Automated grading and screening method and system for deboning fish meat

Abstract

This application provides an automated grading and screening method and system for deboning fish meat. The method includes: segmented spiral extrusion of the fish meat to be processed according to extrusion control parameters to obtain a fish meat extrusion stream and a bone-rich residue stream; thin-layer dispersion treatment of the fish meat extrusion stream to generate a fish meat layer to be screened; controlling a multi-layer vibrating screen to perform particle size classification and screening according to the screening trajectory and screening intensity, and forming particle size outlet materials based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process; detecting the particle size state and residual bone risk of the particle size outlet materials to generate grading and screening results; updating the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen according to the coupling relationship between particle size state and residual bone risk in the grading and screening results, and outputting the graded fish meat material and bone-separated material according to the updated grading and screening path. This application can improve the coordination of deboning and grading, reduce residual bone risk, and improve grading accuracy.

Description

Technical Field

[0001] This application relates to the field of automated control technology for aquatic product processing, and in particular to an automated grading and screening method and system for deboning fish meat. Background Technology

[0002] In fish product processing, deboning and particle size grading are crucial steps affecting product safety, consistent taste, and subsequent molding quality. Current processing techniques typically employ bone and flesh separators, screw extrusion mechanisms, drum screens, or vibrating screens to separate the fish meat from the bone and sieve the particle size. Mechanical extrusion forces the fish meat tissue through the screen openings, while retaining fish skin, bones, and fascia. The fish meat particles are then graded according to the screen opening size. Some production lines also incorporate manual re-inspection or foreign object detection equipment to check for residual bones in the separated fish meat, reducing the risk of bone residue.

[0003] However, existing spiral extrusion and vibrating screening technologies mostly employ fixed process parameters, making it difficult to adaptively adjust based on the fish meat's texture, moisture content, bone distribution, and particle adhesion. When the extrusion intensity is insufficient, the fish meat recovery rate is low; when the extrusion intensity is too high, bones are easily crushed and enter the qualified material along with the fish meat. Meanwhile, traditional vibrating screens primarily screen based on particle size, failing to effectively distinguish between flexible fish meat particles and rigid bones; thin, elongated bones or fragments may be entrained by fish meat agglomerates and pass through the screen. Furthermore, particle size deviations and residual bone risks at the screening end typically cannot be reverse-adjusted to the front-end extrusion parameters, resulting in a lack of closed-loop coordination between deboning, grading, and reprocessing. Therefore, existing fish deboning methods still suffer from insufficient grading accuracy, unstable residual bone control, and weak automation capabilities. Summary of the Invention

[0004] In view of this, embodiments of this application provide an automated grading and screening method and system for deboning fish meat, in order to solve the problems of insufficient coordination between deboning and grading, unstable control of residual bones, and low grading accuracy in the prior art.

[0005] A first aspect of this application provides an automated grading and screening method for deboning fish meat, comprising: acquiring raw material state data of the fish meat to be processed; identifying the physical properties, tissue state, and bone distribution of the raw material state data to generate a raw material state characterization; determining the extrusion control parameters of a screw extrusion mechanism based on the raw material state characterization; and performing segmented screw extrusion on the fish meat to be processed according to the extrusion control parameters to obtain a fish meat extrusion stream and a bone-rich residue stream; performing thin-layer dispersion treatment on the fish meat extrusion stream to generate a fish meat layer to be screened; and controlling a multi-layer vibrating screen to perform grading and screening according to the specified parameters. The fish meat layer to be screened is classified into different particle sizes based on the screening trajectory and screening intensity. During the screening process, the particle size output material is formed based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones. The particle size state and residual bone risk of the particle size output material are detected, and a classification screening result containing particle size grade and residual bone risk grade is generated. According to the coupling relationship between particle size state and residual bone risk in the classification screening result, the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen are updated, and the fish meat graded material and bone separated material are output according to the updated classification screening path.

[0006] A second aspect of this application provides an automated grading and screening system for deboning fish meat, comprising: an acquisition module for acquiring raw material state data of the fish meat to be processed, performing physical property identification, tissue state identification, and bone distribution identification on the raw material state data, and generating a raw material state characterization; a determination module for determining the extrusion control parameters of the screw extrusion mechanism based on the raw material state characterization, and performing segmented screw extrusion on the fish meat to be processed according to the extrusion control parameters to obtain a fish meat extrusion flow and a bone-rich residue flow; a dispersion module for performing thin-layer dispersion treatment on the fish meat extrusion flow to generate a fish meat layer to be screened; and a screening module for controlling multiple... The multi-layer vibrating screen performs particle size classification and screening according to the screening trajectory and screening intensity corresponding to the layer of fish meat to be screened. During the screening process, it forms particle size outlet material based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones. The detection module is used to detect the particle size status and residual bone risk of the particle size outlet material, and generate classification and screening results containing particle size grade and residual bone risk grade. The output module is used to update the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen according to the coupling relationship between particle size status and residual bone risk in the classification and screening results, and output the fish meat graded material and bone separated material according to the updated classification and screening path.

[0007] The above-described technical solutions adopted in the embodiments of this application can achieve the following beneficial effects: By acquiring the raw material state data of the fish meat to be processed, the raw material state data is identified in terms of physical properties, tissue state, and bone distribution to generate a raw material state characterization. Based on the raw material state characterization, the extrusion control parameters of the screw extrusion mechanism are determined, and the fish meat to be processed is subjected to segmented screw extrusion according to the extrusion control parameters to obtain a fish meat extrusion flow and a bone-enriched residue flow. The fish meat extrusion flow is subjected to thin-layer dispersion treatment to generate a fish meat layer to be screened. A multi-layer vibrating screen is controlled to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened. During the screening process, particle size exit material is formed based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bone spurs. Particle size state detection and bone spur risk detection are performed on the particle size exit material to generate a classification screening result containing particle size grade and bone spur risk grade. Based on the coupling relationship between particle size state and bone spur risk in the classification screening result, the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen are updated, and the fish meat graded material and bone spur separated material are output according to the updated classification screening path. This application can improve the synergy of de-sting and grading, reduce the risk of residual stings, and improve grading accuracy. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic flowchart of the automated grading and screening method for deboning fish meat provided in the embodiments of this application; Figure 2 This is a schematic diagram of the structure of the automated grading and screening system for deboning fish meat provided in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0010] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0011] In fish processing, deboning and particle size grading typically rely on equipment such as bone and meat separators, screw extrusion mechanisms, drum screens, or vibrating screens. Current technology generally uses screw extrusion to force fish meat tissue through screen holes, retaining fish skin, bones, fascia, and other residues. A vibrating screen or mesh structure is then used to grade the fish meat according to particle size. Some processing lines also include manual re-inspection or foreign object detection equipment after separation to reduce the risk of residual bones in the fish meat.

[0012] However, existing fish deboning methods typically treat screw extrusion and vibrating screening as independent processes. The screw extrusion and vibrating screening parameters are often fixed, making it difficult to adaptively adjust them based on the fish's tissue structure, moisture content, bone distribution, and particle adhesion. Insufficient extrusion intensity results in low fish recovery; excessive extrusion intensity leads to bone crushing and inclusion in the fish material. Furthermore, traditional vibrating screens primarily rely on particle size for sieving, failing to effectively utilize the difference between the flexible deformation characteristics of fish particles and the rigidity retention characteristics of bones. This allows long, thin bones or fragments to enter the acceptable particle size range along with fish agglomerates. Therefore, existing technologies suffer from insufficient coordination between deboning and grading, unstable control of residual bones, and low grading accuracy.

[0013] To address the aforementioned problems, this application provides an automated grading and screening method for deboning fish meat. The method first acquires raw material state data of the fish meat to be processed, and then performs physical property identification, tissue state identification, and bone distribution identification on the raw material state data to generate a raw material state characterization. Next, based on the raw material state characterization, the extrusion control parameters of the screw extrusion mechanism are determined, causing the fish meat to be processed to form a fish meat extrusion flow and a bone-rich residue flow during the segmented screw extrusion process. Subsequently, the fish meat extrusion flow is subjected to thin-layer dispersion treatment to generate a fish meat layer suitable for screening, reducing the impact of fish meat agglomeration on bone separation and particle size classification.

[0014] Based on this, this application controls a multi-layer vibrating screen to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the layer of fish meat to be screened. During the screening process, the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones are utilized to form particle size outlet materials. Furthermore, particle size status detection and residual bone risk detection are performed on the particle size outlet materials to generate classification screening results containing particle size grade and residual bone risk grade. Based on the coupling relationship between particle size status and residual bone risk in the classification screening results, the extrusion control parameters and screening control parameters are updated in reverse, and the fish meat graded materials and bone separated materials are output according to the updated classification screening path.

