A segmented corn cleaning screen based on bionic structure surface and its intelligent control system

Abstract

A segmented corn cleaning screen based on a bionic structure surface and an intelligent control system thereof, comprising a high-definition camera, a left-moving screen surface a, a right-moving screen surface a, a left-moving screen surface b, a right-moving screen surface b, a left-moving screen surface c, a right-moving screen surface c, a left-moving screen surface d, a right-moving screen surface d, a screen cover, a screen frame, and a push-pull electromagnet. During operation, the high-definition camera shoots the distribution of materials on the screen and transmits the images to a pulse coupled neural network (PCNN) for processing and analysis. In combination with the analysis results, the push-pull electromagnet is used to control the left and right movement of each segment of the screen. The microscopic scale structure of the butterfly wing surface on each segment of the screen is used to make the materials on the screen move in an expected movement form, so as to realize the lateral uniform distribution of the materials on the screen. The invention has the characteristics of simple, reliable and reasonable structure, high cleaning efficiency and good cleaning quality for corn kernels, high intelligence, uniform lateral distribution of materials on the screen, high utilization rate of the screen surface, and the like.

Description

Technical Field

[0001] The invention relates to a segmented corn cleaning screen based on a bionic structure surface and an intelligent control system thereof, belonging to the technical field of agricultural machinery. Background Art

[0002] When operating in the field, the corn grain harvester is affected by the field terrain, and there are large bumps and vibrations, which leads to poor uniformity of material feeding before the screen in the cleaning device. This phenomenon is particularly obvious under conditions of large feed volume operation, causing the material on the screen to accumulate on one side of the screen surface. The screen surface of the existing cleaning device is mostly flat reciprocating motion, lacking lateral motion, and it is difficult to disperse the material accumulated on one side. The screen surface utilization rate is low, affecting the screening efficiency. At the same time, the existing corn kernel cleaning device has a low degree of intelligence, and is unable to reasonably and accurately control the screen surface movement parameters according to the material distribution in each section of the screen surface, thereby reducing the operating quality of the cleaning device. Summary of the invention

[0003] The purpose of the invention is to provide a segmented corn cleaning screen based on a bionic structure surface and its intelligent control system, which utilizes the characteristic that each segment of the screen surface has different friction coefficients in the lateral direction, cooperates with the intelligent control system to control the lateral movement intensity of each segment of the screen surface, so as to achieve uniform lateral distribution of the material on the screen, increase the utilization rate of the screen surface, and improve the screening efficiency and screening quality.

[0004] The purpose of the invention can be achieved through the following scheme: a segmented corn cleaning screen based on a bionic structure surface and its intelligent control system, including a high-definition camera, a left-moving screen surface a, a right-moving screen surface a, a left-moving screen surface b, a right-moving screen surface b, a left-moving screen surface c, a right-moving screen surface c, a left-moving screen surface d, a right-moving screen surface d, a screen cover, a screen frame, and a push-pull electromagnet. It is characterized in that the left-moving screen surface a, the right-moving screen surface a, the left-moving screen surface b, the right-moving screen surface b, the left-moving screen surface c, the right-moving screen surface c, the left-moving screen surface d, and the right-moving screen surface d are all black, and can be laterally slidably installed in the groove between the screen cover and the screen frame from front to back. 8 push-pull electromagnets are fixed on the right side of the screen frame, and their positions correspond to the 8 screen pieces and are respectively fixed together with the corresponding screen pieces, which can drive the corresponding screen pieces to move laterally in the groove between the screen cover and the screen frame. The movement of each push-pull electromagnet is controlled by an independent external control system, and circular holes are opened at the front and back of the screen frame for connecting with an external drive mechanism. The surfaces of the left-moving screen surface a, left-moving screen surface b, left-moving screen surface c, and left-moving screen surface d are right-angled triangle sawtooth structures imitating the microscopic scale structure on the surface of butterfly wings. One of the two right-angled sides of the right-angled triangle structure is parallel to the screen surface, and the other is toward the left side of the screen body, so that the friction coefficient to the right on each screen surface is greater than the friction coefficient to the left; the surfaces of the right-moving screen surface a, right-moving screen surface b, right-moving screen surface c, and right-moving screen surface d are right-angled triangle sawtooth structures imitating the microscopic scale structure on the surface of butterfly wings. One of the two right-angled sides of the right-angled triangle structure is parallel to the screen surface, and the other is toward the right side of the screen body, so that the friction coefficient to the left on each screen surface is greater than the friction coefficient to the right. The high-definition camera is fixed above the screen surface to capture the position of the complete screen surface and connect to the data processing system through a data line to transmit the captured images in real time. The pulse coupled neural network (PCNN) processing technology is used to process the image, identify the material distribution of each section of the screen surface, and control the push-pull electromagnet to drive the screen surface to move horizontally according to demand.

