Renewable resource sorting management system based on image analysis

Abstract

The invention discloses a renewable resource sorting management system based on image analysis. The renewable resource sorting management system comprises an image acquisition module, an image analysis module, a scheduling control module, a real-time monitoring module and a re-analysis module, the image acquisition module comprises a multispectral imaging unit and a 3D point cloud acquisition unit; the image analysis module executes the following steps: (a) matching multispectral data through a material fingerprint database to obtain a material type; (b) calculating the spatial density distribution of the 3D point cloud; (c) fusing the material type and the spatial density distribution to generate renewable resource partition information; the scheduling control module is used for scheduling the sorting equipment according to the partition information; the real-time monitoring module is used for collecting equipment state and sorting information; and the re-analysis module generates equipment management information and sorting evaluation information. According to the system and the method, high efficiency, intelligence and sustainability of renewable resource sorting are realized through an economic oriented strategy, intelligent equipment maintenance and data-driven optimization.

Description

Technical Field

[0001] The present invention relates to the field of renewable resource management systems, and more particularly to a renewable resource sorting and management system based on image analysis. Background Art

[0002] Renewable resource sorting, as a key link in resource recycling, aims to improve the purity and value of recyclables through efficient classification and reduce subsequent processing costs. The traditional sorting mode that relies on manual visual inspection and experience judgment has significant limitations, such as high misjudgment rate of materials, lack of spatial distribution analysis, and lag in equipment failure response, resulting in low resource recovery rate and insufficient economic benefits, making it difficult to meet the requirements of large-scale and refined modern circular economy. Therefore, the industry urgently needs to introduce automated and intelligent technologies to achieve precision, efficiency, and standardization in the sorting process. Therefore, a renewable resource sorting and management system based on image analysis is proposed. Summary of the Invention

[0003] The technical problem to be solved by the present invention is: how to solve the problems of high misjudgment rate of materials, lack of spatial distribution analysis, and lag in equipment failure response existing in the existing management system, and a renewable resource sorting and management system based on image analysis is provided.

[0004] The present invention solves the above technical problems through the following technical solutions. The present invention includes: An image acquisition module, an image analysis module, a scheduling and control module, a real-time monitoring module, and a re-analysis module; The image acquisition module includes a multi-spectral imaging unit and a 3D point cloud acquisition unit; The image analysis module performs: (a): Obtain the material type by matching the multi-spectral data with the material fingerprint library; (b): Calculate the spatial density distribution of the 3D point cloud; (c): Generate renewable resource partition information by fusing the material type and the spatial density distribution. The specific process is as follows: Fuse the material type and the spatial density distribution through a fusion evaluation function, and generate renewable resource partition information by optimizing the weight coefficients of the fusion evaluation function; The real-time monitoring module collects equipment status and sorting information; [[ID=�4]]The re-analysis module generates equipment management information and sorting evaluation information.

[0005] Furthermore, when the image analysis module executes step (c): Construct a fusion evaluation function , and its specific process is as follows: , Where is the confidence level of material identification, is the rationality index of density partition, and are the weight coefficients of the corresponding items, which are used to adjust the importance of the confidence level of material identification and the rationality index of density partition respectively; Select The partition scheme corresponding to the maximum value is the final recycled resource partition information.

[0006] Furthermore, the scheduling control module includes an economic decision sub-module, and the economic decision sub-module is used for: Access the price database to obtain the unit price of materials ; Then calculate the value density of each region , where U is the area of the region, is the mass of the i-th type of material in the corresponding region; Finally, generate the sorting priority in descending order according to the value density of each region ; The scheduling control module is also used to adjust the running speed of the conveyor belt, and receive the instructions of the re-analysis module, and optimize the recycled resource partition scheme by adjusting the weight coefficients of the fusion evaluation function of the image analysis module and .

[0007] Furthermore, the real-time monitoring module is also used for: collecting the air pressure of the pneumatic sorting device , the current of the robotic arm motor , the response time of the color sorter ; Then calculate the air pressure health index , the current health index and the response time index : Air pressure health index : ; Current health index : ; Response time index : ; where is the standard air pressure, is the standard current, is the standard response time; Calculate the equipment health index , and the specific process is as follows: ; When Generate a warning signal when the value is below the threshold, and simultaneously mark the abnormal parameter types, such as abnormal air pressure, abnormal current, or abnormal response time.

[0008] Furthermore, the real-time monitoring module is further configured to: Obtain the sorting qualified quality through the weighing unit and the component detector and the impurity mixing rate ; Count the number of missed inspections through the industrial camera and the number of misclassified items ; Generate a sorting information matrix from the above data according to the time series , and transmit it to the re-analysis module for sorting evaluation, where T represents the transpose of the matrix.