[0015] Through the above technical solution, this application can form a continuous and synergistic processing chain integrating raw material state identification, segmented spiral extrusion, thin-layer dispersion, multi-layer vibrating sieving, particle size and residual bone coupling detection, and feedback control. This links the front-end extrusion process with the back-end sieving results, avoiding bone breakage or insufficient fish meat recovery due to processing with fixed parameters. This application can improve the synergy of deboning and grading, reduce the risk of residual bones, improve the accuracy of fish meat particle size grading, and enhance the automation control and batch adaptability of the fish meat deboning process.

[0016] The technical solution of this application will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0017] Figure 1 This is a schematic flowchart of the automated grading and screening method for deboning fish meat provided in an embodiment of this application. Figure 1 As shown, the method may specifically include: S101, Obtain the raw material state data of the fish meat to be processed, perform physical property identification, tissue state identification and bone distribution identification on the raw material state data, and generate raw material state characterization; S102, Determine the extrusion control parameters of the screw extrusion mechanism based on the raw material state characterization, and perform segmented screw extrusion of the fish meat to be processed according to the extrusion control parameters to obtain the fish meat extrusion stream and the bone and spine enrichment residue stream. S103, the fish meat extrusion flow is subjected to thin-layer dispersion treatment to generate a fish meat layer to be screened; S104 controls the multi-layer vibrating screen to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened, and forms particle size outlet material based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process. S105 performs particle size distribution and residual burr risk detection on the granular outlet material, generating a grading and screening result that includes particle size grade and residual burr risk grade. S106, based on the coupling relationship between particle size and residual bone risk in the grading and screening results, update the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen, and output the fish meat grading material and bone separation material according to the updated grading and screening path.

[0018] In some embodiments, raw material state data of the fish meat to be processed is acquired, and physical property identification, tissue state identification, and bone distribution identification are performed on the raw material state data to generate a raw material state characterization, including: Before the fish meat to be processed enters the spiral extrusion mechanism, multimodal detection data of the fish meat to be processed is collected, and the multimodal detection data is spatiotemporally aligned based on a unified processing coordinate to generate raw material detection data; Input the raw material testing data into the raw material state recognition model to extract the compressibility state, tissue integrity state, and spatial distribution state of the fish meat to be processed; Based on the correlation between compressibility, tissue integrity, and spatial distribution of bone spurs, a raw material state characterization is generated to determine the extrusion control parameters and screening control parameters.

[0019] Specifically, before the fish meat to be processed enters the screw extrusion mechanism, the system collects multimodal detection data of the fish meat through a multimodal detection component installed on the feeding conveyor section. The multimodal detection component may include a shape contour detection unit, a weight detection unit, a compression rebound detection unit, a temperature detection unit, and a bone spur imaging detection unit. The fish meat to be processed can be fish segments that have been deheaded, gutted, and cleaned, or pre-cut fish fillets.

[0020] For example, taking the processing of sea bass fillets as an example, after the fillets enter the detection area along the conveyor belt, the shape contour detection unit acquires the thickness, width and edge contour of the fillets, the weight detection unit acquires the mass of the material per unit length, the compression and rebound detection unit applies a predetermined pressure through a flexible pressure roller and collects the rebound displacement, the temperature detection unit acquires the temperature of the center of the fish meat, and the bone spur imaging detection unit acquires the image of the distribution of bone spurs inside the fillets, thereby forming multimodal detection data that includes geometric state, texture state, temperature state and bone spur imaging state.

[0021] After collecting multimodal detection data, the system establishes a unified processing coordinate system based on the conveyor belt centerline, the feed end reference point, and the material conveying direction, mapping the data collected by different detection units to this unified coordinate system. Due to the installation distance differences between the detection units along the conveying direction, the system performs time-series compensation on the detection data based on the conveying speed, acquisition timestamp, and encoder pulse signal, ensuring that the detection results for the same fish meat area under different detection units correspond to the same coordinate position. For suspected bone spur areas output by the bone spur imaging detection unit, the system performs spatial correction based on the fish fillet edge contour and thickness distribution, eliminating mismatched areas caused by fish fillet warping, overlap, or surface water film, and generating raw material detection data.

[0022] Furthermore, the system inputs the raw material testing data into the raw material state recognition model. The raw material state recognition model may include a feature extraction network and a state fusion network. The feature extraction network extracts representation vectors from the geometric state, compression and rebound state, temperature state, and bone spur imaging state, respectively. The state fusion network generates the compressibility state, tissue integrity state, and bone spur spatial distribution state based on the correspondence between the representation vectors.

[0023] In some examples, compressibility is used to characterize the ease with which fish fillets are extruded through a sieve in a spiral extrusion mechanism; tissue integrity is used to characterize whether the fish fillet fibers are loose, broken, or adhered; and the spatial distribution of bone spurs is used to characterize the dense areas, extension directions, and integration of bone spurs with the fish fillet tissue. For example, when the fish fillet is thick, has a small rebound displacement, and a low center temperature, the model determines that the fish fillet tissue is relatively firm and compressibility is low; when the bone spur imaging area is concentrated near the spine and is distributed in long, continuous strips, the model determines that the risk of bone spur breakage is high.

[0024] After obtaining the above states, the system generates a raw material state characterization based on the correlation between the compressibility state, tissue integrity state, and bone spur spatial distribution state. The raw material state characterization can be recorded in the form of a state vector or a combination of state labels, and a mapping relationship can be established with subsequent extrusion control parameters and screening control parameters. For example, for fish batches with low compressibility but a high risk of bone spur breakage, the raw material state characterization can indicate the use of a lower initial extrusion intensity, a slower feed rate, and a longer screening residence time; for fish batches with loose tissue and strong adhesion, the raw material state characterization can indicate the need to increase the thin-layer dispersion intensity and adjust the screening trajectory of the multi-layer vibrating screen.

[0025] By using the above method, a raw material state characterization oriented towards processing control can be formed before the fish meat enters the spiral extrusion mechanism. This allows the subsequent extrusion control parameters and screening control parameters to no longer rely on fixed empirical values, but to be adapted to the actual tissue state and bone distribution state of the fish meat. This improves the synergy of deboning and grading, reduces the risk of bone crushing entering qualified fish meat material, and enhances the stability of subsequent particle size grading and residual bone control.

[0026] In some embodiments, the extrusion control parameters of the screw extrusion mechanism are determined based on the raw material state characterization, and the fish meat to be processed is subjected to segmented screw extrusion according to the extrusion control parameters to obtain a fish meat extrusion stream and a bone-rich residue stream, including: Based on the characterization of the raw material state, the compressibility range of the fish meat to be processed and the risk of bone breakage are determined, and extrusion control parameters including pre-compression introduction parameters, selective extrusion parameters and residue discharge parameters are generated. According to the extrusion control parameters, the screw extrusion mechanism is controlled to form a progressive extrusion action along the material conveying direction, so that the fish meat tissue is squeezed out through the screen cylinder and the bones are restricted from entering the outside of the screen cylinder along with the fish meat tissue. During the segmented spiral extrusion process, the extrusion operation status is collected, and the extrusion intensity and residue discharge status of the corresponding segments are adjusted according to the extrusion operation status to obtain the fish meat extrusion flow and the bone and spine enriched residue flow.

[0027] Specifically, after the raw material state characterization is generated, the control system calls the extrusion parameter generation rules corresponding to the current batch of fish meat. The raw material state characterization includes extrudability state, tissue integrity state, and bone splinter spatial distribution state. The system first determines the extrudable range of the fish meat to be processed in the spiral extrusion mechanism based on the extrudability state, and then determines the risk of bone splinter breakage based on the bone splinter spatial distribution state.

[0028] For example, the current batch of semi-frozen sea bass fillets has an average thickness of 18mm, small rebound displacement, and concentrated bone fragments near the spine. The system determines that the fish meat tissue is relatively dense and the risk of bone fragment breakage is high. Therefore, the compressible range is limited to a medium-low pressure range, and extrusion control parameters are generated with slow feeding, low shear, and smooth slag discharge as constraints. For fillets with loose tissue and high moisture content, the system reduces the compaction amount in the inlet section and reduces fish meat gelatinization by increasing the intensity of liquid discharge and steady flow control.

