Novel customs port practical screening system and method

Abstract

The invention discloses a novel customs port practical screening system and method.The novel customs port practical screening system comprises a feeding module used for feeding grain samples into a screening module, the material screening module conducts preliminary screening on the samples based on the four-stage screening technology, and materials with different particle sizes and densities are separated out; the identification sample reserving module is used for collecting a sample image, performing seed level identification through a YOLO neural network technology and storing a voucher specimen; the data analysis module is used for performing data analysis on screened data and performing classification, uploading and database updating; the multi-stage screening technology and the multi-channel physical sorting technology are utilized, high-precision sorting of the samples is achieved, the image recognition technology is utilized to control the intelligent mechanical arm to complete retention and labeling of different samples, and finally full-automatic processing of the imported grain samples is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the customs supervision and material sorting technical field, and particularly relates to a novel customs port practical screening system and method. BACKGROUND

[0002] In the current situation of increasingly frequent international trade, the customs port is an important channel for the import and export of goods, and the quality control of imported grain is particularly important. In the process of grain storage and import, in order to ensure the quality and safety of grain, accurate sorting and extraction of grain samples are required.

[0003] The application of the existing customs grain sample sorting and extraction technology improves the efficiency and accuracy of grain sample sorting and extraction, and can to some extent guarantee the safety of grain and improve the efficiency of customs supervision. However, there is still a lack of a multi-system collaborative sorting scheme for a batch of original samples to complete the whole process of automatic processing from sample preliminary screening to sample classification, fine impurity removal, sample screening, automatic identification and sample storage, and finally to data uploading platform and risk estimation. A large amount of manual operation is still required for the current customs entry sample inspection, which is troublesome and time-consuming. Therefore, the present application proposes a novel customs port practical screening system and method to solve the problems in the prior art. SUMMARY

[0004] To solve the above problems, the present application proposes a novel customs port practical screening system and method, which uses multi-stage screening technology and multi-channel physical sorting technology to achieve high-precision sorting of samples, uses image recognition technology to control an intelligent mechanical arm to complete the retention labeling of different samples, and finally realizes the full automation of the entry grain sample processing.

[0005] To achieve the purpose of the present application, the present application realizes the following technical scheme: a novel customs port practical screening system, comprising a feeding module, a material screening module, an identification and sample retention module, and a data analysis module. The feeding module is used to feed the grain sample into the screening module. The material screening module is based on four-stage screening technology to preliminarily screen the sample and separate materials of different particle sizes and densities. The identification and sample retention module is used to collect sample images and perform species identification and save voucher specimens through YOLO neural network technology. The data analysis module is used to analyze the data after screening and perform classification, uploading and database updating.

[0006] Further improvement lies in that the feeding module comprises a main conveying belt feeding sub-module, an auxiliary conveying belt circulating sub-module and a waste impurity disposal sub-module, the main conveying belt feeding sub-module is used for conveying the sample from the inlet to the screening area, the auxiliary conveying belt circulating sub-module is used for ensuring the continuous flow of the sample during the screening process, and the waste impurity disposal sub-module is used for disposing the screened waste and impurities.

[0007] Further improvement lies in that the material screening module comprises a four-stage screening sub-module, a physical sorting sub-module and a color selection extraction sub-module, the four-stage screening sub-module utilizes four layers of sieves with different particle sizes for four-stage screening based on a size particle diameter screen, the physical sorting sub-module further sorts the material based on physical methods of centrifugal screening, vibration negative pressure suction, acoustic standing wave airflow chromatography and electrostatic adsorption cleaning, and the color selection extraction sub-module extracts pure grain particles based on color selection.

[0008] Further improvement lies in that the physical sorting sub-module comprises a centrifugal screening unit, a vibration negative pressure suction unit, an acoustic standing wave airflow chromatography unit and an electrostatic adsorption cleaning unit, the centrifugal screening unit separates different particle sizes of the screened-on material after the three-layer and four-layer screening of the four-stage screening sub-module by using centrifugal force, the vibration negative pressure suction unit further separates and collects particles with different densities by using mechanical vibration and negative pressure suction technology, the acoustic standing wave airflow chromatography unit again accurately separates different size particles by using acoustic standing waves with different frequencies and airflow chromatography technology, and the electrostatic adsorption cleaning unit removes impurities on the surface of the sample based on an electrostatic generator.

[0009] Further improvement lies in that the identification and sample retention module comprises a sample image acquisition sub-module and a neural network identification sub-module, the sample image acquisition sub-module acquires image information of the sample based on a high-resolution camera, and the neural network identification sub-module identifies the sample target in the image and identifies the species level based on a YOLO neural network and then marks and saves.

[0010] Further improvement lies in that the data analysis module comprises a data classification sub-module, a cloud platform data uploading sub-module and a database updating sub-module, the data classification sub-module classifies sample data and impurity data according to the identification result after pre-processing the collected data, the cloud platform data uploading sub-module is used for uploading the data after the classification processing to the cloud platform, and the database updating sub-module is used for updating the sample database and the impurity database according to the classification result.