[0009] Furthermore, the process of the re-analysis module generating the equipment management information includes: Based on the time series data of the sorting information matrix calculate the comprehensive sorting efficiency , where the row vectors of the matrix correspond to the sorting data of different batches, specifically: , λ is the penalty coefficient; When is below the preset threshold, if the missed inspection rate is greater than the preset value, send the equipment management information to the scheduling control module to adjust the weight coefficient of the fusion evaluation function of the image analysis module and ; If the misclassification rate is greater than the preset value, adjust the running speed of the conveyor belt.

[0010] Furthermore, the process of the re-analysis module generating the equipment management information further includes: Calculate the comprehensive equipment efficiency ; Among them, , and are the weight coefficients, is the actual running time, is the total time, is the standard energy consumption, is the actual energy consumption; According to the value, generate equipment management information grading trigger task assignment, maintenance, or shutdown instructions; Calculate the nozzle blockage risk for the pneumatic sorting device , specifically , where k is a constant, is the air pressure change amount, is the air pressure threshold; When is greater than the preset value, nozzle detection requirements are appended; The process of the re-analysis module generating device management information further includes: Based on the independent parameter health index for fault location, the specific process is: when the air pressure health index is less than the preset value, such as when the air pressure health index is less than 0.8, it is determined that the pneumatic system is abnormal, and the air pressure pipeline leakage detection of the pneumatic sorting device is triggered; When the current health index is less than the preset value, such as when the current health index is less than 0.7, it is determined that the manipulator load is abnormal, and the motor current waveform diagram is retrieved to analyze the cause of jamming or overload; When the response time index is greater than the preset value, such as when the response time index is greater than 0.9, the response time is too long, it is determined that the image processing of the color sorter is delayed, and the algorithm time consumption of the image analysis module is automatically optimized.

[0011] Furthermore, the specific process of the re-analysis module generating sorting evaluation information includes: Obtain the sorting information matrix from the real-time monitoring module , and from the sorting information matrix extract the sorting qualified quality , impurity mixing rate , number of missed detections and number of misclassified items ; Obtain the unit price of the material from the scheduling control module ; Calculate the data obtained from the real-time monitoring module and the scheduling control module to obtain the batch value recovery rate , and its specific process is: , where is the quality of the j-th type of input material before sorting, represents the unit price of the j-th type of material in the input material; Then extract the multi-spectral features and 3D contour features of the missed detection materials in the image acquisition module; group the missed detection features through the K-means clustering algorithm, that is, group the multi-spectral features and 3D contour features of the missed detection materials in the image acquisition module, and count the occurrence frequencies of various materials; Mark the top 3 material categories with the highest occurrence frequencies as high-frequency loss materials; For high-frequency loss materials, in the image analysis module, increase the matching weight of their material spectral features by a preset value, and reduce the sensitivity threshold of the 3D point cloud segmentation of this type of material by a preset value; Then, diagnose the reasons for misclassification, parse the sorting logs of the misclassified materials, and classify and count them. The specific counting process is as follows: Material misjudgment: The similarity of image feature matching is between the threshold intervals , and are the threshold values of material matching similarity; Position offset: The Euclidean distance between the actual sorting coordinates and the target coordinates is greater than the preset value; Equipment malfunction: The reading of the robotic arm pressure sensor is lower than the standard value; After that, integrate and classify the batch value recovery rate , high-frequency lost materials and the reasons for misclassification to evaluate information; The classified evaluation information includes the following visual reports: Value recovery rate and its historical trend curve; Radar chart of the feature distribution of high-frequency lost materials; Pie chart of the proportion of reasons for misclassification.

[0012] Furthermore, the diagnosis process of the reasons for misclassification also includes: When the proportion of material misjudgment exceeds 60%, dynamically adjust the material matching threshold , and its specific process is as follows: ; where is the preset adjustment step, the misjudgment rate is the proportion of the number of material misjudgments to the total number of misclassifications, is the original material matching similarity threshold before adjustment, is the new material matching similarity threshold after adjustment; When the proportion of position offset exceeds 30%, send a coordinate calibration instruction to the scheduling control module; When the proportion of equipment malfunction exceeds 20%, mark the equipment in the equipment management information as requiring forced maintenance.

[0013] Furthermore, it also includes a dynamic calibration unit for: Regularly calibrate the spectrometer and 3D camera of the image acquisition module; Reset the standard parameters of the sorting equipment, that is, adjust to the standard air pressure, to the standard current, to the standard response time; Update the matching threshold of the material fingerprint library.