[0029] The extrusion control parameters include pre-compression parameters, selective extrusion parameters, and residue discharge parameters. Pre-compression parameters control the feed rate, inlet pressure, and the adhesion of the fish meat before it enters the screen cylinder, ensuring a stable material layer forms before entering the high-load area. Selective extrusion parameters control the screw shaft speed, the pressure inside the screen cylinder, and the extrusion rhythm of the fish meat through the screen holes, allowing flexible fish meat tissue to pass through the screen cylinder while reducing the probability of bones being crushed near the screen holes. Residue discharge parameters control the discharge port opening, residue conveying resistance, and residue residence time, ensuring that coarse bones, fish skin, fascia, and bone spur accumulations are discharged along the inside of the screen cylinder, rather than entering the outside of the screen cylinder with the extruded fish meat flow.

[0030] In the segmented screw extrusion process, the screw extrusion mechanism sequentially forms a pre-compression inlet section, a selective extrusion section, and a residue discharge section along the material conveying direction. The pre-compression inlet section compacts the fish meat to be processed and adjusts the contact state between the fish meat and the screen cylinder, preventing localized accumulation of fish meat at the inlet. The selective extrusion section applies progressive extrusion according to the compressible range, allowing the fish meat tissue to pass through the screen cylinder under stable pressure, forming a fish meat extrusion flow. The residue discharge section guides the bone fragments that have not passed through the screen cylinder, enriching the residue flow, to the slag discharge channel, and determines whether the residue is allowed to enter the low-pressure reprocessing path based on its meat content.

[0031] Furthermore, the control system continuously collects the extrusion operation status during the segmented screw extrusion process. This status includes screw shaft load, screen cylinder pressure, extrusion flow rate, residue discharge load, and extruded particle size distribution. If the screen cylinder pressure increases and the extrusion flow rate decreases, the system determines that the material is showing signs of blockage in the selective extrusion section and reduces the feed rate and increases the discharge port opening. If the screw shaft load suddenly increases and the residue discharge load decreases, the system determines that bone fragments may be causing blockage inside the screen cylinder and reduces the extrusion intensity to prevent bone fragments from breaking and mixing into the fish meat extrusion stream. If the fish meat content in the residue is high and the risk of bone fragment breakage is low, the system appropriately increases the pressure in the pre-compression inlet section or guides the residue into the low-pressure secondary extrusion path.

[0032] By employing the above method, the segmented operating parameters of the screw extrusion mechanism can be dynamically determined and modified based on the raw material state characterization, enabling stable extrusion of fish meat tissue under controlled pressure and ensuring that bone and spine-enriched residues are concentrated and discharged along the slag discharge path. This embodiment can reduce the risk of bone and spine breakage caused by excessive extrusion, improve the synergy between the fish meat recovery process and the bone and spine separation process, and provide a more stable fish meat extrusion stream with lower particle size and bone and spine risk for subsequent thin-layer dispersion and multi-layer vibrating screening, facilitating continuous and stable production.

[0033] In some embodiments, the fish meat extrusion stream is subjected to a thin-layer dispersion treatment to generate a fish meat layer to be screened, including: The continuity and adhesion state of the fish meat extrusion flow are obtained, and the thin-layer dispersion parameters are determined based on the continuity and adhesion state. The fish meat extrusion flow is controlled and flexibly spread according to the thin-layer dispersion parameters, so that the fish meat extrusion flow is transformed from an agglomerated state to a continuous thin-layer state. The thickness of the fish meat material in a continuous thin layer is balanced and the feeding rhythm is matched to generate a layer of fish meat to be screened that is compatible with the feeding area of ​​the multi-layer vibrating screen.

[0034] Specifically, a fish meat extrusion receiving section is set on the outside of the screen cylinder of the screw extrusion mechanism. After the fish meat extrusion is continuously discharged from the screen cylinder holes, it enters the thin-layer dispersion module. The thin-layer dispersion module may include a discharge status detection unit, a low-temperature material distribution trough, a flexible spreading component, and a feeding matching component. The discharge status detection unit is arranged in the discharge area on the outside of the screen cylinder and is used to collect the flow fluctuation, discontinuity interval, particle size, surface adhesion degree, and local accumulation height of the fish meat extrusion.

[0035] For example, taking the processing of sea bass fillets as an example, when the fish meat extrusion flow is continuous and the particle diameter is concentrated between 8mm and 15mm, the system determines that the discharge continuity is good; when the fish meat extrusion flow forms a blocky accumulation at the discharge port and there is obvious fiber entanglement between adjacent particles, the system determines that the adhesion state is strong and the dispersion treatment intensity needs to be increased.

[0036] The system determines the thin-layer dispersion parameters based on the continuity of discharge and the adhesion state. These parameters may include the conveyor speed of the feeding trough, the downward pressure of the spreading components, the spreading differential speed, the vibration intensity of the spreading material, the intensity of the low-temperature spraying, and the feeding cycle before entering the multi-layer vibrating screen. For fish extrusion streams with high moisture content and strong adhesion, the system reduces the conveyor speed of the feeding trough, increases the spreading differential speed and the vibration intensity of the spreading material, allowing the adhered fish meat clumps to be broken up under low shear. For fish extrusion streams with looser texture and easier breakage, the system reduces the downward pressure of the spreading components and uses intermittent light vibration to prevent further crushing of the fish meat particles, which would affect subsequent particle size classification.

[0037] During the controlled spreading process, the extruded fish meat first falls into a low-temperature spreading trough. The bottom of the trough vibrates slightly according to a preset rhythm, causing the material to spread evenly laterally. Subsequently, a flexible spreading component gently spreads the extruded fish meat. The flexible spreading component can use food-grade elastic rollers or flexible scrapers. The roller speed and the conveying speed of the spreading trough are differentially related, gradually transforming the blocky fish meat clumps into a continuous thin layer. During the spreading process, a low-temperature spray unit applies an atomized cooling medium to the surface of the fish meat to reduce surface adhesion and maintain a stable temperature of the fish meat tissue, preventing the fish meat from gelatinizing, becoming more clump-like, or having bones encased in the fish meat clumps due to temperature rise.

[0038] After the fish meat material forms a continuous thin layer, the system uses a thickness detection unit to detect the thickness of the material layer at different lateral positions and adjusts the posture of the spreading components and the vibration intensity of the feeding trough according to the thickness deviation. For example, when the thickness in the middle of the material layer reaches 25mm while the thickness on both sides is less than 10mm, the system increases the lateral dispersion vibration and adjusts the inclination angle of the flexible scraper to expand the accumulated material in the middle to both sides; when the thickness of the material layer in a certain area is lower than the preset feed thickness, the system reduces the feeding speed of the corresponding area to maintain a stable feed load in the subsequent screening area. After thickness equalization, the fish meat material forms a relatively uniform thickness, reduced agglomeration, and stable feeding direction for screening.

[0039] Furthermore, the feeding matching component adjusts the output rhythm of the fish meat layer to be screened based on the real-time load status of the feeding area of ​​the multi-layer vibrating screen. When the inlet load of the multi-layer vibrating screen increases, the feeding matching component reduces the output speed of the distribution trough to prevent the accumulation of the fish meat layer to be screened; when the inlet load of the multi-layer vibrating screen decreases and there is an unloaded area on the screen surface, the feeding matching component increases the output rhythm to ensure that the fish meat layer to be screened continuously covers the effective area of ​​the screen surface. Through feeding rhythm matching, the fish meat layer to be screened can enter the multi-layer vibrating screen with a stable thickness and continuous state, providing stable feeding conditions for subsequent grading and screening using the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones.

[0040] This embodiment reduces the impact of fish meat agglomeration and adhesion on the screening process before the fish meat extrusion stream enters the multi-layer vibrating screen, transforming the irregular agglomerated state of the fish meat extrusion stream into a continuous thin layer with uniform thickness. This treatment can improve the effective screening area utilization rate of the multi-layer vibrating screen, reduce the probability of bones passing through the screen with fish meat agglomerates, and enhance the stability of particle size classification screening and residual bone risk detection.

[0041] In some embodiments, a multi-layer vibrating screen is controlled to perform particle size classification screening according to a screening trajectory and screening intensity corresponding to the layer of fish meat to be screened, and particle size outlet material is formed based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process, including: The layered screening control parameters of the multi-layer vibrating screen are determined based on the spreading state and granule state of the fish meat layer to be screened. The multi-layer vibrating screen is controlled according to the layered screening control parameters to form a continuously changing screening trajectory, so that the fish meat particles are loosened, turned over and deformed under the action of vibration. Based on the deformable passage state of fish meat particles and the rigid obstruction state of bone spurs, the passing and intercepted materials of each screen layer are diverted to form the particle size outlet material corresponding to the particle size grade.

[0042] Specifically, before the fish meat layer to be screened enters the multi-layer vibrating screen, the control system first acquires the spreading state and particle cluster state output by the thin-layer dispersion module. The spreading state can be obtained by the vision detection unit and thickness detection unit set above the screening inlet, which is used to characterize the coverage, thickness uniformity and feeding continuity of the fish meat layer in the width direction of the screen surface; the particle cluster state can be obtained by image segmentation and contour recognition, which is used to characterize the agglomeration size, adhesion degree and fiber entanglement state of the fish meat particles.