[0011] Further improvement lies in that the data classification sub-module comprises a grain sample data storage unit and an impurity sample data storage unit, the grain sample data storage unit is used for storing the screened grain sample data, and the impurity sample data storage unit is used for storing the impurity sample data after the separation and screening.

[0012] Further improvement lies in that the database updating submodule comprises an impurity sample database updating unit and an effective sample database updating unit, the impurity sample database updating unit is used for storing the identified impurity sample data other than the original database, and the effective sample database updating unit is used for storing the identified effective sample data other than the original database.

[0013] A screening method of a novel customs port practical screening system, comprising the following steps:

[0014] Step one, four-stage screening processing, the sample to be screened is transported to the four-stage screening submodule through the feeding module, four-stage screening is performed through four layers of screens with different particle sizes, straw is screened out, and large-particle weeds and pests and diseases are screened and sampled;

[0015] Step two, centrifugal screening processing, the screen oversize after three-stage and four-stage screening in step one is sent to the centrifugal screening unit, slow centrifugal screening is performed at a speed of 10-100 rpm, and near-center target, intermediate target and edge target are obtained;

[0016] Step three, vibration negative pressure suction processing, the edge target of the rotating disc after centrifugal screening in step two is subjected to mechanical vibration at different frequencies and amplitudes, and is collected by different suction forces, to obtain high-frequency screening target, medium-frequency screening target and low-frequency screening target;

[0017] Step four, acoustic standing wave airflow chromatography processing, each type of target in the three types of targets after vibration negative pressure suction in step three is subjected to resonance screening by sound waves of different frequencies, and is collected by air jet of different speeds at the wave peak position corresponding to the standing wave formed by the sound waves, to obtain high-frequency sound wave screening material, medium-frequency sound wave screening material and low-frequency sound wave screening material;

[0018] Step five, electrostatic adsorption cleaning processing, multiple groups of materials obtained in steps one to four are laid on the conveying belt, and each group of materials is subjected to electrostatic adsorption cleaning by the electrified electrostatic generator to remove surface impurities, and after the cleaned materials are identified and sampled by the identification and sampling module, the impurity-removed materials are obtained;

[0019] Step six, color selection extraction processing, the processed materials are separated according to different characteristics of the materials by the color selection instrument, and finally subsequent image recognition is performed, to complete the screening and sorting of the materials.

[0020] The beneficial effects of this invention are as follows: Through a multi-stage sorting process, from coarse sorting to fine sorting, it gradually removes various impurities such as weed seeds, straw, pests and diseases from grains and forages, effectively ensuring the purity of imported grains and forages, and preventing harmful substances from entering the country with goods, posing a potential threat to the domestic ecological environment and agricultural production. The system also has sample extraction and storage functions, capable of accurately extracting representative grain and forage samples, providing a strong basis for subsequent quality testing, quarantine analysis, and other processes, helping customs departments to carry out supervision work efficiently and accurately, and safeguarding national food and ecological security. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a system architecture diagram of the present invention.

[0022] Figure 2 Schematic diagram of the system flow of the present invention.

[0023] Figure 3 This is a schematic diagram of the four-stage screening principle of the present invention.

[0024] Figure 4 This is the principle diagram of the YOLO neural network of the present invention.

[0025] Figure 5 This is a diagram of the data analysis principle of the present invention.

[0026] Figure 6 This is a schematic diagram of the centrifugal screening principle of the present invention.

[0027] Figure 7 This is a schematic diagram of the vibration negative pressure absorption principle of the present invention.

[0028] Figure 8 This is a schematic diagram of the acoustic standing wave airflow chromatography principle of the present invention.

[0029] Figure 9 This is a schematic diagram of the electrostatic adsorption cleaning principle of the present invention. DETAILED DESCRIPTION

[0030] In order to deepen the understanding of the present invention, the present invention will be further described in detail below with reference to the examples. The examples are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0031] Example 1

[0032] according to Figures 1-5 As shown, this embodiment provides a new practical screening system for customs ports, including a material feeding module, a material screening module, an identification and sample retention module and a data analysis module.

[0033] The feeding module is used for feeding the grain sample into the screening module, and includes a main conveyor feeding sub-module, an auxiliary conveyor circulating sub-module and a waste impurity disposal sub-module. The main conveyor feeding sub-module is used for conveying the sample from the inlet to the screening area. The auxiliary conveyor circulating sub-module is used for ensuring the continuous flow of the sample during the screening process. The waste impurity disposal sub-module is used for disposing the screened waste and impurities.

[0034] The material screening module performs preliminary screening on the sample based on a four-stage screening technology, separates materials of different particle sizes and densities, and includes a four-stage screening sub-module, a physical sorting sub-module and a color selection extraction sub-module. The four-stage screening sub-module uses a four-layer screen with different particle sizes based on the scale particle size screen to perform four-stage screening. The physical sorting sub-module further sorts the materials based on physical methods such as centrifugal screening, mechanical vibration negative pressure suction, acoustic standing wave airflow chromatography and electrostatic adsorption cleaning. The color selection extraction sub-module uses a color sorter to extract pure grain particles based on color differences.