[0014] The present invention has the following advantages compared with the prior art: The regenerated resource sorting management system based on image analysis can accurately identify the material types of regenerated resources and analyze the spatial density distribution, and then generate reasonable zoning information, improving the sorting accuracy and the rationality of resource classification. It determines the sorting priority according to the material value density, can give priority to the processing of high-value materials, improve the economic benefits of sorting operations, and real-time monitor the equipment status parameters and calculate the health index, which can timely warn of equipment abnormalities, ensure the stable operation of the equipment, and reduce breakdown downtime. Through comprehensive analysis and evaluation based on sorting data, it can dynamically optimize and fuse evaluation functions, equipment parameters and sorting strategies, continuously improve the sorting efficiency and quality, regularly calibrate and update the equipment parameters to ensure the system maintains high-precision operation in the long term, and presents sorting evaluation information through visual reports, facilitating managers to intuitively grasp the system operation status and achieve refined management and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is the system block diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following details the embodiments of the present invention. The embodiments are implemented on the premise of the technical solution of the present invention, and detailed implementation manners and specific operation processes are given. However, the protection scope of the present invention is not limited to the following embodiments.

[0017] As Figure 1 shown, this embodiment provides a technical solution: A regenerated resource sorting management system based on image analysis, including: An image acquisition module, an image analysis module, a scheduling control module, a real-time monitoring module, and a re-analysis module; the image acquisition module includes a multi-spectral imaging unit and a 3D point cloud acquisition unit; The image analysis module performs: (a): Obtain the material type by matching the multi-spectral data through the material fingerprint library; (b): Calculate the spatial density distribution of the 3D point cloud; (c): Fuse the material type and the spatial density distribution to generate the regenerated resource zoning information. The specific process is as follows: Fuse the material type and the spatial density distribution through a fusion evaluation function, and generate the regenerated resource zoning information by optimizing the weight coefficients of the fusion evaluation function; The scheduling control module schedules the sorting equipment according to the zoning information; The real-time monitoring module collects equipment status and sorting information; The re-analysis module generates equipment management information and sorting evaluation information.

[0018] When the image analysis module executes step (c): Construct a fusion evaluation function , and the specific process is as follows: , where is the confidence level of material recognition, is the rationality index of density zoning, and are the weight coefficients of the corresponding items, which are used to adjust the importance of the confidence level of material recognition and the rationality index of density zoning respectively; Select The zoning scheme corresponding to the maximum value is the final recycled resource zoning information.

[0019] By constructing a fusion evaluation function, the confidence level of material recognition and the rationality index of density zoning are weighted and fused, which can comprehensively consider the material credibility and spatial distribution rationality of recycled resources, avoid the one-sidedness of single-dimensional evaluation, and thus generate more accurate and reliable recycled resource zoning information, improving the adaptability and decision-making scientificity of the sorting system to complex material scenarios; For example, assume that in a certain pile of recycled resources, the multispectral imaging unit identifies the material in area A as metal with a high confidence level, = 0.9, but the 3D point cloud shows that its spatial density distribution is scattered and the rationality index is low, = 0.5; The confidence level of material recognition in area B is medium, = 0.7, but the density distribution is compact and uniform = 0.8; If only sorted by material recognition, area A may be preferentially zoned, but there may be mixed materials due to the actual scattered density; If only sorted by density distribution, area B may be better, but the credibility of material recognition is insufficient.

[0020] Through the fusion evaluation function calculation, assume = = 0.5; For area A, S = 0.5 × 0.9 + 0.5 × 0.5 = 0.7; For area B, S = 0.5 × 0.7 + 0.5 × 0.8 = 0.75; Finally, area B with a larger S is selected as the preferential zoning, balancing the material credibility and distribution rationality and avoiding misjudgment caused by a single index.

[0021] The scheduling control module includes an economic decision sub-module, and the economic decision sub-module is used for: Access the price database to obtain the unit price of materials ; Then calculate the value density of each area , where U is the area of the region, is the mass of the i-th type of material in the corresponding region; Finally, generate the sorting priority in descending order according to the value density of each region ; The scheduling control module is also used to adjust the running speed of the conveyor belt, receive instructions from the re-analysis module, and optimize the waste resource zoning plan by adjusting the weight coefficients of the fusion evaluation function of the image analysis module and ; By accessing the price database through the economic decision sub-module, calculating the value density by combining the unit price of the material and the regional quality, and generating the sorting priority in descending order according to the value density, the sorting system can give priority to processing high-value waste resources, optimize the economic benefits of resource sorting, avoid low-value materials occupying sorting resources, and improve the input-output ratio and resource recovery efficiency of the overall sorting operation; assume that there are two material regions in the waste resource sorting scenario: Region 1: with an area of 1 square meter, containing 50 kg of scrap iron with a unit price of 1 yuan / kg; 30 kg of waste plastic with a unit price of 0.5 yuan / kg; Value density =(50×1 + 30×0.5) / 1 = 65 yuan / square meter; Region 2: with an area of 1.5 square meters, containing 20 kg of waste copper with a unit price of 10 yuan / kg; 10 kg of waste aluminum with a unit price of 5 yuan / kg; Value density =(20×10 + 10×5) / 1.5 = 166.67 yuan / square meter.