[0043] For example, taking sea bass meat processing as an example, when the thickness of the inlet fish meat layer is stable at 12mm to 18mm and most of the particle outlines are clear, the system determines that the spreading state is good and conventional layered screening intensity can be used; when there is local accumulation in the inlet area with a thickness of more than 25mm and there are long fibers connecting the fish meat particles, the system determines that the particle state is too sticky and it is necessary to increase the loosening strength and residence time of the primary screening layer.

[0044] The system determines the stratified screening control parameters of the multi-layer vibrating screen based on the spreading and agglomeration states. These parameters include the vibration frequency, amplitude, screen inclination angle, material residence time, lateral disturbance intensity, and outlet flow guidance for each screen layer. The first screen layer primarily handles loosening and coarse grading. The control system creates a large longitudinal throwing component in the first screen layer, causing fish meat agglomerates to detach from the screen surface and fall back down, thus weakening the adhesion between fish meat particles. The second screen layer primarily handles particle size stabilization and grading. The system reduces the throwing component and increases lateral disturbance, causing fish meat particles to roll and tumble on the screen surface. The third screen layer primarily handles fine particle screening and retaining residual hard materials. The system extends the residence time on the screen surface, allowing fine fish meat particles to pass fully through the target screen openings.

[0045] In the actual screening process, the multi-layer vibrating screen forms a continuously changing screening trajectory according to the layered screening control parameters. The fish meat layer to be screened is first subjected to strong vibration in the first screen layer, and the lumps of fish meat gradually loosen into smaller particles. Then, when entering the second screen layer, the fish meat particles undergo orientation flipping under the combined vibration, so that fish meat particles in different directions have the opportunity to align with the screen holes. After entering the third screen layer, the flexible fish meat particles undergo slight deformation due to their own water content and fiber elasticity, and enter the next collection channel through the corresponding screen holes. For rigid or semi-rigid materials such as bones, bone fragments, and fascia with bones, due to their slender shape and weak bending deformation ability, they are prone to overlapping, lying horizontally, or getting stuck with the edge of the screen holes under the continuously changing screening trajectory, thus remaining on the surface of the corresponding screen layer or being guided to the interception outlet.

[0046] Furthermore, to improve the stability of the flow distribution, each screen layer outlet is equipped with a through material channel and a retaining material channel. The system adjusts the position of the outlet guide plate according to the through load, retaining load, and suspected bone splinter aggregation status of each screen layer. The first screen layer retains mainly larger fish meat clumps, fish skin fascia, and coarse bone fragments, which can be guided into the coarse particle return screening or bone splinter enrichment and separation path. The second screen layer passes through to form medium-sized outlet material, and the retained material can enter the coarse fish meat channel when the risk of residual bones is low, and enter the reprocessing channel when the risk of residual bones is high. The third screen layer passes through to form fine-sized outlet material, and the retained material enters the residual bone verification or bone splinter separation path. Thus, the through material and retained material of each screen layer are separated according to particle size and rigid-flexible motion differences, forming coarse, medium, fine, and bone splinter-enriched outlet materials.

[0047] In this embodiment, the multi-layer vibrating screen no longer relies solely on fixed screen aperture size for passive screening. Instead, it dynamically generates layered screening control parameters based on the actual spreading and agglomeration state of the fish meat layer to be screened. This allows the fish meat particles to loosen, tumble, and deform during vibration, while bones are trapped or expelled due to rigidity. This method improves the accuracy of fish meat particle size grading, reduces the risk of bones entering the qualified particle size range with fish meat agglomerates, and enhances the adaptability of the vibrating screening process to different batches of fish meat.

[0048] In some embodiments, the multi-layer vibrating screen is controlled according to the layered screening control parameters to form a continuously changing screening trajectory, so that the fish meat particles are loosened, turned over, and deformed under the vibration and pass through, including: Based on the thickness of the fish meat layer to be screened on the adjacent screen layers and the passing load, determine the vibration excitation parameters that vary continuously along the conveying direction of the screen surface; Drive the multi-layer vibrating screen according to the vibration excitation parameters to form a composite screening trajectory, so that the fish meat particles will sequentially produce a loose state, a flipped state, and a deformed passing state under the action of the composite screening trajectory. Based on the difference in movement between fish meat particles and bones under the composite screening trajectory, the bones are controlled to remain in the corresponding screening layer or guided to the corresponding interception outlet.

[0049] Specifically, the multi-layer vibrating screen can adopt a screening structure driven by partitions along the material conveying direction. Each screen layer is equipped with an independent vibration excitation unit, a screen surface load detection unit, and an outlet flow guiding unit. After the fish meat layer enters the first screen layer, the system obtains the spreading thickness of the fish meat layer in the screen width direction and the conveying direction through the inlet thickness detection unit, and obtains the through load between adjacent screen layers through the underflow detection unit.

[0050] For example, taking sea bass meat processing as an example, when the thickness of the first screen layer inlet is 20mm to 28mm and the underpass throughput is lower than the target throughput, the system determines that the meat layer has a local accumulation and adhesion state, and it is necessary to increase the vibration excitation of the inlet section; when the second screen layer has an excessively high load and the inlet of the third screen layer forms a continuous pile, the system determines that the speed of the previous screen layer is too fast, and it is necessary to reduce the conveying speed in the middle section and extend the turning and screening time.

[0051] The system determines continuously varying vibration excitation parameters along the conveying direction of the screen surface based on the spreading thickness and the load. These parameters can include the excitation frequency, vibration amplitude, vibration direction angle, longitudinal conveying component, and transverse disturbance component for different screen areas. In the inlet area, to meet the requirement of loose fish meat, the system sets a larger vertical throwing component, causing fish meat clumps to repeatedly detach from the screen surface and fall back down, thereby weakening the adhesion between fish meat fibers. In the middle area, to meet the need for posture adjustment, the system reduces the vertical throwing component and increases the transverse disturbance component, causing fish meat particles to roll, deflect, and flip on the screen surface. In the outlet area, to meet the need for fine passage, the system reduces strong impact vibration and extends the residence time on the screen surface, allowing flexible fish meat particles to deform moderately under the action of the screen hole edges and pass through the corresponding screen holes.

[0052] During the driving process, the multi-layer vibrating screen forms a composite screening trajectory according to the vibration excitation parameters. This composite screening trajectory is not a single up-and-down vibration trajectory, but a continuous trajectory formed by throwing, lateral movement, buffer contact, and directional conveying. Fish meat particles first change from an agglomerated state to a loose state in the inlet area, then undergo posture changes with the composite movement of the screen surface in the middle area, and finally deform and pass through the outlet area according to their own flexibility and water content. For long, thin bone fragments, bone chips, and fascia with bone, due to their high rigidity and weak bending deformation ability, they are difficult to deform and pass through the screen holes under the action of the composite screening trajectory, and are easily held on the screen surface by being placed horizontally, overlapping, or stuck at the ends.

[0053] Furthermore, the system controls the retention path based on the difference in motion between fish meat particles and bones along the composite screening trajectory. If a long, hard target is detected on the screen surface and moves at a speed lower than the average speed of the fish meat particles along the conveying direction, and the target fails to pass through the screen holes multiple times, the target is identified as a bone retention target, and the vibration direction of the corresponding area is deflected towards the retention outlet, guiding the bone to the retention outlet. If the retained material has a high fish meat content and a low risk of residual bones, the system guides the retained material into the return screening path for redispersing; if the density of hard targets in the retained material is high, the system guides the retained material into the bone enrichment and separation path to avoid repeated impacts on the screen surface that could cause the bone to break.

[0054] This embodiment allows for the continuous variation of vibration excitation parameters based on the spreading thickness of adjacent screen layers and the dynamic load, enabling fish meat particles to sequentially undergo loosening, tumbling, and deformation on a multi-layer vibrating screen. This also ensures that bones are held in the screen layers or guided to the outlet due to rigidity. This method improves the stability of fish meat particle size grading, reduces the risk of long, thin bones entering the qualified particle size range along with the fish meat particles, and enhances the adaptability of the multi-layer vibrating screen to fish meat materials with different adhesion states and particle sizes.

