Blockchain-based cable material full-life-cycle traceability management method and system

Abstract

The application provides a cable material full life cycle traceability management method and system based on a blockchain, relates to the technical field of quality control, and comprises the following steps: assigning an identifier to each cable material and registering the identifier in a blockchain network; collecting key parameter data in real time during the processing of the cable material; converting the key parameter data into a time series data point set to establish parameter distribution characteristics, and dividing the cable material into a conductor core section, an insulation layer section, a shielding layer section and an outer sheath section; setting three quality points in each section to construct a spatial topology network, establishing a multi-level grid structure according to the spatial topology network, dynamically optimizing the multi-level grid structure by using the parameter distribution characteristics, and forming a data unit. Through real-time collection of cable key parameters and calibration and optimization verification, and in combination with the non-tamperable characteristics of the blockchain, the application realizes traceability and quality control of the full life cycle of the cable material.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of quality control, in particular to a cable material full life cycle traceability management method and system based on a block chain. BACKGROUND

[0002] In the process of cable product supply chain management, especially for power cables used in complex environments, establishing a reliable product quality traceability system is a problem of industry concern. Currently, the full life cycle information management of such cable products mainly adopts a traditional centralized data storage method, which records the information of each link from production to use in the independent systems of different participants.

[0003] Taking the power cable used in coal mine as an example, its quality traceability involves multiple stages, that is, the process parameters of the production link are recorded by the manufacturer, the quality report issued by the testing agency is independently stored, and the operation record maintained by the user is also a separate system. When the cable has quality problems during use and needs to be traced, since the data systems of each party are independent of each other and adopt a centralized storage mode, some challenges may be faced in building a complete and reliable traceability evidence chain.

[0004] In the process of analyzing quality problems that require the participation of manufacturers, purchasers, testing agencies and other parties, the existing data management method has room for improvement in ensuring the integrity and credibility of the full chain data, especially in scenarios involving product quality responsibility identification. When cross-verified, the data independently maintained by each party may face coordination difficulties, which brings some difficulty to establishing a reliable product quality traceability mechanism. SUMMARY

[0005] The technical problem to be solved by the present application is to provide a cable material full life cycle traceability management method and system based on a block chain, which improves the efficiency of traceability management.

[0006] To solve the above technical problems, the technical scheme of the present application is as follows:

[0007] In the first aspect, the cable material full life cycle traceability management method based on a block chain comprises:

[0008] An identifier is assigned to each cable material and registered in a block chain network, and key parameter data is collected in real time in the processing process of the cable material;

[0009] The key parameter data is converted into a set of time series data points to establish parameter distribution characteristics, and the cable material is divided into a conductor core section, an insulation layer section, a shielding layer section and an outer sheath section;

[0010] Three mass points are set in each segment to construct a spatial topology network, and a multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units.

[0011] Based on the spatial constraints of the data units, time series data points are mapped to the corresponding data units, the spatial distribution characteristic values ​​of the data points in each data unit are calculated, and parameter calibration coefficients are generated based on the spatial distribution characteristic values.

[0012] The key parameter data is calibrated at multiple levels using parameter calibration coefficients to obtain calibrated key parameter data. Signal processing is then performed on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf Optimization Algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, thus obtaining verified feature data.

[0013] The verified feature data is bound to the identifier to generate a digital fingerprint, and the digital fingerprint and key parameter data are uploaded to the blockchain network to complete the traceability record of the entire life cycle of the cable material.

[0014] Secondly, a blockchain-based cable material lifecycle traceability management system includes:

[0015] The data acquisition module is used to assign an identifier to each cable material and register the identifier to the blockchain network; during the processing of the cable material, it collects key parameter data in real time.

[0016] The segmentation module is used to convert key parameter data into a set of time-series data points to establish parameter distribution characteristics and to divide the cable material into conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment.

[0017] A module is established to set three mass points in each segment to construct a spatial topology network. A multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units.