[0022] According to the descending order of value density, the system preferentially schedules the sorting equipment to process Region 2, that is, the sorting of high-value waste copper and waste aluminum, to ensure the priority recovery of high-yield materials and improve the economy of the sorting operation.

[0023] The real-time monitoring module is also used for: Collecting the air pressure of the pneumatic sorting device , the current of the robotic arm motor , and the response time of the color sorter ; Then calculate the air pressure health index , the current health index and the response time index : Air pressure health index : ; Current health index : ; Response time index : ; where is the standard air pressure, is the standard current, is the standard response time; Calculate the device health index , and the specific process is as follows: ; When is lower than the threshold, a warning signal is generated, and the abnormal parameter type is synchronously marked, such as abnormal air pressure, abnormal current, or abnormal response time; By collecting key operating parameters of the device in real time, such as air pressure, current, and response time, calculating the standardized health index and dynamically warning, it can achieve: Precise monitoring of device status: Convert complex device parameters into an intuitive health index to reflect the device operation quality in real time; Early warning of faults: Trigger a warning before the device performance deteriorates or fails to avoid sudden shutdowns affecting the sorting efficiency; Efficient abnormal location: Synchronously mark the specific abnormal parameter type, shorten the fault troubleshooting time, and reduce the maintenance cost; For example, assume that in a certain sorting line: Standard air pressure = 0.6 MPa, and the measured air pressure at a certain moment = 0.5 MPa, Air pressure health index =1 - ∣0.5 - 0.6∣ / 0.6 ≈ 0.83, and the threshold is assumed to be 0.8. At this time, no warning is triggered; Subsequently, the measured air pressure continues to drop to = 0.48 MPa, HP = 1 - ∣0.48 - 0.6∣ / 0.6 = 0.8, reaching the threshold. The system generates an air pressure abnormality warning and automatically triggers the air pressure pipeline leakage detection of the pneumatic sorting device, prompting the maintenance personnel to check the air leakage point in time to avoid the sorting action failure caused by insufficient air pressure.

[0024] The real-time monitoring module is also used for: Obtain the sorting qualified quality and the impurity mixing rate through the weighing unit and the component detector; Count the number of missed inspections and the number of misclassified items through the industrial camera; Generate a sorting information matrix for the above data in time series , and transmit it to the re-analysis module for sorting evaluation, where T represents the transpose of the matrix; By collecting multi-dimensional data such as sorted qualified quality, impurity mixing rate, number of missed inspections, and number of misclassifications in real time, generating a sorting information matrix according to the time series and transmitting it to the analysis module, full-process data monitoring can be realized, converting the key performance indicators of the sorting operation into structured data, and reflecting the sorted qualified quality and efficiency in real time; It provides raw data support for subsequent comprehensive sorting efficiency calculation, equipment effectiveness analysis, and sorting strategy optimization, traces the time nodes and characteristics of sorting anomalies through historical data sequences, and assists in locating the root causes of problems such as missed inspections and misclassifications; Suppose that within a certain period, the real-time monitoring module collects the following 3 batches of sorting data: Batch 1: Sorted qualified quality W1 = 1000 kg, impurity mixing rate Z1 = 3%, number of missed inspections = 5, number of misclassifications = 3; Batch 2: W2 = 1200 kg, Z2 = 5%, = 10, = 8; Batch 3: W3 = 900 kg, Z3 = 2%, = 2, = 1; The generated sorting information matrix is: ; By analyzing the matrix, it is found that the impurity mixing rate and the number of missed inspections / misclassifications in Batch 2 are significantly higher. The re-analysis module can specifically retrieve the image data and equipment operation logs of this batch to trace whether the decline in sorted qualified quality is caused by misjudgment of the image analysis algorithm or equipment parameter drift, and then optimize the fusion evaluation function or calibrate the equipment parameters.

[0025] The process by which the re-analysis module generates equipment management information includes: Based on the time series data of the sorting information matrix , calculate the comprehensive sorting efficiency , where the row vector of the matrix corresponds to the sorting data of different batches, specifically: , λ is the penalty coefficient; When is lower than the preset threshold, if the missed inspection rate is greater than the preset value, send equipment management information to the scheduling control module to adjust the weight coefficients and of the fusion evaluation function of the image analysis module; If the misclassification rate is greater than the preset value, adjust the running speed of the conveyor belt, that is, reduce the conveyor belt speed to increase the recognition time to reduce the misclassification rate; The missed inspection rate is the ratio of the number of missed inspections to the total sorting volume, and the misclassification rate is the ratio of the number of misclassifications to the total sorting volume; By calculating the comprehensive sorting efficiency Compare with the threshold value to dynamically evaluate the actual efficiency of the sorting operation, accurately locate the batches or links with declining efficiency based on time series data, automatically trigger the optimization of the fusion evaluation function or the adjustment of the conveyor belt speed, realize the intelligent dynamic calibration of the sorting strategy, avoid the efficiency bottleneck caused by fixed parameters, and improve the overall efficiency and stability of the sorting of renewable resources.