[0055] In some embodiments, particle size distribution and residual burr risk are detected in the effluent material to generate a grading and screening result that includes particle size class and residual burr risk class, including: Collect online detection data of the outlet material of each particle size, and correlate the online detection data with the corresponding sieve layer and outlet position to generate an outlet material detection characterization; The particle size distribution of fish meat particles is identified based on the detection and characterization of exported materials, and the corresponding particle size level is determined according to the preset particle size determination rules. Based on the detection and characterization of exported materials, suspected bone spur features are identified, and the risk assessment of suspected bone spur features is carried out in combination with particle size classification to determine the corresponding residual bone spur risk level. Based on the matching relationship between particle size class and residual burr risk level, a grading and screening result is generated for material diversion and parameter feedback.

[0056] Specifically, online detection components are installed at the outlets of each particle size group of the multi-layer vibrating screen. These components include an outlet weighing unit, an image acquisition unit, a transmission detection unit, and a material position tracking unit. After exiting the corresponding screen layer, the material from each particle size group first passes through a short-distance flat conveying section, transforming the fish meat particles from an accumulated state to a detectable state. Taking sea bass processing as an example, the coarse particle size outlet mainly outputs larger fish meat particles and a small amount of fascia residue; the medium particle size outlet mainly outputs fish meat particles with relatively stable particle sizes; and the fine particle size outlet mainly outputs finely broken fish meat and scrambled fish material. The system acquires material images, transmission responses, output mass per unit time, and outlet position signals at each outlet location, and correlates the detection data with the screen layer number, outlet channel, acquisition time, and corresponding processing batch to generate an outlet material detection characterization.

[0057] During particle size distribution detection, the system performs particle segmentation and contour analysis on the outgoing material to extract the equivalent particle size, area distribution, length-to-width ratio, edge continuity, and agglomeration degree of the fish meat particles. Based on preset particle size determination rules, the system determines the corresponding particle size grade. For example, if the equivalent particle size of most fish meat particles in the medium-grained outgoing material is within the preset medium-grained range, and the proportion of oversized particles is below the allowable threshold, the system determines that the outgoing material meets the medium-grained grade. When many large particles or fiber-linked clusters appear in the fine-grained outgoing material, the system determines that the outgoing material has a coarse particle size and writes a particle size deviation indicator in the grading and screening results. For materials with unclear particle boundaries due to high moisture content, the system combines transmission response and changes in discharge quality to correct the image segmentation results, avoiding misjudging adhered fish meat as single large particles.

[0058] During the residual bone risk detection process, the system identifies suspected bone spur features based on the characteristics of the exported material. Suspected bone spur features can be comprehensively identified through elongated hard edges, localized high transmission attenuation, tip contours, and movement and retention characteristics different from fish meat particles. For coarse-grained outlets, the system focuses on whether elongated bone spurs are encased in large fish meat clumps; for medium-grained outlets, the system focuses on whether bone fragments pass through the sieve layer with the fish meat particles; for fine-grained outlets, the system focuses on whether small bone spurs or broken bone spur segments are mixed with fine fish meat particles. The system combines suspected bone spur features with particle size classification for risk assessment. If fine-grained outlets have small, high-attenuation targets, even if the number of targets is small, they will be assigned a higher residual bone risk level; if coarse-grained outlets have low-density suspected targets and low fish meat aggregation, they can be determined to have a medium-low residual bone risk level and enter the review path.

[0059] Furthermore, the system generates grading and screening results based on the matching relationship between particle size class and residual bone risk level. The grading and screening results can include particle size outlet identifier, particle size class, residual bone risk level, particle size deviation status, residual bone deviation status, and recommended diversion path. When a certain outlet material simultaneously meets the target particle size class and low residual bone risk level, the system marks the material as qualified fish meat grading material; when the material particle size meets the requirements but the residual bone risk is high, the system marks the corresponding material as residual bone risk rework material; when the material particle size deviates and the residual bone risk is low, the system marks the corresponding material as particle size adjustment rework material; when the material simultaneously has abnormal particle size and high residual bone risk, the system imports the corresponding material into the bone enrichment and separation or manual verification path.

[0060] This embodiment allows for the simultaneous acquisition of particle size and residual bone risk status at the outlet of each particle size class. The detection results are then correlated with the sieve layer and outlet position, ensuring that the grading and screening results reflect not only the size of the fish meat particles but also the safety status of residual bones in the corresponding particle size class. This method improves the reliability of fish meat grading results, reduces the risk of bones being output with qualified material, and provides accurate data for closed-loop feedback of subsequent extrusion control parameters and screening control parameters.

[0061] In some embodiments, based on the coupling relationship between particle size distribution and residual bone risk in the grading and screening results, the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen are updated, and the fish meat grading material and bone-separated material are output according to the updated grading and screening path, including: Based on the grading and screening results, identify the particle size deviation and residual burr risk deviation of the outlet material of each particle size class, and generate screening feedback characterization. Input the screening feedback characterization into the closed-loop control model to determine the extrusion correction amount corresponding to the screw extrusion mechanism and the screening correction amount corresponding to the multi-layer vibrating screen. The extrusion control parameters are updated based on the extrusion correction amount, and the screening control parameters are updated based on the screening correction amount. Adjust the grading and screening path according to the updated extrusion control parameters and screening control parameters. The material that meets the matching conditions of particle size grade and residual bone risk level will be output as fish meat graded material, and the material that does not meet the matching conditions will be output as bone and spine separation material or rework material.

[0062] Specifically, after particle size distribution and residual burr risk detection are completed for each particle size class of material at the outlet, the control system correlates the grading and screening results with the screen layer number, outlet channel, processing time segment, and current screw extrusion operating parameters to form basic data for closed-loop control. The grading and screening results include both the particle size grades corresponding to coarse, medium, and fine particle sizes, and the residual burr risk grades corresponding to each particle size class. The system first compares the deviation between the actual particle size grade and the target particle size grade, and then compares the deviation between the actual residual burr risk grade and the target residual burr risk grade, thereby identifying the particle size deviation and residual burr risk deviation states.

[0063] For example, during the processing of sea bass, if the proportion of fish meat clumps exceeding the target particle size range increases in the medium-particle stage outlet, while the risk of residual bones remains low, the system determines that there is a deviation mainly due to coarse particle size; if small, high-attenuation hard targets appear in the fine-particle stage outlet, and the particle size distribution basically meets the requirements, the system determines that there is a deviation mainly due to a high risk of residual bones; if large aggregates and long strips of suspected bone spurs are present in the coarse-particle stage outlet, the system determines that there is a combined deviation.

[0064] After identifying the deviation, the system generates a screening feedback characterization based on the degree of deviation, duration of deviation, and corresponding screen layer position of the material exiting each particle size. The screening feedback characterization can be recorded in the form of a feedback vector, including the direction of particle size deviation, the level of residual burr risk deviation, the abnormal exit position, the proportion of abnormal material, and the current processing load. The screening feedback characterization is used to indicate whether the current deburring and grading process is caused by abnormal front-end extrusion, insufficient thin-layer dispersion, or mismatched screening intensity at the rear end. If the residual burr risk continuously increases at the fine particle size exit while the discharge of bone-rich residue decreases, the system determines that the front-end selective extrusion section may have excessive extrusion or poor slag discharge; if the particle size at the coarse particle size exit is relatively coarse but the residual burr risk is low, the system determines that the thin-layer dispersion or the primary screen layer's looseness intensity is insufficient.

[0065] Furthermore, the system inputs the screening feedback characteristics into the closed-loop control model. Based on historical batch processing records and the current screening feedback characteristics, the closed-loop control model outputs the extrusion correction amount corresponding to the screw extrusion mechanism and the screening correction amount corresponding to the multi-layer vibrating screen. The extrusion correction amount is used to adjust the screw shaft speed, selective extrusion section pressure, feed cycle time, and discharge port opening; the screening correction amount is used to adjust the vibration frequency, vibration amplitude, screen surface conveying speed, and outlet flow guidance state of each screen layer.

[0066] For example, when the risk of residual spikes at the fine-particle outlet is high, the closed-loop control model outputs corrections to reduce the selective extrusion section pressure, increase the residue discharge outlet opening, and extend the intermediate screen layer residence time, in order to reduce the amount of spikes that are crushed and enter the fine-particle material. When the particle size at the intermediate-particle outlet is coarse and the risk of residual spikes is low, the model outputs corrections to increase the loosening strength of the first screen layer and moderately increase the matching speed of the thin-layer material distribution, in order to improve particle dispersion and throughput stability.

[0067] The control system updates the extrusion control parameters based on the extrusion correction amount and the screening control parameters based on the screening correction amount. Parameter updates can be performed using a tiered, limited-amplitude approach, that is, adjusting parameters gradually while ensuring that the fish meat tissue is not excessively sheared and the bones are not continuously crushed, avoiding sudden changes in material state caused by a large-scale adjustment at once. After the update, the system synchronously adjusts the grading and screening paths. Materials that meet the matching conditions of particle size grade and residual bone risk level are guided to the corresponding fish meat grading material outlet; materials with qualified particle size but high residual bone risk are guided to the residual bone verification or secondary screening path; materials with deviated particle size but low residual bone risk are guided to the return screening path; materials with high residual bone risk and obvious bone enrichment characteristics are guided to the bone separation path.