[0018] The calculation module is used to map time series data points to corresponding data units according to the spatial constraints of the data units, calculate the spatial distribution characteristic value of the data points in each data unit, and generate parameter calibration coefficients based on the spatial distribution characteristic value.

[0019] The calibration module is used to perform multi-level calibration on key parameter data using parameter calibration coefficients to obtain calibrated key parameter data. Then, it performs signal processing on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf optimization algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, thus obtaining verified feature data.

[0020] The binding module is used to bind verified feature data with identifiers, generate digital fingerprints, and upload digital fingerprints and key parameter data to the blockchain network to complete the traceability record of the entire life cycle of cable materials.

[0021] Thirdly, a computing device includes:

[0022] One or more processors;

[0023] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0024] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0025] The above-described solution of the present invention has at least the following beneficial effects:

[0026] By deeply integrating blockchain technology with physical data collection, a decentralized and tamper-proof traceability ledger has been constructed. This effectively overcomes the technical drawbacks of data silos and difficulties in cross-verification under the existing centralized storage model. It provides a unique and reliable full-lifecycle data evidence chain for multiple participants, including manufacturers, testing institutions, and users, fundamentally improving the credibility of quality traceability results. By dividing cable materials into multiple physical segments according to function and constructing a spatial topology network and multi-level grid structure, this method changes the past mode of general monitoring of the entire product. It can characterize and track the distribution and changes of key parameters in the internal spatial structure of the material, improving the accuracy of quality anomaly location and the fine granularity of traceability. By generating parameter calibration coefficients and performing multi-level calibration, spatial deviations and interferences in the data collection process are effectively suppressed. Furthermore, the Grey Wolf optimization algorithm is used to adaptively extract and verify key features, ensuring that the final on-chain data can truly and timely reflect the essential state of the cable material. From dynamically optimizing the grid structure based on parameter distribution to using algorithms for data feature extraction and verification, the reliance on manual intervention has been reduced, improving the efficiency of traceability management. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating the blockchain-based cable material lifecycle traceability management method provided in an embodiment of the present invention.

[0028] Figure 2 This is a schematic diagram of a blockchain-based cable material lifecycle traceability management system provided in an embodiment of the present invention. Detailed Implementation

[0029] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0030] like Figure 1 As shown, embodiments of the present invention propose a blockchain-based method for full lifecycle traceability management of cable materials, the method comprising the following steps:

[0031] Step 1: Assign an identifier to each cable material and register the identifier in the blockchain network; during the processing of the cable materials, collect key parameter data in real time;

[0032] Step 2: Convert the key parameter data into a time-series data point set to establish parameter distribution characteristics, and divide the cable material into conductor core section, insulation layer section, shielding layer section and outer sheath section;

[0033] Step 3: Set three mass points in each segment to construct a spatial topology network, and establish a multi-level grid structure based on the spatial topology network. Utilize parameter distribution characteristics to dynamically optimize the multi-level grid structure and form data units.

[0034] Step 4: Based on the spatial constraints of the data units, map the time series data points to the corresponding data units, calculate the spatial distribution characteristic values ​​of the data points in each data unit, and generate parameter calibration coefficients based on the spatial distribution characteristic values.

[0035] Step 5: Perform multi-level calibration on the key parameter data using parameter calibration coefficients to obtain calibrated key parameter data. Then, perform signal processing on the calibrated key parameter data to obtain signal-processed data. Finally, use the Grey Wolf optimization algorithm to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, and obtain verified feature data.

[0036] Step 6: Bind the verified feature data with the identifier to generate a digital fingerprint, and upload the digital fingerprint and key parameter data to the blockchain network to complete the traceability record of the cable material throughout its entire life cycle.