[0026] If the comprehensive sorting efficiency within a certain period continues to be lower than the preset threshold value, such as the comprehensive sorting efficiency continues to be lower than 90%, the system determines that the current fusion evaluation function or the conveyor belt speed is unreasonable, and automatically adjusts the partition fusion weight of the image analysis module, such as increasing the weight of the material recognition confidence , or reducing the conveyor belt running speed to extend the material recognition time, so that the sorting efficiency of subsequent batches can return to a reasonable range.

[0027] The process of the re-analysis module generating equipment management information also includes: Calculate the comprehensive equipment efficiency ; Among them, , and are weight coefficients, is the actual running time, is the total time, is the standard energy consumption, is the actual energy consumption; Generate equipment management information classification trigger task allocation, maintenance or shutdown instructions according to the value; Calculate the nozzle blockage risk for the pneumatic sorting device , specifically , where k is a constant, is the air pressure change, is the air pressure threshold; When is greater than the preset value, append the nozzle detection requirement; The process of the re-analysis module generating equipment management information also includes: Fault location based on the independent parameter health index, and the specific process is: when the air pressure health index is less than the preset value, such as when the air pressure health index is less than 0.8, it is determined that the pneumatic system is abnormal, and the air pressure pipeline leakage detection of the pneumatic sorting device is triggered; When the current health index is less than the preset value, such as when the current health index is less than the preset value of 0.7, it is determined that the manipulator load is abnormal, and the motor current waveform diagram is retrieved to analyze the cause of jamming or overload; When the response time index is greater than the preset value, such as when the response time index is greater than 0.9, the response time is too long, and it is determined that there is a delay in the image processing of the color sorter, and the algorithm time consumption of the image analysis module is automatically optimized; By integrating the equipment health index, the actual operating time ratio, and the energy consumption efficiency, a comprehensive quantitative analysis of the equipment performance is realized, avoiding single-index deviation, and assisting in formulating a hierarchical maintenance strategy, such as task allocation, maintenance, or shutdown.

[0028] An early warning of potential failures of pneumatic equipment is given in advance through the nozzle blockage risk model, reducing sudden shutdowns; based on the independent parameter health index, such as air pressure, current, and response time, the fault type is accurately located, shortening the troubleshooting time.

[0029] Automatic triggering of algorithm optimization according to parameter anomalies, such as optimizing the image analysis algorithm when there is a delay in the image processing of the color sorter, improving the system self-adaptability; Such as equipment comprehensive efficiency evaluation: Suppose the parameters of a certain sorting robotic arm are as follows: The equipment health index H = 0.7, the threshold is 0.6, which is normal; The actual operating time ratio = 0.8, higher than the standard of 0.7; The energy consumption efficiency = 0.9, lower than the standard of 1.0, with high energy consumption; The weight coefficient = 0.5, = 0.3, = 0.2; Calculate the comprehensive efficiency: K = 0.5×0.7 + 0.3×0.8 + 0.2×0.9 = 0.77; Since K is higher than the threshold of 0.7, the equipment is operating normally. If K continues to be lower than the threshold, the system triggers an energy consumption anomaly warning, prompting to check the motor load or conveyor belt resistance.

[0030] Such as nozzle blockage risk warning: The standard air pressure of the pneumatic sorting device = 0.6 MPa, and the air pressure change at a certain moment = 0.15 MPa, = 0.1 MPa, and the constant k = 5, calculate the blockage risk: ; When the preset blockage risk threshold is 0.5, > 0.5, the system adds a nozzle detection task, clears potential blockages in advance, and avoids the failure of sorting actions; Such as fault precise positioning: The air pressure health index = 0.75, and the preset threshold is 0.8. That is, it is determined that the pneumatic system is abnormal, and the air pressure pipeline leakage detection is triggered; Current health index = 0.6, and the preset threshold is 0.7. That is, it is determined that the load of the robotic arm is abnormal, and the current waveform diagram is retrieved to analyze whether overload is caused by jamming; Response time index = 0.95, and the preset threshold is 0.9. The response time is relatively long. That is, it is determined that the image processing of the color sorter is delayed, and the time-consuming of the image analysis algorithm is automatically optimized.