[0068] This embodiment allows for the coordinated adjustment of the screw extrusion mechanism and multi-layer vibrating screen by using particle size distribution and residual bone risk status as common feedback criteria. This enables dynamic correction of the front-end extrusion intensity, residue discharge status, and rear-end screening trajectory based on actual screening results. This method improves the synergistic control between deboning and grading, reduces the risk of bone crushing or inclusion in qualified materials, and enhances grading accuracy and continuous production stability across different batches of fish.

[0069] In some embodiments, materials that do not meet the matching criteria are output as bone splinter separation materials or rework materials, including: Based on the particle size deviation, residual bone risk deviation, and fish meat recyclability status of materials that do not meet the matching conditions, an abnormal material diversion characterization is generated. Based on the abnormal material diversion characterization, it is determined whether the corresponding material belongs to the fish meat enrichment and rework state or the bone and spine enrichment and separation state. When the corresponding material is in a fish meat enrichment and rework state, the corresponding material is introduced into the low-pressure secondary extrusion path or the secondary vibrating screening path. When the corresponding material is in a state of bone spike enrichment and separation, the corresponding material is guided into the bone spike separation path and the corresponding material is restricted from re-entering the screw extrusion mechanism.

[0070] Specifically, after the grading and screening results are generated, the control system re-collects and identifies the status of each abnormal outlet material. Abnormal outlet materials can come from coarse-grained outlets, medium-grained outlets, fine-grained outlets, intercepted outlets, or residual bone verification outlets. The system first reads the particle size deviation status, residual bone risk deviation status, and fish meat recyclability status corresponding to the abnormal outlet materials. Among them, the particle size deviation status is used to characterize whether the abnormal material is coarser, finer, or agglomerated relative to the target particle size; the residual bone risk deviation status is used to characterize the number of suspected bone spurs, the density of hard targets, and the degree of bone spur enrichment; and the fish meat recyclability status is used to characterize the proportion of usable fish meat tissue in the abnormal material, the integrity of the fish meat tissue, and the feasibility of reprocessing.

[0071] For example, taking sea bass meat processing as an example, when the particle size of the middle-grade outlet material is relatively coarse but has few suspected bone-like features, and the fish meat tissue still has an intact granular structure, the system determines that the abnormal material has a high fish meat recyclability. When there are many small, high-attenuation hard targets in the fine-grade outlet material, and the fish meat tissue has become fragmented, the system determines that the abnormal material has a high risk of residual bones and a low fish meat recyclability.

[0072] The control system generates an abnormal material diversion characterization based on particle size deviation, residual bone risk deviation, and fish meat recyclability. This characterization can be recorded as a state vector and correlated with the abnormal outlet location, processing batch, front-end extrusion parameters, and screening control parameters. If the abnormal material primarily exhibits particle size deviation and the residual bone risk is below the set risk level, the system marks the abnormal material diversion characterization as a fish meat enrichment and rework state. If the abnormal material has a high density of hard targets and suspected bone spikes are concentrated or continuously distributed, the system marks the abnormal material diversion characterization as a bone spike enrichment and separation state. If the abnormal material simultaneously has a high fish meat content and a high residual bone risk, the system further determines whether to enter the low-pressure rework path or directly enter the bone spike separation path based on the fish meat recyclability and bone spike breakage risk.

[0073] When abnormal material is in a rework state due to fish meat enrichment, the control system selects either a low-pressure secondary extrusion path or a secondary vibrating screening path based on the particle size deviation. For materials with coarser particle sizes and intact fish meat tissue, the system guides the abnormal material into the secondary vibrating screening path. By reducing the feed thickness, extending the screen residence time, and increasing the loosening vibration intensity, the fish meat agglomerates are redispersed and reclassified. For materials with a large amount of fish meat entrained in the residue but a low risk of bone breakage, the system guides the abnormal material into the low-pressure secondary extrusion path. Under pressure conditions lower than the initial selective extrusion section, the material is slowly extruded, allowing the remaining fish meat tissue to be discharged through the screen cylinder while preventing further crushing of bones. The material after secondary processing re-enters particle size detection and residual bone risk detection, and is not directly incorporated into the qualified material outlet.

[0074] When abnormal material falls into the bone spike enrichment and separation state, the control system guides the abnormal material into the bone spike separation path. The bone spike separation path can include a bone spike enrichment collection channel, a hard foreign object verification channel, and a waste discharge channel. For materials with high bone spike density and low fish meat recyclability, the system directly guides the material to the bone spike enrichment collection channel and generates a "no rework" flag. After the "no rework" flag is written into the batch processing record, the control system restricts the material from re-entering the screw extrusion mechanism to prevent bone spikes from breaking into smaller fragments during repeated extrusion and mixing into the fish meat extrusion stream. For materials that still contain a small amount of recyclable fish meat but whose bone spike risk is uncertain, the system can guide them into the hard foreign object verification channel for verification before determining whether to enter the low-pressure rework path.

[0075] This embodiment allows for differentiated sorting of abnormal materials based on particle size deviation, residual bone risk, and fish meat recyclability, avoiding the direct re-compression or simple rejection of all abnormal materials. This method improves the utilization rate of recyclable fish meat, reduces the risk of bones breaking and mixing into qualified materials due to repeated compression, and enhances the safety, stability, and automated coordination of rework and bone separation processes in fish meat deboning.

[0076] In some embodiments, after outputting the graded fish meat and separated bone material according to the updated grading and screening path, the method further includes: Obtain the raw material status characterization, extrusion control parameters, screening control parameters, and grading screening results corresponding to the current processing batch, and generate a batch processing status record; Based on the consistency between the residual burr risk level and the material diversion results in the batch processing status record, the parameter adaptation evaluation results are determined. The process parameter index corresponding to the parameter adaptation evaluation results and the raw material state characterization is updated so that subsequent processing batches can call the updated process parameter index to determine the corresponding extrusion control parameters and screening control parameters.

[0077] Specifically, after the current processing batch completes the output of fish meat grading and bone separation materials, the control system archives the status of the batch processing process. The system uses the processing batch number as an index to obtain the raw material status characterization generated before the batch enters the screw extrusion mechanism, the actual extrusion control parameters executed, the screening control parameters executed by the multi-layer vibrating screen, and the grading and screening results corresponding to each particle size outlet, and generates a batch processing status record.

[0078] Raw material condition characterization is used to record the compressibility, tissue integrity, and spatial distribution of bone fragments in the fish meat to be processed; extrusion control parameters are used to record the feed cycle time, segmented extrusion intensity, screw shaft speed, and residue discharge status; screening control parameters are used to record the stratified screening trajectory, screening intensity, screen surface residence time, and outlet flow guidance status; grading screening results are used to record the particle size class, residual bone risk level, rework status, and final output path of each particle size material.

[0079] For example, taking sea bass fillet processing as an example, the current batch of raw material characteristics shows that the fish meat tissue is relatively firm, the bones are concentrated on the spine, and the risk of bone breakage is high. During processing, the system uses a lower selective extrusion pressure, a larger discharge port opening, and a longer intermediate screening residence time. After processing, the batch processing status record shows that the risk level of residual bones at the fine particle outlet is low, the particle size at the medium particle outlet meets the target range, the density of hard target material in the bone separation path is high, and the proportion of reworked material is within the allowable range. Based on this, the system determines that the current extrusion control parameters and screening control parameters are highly compatible with the characteristics of this type of raw material. If the record shows a continuous high risk level of residual bones at the fine particle outlet, and the output of the bone separation path is insufficient, the system determines that the extrusion intensity or discharge status does not match the characteristics of this type of raw material, and it is necessary to reduce the selective extrusion pressure and increase the residue discharge capacity.

[0080] When determining the parameter adaptation evaluation results, the system uses the consistency between the residual spine risk level and the material diversion results as the primary evaluation criterion. If high-risk residual spine materials mainly enter the bone spine separation path or the verification path, and low-risk residual spine materials mainly enter the corresponding fish meat grading material outlet, the system determines that the diversion results are consistent with the residual spine risk level. If a large amount of low-risk residual spine materials enter the rework path, it indicates that the screening path is overly conservative. If high-risk residual spine materials enter the fish meat grading material outlet, it indicates that the parameter adaptation is insufficient and triggers risk correction. The system can also supplement and correct the parameter adaptation evaluation results by combining particle size deviation, rework frequency, and the state of fish meat residue in the bone spine enrichment residue stream, avoiding parameter adjustments based solely on a single outlet result.