[0037] In this embodiment of the invention, by deeply integrating blockchain technology with data collection from the physical world, a decentralized and tamper-proof traceability ledger is constructed. This effectively overcomes the technical drawbacks of data silos and difficulties in cross-verification under the existing centralized storage model, providing a unique and reliable full lifecycle data evidence chain for multiple participants such as manufacturers, testing institutions, and users, fundamentally improving the credibility of quality traceability results. The cable material is divided into multiple physical segments according to function, and by constructing a spatial topology network and a multi-level grid structure, this method changes the past model of general monitoring of the entire product. It can characterize and track the distribution and changes of key parameters in the internal spatial structure of the material, improving the accuracy of quality anomaly location and the fineness of traceability. By generating parameter calibration coefficients and performing multi-level calibration, spatial deviations and interference during the data collection process are effectively suppressed. Furthermore, the Grey Wolf optimization algorithm is used to adaptively extract and verify key features, ensuring that the final on-chain data can truly and timely reflect the essential state of the cable material. From dynamically optimizing the grid structure based on parameter distribution to using algorithms for data feature extraction and verification, the reliance on manual intervention is reduced, improving the efficiency of traceability management.

[0038] In a preferred embodiment of the present invention, an identifier is assigned to each cable material, and the identifier is registered in a blockchain network; during the processing of the cable material, key parameter data is collected in real time, including:

[0039] Step 100: Collect conductor core characteristic parameters, including conductor core resistance, conductor core diameter, and conductor core surface roughness, using a first sensor array deployed on the conductor core processing equipment. Specifically, this involves deploying the first sensor array at the corresponding workstations of key processing equipment such as wire drawing, annealing, and stranding of the conductor core, aligning it centrally along the processing trajectory and ensuring interference-free detection. This array consists of a four-point probe sensor, a laser rangefinder, and a contact roughness meter. The sensors and processing equipment are linked in real-time via a PLC control system. The sensors are synchronously activated when the equipment starts and automatically go into sleep mode when the equipment stops, ensuring consistency between detection and processing rhythms. When collecting the conductor core resistance value, the four probes of the four-point probe sensor... The probes are evenly spaced (1.5 times the nominal diameter of the conductor core) against the surface of the conductor core to prevent slippage or poor contact. A constant DC current of 1mA to 10mA is applied to the two outer probes (5mA is used for copper conductors) based on the conductivity of the conductor core material. Simultaneously, the voltage drop between the two inner probes is measured using a high-precision voltmeter to ensure microvolt-level accuracy. The conductor core resistance is then calculated using Ohm's law: conductor core resistance = measured voltage drop ÷ applied constant DC current. When acquiring the conductor core diameter, two laser rangefinders are symmetrically positioned on either side of the conductor core, with the sensor probes perpendicular to the conductor core axis. The reference distance between the two sensor probes is preset according to the nominal diameter of the conductor core. For example, for a conductor with a nominal diameter of 10mm, the reference distance is set to 30mm to allow sufficient detection margin. Both sensors simultaneously emit laser signals and receive reflected signals to accurately measure the distance from each probe to the surface of the conductor core and calculate the conductor core diameter. The calculation formula is: Conductor core diameter = Reference distance between the two sensor probes - Measurement distance of the left sensor - Measurement distance of the right sensor. Data is collected every 0.5 seconds during the measurement process to ensure coverage of all points around the conductor circumference. When collecting the surface roughness of the conductor core, the probe of the contact roughness meter is made of diamond material with a probe tip radius of 2μm, moving along the conductor core axis at a speed of 1mm / s. The probe moves across the surface at a constant speed, and the vertical displacement change of the probe is recorded in real time (the displacement measurement accuracy reaches the nanometer level), forming a continuous surface profile curve. A continuous section without joints or obvious defects is selected as the sampling section, and the length of the sampling section is specified as 0.8 mm (which meets the sampling requirements for slender cable components in the surface roughness measurement standard). The arithmetic mean of the absolute values ​​of the deviation of the profile from the center line within the sampling section is calculated, which is the surface roughness of the conductor core. The calculation formula is: surface roughness of conductor core = (1 ÷ sampling section length) × the sum of the absolute values ​​of the deviation of the profile from the center line at each point within the sampling section. Three sampling sections at different locations are randomly selected for each processing batch, and the arithmetic mean of the three measurement results is taken as the final surface roughness data of the conductor core.