[0005] The beneficial effects created by the present invention are: through image processing technology and intelligent control technology, according to the lateral distribution of materials on the cleaning screen, the screen plates with bionic structure surfaces are controlled to make different lateral movements, so as to accurately control the lateral distribution of materials on the screen. The present invention has the characteristics of simple and reliable structure, high screen surface utilization rate, high operating efficiency, good operating quality, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0006] Figure 1 It is a schematic diagram of the overall structure of the present invention.

[0007] Figure 2 yes Figure 1 Schematic diagram of the structure after removing the screen cover.

[0008] Figure 3 yes Figure 1 Side view of the left-shifted screen surfaces a, b, c, and d.

[0009] Figure 4 yes Figure 1 Side view of the right-moving screen surfaces a, b, c, and d.

[0010] Figure 5 It is a schematic diagram of the working process of the intelligent control system.

[0011] Explanation of the part numbers in the figure: 1. HD camera, 2. Left-moving screen surface a, 3. Right-moving screen surface a, 4. Left-moving screen surface b, 5. Right-moving screen surface b, 6. Left-moving screen surface c, 7. Right-moving screen surface c, 8. Left-moving screen surface d, 9. Right-moving screen surface d, 10. Screen cover, 11. Screen frame, 12. Push-pull electromagnet, 13. Corn extract. DETAILED DESCRIPTION

[0012] The invention is further described below with reference to the accompanying drawings and examples. A segmented corn cleaning screen based on a bionic structure surface and its intelligent control system, comprising a high-definition camera 1, a left-moving screen surface a 2, a right-moving screen surface a 3, a left-moving screen surface b 4, a right-moving screen surface b 5, a left-moving screen surface c 6, a right-moving screen surface c 7, a left-moving screen surface d 8, a right-moving screen surface d 9, a screen cover 10, a screen frame 11, and a push-pull electromagnet 12. The invention is characterized in that the left-moving screen surface a 2, the right-moving screen surface a 3, the left-moving screen surface b 4, the right-moving screen surface b 5, the left-moving screen surface c 6, the right-moving screen surface c 7, the left-moving screen surface d 8, the right-moving screen surface d 9 are all black and can be installed in the groove between the sieve cover 10 and the sieve frame 11 in a transversely slidable manner from front to back. Eight push-pull electromagnets 12 are fixedly installed on the right side of the sieve frame 11. Their positions correspond to the eight sieve sheets and are fixedly installed with their corresponding sieve sheets, respectively, so as to drive the corresponding sieve sheets to move transversely in the groove between the sieve cover 10 and the sieve frame 11. The movement of each push-pull electromagnet 12 is controlled by an independent external control system. Round holes are opened at the front and rear of the sieve frame 11 for connecting with an external driving mechanism. The surfaces of left-moving screen surface a 2, left-moving screen surface b 4, left-moving screen surface c 6, and left-moving screen surface d 8 are right-angled triangle serrated structures imitating the microscopic scale structure on the surface of butterfly wings. One of the two right-angled sides of the right-angled triangle structure is toward the screen surface and is parallel to the screen surface, and the other is toward the left side of the screen body, so that the friction coefficient to the right on each screen surface is greater than the friction coefficient to the left; the surfaces of right-moving screen surface a 3, right-moving screen surface b 5, right-moving screen surface c 7, and right-moving screen surface d 9 are right-angled triangle serrated structures imitating the microscopic scale structure on the surface of butterfly wings. One of the two right-angled sides of the right-angled triangle structure is toward the screen surface and is parallel to the screen surface, and the other is toward the right side of the screen body, so that the friction coefficient to the left on each screen surface is greater than the friction coefficient to the right. The high-definition camera 1 is fixed above the screen surface and can capture the position of the entire screen surface and connect to the data processing system through a data cable to transmit the captured images in real time. The pulse coupled neural network (PCNN) processing technology is used to process the image, identify the material distribution of each section of the screen surface, and control the push-pull electromagnet 12 to drive the screen surface to move horizontally according to demand.