[0035] The four-stage screening sub-module: the sample enters the scale particle size screen through the feeding port, and the sample is spread flat by the rotating rake to ensure that each particle has a chance to pass through the screen hole, thereby improving the screening efficiency. Then, the scale particle size screen uses a four-layer screen with different particle sizes to perform four-stage screening, which can screen out straw and select samples of large-particle weeds and pests. The screen adopts vibration technology, so that the spread sample jumps and vibrates on the screen. During the movement, particles smaller than the screen aperture will fall, and particles larger than the screen aperture will remain on the screen. The four-layer screen apertures are arranged from large to small from top to bottom. The material on each layer of the screen is transported out of the screen by the conveying device after screening is completed. The four-stage screening screens out:

[0036] First-stage screening: the material on the screen is straw, large-stem weeds and pests, and the material under the screen enters the second-stage screening;

[0037] Second-stage screening: the material on the screen is complete beans of 5 mm or larger and weed seeds, and the material under the screen enters the third-stage screening;

[0038] Third-stage screening: the material on the screen is incomplete beans of 3-5 mm and weed seeds, and the material under the screen enters the fourth-stage screening;

[0039] Fourth-stage screening: the material on the screen is weed seeds and impurities of 0.5-3 mm, and the material under the screen enters the waste conveying module.

[0040] Through this multi-stage screening, most of the impurities can be effectively removed, and the purity of the sample is improved, providing high-quality input for subsequent centrifugal screening and mechanical vibration suction steps. The principle diagram of the four-stage screening is shown in the accompanying drawings. Figure 3

[0041] ​Color selection extraction sub-module: the oversize after secondary particle size screening is arranged in a column into the color selection instrument, separating soybeans from other impurities, and the impurities are centrifugally screened with the oversize of the third and fourth levels to image recognition processing. The color selection instrument separates and screens materials according to their different characteristics through optical equipment irradiation, with high efficiency, accurate classification, good product quality, and low damage rate.

[0042] Color selection extraction is an important step in the automatic sorting system, which separates soybeans from other impurities to ensure the purity and quality of soybeans. In the color selection extraction process, the oversize after secondary particle size screening is arranged in a column into the color selection instrument. The color selection instrument separates and screens materials according to their different characteristics through optical equipment irradiation. The color selection instrument analyzes the color, shape, and texture of each particle through high-precision optical sensors and image processing technology, effectively separating soybeans from other impurities. The separated impurities are centrifugally screened with the oversize of the third and fourth levels to image recognition and other subsequent steps. The color selection instrument has high efficiency, accurate classification, good product quality, and low damage rate, which can effectively improve the accuracy and efficiency of sorting. Through this color selection extraction, the purity and quality of soybeans can be ensured, providing high-quality samples for customs supervision and subsequent processing.

[0043] The physical sorting sub-module includes a centrifugal screening unit, a vibration negative pressure suction unit, a sound standing wave airflow chromatography unit, and an electrostatic adsorption cleaning unit. The centrifugal screening unit separates particles of different sizes using centrifugal force on the oversize after three and four levels of screening by the fourth screening sub-module. The vibration negative pressure suction unit further separates and collects particles of different densities using mechanical vibration and negative pressure suction technology. The sound standing wave airflow chromatography unit uses sound wave standing waves of different frequencies and airflow chromatography technology to accurately separate particles of different sizes. The electrostatic adsorption cleaning unit removes impurities on the surface of the sample based on the electrostatic generator.

[0044] The identification and sample retention module is used to collect sample images and identify species and save voucher specimens through YOLO neural network technology. It includes a sample image acquisition sub-module and a neural network recognition sub-module. The sample image acquisition sub-module acquires image information of the sample based on a high-resolution camera. The neural network recognition sub-module identifies and identifies the species of the sample target in the image based on the YOLO neural network and marks and saves it after identification.

[0045] Specifically, after electrostatic adsorption cleaning, multiple groups of targets and oversize samples are laid on the target plate. The target detection algorithm based on artificial intelligence is used to identify the species of the weed seeds in the sample, and the mechanical arm is used to suck and package the voucher specimens and label them for storage. The target detection algorithm can accurately identify and classify different types of weed seeds, ensuring the accuracy and integrity of the sample.

[0046] Image recognition and voucher specimen retention are key steps in the automated sorting system for species-level identification of weed seeds in samples and preservation of voucher specimens. In the image recognition process, multiple groups of target and primary screen material samples after electrostatic adsorption cleaning are laid flat on a target plate, and high-resolution cameras are used to take images of the samples. These images are then transmitted to an artificial intelligence target detection algorithm, which uses a deep learning model to identify weed seeds in the samples at the species level. The target detection algorithm can accurately identify and classify different types of weed seeds, ensuring the accuracy and integrity of each sample. The identified samples are taken by a robotic arm and packaged, each sample is labeled and stored in a special storage container for subsequent analysis and verification.