[0081] Furthermore, the system updates the process parameter index corresponding to the raw material state characterization based on the parameter adaptation evaluation results. The process parameter index can establish a mapping relationship according to the raw material state category, the compressibility range of fish meat, the risk level of bone fragment breakage, and the target particle size level. When the parameter adaptation evaluation result is good, the system increases the calling priority of the current parameter combination under the same raw material state characterization; when the parameter adaptation evaluation result is poor, the system writes the extrusion correction amount and screening correction amount into the process parameter index and reduces the calling priority of the original parameter combination; when a high risk of residual bone fragments is misdiverted, the system sets the corresponding parameter combination to a restricted calling state to prevent subsequent batches from directly using it. When subsequent processing batches enter the production line, the system first retrieves the updated process parameter index based on the raw material state characterization of the new batch, and then determines the corresponding extrusion control parameters and screening control parameters.

[0082] This embodiment allows for closed-loop recording of the raw material state, extrusion process, screening process, and diversion results within a single processing batch. Parameter adaptability is evaluated based on the consistency between residual bone risk level and material diversion results. This method continuously optimizes the process parameter index under different raw material states, improves the accuracy of parameter calls for subsequent batches, reduces repeated debugging costs, and enhances the adaptability of fish deboning and grading processing to raw material differences and batch fluctuations.

[0083] The following are system embodiments of this application, which can be used to execute the method embodiments of this application. For details not disclosed in the system embodiments of this application, please refer to the method embodiments of this application.

[0084] Figure 2 This is a schematic diagram of the structure of the automated grading and screening system for deboning fish meat provided in an embodiment of this application. Figure 2 As shown, the system includes: The acquisition module 201 is used to acquire the raw material state data of the fish meat to be processed, and to perform physical property identification, tissue state identification and bone distribution identification on the raw material state data to generate a raw material state characterization. The determination module 202 is used to determine the extrusion control parameters of the screw extrusion mechanism based on the raw material state characterization, and to perform segmented screw extrusion of the fish meat to be processed according to the extrusion control parameters to obtain the fish meat extrusion flow and the bone enrichment residue flow. The dispersion module 203 is used to perform thin-layer dispersion processing on the fish meat extrusion flow to generate a layer of fish meat to be screened. Screening module 204 is used to control the multi-layer vibrating screen to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened, and to form particle size outlet material based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process. The detection module 205 is used to detect the particle size status and residual burr risk of the particle size outlet material, and generate a grading and screening result that includes the particle size grade and the residual burr risk grade. The output module 206 is used to update the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen according to the coupling relationship between particle size and residual bone risk in the grading and screening results, and output the fish meat grading material and bone separation material according to the updated grading and screening path.

[0085] In some embodiments, Figure 2 Before the fish meat to be processed enters the screw extrusion mechanism, the acquisition module 201 collects multimodal detection data of the fish meat to be processed, and performs spatiotemporal alignment of the multimodal detection data based on unified processing coordinates to generate raw material detection data; inputs the raw material detection data into the raw material state recognition model to extract the compressibility state, tissue integrity state and bone splinter spatial distribution state of the fish meat to be processed; and generates a raw material state characterization for determining extrusion control parameters and screening control parameters based on the correlation between the compressibility state, tissue integrity state and bone splinter spatial distribution state.

[0086] In some embodiments, Figure 2 The determination module 202 determines the compressible range of the fish meat to be processed and the risk of bone breakage based on the raw material state characterization, and generates extrusion control parameters including pre-compression parameters, selective extrusion parameters, and residue discharge parameters; according to the extrusion control parameters, the screw extrusion mechanism is controlled to form a progressive extrusion action along the material conveying direction, so that the fish meat tissue is extruded through the screen cylinder and the bones are restricted from entering the outside of the screen cylinder with the fish meat tissue; during the segmented screw extrusion process, the extrusion operation status is collected, and the extrusion intensity and residue discharge status of the corresponding segment are adjusted according to the extrusion operation status to obtain the fish meat extrusion flow and the bone enrichment residue flow.

[0087] In some embodiments, Figure 2 The dispersion module 203 acquires the discharge continuity and adhesion state of the fish meat extrusion flow, and determines the thin-layer dispersion parameters based on the discharge continuity and adhesion state; according to the thin-layer dispersion parameters, the fish meat extrusion flow is controlled and flexibly spread, so that the fish meat extrusion flow is transformed from an agglomerated state to a continuous thin-layer state; the thickness of the fish meat material in the continuous thin-layer state is balanced and the feeding rhythm is matched to generate a fish meat layer to be screened that is compatible with the feeding area of ​​the multi-layer vibrating screen.

[0088] In some embodiments, Figure 2The screening module 204 determines the layered screening control parameters of the multi-layer vibrating screen based on the spreading state and particle state of the fish meat layer to be screened; it controls the multi-layer vibrating screen to form a continuously changing screening trajectory according to the layered screening control parameters, so that the fish meat particles are loosened, turned over and deformed under the vibration; based on the deformed passing state of the fish meat particles and the rigid obstruction state of the bones, the passing material and the intercepted material of each screening layer are diverted to form the particle size outlet material corresponding to the particle size grade.

[0089] In some embodiments, Figure 2 The screening module 204 determines the vibration excitation parameters that change continuously along the conveying direction of the screen surface based on the spreading thickness of the fish meat layer to be screened on the adjacent screen layers and the passing load; it drives the multi-layer vibrating screen to form a composite screening trajectory according to the vibration excitation parameters, so that the fish meat particles sequentially produce a loose state, a flipped state, and a deformed passing state under the action of the composite screening trajectory; based on the difference in movement between the fish meat particles and bones under the composite screening trajectory, it controls the bones to remain on the corresponding screen layer or guide them to the corresponding interception outlet.

[0090] In some embodiments, Figure 2 The detection module 205 collects online detection data of the outlet material of each particle size and associates the online detection data with the corresponding sieve layer and outlet position to generate an outlet material detection characterization; based on the outlet material detection characterization, it identifies the particle size distribution of fish meat particles and determines the corresponding particle size level according to the preset particle size determination rules; based on the outlet material detection characterization, it identifies suspected bone spur features and performs risk assessment on suspected bone spur features in combination with particle size level to determine the corresponding residual bone risk level; based on the matching relationship between particle size level and residual bone risk level, it generates a graded screening result for material diversion and parameter feedback.

[0091] In some embodiments, Figure 2 The output module 206 identifies the particle size deviation and residual thorn risk deviation of the outlet material of each particle size class according to the grading and screening results, and generates a screening feedback characterization; the screening feedback characterization is input into the closed-loop control model to determine the extrusion correction amount corresponding to the screw extrusion mechanism and the screening correction amount corresponding to the multi-layer vibrating screen; the extrusion control parameters are updated based on the extrusion correction amount, and the screening control parameters are updated based on the screening correction amount; the grading and screening path is adjusted according to the updated extrusion control parameters and screening control parameters, and the material that meets the matching conditions of particle size class and residual thorn risk level is output as fish meat grading material, and the material that does not meet the matching conditions is output as bone and thorn separation material or rework material.

[0092] In some embodiments, Figure 2The output module 206 generates an abnormal material diversion characterization based on the particle size deviation state, residual bone risk deviation state, and fish meat recyclability state of the materials that do not meet the matching conditions. Based on the abnormal material diversion characterization, it determines whether the corresponding material belongs to the fish meat enrichment and rework state or the bone enrichment and separation state. When the corresponding material belongs to the fish meat enrichment and rework state, it is introduced into the low-pressure secondary extrusion path or the secondary vibration screening path. When the corresponding material belongs to the bone enrichment and separation state, it is introduced into the bone separation path and the corresponding material is restricted from re-entering the screw extrusion mechanism.

[0093] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0094] Figure 3 This is a schematic diagram of the electronic device 3 provided in an embodiment of this application. Figure 3 As shown, the electronic device 3 of this embodiment includes: a processor 301, a memory 302, and a computer program 303 stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program 303, it implements the steps in the various method embodiments described above. Alternatively, when the processor 301 executes the computer program 303, it implements the functions of each module / unit in the various system embodiments described above.

[0095] Electronic device 3 can be a desktop computer, laptop, handheld computer, cloud server, or other electronic device. Electronic device 3 may include, but is not limited to, processor 301 and memory 302. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or different components.

[0096] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0097] The memory 302 can be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. The memory 302 can also be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. The memory 302 can also include both internal and external storage units of the electronic device 3. The memory 302 is used to store computer programs and other programs and data required by the electronic device.

