sets of active rollers to the preset conveyor speed of the conveyor line. Set the initial speed of both sets of rollers to the preset conveyor speed of the conveyor line, such as 100 rpm. This speed is consistent with the speed of other common conveyor rollers on the conveyor line, ensuring that the conveyed object can be transported at a constant speed along the preset path without any posture deviation. After initialization, the drive assembly is in a standby state, waiting for the input of force control data in real time.

[0075] When the force applied to the first controlled portion by the conveyor line is adjusted based on the first force control data, the speed increment of the active roller corresponding to the first controlled portion is adjusted in a positive correlation with the first force control data to adjust the driving force applied to the first controlled portion. When the system outputs the first force control data for adjusting the first controlled portion, the speed increment of the active roller corresponding to the first controlled portion is adjusted in a positive correlation with the first force control data. Specifically, the magnitude of the first force control data is directly proportional to the speed increment. If the first force control data is positive and its absolute value increases, it indicates that the driving force applied to the first controlled portion needs to be increased. In this case, the drive assembly of the corresponding active roller is controlled to increase its speed, and the speed increment increases with the increase in the force control data. For example, for every 1N increase in the force control data, the speed increment increases by 5r / min. If the first force control data is negative, it indicates that the driving force applied to the first controlled portion needs to be reduced. In this case, the drive assembly reduces the speed of the corresponding roller, and the speed increment is negative and its absolute value increases with the increase in the absolute value of the force control data. By changing the rotational speed, the friction force (driving force) between the active roller and the bottom of the conveying target is changed, thereby achieving precise adjustment of the force on the first controlled part.

[0076] When the force applied to the second controlled portion by the conveyor line is adjusted based on the second force control data, the speed increment of the active roller below the second controlled portion is adjusted in direct correlation with the second force control data to adjust the driving force applied to the second controlled portion. When the system outputs the second force control data for adjusting the second controlled portion, the speed increment of the active roller below the second controlled portion is adjusted in direct correlation with the second force control data. This adjustment logic is consistent with that for the first controlled portion: the value of the second force control data is directly proportional to the speed increment of the corresponding roller, and the driving force applied to the second controlled portion is altered by changes in speed.

[0077] The speed adjustment trends of the active rollers corresponding to the first controlled area and the second controlled area are opposite. For example, if the conveying target is tilting to the left and needs to be corrected to the right, the first force control data is positive, and the left driving force needs to be increased, driving the speed of the active roller corresponding to the first controlled area to increase, for example, from 100r / min to 110r / min. At the same time, the second force control data is negative, and the right driving force needs to be reduced, driving the speed of the active roller corresponding to the second controlled area to decrease, for example, from 100r / min to 90r / min. Conversely, if the conveying target is tilting to the right, the speed of the left roller decreases while the speed of the right roller increases. The speed difference between the two sets of rollers creates a clear direction of correction torque, pushing the conveying target to gradually return to its standard posture.

[0078] Through the above process, the first adjustment method utilizes two sets of independently controlled active rollers. Based on the preset conveying speed, the force control data is adjusted in a positive correlation with the speed increment to achieve differential control of the driving force on both sides, with opposite adjustment trends. This design not only achieves real-time deviation correction through rapid speed response, but also ensures accurate posture adjustment through precise control of the speed difference. This effectively solves the problem of uneven force caused by roller synchronization deviation and ensures that the conveying target maintains a stable posture throughout the entire conveying process.

[0079] The second adjustment method involves installing a second controlled conveyor roller at a set position on the conveyor line. The second controlled conveyor roller consists of two sets of independent driven rollers, each connected to an adjustment component that can be independently controlled to raise and lower the rollers. The two sets correspond to the first and second controlled locations of the conveying target, respectively. The two sets of driven rollers are arranged horizontally and parallel to the conveyor line, with their axes perpendicular to the conveying direction and their spacing aligned with the width of the conveying target (such as a batching frame) to ensure they can act on the bottom of each side of the target. Each set of driven rollers is independently connected to an adjustment component, such as an electric push rod or a pneumatic cylinder-driven lifting mechanism. This component receives control signals and independently adjusts the height and duration of the corresponding roller's lift. The two adjustment components are not mechanically linked, ensuring independent lifting control.

[0080] Initialize the height of the two sets of driven rollers lower than the other conveyor rollers on the conveyor line; set the initial height of both sets of rollers to be lower than the height of other common conveyor rollers on the conveyor line, for example, 5-10mm lower. This prevents the driven rollers from contacting the bottom of the conveyed object, preventing additional friction interference during the conveying process when there is no posture deviation, and ensures that the conveyed object is supported only by the other conveyor rollers and maintains a constant speed.