[0047] Through this image recognition and voucher specimen retention, detailed records and backups of each sample can be ensured, providing strong support for customs supervision and scientific research. The principle of the YOLO neural network used in image recognition is shown in the accompanying drawings. Figure 4

[0048] The data analysis module is used for data analysis of the screened data and classification, uploading and database updating, including a data classification sub-module, a cloud platform data uploading sub-module and a database updating sub-module. The data classification sub-module classifies the collected data after preprocessing according to the identification results, the cloud platform data uploading sub-module is used to upload the classified data to the cloud platform, and the database updating sub-module is used to update the sample database and impurity database according to the classification results.

[0049] The data classification sub-module includes a grain sample data storage unit and an impurity sample data storage unit, the grain sample data storage unit is used to store the screened grain sample data, and the impurity sample data storage unit is used to store the separated and screened impurity sample data.

[0050] The database updating sub-module includes an impurity sample database updating unit and an effective sample database updating unit, the impurity sample database updating unit is used to store the impurity sample data identified from the original database, and the effective sample database updating unit is used to store the effective sample data identified from the original database.

[0051] The detailed process of the data analysis module is as follows:

[0052] (1) Data preprocessing: Before the processing results of the identification module are transmitted to the data analysis module, the data needs to be preprocessed, which includes data cleaning, format conversion and standardization, etc. to ensure the quality and consistency of the data.

[0053] ​(2) Sample and impurity classification: The pre-processed results will be transmitted to the data analysis module through the File Transfer Protocol system, and the sample data and impurity data will be classified according to the recognition results of the recognition module. The classification results will be used for subsequent updating and saving of sample and impurity data.

[0054] (3) Gimbal data upload: The classified data needs to be uploaded to the cloud platform. This usually involves the following steps:

[0055] Evaluate the data type and size, and choose a suitable cloud service provider.

[0056] Ensure a stable and sufficient bandwidth internet connection.

[0057] Classify and organize the data, and choose appropriate upload tools such as web upload, dedicated client software or command line tools.

[0058] Encrypt the data when uploading sensitive data.

[0059] (4) Database update: Update the sample and impurity database according to the classification results. This includes inserting new records, updating existing records or deleting outdated records. The update process needs to ensure the consistency and integrity of the data, and handle concurrent operations through transaction management.

[0060] (5) Result verification and feedback: After data upload and database update, verify the results to ensure data accuracy. The system provides a feedback mechanism for users to understand the status of data processing and intervene if necessary.

[0061] (6) Monitoring and maintenance: The data analysis module has monitoring functions to track possible problems during data processing. Regular maintenance helps optimize system performance and ensure efficient and stable data processing.

[0062] Through the above process, the data analysis module can effectively process multiple types of data, classify samples and impurities, ensure accurate data upload and timely database update. The data analysis principle is shown in the description Figure 5 .

[0063] Example 2

[0064] According to Figures 6-9 , this embodiment provides a theoretical support for a physical sorting sub-module in a new type of customs port practical screening system, and a principle description of YOLO neural network.

[0065] Centrifugal screening unit: the oversize after three or four layers of screening enters the centrifugal screening unit for slow centrifugal screening. The rotating disc surface is rough, and the centrifugal force is generated by the uniform rotation controlled by the motor, with a rotating speed controlled within the range of 10-100 rpm. The sample on the rotating disc surface will move to the edge of the rotating disc under the action of centrifugal force. The radial distribution of the sample on the rotating disc will be affected by the roughness of the sample surface, the shape of the sample, and the density of the sample. Finally, the sample is distributed along the radial direction to different areas with different centrifugal radii. If the radius of the rotating disc used is r, the target after centrifugation is divided into three categories according to the radius:

[0066] Near-centre target (0-1 / 3r): targets with small density, rough surface, small particle size and irregular shape, such as impurities;

[0067] Intermediate target (1 / 3r-2 / 3r): targets with moderate density, smooth surface, moderate particle size and regular shape, such as light weed seeds and incomplete grain and grass samples;

[0068] Edge target (2 / 3r-r): targets with large density, smooth surface, large particle size and regular shape, such as round weed seeds.

[0069] The annular baffle after centrifugation separates the rotating disc area into three areas, and the pneumatic cleaning module is started to send the samples in the three areas to three containers along three discharge ports and to the next sorting module along the conveyor belt.

[0070] Through this centrifugal screening, different density particles can be effectively separated to provide classified samples for further mechanical vibration and suction. Centrifugal screening not only improves the accuracy of sorting, but also reduces the complexity of subsequent processing, ensuring that each classified sample can be processed specifically. The centrifugal screening principle diagram is shown in the description attached Figure 6 .