[0031] The specific process of the re-analysis module generating the sorting evaluation information includes: Obtain the sorting information matrix from the real-time monitoring module , from the sorting information matrix Extract the sorting qualified quality , impurity mixing rate , number of missed detections and number of misclassifications ; Obtain the unit price of the material from the scheduling control module ; Calculate the data obtained from the real-time monitoring module and the scheduling control module to obtain the batch value recovery rate , and its specific process is: , where is the quality of the j-th type of input material before sorting, refers to the unit price of the j-th type of material in the input materials; Extract the multi-spectral features and 3D contour features of the missed-detection materials in the image acquisition module; Group the missed-detection features through the K-means clustering algorithm, that is, group the multi-spectral features and 3D contour features of the missed-detection materials in the image acquisition module, and count the occurrence frequencies of various materials; Mark the top 3 material categories with the highest occurrence frequencies as high-frequency loss materials; For high-frequency loss materials, in the image analysis module, increase the matching weight of their material spectral features by a preset value, such as increasing by 20%, and decrease the sensitivity threshold of the 3D point cloud segmentation of this type of material by a preset value, such as decreasing by 10%; Then diagnose the reasons for misclassification, parse the sorting logs of misclassified materials, and classify and count them. The specific counting process is: Material misjudgment: The similarity of image feature matching is between the threshold intervals , and is the material matching similarity threshold; Position offset: The Euclidean distance between the actual sorting coordinates and the target coordinates is greater than the preset value; Equipment malfunction: The reading of the robotic arm pressure sensor is lower than the standard value; After that, the batch value recovery rate , high-frequency lost materials and misclassification reasons are integrated and classified to evaluate information; The classified evaluation information includes a visual report of the following content: Value recovery rate And its historical trend curve; Radar chart of the characteristic distribution of high-frequency lost materials; The resource value retention ability of the sorting operation is comprehensively measured by the batch value recovery rate. Combining the data of undetected / misclassified items, a multi-dimensional quantitative evaluation of the sorting quality is realized, avoiding the one-sidedness of relying solely on a single indicator. The K-means clustering algorithm is used to analyze the image features of undetected materials to identify high-frequency lost material categories, and the matching weights and segmentation sensitivities of the image analysis module are optimized accordingly to reduce the repeated undetected rate of the same type of materials. Through the analysis of the sorting log, the misclassification reasons are classified as material misjudgment, position deviation or equipment malfunction. Combining the proportion statistics to trigger dynamic adjustments, such as threshold correction, coordinate calibration or equipment maintenance, to improve the self-adaptability and stability of the sorting system. Through visual reports such as the value recovery rate trend curve, radar chart of the characteristics of high-frequency lost materials, and pie chart of misclassification reasons, an intuitive sorting efficiency analysis tool is provided for managers to assist in quickly formulating optimization strategies; For example, when a certain renewable resource sorting line processes mixed materials, such as a mixture of waste plastic bottles, aluminum cans, and cardboard boxes, it is found that the batch value recovery rate C is lower than the industry average for 3 consecutive times, and the number of undetected and misclassified items is relatively high. At this time, the value recovery rate is calculated as follows: Input materials before sorting: 1000 kg of waste plastic bottles, unit price 0.8 yuan / kg; 500 kg of aluminum cans, unit price 2 yuan / kg; 800 kg of cardboard boxes, unit price 0.5 yuan / kg, Total input value = 1000×0.8 + 500×2 + 800×0.5 = 2200 yuan; Qualified quality after sorting: 900 kg of waste plastic bottles, impurity mixing rate 5%; 450 kg of aluminum cans, impurity mixing rate 3%; 750 kg of cardboard boxes, impurity mixing rate 2%; Qualified value = 900×0.8×(1 - 5%) + 450×2×(1 - 3%) + 750×0.5×(1 - 2%) = 1812 yuan; Batch value recovery rate C = 1812 / 2200 ≈ 82.4% < 85%, it is determined that the sorting efficiency does not meet the standard.

[0032] Identification of high-frequency lost materials: After the image features of undetected materials are extracted, it is found through K-means clustering that transparent PET plastic bottles and special-shaped aluminum cans account for 70% of the total undetected items and are marked as high-frequency lost materials; The system automatically increases the material spectrum matching weight of PET plastic bottles by 25% and reduces the 3D point cloud segmentation sensitivity threshold of special-shaped aluminum cans by 12%, enhancing the recognition accuracy of the two types of materials. Diagnosis of misclassification reasons: Analysis of the misclassification log reveals that 60% of the misclassifications are due to incorrect material judgments. For example, the spectral similarity between transparent PET plastic bottles and glass bottles falls within the threshold range [0.6, 0.8], 30% is due to position offset, and the sorting coordinate deviation > 5 cm. System dynamic adjustment of material matching threshold: = - ×(60% / 100%), assuming = 0.1, then the threshold is adjusted from 0.7 to 0.64, and at the same time, a coordinate calibration instruction is sent to the scheduling module to reduce misclassifications caused by position offset.

[0033] Generate a historical trend curve of value recovery rate, display a radar chart of the characteristics of high-frequency loss materials with fluctuating C values in the past week, highlight the spectral / contour characteristics of transparent PET plastic bottles and special-shaped aluminum cans, and a pie chart of misclassification reasons, with material misjudgment accounting for 60%, assisting management personnel to quickly locate problems and initiate optimization strategies.