[0081] When the force applied to the first controlled portion of the conveyor line is adjusted based on the first force control data, the time it takes for the driven roller below the first controlled portion to rise to a set height is adjusted in direct correlation with the first force control data to adjust the friction duration of the first controlled portion. When the system outputs the first force control data for adjusting the first controlled portion, the time it takes for the driven roller below the first controlled portion to rise to a set height is adjusted in direct correlation with the first force control data. The set height is flush with the top surfaces of the other conveyor rollers, ensuring that the roller can contact and generate friction with the bottom of the conveyed object. The rise time refers to the duration it takes for the roller to rise from its initial low position to the set height and maintain contact. The specific adjustment logic is as follows: the numerical value of the first force control data is directly proportional to the lifting time. If the first force control data is positive and its absolute value increases, it indicates that the friction force on the first controlled part needs to be increased. In this case, the corresponding adjustment component is controlled to extend the lifting time. For example, for every 1N increase in the force control data, the lifting time increases by 0.5 seconds. If the first force control data is negative, it indicates that the friction force needs to be reduced. In this case, the lifting time is shortened and can be zero, that is, the small roller remains in the low position without contact. By changing the contact time between the driven small roller and the conveying target, the cumulative effect of the friction force is adjusted, and flexible adjustment of the force on the first controlled part is achieved.

[0082] When the force applied to the second controlled portion by the conveyor line is adjusted based on the second force control data, the time it takes for the driven roller below the second controlled portion to rise to the set height is adjusted in a positive correlation with the second force control data to adjust the friction duration on the second controlled portion. When the system outputs the second force control data for adjusting the second controlled portion, the time it takes for the driven roller below the second controlled portion to rise to the set height is adjusted in a positive correlation with the second force control data. The adjustment logic is consistent with that for the first controlled portion: the value of the second force control data is directly proportional to the time it takes for the corresponding driven roller to rise, and the cumulative friction effect on the second controlled portion is altered by varying the contact duration.

[0083] The rise time adjustment trends of the driven rollers corresponding to the first controlled area and the second controlled area are opposite. For example, when the conveying target is tilted to the left and needs to be corrected to the right, the first force control data is positive, increasing the left friction effect, driving the driven roller corresponding to the first controlled area to extend the rise time, such as from 0.3 seconds to 0.8 seconds. At the same time, the second force control data is negative, reducing the right friction effect, driving the driven roller corresponding to the second controlled area to shorten the rise time, such as from 0.3 seconds to 0 seconds. Conversely, when the conveying target is tilted to the right, the left roller shortens the rise time while the right roller prolongs it. The difference in contact time between the two sets of rollers creates an unbalanced friction force on both sides, generating a steady corrective torque that gradually pushes the conveying target back to the right.

[0084] Through the above process, the second adjustment method utilizes two sets of independently rising and falling driven rollers, using the initial low position as a reference. Through the positive correlation between force control data and rise time, differential control of the duration of friction force on both sides is achieved, with opposite adjustment trends. This design not only avoids interference when no adjustment is required through the non-contact initial state, but also achieves flexible correction through precise control of contact duration. This effectively reduces the possibility of conveying target shaking or material spillage caused by rigid adjustment. It is particularly suitable for conveying posture adjustment scenarios for fragile materials such as traditional Chinese medicine raw materials, ensuring adjustment accuracy while improving the stability of the conveying process.

[0085] An embodiment of the present application further discloses a system for automatically adjusting the conveying posture of a raw material batching device, comprising a processor, wherein the processor executes the steps of the method for automatically adjusting the conveying posture of a raw material batching device as described in any one of the above.

[0086] An embodiment of the present application further discloses a storage medium, in which a program is stored. When the program is executed by a processor, the steps of the method for automatically adjusting the conveying posture of the raw material batching device described above are implemented.

[0087] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for automatically adjusting the conveying posture of a raw material batching device, characterized in that: The steps include: Based on the raw material batching instructions, obtain real-time images of the conveyor line; Extract the conveying target and conveying line from the real-time image; Calculate the real-time posture data of the conveying target and the conveying line mark of the conveying line through the image recognition algorithm; Match the posture reference data from the preset posture database according to the conveyor line mark; Match the conveying direction data from the preset conveying database according to the conveying line mark; Setting the two sides of the transport target along the transport direction as a first controlled part and a second controlled part according to the transport direction data; Calculate the posture deviation data through the posture comparison algorithm based on the real-time posture data and the posture reference data; The first force control data is calculated based on the posture deviation data through the deviation control algorithm, and the force applied by the conveyor line to the first controlled part is adjusted based on the first force control data; and / or, the second force control data is calculated based on the posture deviation data through the posture force algorithm, and the force applied by the conveyor line to the second controlled part is adjusted based on the second force control data, and the control trend of the first force control data is opposite to the control trend of the second force control data.

2. The method for automatically adjusting the conveying posture of the raw material batching device according to claim 1, characterized in that: Image recognition algorithms include: Extract the contour line of the part of the conveying target close to the conveying line; The intersection center and intersection angle are calculated based on the contour line, and the intersection center and intersection angle correspond one to one; Extract the conveying edge line of the conveying line; fit the conveying line mark according to the conveying edge line; Select a pair of intersection centers and intersection angles that are closest to the conveying edge line; Calculate the offset distance between the conveying edge line and the center of the intersection; Calculate the offset data based on the offset distance and the preset reference distance; Calculate the posture angle between the conveying edge line and the median line of the intersection angle; Calculating an angle difference based on the posture angle and a preset first reference angle; The real-time attitude data is formed based on the angle difference and offset data and output.