[0071] Centrifugal screening is based on the Stokes law at low Reynolds number and the balance of centrifugal force and friction:

[0072] Particle Reynolds number Re p is a dimensionless number that describes the characteristics of particle motion in a fluid, which can help determine whether the fluid flow is laminar or turbulent. The formula for calculating the particle Reynolds number is:

[0073]

[0074] where p f is the density of the gas in the centrifuge, D pis the diameter of the sample particle, u is the velocity of the sample particle relative to the gas, and μ is the dynamic viscosity of the gas in the centrifuge. By calculating the particle Reynolds number, it can be determined whether the flow is laminar, transitional or turbulent, which is crucial for the study and application of fluid dynamics. The Reynolds number is proportional to the velocity of the sample, and the Reynolds number of the system is small because the centrifugation is slow, and the flow of air in the centrifuge is laminar. The velocity of the sample when it is balanced can be divided into two categories. One category is light mass targets that float up during centrifugation, and the velocity of this category of targets in the laminar air can be calculated by Stokes' law:

[0075]

[0076] where ρ p is the density of the sample particle, η is the viscosity of the gas in the centrifuge, r is the radius of the sample particle, and g is the acceleration of gravity. On the other hand, the velocity of the rotating sample is affected by the radius of rotation R and the rotational speed ω: v = ωR.

[0077] At a constant rotational speed, the greater the density, the faster the velocity, and the greater the radius of rotation, which is located at the periphery of the rotating disc. The other category of targets moves on the rough rotating disc during centrifugation, and the velocity can be given by the centrifugal force formula:

[0078]

[0079] where the friction force is mainly determined by the roughness of the target surface, defined as f 摩擦 = μmg, and the velocity required for equilibrium can be obtained as The rougher the sample surface, the greater the velocity required for equilibrium, the more difficult it is to move, and it is located near the center of the rotating disc, while a smooth surface cannot be balanced at a lower speed and will move to the edge of the rotating disc.

[0080] Vibrating negative pressure suction unit: the targets at the edge of the rotating disc after centrifugal screening are subjected to mechanical vibration at different frequencies and amplitudes, and are collected using different suction forces. The vibration frequency is controlled between 0-30Hz, the amplitude is controlled within the range of 0-5mm, and the negative pressure suction force is controlled within the range of 1-4kPa. Different combinations of frequency, amplitude and suction force form three modes, which divide the targets into three categories:

[0081] Mode one: high frequency vibration, low suction force, suitable for light impurities;

[0082] Mode two: medium frequency vibration, medium suction force, suitable for medium density weed seeds;

[0083] Mode three: low frequency vibration, high suction force, suitable for heavy impurities.

[0084] Mechanical vibration and suction technology is a key step in automated sorting systems, used for further separation and collection of particles of different densities. During mechanical vibration, the targets at the edge of the rotating disc are laid on a vibrating platform, which vibrates at different frequencies and amplitudes. High-frequency vibration is suitable for light impurities, which are easily separated under high-frequency vibration and collected by suction devices with low suction force. Medium-frequency vibration is suitable for medium-density weed seeds, which are separated under medium-frequency vibration and collected by suction devices with medium suction force. Low-frequency vibration is suitable for heavy impurities, which are separated under low-frequency vibration and collected by suction devices with high suction force.

[0085] Through this combination of mechanical vibration and suction, particles of different densities can be accurately separated, ensuring that each classified sample can be effectively processed. This technology not only improves the efficiency of sorting, but also reduces human intervention and improves the degree of automation of the system. The schematic diagram of vibration negative pressure suction is shown in the description Figure 7 .

[0086] Acoustic standing wave air flow chromatography unit: After vibration negative pressure suction, three types of targets are proposed, and different frequencies of sound waves are used for resonance screening for each type of target. In the position corresponding to the wave peak of the standing wave formed by the sound wave, different speeds of air jet are used for collection. The frequency range of the sound wave is controlled between 10-40MHz, and the speed of the air jet is controlled between 0-10m / s. According to different sound wave frequencies and jet speeds, each group of targets is divided into three categories:

[0087] High-frequency sound wave low-speed jet: suitable for small weed seeds;

[0088] Medium-frequency sound wave medium-speed jet: suitable for medium-sized weed seeds;

[0089] Low-frequency sound wave high-speed jet: suitable for larger weed seeds.

[0090] Acoustic standing wave air flow chromatography technology is an advanced sorting method that uses the resonance characteristics of sound waves to further screen particles of different sizes. During acoustic vibration, the sound wave generator generates sound waves of different frequencies to form standing waves in the cavity. The sample will resonate under the action of high-frequency sound waves and concentrate at the wave peak position formed by the sound wave to be separated. At each wave peak position, different suction air flows are used for suction to collect the separated particles.

[0091] Through this combination of acoustic vibration and air flow chromatography suction, particles of different sizes can be accurately separated, ensuring that each classified sample can be effectively processed. This technology not only improves the accuracy of sorting, but also reduces the complexity of subsequent processing, improving the overall efficiency of the system. The principle of acoustic standing wave air flow chromatography is shown in the description Figure 8 .