[0098] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0099] If integrated modules / units are implemented as software functional units and sold or used as independent products, they can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program may include computer program code, which may be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium may include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0100] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An automated grading and screening method for deboning fish meat, characterized in that, include: Obtain raw material state data of fish meat to be processed, and perform physical property identification, tissue state identification and bone distribution identification on the raw material state data to generate raw material state characterization; The extrusion control parameters of the screw extrusion mechanism are determined based on the raw material state characterization, and the fish meat to be processed is subjected to segmented screw extrusion according to the extrusion control parameters to obtain the fish meat extrusion stream and the bone and spine enrichment residue stream. The fish meat extrusion stream is subjected to a thin-layer dispersion treatment to generate a fish meat layer to be screened; The multi-layer vibrating screen is controlled to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened, and the particle size outlet material is formed based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process. The particle size distribution material is subjected to particle size state detection and residual burr risk detection to generate a grading and screening result that includes particle size grade and residual burr risk grade. Based on the coupling relationship between particle size and residual bone risk in the grading and screening results, the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen are updated, and the fish meat grading material and bone separation material are output according to the updated grading and screening path.

2. The method according to claim 1, characterized in that, The process of acquiring raw material state data of the fish meat to be processed, and performing physical property identification, tissue state identification, and bone and spine distribution identification on the raw material state data to generate a raw material state characterization includes: Before the fish meat to be processed enters the spiral extrusion mechanism, multimodal detection data of the fish meat to be processed is collected, and the multimodal detection data is spatiotemporally aligned based on a unified processing coordinate to generate raw material detection data; The raw material testing data is input into the raw material state recognition model to extract the compressibility state, tissue integrity state, and spatial distribution state of the fish meat to be processed. Based on the correlation between the compressibility state, tissue integrity state, and bone spur spatial distribution state, a raw material state characterization is generated to determine the extrusion control parameters and screening control parameters.

3. The method according to claim 1, characterized in that, The process of determining the extrusion control parameters of the screw extrusion mechanism based on the raw material state characterization, and performing segmented screw extrusion of the fish meat to be processed according to the extrusion control parameters to obtain a fish meat extrusion stream and a bone-rich residue stream, includes: Based on the raw material state characterization, the compressibility range of the fish meat to be processed and the risk of bone breakage are determined, and extrusion control parameters including pre-compression introduction parameters, selective extrusion parameters and residue discharge parameters are generated. According to the extrusion control parameters, the spiral extrusion mechanism is controlled to form a progressive extrusion action along the material conveying direction, so that the fish meat tissue is squeezed out through the screen cylinder and the bones are restricted from entering the outside of the screen cylinder along with the fish meat tissue. During the segmented spiral extrusion process, the extrusion operation status is collected, and the extrusion intensity and residue discharge status of the corresponding segments are adjusted according to the extrusion operation status to obtain the fish meat extrusion flow and the bone and spine enriched residue flow.

4. The method according to claim 1, characterized in that, The process of thin-layer dispersion of the extruded fish meat to generate a layer of fish meat to be screened includes: The discharge continuity and adhesion state of the fish meat extrusion flow are obtained, and the thin-layer dispersion parameters are determined based on the discharge continuity and adhesion state. The fish meat extrusion flow is controlled and flexibly spread according to the thin-layer dispersion parameters, so that the fish meat extrusion flow is transformed from an agglomerated state to a continuous thin-layer state. The thickness of the fish meat material in a continuous thin layer is balanced and the feeding rhythm is matched to generate a layer of fish meat to be screened that is compatible with the feeding area of ​​the multi-layer vibrating screen.

5. The method according to claim 1, characterized in that, The controlled multi-layer vibrating screen performs particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened, and forms particle size outlet material based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process, including: Based on the spreading state and granule state of the fish meat layer to be screened, determine the layer screening control parameters of the multi-layer vibrating screen. The multi-layer vibrating screen is controlled according to the layered screening control parameters to form a continuously changing screening trajectory, so that the fish meat particles are loosened, turned over and deformed under the action of vibration. Based on the deformable passage state of fish meat particles and the rigid obstruction state of bone spurs, the passing and intercepted materials of each screen layer are diverted to form the particle size outlet material corresponding to the particle size grade.

6. The method according to claim 5, characterized in that, The step of controlling the multi-layer vibrating screen to form a continuously changing screening trajectory according to the layered screening control parameters, so that the fish meat particles are loosened, turned over, and deformed under the vibration, includes: Based on the thickness of the fish meat layer to be screened on the adjacent screen layers and the passing load, the vibration excitation parameters that change continuously along the conveying direction of the screen surface are determined. Drive the multi-layer vibrating screen to form a composite screening trajectory according to the vibration excitation parameters, so that the fish meat particles will sequentially produce a loose state, a flipped state and a deformed passing state under the action of the composite screening trajectory. Based on the difference in movement between fish meat particles and bones under the composite screening trajectory, the bones are controlled to remain in the corresponding screening layer or be guided to the corresponding interception outlet.

7. The method according to claim 1, characterized in that, The process of performing particle size distribution and residual burr risk detection on the effluent material to generate a grading and screening result containing particle size class and residual burr risk class includes: Collect online detection data of the outlet material of each particle size, and associate the online detection data with the corresponding sieve layer and outlet position to generate an outlet material detection characterization; Based on the export material detection and characterization, the particle size distribution of fish meat particles is identified, and the corresponding particle size level is determined according to the preset particle size determination rules. Based on the aforementioned export material detection and characterization, suspected bone spur features are identified, and the risk assessment of suspected bone spur features is performed in conjunction with the aforementioned particle size classification to determine the corresponding residual bone spur risk level. Based on the matching relationship between the particle size class and the residual burr risk level, a grading and screening result is generated for material diversion and parameter feedback.

8. The method according to claim 1, characterized in that, The process involves updating the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen based on the coupling relationship between particle size distribution and residual bone risk in the grading and screening results, and outputting the graded fish meat material and bone-separated material according to the updated grading and screening path, including: Based on the grading and screening results, the particle size deviation and residual burr risk deviation of the outlet material of each particle size class are identified, and a screening feedback characterization is generated. The screening feedback characterization is input into the closed-loop control model to determine the extrusion correction amount corresponding to the spiral extrusion mechanism and the screening correction amount corresponding to the multi-layer vibrating screen. The extrusion control parameters are updated based on the extrusion correction amount, and the screening control parameters are updated based on the screening correction amount; The grading and screening path is adjusted according to the updated extrusion control parameters and screening control parameters. Materials that meet the matching conditions of particle size grade and residual bone risk grade are output as fish meat graded materials, while materials that do not meet the matching conditions are output as bone and spine separation materials or rework materials.

9. The method according to claim 8, characterized in that, The step of outputting materials that do not meet the matching conditions as bone splinter separation materials or rework materials includes: Based on the particle size deviation state, residual bone risk deviation state, and fish meat recyclability state of the materials that do not meet the matching conditions, an abnormal material diversion characterization is generated. Based on the abnormal material diversion characterization, it is determined that the corresponding material belongs to the fish meat enrichment and rework state or the bone and spine enrichment and separation state. When the corresponding material is in a fish meat enrichment and rework state, the corresponding material is introduced into a low-pressure secondary extrusion path or a secondary vibrating screening path. When the corresponding material is in a state of bone spike enrichment and separation, the corresponding material is guided into the bone spike separation path and the corresponding material is restricted from re-entering the spiral extrusion mechanism.

10. An automated grading and screening system for deboning fish meat, characterized in that, include: The acquisition module is used to acquire raw material state data of fish meat to be processed, and to perform physical property identification, tissue state identification and bone distribution identification on the raw material state data to generate a raw material state characterization. The determination module is used to determine the extrusion control parameters of the screw extrusion mechanism based on the raw material state characterization, and to perform segmented screw extrusion of the fish meat to be processed according to the extrusion control parameters to obtain the fish meat extrusion flow and the bone and spine enrichment residue flow. The dispersion module is used to perform thin-layer dispersion processing on the fish meat extrusion flow to generate a layer of fish meat to be screened. The screening module is used to control the multi-layer vibrating screen to perform particle size classification screening according to the screening trajectory and screening intensity corresponding to the fish meat layer to be screened, and to form particle size outlet material based on the deformation characteristics of fish meat particles and the rigidity retention characteristics of bones during the screening process. The detection module is used to detect the particle size distribution and residual burr risk of the particle size outlet material, and generate a grading and screening result that includes the particle size grade and the residual burr risk grade. The output module is used to update the extrusion control parameters and the screening control parameters of the multi-layer vibrating screen according to the coupling relationship between particle size and residual bone risk in the grading and screening results, and output the fish meat grading material and bone separation material according to the updated grading and screening path.

Images