3. The method for automatically adjusting the conveying posture of a raw material batching device according to claim 1, characterized in that: Image recognition algorithms include: Extract the conveying edge line of the conveying line; Extract the conveying center line according to the conveying edge line; Extract the target contour of the conveying target on the conveying line; According to the relative position of the conveying center line and the target contour, the projection line of the conveying center line on the target contour surface is fitted; Calculating the volumes of the target contour on both sides of the projection line as the first volume and the second volume; calculating a volume difference between the first volume and the second volume; Calculating offset data based on the volume difference and a preset volume reference value; Extract the center line of the target contour, which is parallel to the conveying center line when the conveying target is in a standard posture; Calculate the posture angle between the conveying center line and the contour center line; Calculating an angle difference based on the posture angle and a preset second reference angle; The real-time attitude data is formed based on the angle difference and offset data and output.

4. The method for automatically adjusting the conveying posture of a raw material batching device according to claim 2 or 3, characterized in that: The pose comparison algorithm includes: Extract angle difference and offset data from real-time posture data; Extracting angle reference difference and offset reference data from attitude reference data; Calculating first deviation data according to the angle difference and the angle reference difference; Calculating second deviation data based on the offset data and the offset reference data; The posture deviation data is calculated according to the first deviation value and the second deviation value.

5. The method for automatically adjusting the conveying posture of the raw material batching device according to claim 4, characterized in that: Deviation control algorithms include: The difference between the calculated attitude deviation data and the preset deviation control data is the control difference; A control average value is obtained based on multiple control differences using a preset filtering algorithm; The control data is calculated based on the control average value and the preset correction reference value; If the control data is less than the preset first reference data, and the control data is used to adjust the first force-bearing part, the control data is used as the first force control data; If the control data is greater than the preset second reference data, and the control data is used to adjust the second force-bearing part, the control data is used as the second force control data; The second reference data is greater than the first reference data.

6. The method for automatically adjusting the conveying posture of a raw material batching device according to claim 5, characterized in that: Adjusting the force exerted by the conveyor line on the first controlled part and / or the second controlled part includes the following steps: A first controlled conveying roller is provided at a set position on the conveying line. The first controlled conveying roller includes two sets of active small rollers. The two sets of active small rollers are independently provided and are respectively connected to a driving assembly with independently controllable rotational speed. Initialize the speed of the two sets of active small rollers to the preset conveying speed of the conveyor line; When the force exerted by the conveyor line on the first controlled part is adjusted according to the first force control data, the rotation speed increment of the active small roller corresponding to the first controlled part is adjusted in a positive correlation with the first force control data to adjust the driving force exerted on the first controlled part; When the force exerted by the conveyor line on the second controlled part is adjusted according to the second force control data, the rotation speed increment of the active small roller corresponding to the second controlled part is adjusted in a positive correlation with the second force control data to adjust the driving force exerted on the second controlled part; The rotation speed adjustment trends of the active small roller corresponding to the first controlled part and the active small roller corresponding to the second controlled part are opposite.

7. The method for automatically adjusting the conveying posture of a raw material batching device according to claim 5, characterized in that: Adjusting the force applied by the conveyor line to the first controlled portion and / or the second controlled portion comprises the following steps: A second control conveying roller is provided at a set position on the conveying line. The second control conveying roller includes two sets of driven small rollers. The two sets of driven small rollers are independently provided and are respectively connected to an adjustment component that can be independently controlled to rise and fall. Initialize the height of the two sets of driven small rollers to be lower than the height of other conveyor rollers on the conveyor line; When the force exerted by the conveyor line on the first controlled part is adjusted according to the first force control data, the time for the driven small roller corresponding to the first controlled part to be raised to the set height is adjusted in a positive correlation with the first force control data to adjust the friction time on the first controlled part; When the force exerted by the conveyor line on the second controlled part is adjusted according to the second force control data, the time for the driven small roller corresponding to the second controlled part to be raised to the set height is adjusted in a positive correlation with the second force control data to adjust the friction time on the second controlled part; The rising time adjustment trends of the driven small roller corresponding to the first controlled portion and the driven small roller corresponding to the second controlled portion are opposite.

8. A conveying posture automatic adjustment system for a raw material batching device, characterized in that: The method comprises a processor, wherein the processor executes the steps of the method for automatically adjusting the conveying posture of the raw material batching device according to any one of claims 1 to 7.

9. A storage medium, characterized in that: The storage medium stores a program, and when the program is executed by the processor, the steps of the method for automatically adjusting the conveying posture of the raw material batching device according to any one of claims 1 to 7 are implemented.

Images