[0092] The principle of standing wave in cavity is the theoretical support of this system.

[0093] In the acoustic standing wave airflow chromatography unit, the acoustic wave generated by the acoustic wave generator propagates in the cavity. Due to the existence of reflected waves, the waves in the space have both forward propagation components and backward propagation components. They superimpose to form a fixed waveform at certain positions, which is called standing wave. In these waveforms, some points have always zero amplitude, which are called wave troughs. Some points have superimposed maximum amplitude, which are called wave crests. The condition for the formation of standing wave is that there are two columns of waves with the same frequency and amplitude but opposite propagation directions. The mathematical explanation of its formation is as follows:

[0094] Suppose there is an acoustic wave propagating along the positive direction of the x-axis, whose expression is y1=Acos(kx-ωt). The expression of the reflected wave is y2=Acos(kx+ωt). The total displacement after superposition of the two waves is:

[0095] y=y1+y2=Acos(kx-ωt)+Acos(kx+ωt)

[0096] =2Acos(kx)cos(ωt)

[0097] From this formula, it can be obtained that the spatial position of the wave crest of the standing wave is determined by the cosine function cos(kx). Its maximum value appears at the odd multiples of π / 2, i.e.:

[0098]

[0099] Since After substitution, it can be obtained that the wave crest position x satisfies

[0100]

[0101] That is, the wave crest appears at the odd multiples of one quarter of the wavelength, i.e. The positions are the same positions, at which the wave peaks and wave troughs of the two waves propagating in opposite directions meet, resulting in superimposed amplitude and forming the wave crest.

[0102] Electrostatic adsorption cleaning unit: the above-mentioned several tens of classified group samples are laid on the conveying belt through the electrostatic generator, and then the samples carrying static electricity are transmitted into the electrostatic adsorption cleaning device through the conveying belt with voltage added on both sides. The micro dust particles carrying static electricity are adsorbed onto the conveying belts on both sides. Different voltages are added on both sides of the conveying belt in different intervals. The voltage range is controlled between 5kV-35kV. The current range between the conveying belts is 0-400μA. The distance between the charged conveying belts is 5mm-50mm. Electrostatic adsorption cleaning is performed on each group of samples. Then, the dust cleaning brush head is used to clean the dust on the conveying belts on both sides to ensure the recycling use of the conveying belts on both sides. The cleaned samples are left on the original conveying belt to enter the subsequent identification and sampling module.

[0103] The electrostatic adsorption technology can effectively remove the micro impurities on the surface of the samples, prepare for the image recognition and voucher specimen retention of the samples, and improve the purity and image recognition accuracy of the samples. The electrostatic adsorption cleaning principle is shown in the description. Figure 9

[0104] The electrostatic induction, electrical conductivity, charge adsorption force and van der Waals force serve as the theoretical support of this module.

[0105] Electrostatic induction

[0106] In the electrostatic adsorption cleaning, the electrostatic induction plays an important role. When the impurities such as weeds in the grain materials approach the electrostatic field, the electrostatic induction phenomenon occurs. The electrostatic induction refers to when a charged body approaches a conductor, the free charges inside the conductor will redistribute under the action of the electric field force, so that one end of the conductor has one kind of charge, and the other end has another kind of charge, and the charge amount is equal in size and opposite in sign. The formula is:

[0107] q 感应 =-q 外

[0108] wherein q 感应 is the induced charge amount on the conductor, and q 外 is the charge amount of the external charged body.

[0109] Electrical conductivity

[0110] The electrical conductivity is a physical quantity describing the electrical conductivity of a substance, and has a certain influence on the sorting of the impurities in the grain materials. The formula of the electrical conductivity is:

[0111]

[0112] wherein σ is the electrical conductivity, the unit is Siemens per meter (S / m); I is the current passing through the conductor, the unit is ampere (A); and E is the electric field intensity, the unit is volt per meter (V / m).

[0113] Charge adsorption force​

[0114] During the sorting process, the charge adsorption force helps to adsorb impurities onto specific collection devices. Charge adsorption force refers to the interaction force between charged bodies, and its formula is Coulomb's law:

[0115]

[0116] where F is the charge adsorption force, with units of Newton (N); k is Coulomb's constant, with a value of 9.0 × 10 9 N·m 2 / C 2 ; q1 and q2 are the charge amounts of the two charged bodies, both in Coulombs (C); r is the distance between the two charged bodies, in meters (m).

[0117] Van der Waals force

[0118] Van der Waals force is a weaker interaction force that exists between neutral molecules or atoms, and also plays a supporting role in the adsorption and sorting process of impurities in grain materials. The formula for Van der Waals force is:

[0119]

[0120] where F vdW is the Van der Waals force, with units of Newton (N); C is the Van der Waals constant, whose value depends on the type of molecule or atom; r' is the distance between molecules or atoms, in meters (m).