[0040] Step 101 involves collecting insulation layer characteristic parameters, including insulation layer thickness, insulation layer resistance, and insulation layer dielectric constant, using a second sensor array deployed on the insulation layer processing equipment. Specifically, this includes deploying a second sensor array at the outlet and cooling station of the insulation layer extrusion and curing equipment. This array consists of an ultrasonic thickness sensor, a high-resistance measuring instrument, and a capacitance bridge tester. The sensor mounting brackets are made of insulating material to avoid conductive interference with the equipment's metal components. The sensor probes are vertically aligned with the insulation layer surface, with the distance between the probes and the insulation layer surface controlled between 5mm and 10mm to ensure stable detection signals and prevent contact with the insulation layer during processing. When collecting insulation layer thickness data, six ultrasonic thickness sensors are evenly distributed along the circumference of the insulation layer (with a 60-degree spacing between adjacent sensors). Simultaneously, they emit 5MHz ultrasonic signals into the insulation layer. After penetrating the insulation layer, the signals are reflected by the surface of the internal conductive core. The sensors receive the reflected signals and record the round-trip propagation time of the ultrasonic waves. This data is then combined with the ultrasonic propagation speed corresponding to the insulation material (based on the type of insulation material). The model is preset, such as the ultrasonic propagation speed of PVC insulation material being 2300m / s and cross-linked polyethylene being 2200m / s. The insulation layer thickness measured by a single sensor is calculated (single sensor thickness = ultrasonic propagation speed × propagation time ÷ 2). Then, the arithmetic mean of the measurement results of 6 sensors is taken to eliminate the error caused by the uneven thickness of the insulation layer in the circumferential direction. The calculation formula is insulation layer thickness = (sum of thicknesses measured by each ultrasonic sensor) ÷ number of sensors. When collecting the insulation layer resistance value, one electrode of the high-resistance measuring instrument is fixed to the exposed end of the conductor core by a conductive clamp, and the other electrode is tightly attached to the outer surface of the insulation layer by a ring conductive sheet to ensure that the electrode is in full contact with the measurement surface without damaging the insulation layer. A stable DC high voltage of 500V is applied between the two electrodes. The leakage current through the insulation layer is measured by the micro-current sensor built into the high-resistance measuring instrument (the measurement accuracy reaches the picoampere level). The insulation layer resistance value is calculated according to Ohm's law. The calculation formula is insulation layer resistance value = applied DC high voltage ÷ measured leakage current. When collecting the dielectric constant of the insulating layer, the two plates of the capacitance bridge tester are respectively attached to the conductor core and the outer surface of the insulating layer to form a parallel plate capacitor with the insulating layer as the dielectric. The capacitance value of this capacitor is measured, and combined with the measured insulation layer thickness and the effective contact area between the conductor core and the insulating layer (effective contact area = π × the square of the conductor core radius, which is calculated from the conductor core diameter measured in step 100), and the vacuum dielectric constant (fixed value 8.85 × ... The dielectric constant of the insulating layer is calculated using the formula: dielectric constant of insulating layer = (measured capacitance value × insulation layer thickness) ÷ (vacuum dielectric constant × effective contact area).