[0034] The diagnosis process of the misclassification reasons also includes: When the proportion of material misjudgment exceeds 60%, dynamically adjust the material matching threshold , and its specific process is: ; where is the preset adjustment step size, the misjudgment rate is the proportion of the number of material misjudgments to the total number of misclassifications, is the original material matching similarity threshold before adjustment, is the new material matching similarity threshold after adjustment; When the proportion of position offset exceeds 30%, send a coordinate calibration instruction to the scheduling control module; When the proportion of equipment malfunction exceeds 20%, mark in the equipment management information that the equipment needs to be forced to maintain; Through quantitative analysis of the proportion of misclassification reasons, trigger targeted automated optimization strategies to achieve dynamic self-calibration of the sorting system. When the proportion of material misjudgment is too high, adaptively adjust the matching threshold to reduce misclassification caused by similar spectral characteristics. Correct the equipment coordinates in a timely manner for position offset problems, improve sorting positioning accuracy, and mark the maintenance requirements according to the proportion of equipment malfunction to prevent sorting abnormalities caused by mechanical performance degradation and reduce the risk of systematic failures.

[0035] For example, in a continuous operation of a certain renewable resource sorting line, the statistics of misclassified materials are as follows: Ratio of material misjudgment: 65%, exceeding the 60% threshold; Ratio of position deviation: 25%, not exceeding the 30% threshold; Ratio of equipment malfunction: 10%, not exceeding the 20% threshold; At this time, automatic optimization is triggered, and the specific process of automatic optimization is as follows: Adjust the material matching threshold: The original material matching threshold range is [0.6, 0.8], and the misjudgment rate corresponds to the adjustment step = 0.1; Then calculate the threshold: ; The system automatically tightens the material matching standard to improve the discrimination accuracy of materials with small spectral feature differences, such as transparent PET plastic and transparent glass.

[0036] Since the ratios of position deviation and equipment malfunction do not reach the threshold, coordinate calibration or forced maintenance is not triggered temporarily, and only subsequent data is continuously monitored.

[0037] The system also includes a dynamic calibration unit for: Regularly calibrating the spectrometers and 3D cameras of the image acquisition module; Resetting the standard parameters of the sorting equipment, that is, adjusting to the standard air pressure, to the standard current, to the standard response time; Updating the matching threshold of the material fingerprint library; The above process prevents parameter drift of image acquisition devices, such as spectrometers and 3D cameras, caused by usage wear or environmental changes, ensures the accuracy of material identification and spatial density analysis, unifies the operating parameters of sorting equipment, such as air pressure, current, and response time, avoids inconsistent sorting actions or equipment failures caused by parameter deviations, updates the matching threshold of the material fingerprint library, adapts to changes in the types of renewable resources or the introduction of new materials, and maintains high sorting accuracy.

[0038] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0039] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples", etc. means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0040] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A regenerated resource sorting and management system based on image analysis, characterized in that Including: An image acquisition module, an image analysis module, a scheduling control module, a real-time monitoring module, and a re-analysis module; The image acquisition module includes a multi-spectral imaging unit and a 3D point cloud acquisition unit; The image analysis module performs: (a): Obtain the material type by matching the multi-spectral data with a material fingerprint library; (b): Calculate the spatial density distribution of the 3D point cloud; (c): Generate renewable resource zoning information by fusing the material type and the spatial density distribution. The specific process is as follows: Fuse the material type and the spatial density distribution through a fusion evaluation function, and generate renewable resource zoning information by optimizing the weight coefficients of the fusion evaluation function; The scheduling control module schedules the sorting equipment according to the zoning information; The real-time monitoring module collects equipment status and sorting information; The re-analysis module generates equipment management information and sorting evaluation information.

2. The regenerated resource sorting and management system based on image analysis according to claim 1, characterized in that: When the image analysis module executes step (c): Construct a fusion evaluation function , and the specific process is as follows: ； wherein is the confidence level of material identification, is the rationality index of density partition, and are the weight coefficients of the corresponding items, respectively used to adjust the importance of the confidence level of material identification and the rationality index of density partition; Select The partition scheme corresponding to the maximum value is the final renewable resource partition information.

3. The a regenerative resource sorting management system based on image analysis according to claim 2, characterized in that: The scheduling control module includes an economic decision sub-module, and the economic decision sub-module is used for: Access the price database to obtain the unit price of materials , and then calculate the value density of each region , and finally generate the sorting priority in descending order according to the value density of each region ; The scheduling control module is further configured to adjust the running speed of the conveyor belt, receive instructions from the re-analysis module, and optimize the regeneration resource zoning plan by adjusting the weight coefficients of the fusion evaluation function of the image analysis module. and , optimize the regeneration resource zoning plan.