[0121] Principle of YOLO neural network:

[0122] Grid division: YOLO first divides the input image into a fixed size S × S grid. Each grid is responsible for detecting objects in that region. For example, in YOLOv1, the original image is divided into a 7 × 7 grid.

[0123] Boundary box prediction: For each grid, the YOLO algorithm predicts multiple bounding boxes. These bounding boxes are represented by center coordinates (x, y), width (w), and height (h). Each bounding box also contains a confidence score indicating whether the box contains an object and its accuracy.

[0124] Object classification: For each bounding box, YOLO uses a classifier to predict the class of the object. Usually, a convolutional neural network (CNN) is used to extract features, and a fully connected layer is used for classification.

[0125] Confidence evaluation: Each bounding box also predicts a confidence score indicating the probability of an object existing in the bounding box and the accuracy of the bounding box.

[0126] Non-maximum suppression (NMS): For each class, non-maximum suppression is used to remove overlapping bounding boxes. The bounding box with the highest confidence is selected, and bounding boxes with high overlap are discarded.

[0127] The network output of YOLO is an SxSx(5xB+C) tensor, where S is the size of the grid; B is the number of bounding boxes predicted per grid; and C is the number of classes.

[0128] For each bounding box, YOLO predicts the following parameters:

[0129] x, y: relative coordinates of the center point of the bounding box, ranging from 0 to 1;

[0130] w, h: width and height of the bounding box, normalized with respect to the size of the entire image;

[0131] confidence: probability that the bounding box contains an object and accuracy of the bounding box prediction.

[0132] Embodiment 3

[0133] The present embodiment provides an application example of a screening method of a new type of customs port practical screening system.

[0134] Suppose there is a batch of grain and grass samples containing straw, weeds, weed seeds, diseased beans, insect carcasses and insect eggs, and impurities such as sand. The above-mentioned automatic sorting system will be used to sort this batch of samples to remove impurities and extract pure grain and grass.

[0135] (1) The grain and grass samples are laid flat on the vibrating screening platform to improve the purity of the samples through four-stage screening. The first-stage screening removes straw and large weeds, the second-stage screening separates 5mm and above beans and weed seeds, the third-stage screening removes 3mm-5mm incomplete particles, and the fourth-stage screening removes 0.5mm-3mm small weed seeds and impurities. This process ensures uniform distribution of the sample and improves the efficiency of the screening.

[0136] (2) The 3mm-5mm incomplete beans and weed seeds screened by the third and fourth stages are placed in a rotating disc centrifuge, which uses centrifugal force to separate the particles into three categories: the near-center target is the relatively heavy impurities, the middle target is the moderate weed seeds, and the edge target is the relatively light impurities such as insect carcasses and insect eggs.

[0137] (3) The three types of targets after centrifugal screening are placed on the vibrating platform. High-frequency vibration is suitable for light impurities (such as insect carcasses and insect eggs), which are collected by low suction devices; medium-frequency vibration is suitable for moderate-density weed seeds, which are collected by medium suction devices; and low-frequency vibration is suitable for heavy impurities (such as sand), which are collected by high suction devices.

[0138] (4) The target after mechanical vibration absorption is laid on the vibration platform, and resonance screening is performed using sound waves of different frequencies and amplitudes. High-frequency sound waves separate small weed seeds, medium-frequency sound waves separate medium-sized seeds, and low-frequency sound waves separate larger weed seeds. The separated particles are collected by air flow with different suction.

[0139] (5) The classified multiple groups of samples are laid on a conveyor belt, and surface micro-impurities are removed by electrostatic adsorption cleaning to improve purity and identification accuracy. The cleaned samples are photographed by a high-resolution camera, and the image is transmitted to an AI target detection algorithm for weed seed grade identification. Finally, the samples are sucked, packaged, and labeled by a mechanical arm, and stored for subsequent analysis.

[0140] (6) The complete beans and weed seeds after secondary screening enter the color sorter, and are accurately separated from impurities by optical sensor analysis of color, shape, and texture. The separated impurities and other screen overs are subjected to centrifugal screening and subsequent image recognition processing to improve sorting efficiency and quality.

[0141] Through the above multi-step sorting and processing, the customs can effectively remove straw, weeds, weed seeds, diseased beans, insect carcasses and eggs, sand, and other impurities in the grain and grass samples, extract pure grain and grass, and ensure the quality and safety of imported grain and grass.

[0142] The above shows and describes the basic principles, main features, and advantages of the present application. Those skilled in the art should understand that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A new type of practical screening system for customs ports, characterized by: It includes a feeding module, a material screening module, an identification and sampling module and a data analysis module. The feeding module is used to feed grain samples into the screening module. The material screening module performs preliminary screening of samples based on four-level screening technology to separate materials with different particle sizes and densities. The identification and sampling module is used to collect sample images and perform species identification and save voucher specimens through YOLO neural network technology. The data analysis module is used to analyze the screened data and perform classification, upload and database update.