[0041] Step 102 involves collecting shielding layer characteristic parameters, including shielding layer coverage, shielding layer resistance, and shielding layer continuity, using a third sensor array deployed on the shielding layer processing equipment. Specifically, this includes deploying a third sensor array at the processing station and finished product inspection station of the shielding layer braiding, winding, or metal strip longitudinal wrapping equipment. This array consists of an image acquisition sensor, a four-point probe sensor, and a continuity test sensor. The sensor detection range completely covers the shielding layer processing area. The vertical distance between the image acquisition sensor lens and the shielding layer surface is set to 30cm to ensure a complete field of view and clear details. The probes of the four-point probe sensor and the continuity test sensor employ an elastic contact design to avoid damaging the shielding layer surface. When collecting shielding layer coverage data, the image acquisition sensor... The device vertically photographs the surface of the shielding layer at a resolution of 1920×1080 pixels and a frequency of 10 frames per second. The photographing area corresponds to a continuous segment of the shielding layer (5cm in length). The captured image is converted into a pixel matrix in grayscale mode. An image thresholding algorithm is used to divide the covered and uncovered areas, with a preset grayscale threshold of 100 (based on the grayscale difference between the shielding material and the background; the shielding material is mostly metal, with a grayscale value below 100, while the background and uncovered areas have grayscale values ​​above 100). Based on this, the covered and uncovered areas of the shielding material are distinguished. The number of pixels in the covered area and the total number of pixels in the image are counted to calculate the shielding layer coverage rate. The calculation formula is: Shielding layer coverage rate = (Number of pixels in the covered area ÷ Number of pixels in the image). (Total number of pixels) × 100%. For each processing batch, images from 5 different locations are captured, and the average value is used as the final coverage data. When collecting the shielding layer resistance value, the same four-point probe method as for collecting the conductor core resistance value is used. The probe spacing of the four-point probe sensor is set to 2mm, uniformly contacting the shielding layer surface. A constant DC current of 3mA is applied to the two outer probes, and the voltage drop between the two inner probes is measured. The shielding layer resistance value is calculated according to Ohm's law: Shielding layer resistance value = Measured voltage drop ÷ Applied constant DC current. Three measurement points are selected along the shielding layer axis during measurement, and the arithmetic mean of the three measurement results is taken. When collecting the shielding layer continuity data, 10 continuity test sensors are arranged at equal intervals along the shielding layer axis. Test contacts (5cm spacing between adjacent contacts) are used. Each contact is reliably connected to the shielding layer surface via an elastic probe. Two adjacent test contacts are connected sequentially, and the resistance between the two points is measured using a high-precision ohmmeter. The preset conduction threshold is set to 1Ω for copper shielding, 3Ω for aluminum foil shielding, and 2Ω for braided shielding (tin-plated copper wire). The conductivity is calibrated according to the conductivity of different shielding materials to ensure that the conduction judgment meets the actual use requirements. The number of adjacent contact pairs with resistance values ​​less than the corresponding material's conduction threshold is counted. Combined with the total number of test contact pairs (9 pairs, 10 contacts forming 9 adjacent combinations), the shielding layer continuity is calculated. The calculation formula is: Shielding layer continuity = (Number of adjacent contact pairs with resistance values ​​less than the conduction threshold ÷ Total number of test contact pairs) × 100%.