4. The regenerated resource sorting and management system based on image analysis according to claim 3, characterized in that: The real-time monitoring module is also used for: Collect the air pressure of the pneumatic sorting device , the current of the robotic arm motor , the response time of the color sorter ; After that, the air pressure health index is calculated , the current health index and the response time index ; Then calculate the air pressure health index , the current health index and the response time index to obtain the device health index ; When Generate a warning signal when it is lower than the threshold, and synchronously mark the abnormal parameter type.

5. The regenerated resource sorting and management system based on image analysis according to claim 4, wherein: The real-time monitoring module is also used for: Obtain the sorted qualified quality through the weighing unit and the component detector and the impurity mixing rate ; Statistical number of undetected parts by industrial camera and misclassification number ; Generate a sorting information matrix for the above data in a time series , and transmit it to the re-analysis module for sorting evaluation, where T represents the transpose of the matrix.

6. The regenerated resource sorting and management system based on image analysis according to claim 5, wherein: The process by which the re-analysis module generates equipment management information includes: Based on the sorting information matrix of time series data, calculate the comprehensive sorting efficiency , where the row vectors of the matrix correspond to the sorting data of different batches; When is lower than the preset threshold, if the missed detection rate is greater than the preset value, send device management information to the scheduling control module to adjust the weight coefficient of the fusion evaluation function of the image analysis module and ; When the misclassification rate is greater than a preset value, adjust the running speed of the conveyor belt.

7. The regenerated resource sorting and management system based on image analysis according to claim 6, wherein: The process by which the re-analysis module generates equipment management information also includes: Comprehensive efficiency of computing devices , based on value to generate task assignments, maintenance, or shutdown instructions for hierarchical triggering of device management information Calculate the nozzle blockage risk for the pneumatic sorting device , specifically , where k is a constant, is the air pressure change amount, is the air pressure threshold value; When is greater than a preset value, additional nozzle detection requirements are appended; The process by which the re-analysis module generates equipment management information also includes: Fault location is carried out based on the independent parameter health index. The specific process is as follows: when the air pressure health index is less than the preset value, it is determined that the pneumatic system is abnormal, and the air pressure pipeline leakage detection of the pneumatic sorting device is triggered; When the current health index is less than the preset value, it is determined that the load of the robotic arm is abnormal, and the motor current waveform diagram is retrieved to analyze the reasons for jamming or overload; When the response time index is greater than the preset value, it is determined that there is an image processing delay in the color sorter, and the algorithm time consumption of the image analysis module is automatically optimized.

8. The regenerated resource sorting and management system based on image analysis according to claim 7, characterized in that: The specific process by which the re-analysis module generates sorting evaluation information includes: Obtain the sorting information matrix from the real-time monitoring module , from the sorting information matrix extract the sorting qualified quality , impurity mixing rate , number of undetected items and number of misclassified items ; Obtain the unit price of materials from the scheduling control module ; Calculate the data obtained from the real-time monitoring module and the scheduling control module to obtain the batch value recovery rate , and then extract the multi-spectral features and 3D contour features of the undetected materials in the image acquisition module; Group the missed detection features through the K-means clustering algorithm, that is, group the multi-spectral features and 3D contour features of the missed detection materials in the image acquisition module, and count the occurrence frequencies of various materials; Mark the top 3 material categories with the highest occurrence frequencies as high-frequency lost materials; For high-frequency lost materials, in the image analysis module, increase the matching weight of their material spectral features by a preset value, and reduce the 3D point cloud segmentation sensitivity threshold of this type of material by a preset value; Then conduct a diagnosis of the reasons for misclassification, analyze the sorting logs of the misclassified materials, and classify and count them. The specific counting process is as follows: Material misjudgment: The similarity of image feature matching falls within the threshold range , and are the threshold values of material matching similarity; Position offset: The Euclidean distance between the actual sorting coordinates and the target coordinates is greater than a preset value; Equipment malfunction: The reading of the robotic arm pressure sensor is lower than the standard value; After that, the batch value recovery rate , high-frequency loss materials, and misclassification reasons are integrated and classified to evaluate information.

9. The regenerated resource sorting and management system based on image analysis according to claim 8, wherein: The diagnosis process of the reasons for misclassification also includes: When the proportion of material misjudgment exceeds the preset value, dynamically adjust the material matching threshold , and the specific process is as follows: ； wherein is a preset adjustment step size, and the misjudgment rate is the ratio of the number of material misjudgments to the total number of misclassified scores, is the original material matching similarity threshold before adjustment, is the new material matching similarity threshold after adjustment; When the proportion of position offset exceeds the corresponding threshold, send a coordinate calibration instruction to the scheduling control module; When the proportion of equipment malfunction exceeds the corresponding threshold, mark in the equipment management information that the equipment needs to be forced to be maintained.

Images