2. A novel practical screening system for customs ports according to claim 1, characterized in that: The feeding module includes a main conveyor belt feeding submodule, an auxiliary conveyor belt circulation submodule and a waste and impurity disposal submodule. The main conveyor belt feeding submodule is used to transport samples from the entrance to the screening area, the auxiliary conveyor belt circulation submodule is used to ensure the continuous flow of samples during the screening process, and the waste and impurity disposal submodule is used to dispose of the screened waste and impurities.

3. The novel practical screening system for customs ports according to claim 1 is characterized by: The material screening module includes a four-stage screening submodule, a physical sorting submodule and a color sorting and extraction submodule. The four-stage screening submodule uses four layers of sieves with different particle sizes based on a scale particle size screener to perform four-stage screening. The physical sorting submodule further sorts the material based on the physical methods of centrifugal screening, vibration negative pressure absorption, acoustic standing wave airflow chromatography and electrostatic adsorption cleaning. The color sorting and extraction submodule uses color differences based on a color sorter to screen and extract pure grain particles.

4. A novel practical screening system for customs ports according to claim 3, characterized in that: The physical sorting submodule includes a centrifugal screening unit, a vibration negative pressure suction unit, an acoustic standing wave airflow chromatography unit and an electrostatic adsorption cleaning unit. The centrifugal screening unit uses centrifugal force to separate particles of different sizes on the sieve material after screening the third and fourth layers of the four-stage sieve molecular module. The vibration negative pressure suction unit uses mechanical vibration and negative pressure suction technology to further separate and collect particles of different densities. The acoustic standing wave airflow chromatography unit uses acoustic standing waves of different frequencies and airflow chromatography technology to accurately separate particles of different sizes again. The electrostatic adsorption cleaning unit generates electrostatic adsorption based on an electrostatic generator to remove impurities on the surface of the sample.

5. The novel practical screening system for customs ports according to claim 1 is characterized by: The sample identification and retention module includes a sample image acquisition submodule and a neural network recognition submodule. The sample image acquisition submodule collects image information of the sample based on a high-resolution camera, and the neural network recognition submodule recognizes the sample target in the image based on the YOLO neural network and identifies the species level and then marks and saves it.

6. The novel practical screening system for customs ports according to claim 1, characterized in that: The data analysis module includes a data classification submodule, a pan-tilt data upload submodule and a database update submodule. The data classification submodule pre-processes the collected data and classifies the sample data and impurity data according to the recognition results. The pan-tilt data upload submodule is used to upload the classified data to the pan-tilt. The database update submodule is used to update the sample database and impurity database according to the classification results.

7. A novel practical screening system for customs ports according to claim 6, characterized in that: The data classification submodule includes a grain sample data storage unit and an impurity sample data storage unit. The grain sample data storage unit is used to store the screened grain sample data, and the impurity sample data storage unit is used to store the impurity sample data after separation and screening.

8. The novel practical screening system for customs ports according to claim 6, characterized in that: The database updating submodule includes an impurity sample database updating unit and a valid sample database updating unit. The impurity sample database updating unit is used to store impurity sample data outside the original database, and the valid sample database updating unit is used to store valid sample data outside the original database.

9. A screening method for a new type of practical screening system for customs ports according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step 1: Four-stage screening treatment: the sample to be screened is transported to the four-stage screening submodule through the feeding module, and four-stage screening is performed through four layers of sieves with different particle sizes to screen out the straw and screen out large weeds and pests; Step 2: Centrifugal screening treatment: the oversize material after the third and fourth layers of screening in step 1 is sent to the centrifugal screening unit and slowly centrifuged at a speed of 10-100 rpm to obtain the near-center target, the middle target and the edge target; Step 3: Vibration negative pressure suction treatment: the targets on the edge of the turntable after centrifugal screening in step 2 are mechanically vibrated at different frequencies and amplitudes, and collected using different suction forces to obtain high-frequency screening targets, medium-frequency screening targets, and low-frequency screening targets; Step 4: Acoustic standing wave airflow chromatography treatment, using acoustic waves of different frequencies to resonate and screen each of the three types of targets proposed after the vibration negative pressure absorption in step 3, and using air jets of different speeds from an air jet tube to collect the peak position corresponding to the standing wave formed by the acoustic wave to obtain high-frequency acoustic wave screened materials, medium-frequency acoustic wave screened materials, and low-frequency acoustic wave screened materials; Step 5: Electrostatic adsorption cleaning treatment: The multiple groups of materials obtained in steps 1 to 4 are spread on a conveyor belt, and each group of materials is electrostatically adsorbed and cleaned by an energized electrostatic generator to remove tiny impurities on the surface. The cleaned materials are identified and retained by the identification sample module to obtain the impurity-free materials; Step 6: Color sorting and extraction processing: the processed materials are separated according to their different characteristics through a color sorter, and finally subsequent image recognition is performed to complete the screening and sorting of the materials.

Images