[0042] Step 103 involves collecting outer sheath characteristic parameters, including outer sheath thickness, hardness, and abrasion resistance, using a fourth sensor array deployed on the outer sheath processing equipment. Specifically, this includes deploying the fourth sensor array at the outlet of the outer sheath extrusion and cooling / shaping equipment and at the performance testing station. This array consists of an ultrasonic thickness sensor, a Shore hardness tester, and an abrasion tester. The sensors are installed away from the cooling spray area on the outer sheath surface, at least 50cm from the spray nozzle, to ensure a dry and stable testing environment. The sensor brackets are equipped with shock-absorbing devices to reduce the impact of equipment vibration on the test results. When collecting the outer sheath thickness, a circular deployment method consistent with the insulation layer thickness collection is adopted. Six ultrasonic thickness sensors are evenly distributed along the circumference of the outer sheath. The ultrasonic signals emitted by the sensors penetrate the outer sheath and are reflected by the internal shielding layer surface. The ultrasonic propagation time is measured and combined with the ultrasonic propagation speed of the outer sheath material (e.g., the propagation speed of polyethylene outer sheath material is 2250m / s), to calculate the outer sheath thickness measured by a single sensor (thickness measured by a single sensor = ultrasonic propagation speed × propagation time ÷ 2). The arithmetic mean of the measurement results from the six sensors is taken, calculated using the formula: Outer sheath thickness = (sum of thicknesses measured by each ultrasonic sensor) ÷ number of sensors. When collecting the hardness data of the outer sheath, a Shore A hardness tester (for soft outer sheath materials) is used. The tester probe is vertically pressed into the outer sheath surface to a preset depth of 1 mm, applying a constant pressure of 1 kgf. After maintaining this pressure for 3 seconds, the reaction force on the probe is measured. The hardness tester has a built-in preset calibration curve (calibration curve coefficient is 60HA / N, zero-point correction value is 20HA, for example, a reaction force of 0.5). At N, the hardness = 60 × 0.5 + 20 = 50HA; at a reaction force of 1N, the hardness = 60 × 1 + 20 = 80HA. Based on the calibration curve, the reaction force is converted into the outer sheath hardness value. The conversion logic is: outer sheath hardness = calibration curve coefficient × measured reaction force + zero-point correction value. Four evenly distributed measurement points are selected along the circumference of the outer sheath, and the average value is taken as the final hardness data. When collecting the wear resistance data of the outer sheath, the friction head of the wear resistance tester is made of tungsten carbide (hardness HRC60 or higher), and the contact area between the friction head and the surface of the outer sheath is 1... The outer sheath was subjected to reciprocating friction at a constant speed of 50 mm / s along its axial direction. The preset number of friction cycles was 1000. During the friction process, a constant pressure of 2 kgf was maintained. After the friction was completed, the mass of the friction area of ​​the outer sheath was measured using an electronic balance with an accuracy of 0.1 mg (the mass before and after friction was measured separately). The change in mass before and after friction was calculated. The wear resistance of the outer sheath was characterized by the mass loss per unit number of friction cycles and per unit pressure. The calculation formula is: wear resistance of outer sheath = change in mass before and after friction ÷ (number of friction cycles × constant pressure applied). Two non-overlapping friction areas were selected for each sample for testing, and the arithmetic mean of the two test results was taken.

[0043] Step 104: Collect environmental parameters, including ambient temperature, ambient humidity, and ambient cleanliness, using environmental monitoring equipment. Specifically, this involves deploying environmental monitoring equipment in the surrounding areas of each processing station, including the conductor core, insulation layer, shielding layer, and outer sheath, according to the principle of uniform distribution and close proximity to the processing surface. This equipment consists of thermocouple temperature sensors, capacitive humidity sensors, and laser particle counters. The installation height is consistent with the processing surface of each station (approximately 1.5m). All sensor probes are equipped with dustproof protective covers, while avoiding areas directly exposed to sunlight and areas directly impacted by airflow. The distance from the workshop ventilation openings should be at least 1m to ensure a stable and interference-free testing environment. When collecting ambient temperature data, a type K thermocouple temperature sensor is selected, with a measurement range of -20℃ to 200℃, meeting the ambient temperature detection requirements of various processing stages in cable production. The sensor probe adopts an open installation design to ensure full contact with air, while being fixed by a bracket to avoid high-temperature components of the equipment, such as heating modules and motor housings, to prevent localized high temperatures from affecting measurement accuracy. The sensor's calibration coefficient is determined through factory calibration and periodic re-inspection. At the factory, the manufacturer compares the type K thermocouple one-to-one with a metrologically certified standard temperature source, selecting five uniform calibration points (0℃, 25℃, 50℃, 75℃, 100℃) in the 0℃ to 100℃ range. After each point stabilizes for 30 minutes, the corresponding output voltage V0 to V4 is recorded, and the standard temperature change ΔT (all at 25℃) and voltage change ΔV1 to V4 are calculated for adjacent points. Based on the linear characteristics of the K-type thermocouple, a linear equation ΔT = k × ΔV is established, where ΔT is the standard temperature change, ΔV is the sensor output voltage change, and k is the calibration coefficient to be determined. This equation is fitted using the least squares method to obtain the slope k, which is the final calibration coefficient. During use, the system is re-inspected every 3 months, repeating the above process to obtain new coefficients. If the deviation from the initial value exceeds ±0.002mV / ℃, the coefficients are updated to ensure measurement accuracy. When the thermocouple is working, it converts the ambient temperature change into a corresponding electrical signal voltage. The measurement accuracy of this voltage is 0.01mV. Combining the calibration coefficients determined above and the zero-point correction value (the standard output voltage value corresponding to the sensor at 0℃, set to 0mV), the ambient temperature is calculated using the formula: Ambient Temperature = Sensor Calibration Coefficient × Converted Electrical Signal Voltage + Zero-Point Correction Value. To reflect the dynamic changes in ambient temperature in real time, temperature data is collected every minute to ensure real-time reflection of ambient temperature changes.