[0013] When the cleaning operation is performed, the circular holes at the front and rear ends of the screen frame 11 are connected to the driving mechanism of the harvester cleaning device (not shown in the figure), driving the cleaning screen as a whole to make a planar reciprocating motion. After the material is fed into the cleaning device, a high-definition camera 1 fixed above the screen surface is used to capture a panoramic image of the screen surface when it is working, and the image is transmitted to the pulse coupled neural network (PCNN). Each PCNN neuron is mainly composed of two functional units, namely a feedback input unit and a coupling connection input unit. The coupling connection input functional unit of the neuron is the linear weighted sum of the adjacent neurons, and its synaptic connection weight is W. The mathematical operation formula is:

[0014]

[0015] Where: L ij ——The coupled connection input of the (i,j)th neuron;

[0016] V L - the gain factor of the coupled input unit;

[0017] α L ——L ij The decay time constant of

[0018] Y kl ——The output of the neuron at (n-1) iterations.

[0019] The linear weighted sum of neighboring neurons and the external stimulus S, i.e., the pixel gray value, which is forced to excite the neuron, constitute the feedback input functional unit of the neuron. Its synaptic connection weight is M, and the mathematical operation formula is:

[0020]

[0021] Where F ij ——Feedback input of the (i,j)th neuron;

[0022] V F ——Amplification factor of the feedback input unit;

[0023] α F ——F ij The decay time constant.

[0024] The connection weight coefficient matrices M and W indicate the degree to which the central neuron is affected by the surrounding neurons, reflecting the strength of the information transmitted by the neighboring neurons to the central neuron. The two functional units M and W also need to perform appropriate number of iterations and decay according to the exponential law during the iteration process. The internal activity items of the neuron include the feedback input functional unit and the coupling connection input functional unit, which are composed of nonlinear multiplication. β is the connection strength coefficient between the internal activity items (synapses):

[0025] U ij [n] = F ij [n]{1+βL ij [n]} (3)

[0026] The output timing pulse train formula of a neuron is:

[0027]

[0028] The dynamic threshold T decays during the iteration process. When the neuron is excited (U>T), the dynamic threshold increases immediately, and then gradually decays according to the exponential law until the neuron is excited again. This process can be described as:

[0029]

[0030] Where V T Generally, a relatively large value is taken, which is one order of magnitude larger than the mean value of U. The array of the above neurons eventually forms a complete PCNN. After applying the pulse coupled neural network (PCNN) to the image for 5 iterative processes, start processing the newly generated black and white binary image. Divide the obtained binary image of the screen surface into 5 strip areas along the x-axis direction, from left to right, they are area 1, area 2, area 3, area 4, and area 5, and divide it into 4 sections along the y-axis direction (every 2 sieve plates correspond to 1 section, and each section contains a left-moving sieve plate and a right-moving sieve plate), from top to bottom, they are section 1, section 2, section 3, and section 4. As Figure 5As shown, for the sub-image matrix in each segment, the total value of the number of white pixels in each equally divided area in the segment is calculated respectively, a histogram of the total white pixels in each equally divided area is formed, and the average value of the total white pixels in the five areas is calculated. According to the average value, it is judged which of the five areas has a smaller total white pixel, and the judgment result is fed back according to different distributions, and a corresponding adjustment control signal is given. For example, if the total white pixel value of areas 1 and 2 in the third segment is too small, it means that the distribution of the seed mixture is concentrated in areas 3, 4, and 5, then a command is given to the left-moving screen corresponding to the third segment, that is, the left-moving screen c6, to move. The push-pull electromagnet 12 connected to the left-moving screen c6 drives the left-moving screen c6 to do a lateral reciprocating motion. Due to the microscopic scale structure of the butterfly wing surface on the left-moving screen c6, the left-moving screen c6 is moved. 6 When moving to the left, the friction force on the corn off-screen 13 is greater than the friction force when moving to the right, so the corn off-screen 13 is driven to move to the left. At the same time, the high-definition camera continues to shoot and continuously transmits the captured pictures to the pulse coupled neural network (PCNN) for processing. When it is detected that the total value of white pixels in each area of ​​the section is equal, it means that the screen surface material in this area has completed the horizontal uniform distribution. At this time, the push-pull electromagnet 12 is controlled to stop moving, and the above operation is continuously repeated to achieve the horizontal uniform distribution of the material on the screen.

[0031] The present invention designs a segmented corn cleaning screen based on a bionic structure surface and its intelligent control system, which has the characteristics of simple, reliable and reasonable structure, high corn kernel cleaning efficiency, good cleaning quality, high intelligence, uniform lateral distribution of materials on the screen, high screen surface utilization, etc.

Claims

1. A segmented corn cleaning screen based on a bionic structure surface and an intelligent control system thereof, comprising a high-definition camera (1), a left-moving screen surface a (2), a right-moving screen surface a (3), a left-moving screen surface b (4), a right-moving screen surface b (5), a left-moving screen surface c (6), a right-moving screen surface c (7), a left-moving screen surface d (8), a right-moving screen surface d (9), a screen cover (10), a screen frame (11), and a push-pull electromagnet (12). The screen is characterized in that the left-moving screen surface a (2), the right-moving screen surface a (3), the left-moving screen surface b (4), the right-moving screen surface b (5), the left-moving screen surface c (6), and the right-moving screen surface c (7) The left-moving screen surface d (8) and the right-moving screen surface d (9) are all black and are installed in the groove between the screen cover (10) and the screen frame (11) in a manner that they can slide horizontally from front to back. Eight push-pull electromagnets (12) are fixedly installed on the right side of the screen frame (11). The positions correspond to the geometric center lines of the eight screen sheets and are fixedly installed with the corresponding screen sheets. The corresponding screen sheets can be driven to move horizontally in the groove between the screen cover (10) and the screen frame (11). The movement of each push-pull electromagnet (12) is controlled by an independent external control system. The screen frame (11) has round holes at the front and rear for connecting with an external drive. The left-moving screen surface a (2), the left-moving screen surface b (4), the left-moving screen surface c (6), and the left-moving screen surface d (8) have a right-angled triangle serrated structure imitating the microscopic scale structure of the surface of a butterfly wing. One of the two right-angled sides of the right-angled triangle structure is parallel to the screen surface, and the other is toward the left side of the screen body, so that the friction force on the screened material when it moves to the right on the screen surface is greater than the friction force on it when it moves to the left; the right-moving screen surface a (3), the right-moving screen surface b (5), the right-moving screen surface c (7), and the right-moving screen surface d (9) have a right-angled triangle structure imitating the microscopic scale structure of the surface of a butterfly wing. The sawtooth structure has two right-angled sides, one of which faces the screen surface and is parallel to the screen surface, and the other faces the right side of the screen body, so that the friction force on the screened material when it moves to the left on the screen surface is greater than the friction force on it when it moves to the right. The high-definition camera (1) is fixed above the screen surface and can capture the position of the entire screen surface and is connected to the data processing system via a data cable. The captured image is transmitted in real time. The pulse coupled neural network (PCNN) processing technology is used to process the image, identify the distribution of materials on each section of the screen surface, and control the push-pull electromagnet (12) according to demand to drive the screen surface to move horizontally.

Images