[0044] When collecting ambient humidity data, the probe of the capacitive humidity sensor is exposed to the air, and its capacitance value changes linearly with the ambient humidity. First, the sensor's capacitance value in the current environment is measured. Then, combined with the sensor's factory-calibrated dry reference capacitance value (capacitance value at 0% relative humidity in a dry state, set to 100pF) and saturation maximum capacitance value (capacitance value at 100% relative humidity in a saturated state, set to 300pF), the ambient relative humidity is calculated using the formula: Ambient Humidity = (Measured Capacitance Value - Dry Reference Capacitance Value) ÷ (Saturation Maximum Capacitance Value - Dry Reference Capacitance Value) × 100%. Humidity measurement accuracy is controlled within ±2%RH. The preset humidity warning threshold is 65%RH (exceeding this threshold may affect the curing of the insulation layer and the quality of the outer sheath molding). When collecting ambient cleanliness data, the laser particle counter emits light at a wavelength of 650nm. A laser beam is used. Air passes through the laser beam channel at a flow rate of 1L / min via a sampling pump built into the counter. When airborne particles (dust, fibers, etc.) pass through the laser beam, they generate scattered light. The sensor receives the scattered light signal and converts it into an electrical pulse signal. The particle size is determined by the amplitude of the pulse signal. The preset particle size threshold is 0.5μm (a key particle size indicator for industrial environmental cleanliness detection; particles larger than 0.5μm are prone to adhering to cable surfaces and affecting product quality). The number of particles larger than 0.5μm in a unit volume (1 liter) is counted to determine the environmental cleanliness. The calculation method is: Environmental cleanliness = total number of particles that meet the particle size requirement in a unit volume. Data is collected every 5 minutes. At the same time, the preset cleanliness warning threshold is 35,000 particles / liter (if this threshold is exceeded, the workshop purification system needs to be activated). The trend of environmental cleanliness changes is recorded.

[0045] This embodiment, through the targeted deployment of sensor arrays and environmental monitoring equipment, achieves comprehensive collection of characteristic parameters of the entire structural layer of cable material and production environment parameters, making up for the shortcomings of traceability that only focuses on a single link or a few parameters. The collection of key characteristic parameters of each structural layer covers the core performance indicators of the conductor core, insulation layer, shielding layer, and outer sheath, avoiding the problem of incomplete traceability due to missing parameters. Through standardized measurement methods and clear calculation logic, the accuracy and consistency of the collected data are ensured, reducing human operation errors and equipment detection deviations, and solving the pain point of insufficient data reliability in traceability. The synchronous collection of environmental parameters incorporates external influencing factors in the production process into the traceability system, avoiding traceability deviations caused by relying solely on the product's own parameters, and making the determination of quality responsibility more scientific.

[0046] In a preferred embodiment of the present invention, key parameter data is converted into a set of time-series data points to establish parameter distribution characteristics, and the cable material is divided into a conductor core segment, an insulation layer segment, a shielding layer segment, and an outer sheath segment, including:

[0047] Step 200: Convert the key parameter data into a time-series data point set according to time order. Each data point contains a timestamp and corresponding spatial location information. Specifically, this includes: systematically classifying and organizing all key parameter data collected in steps 100 to 104, dividing them into two main categories: structural layer characteristic parameters and environmental parameters. Structural layer characteristic parameters are further subdivided into conductor core characteristic parameters (conductor core resistance value, conductor core diameter, conductor core surface roughness), insulation layer characteristic parameters (insulation layer thickness, insulation layer resistance value, insulation layer dielectric constant), and shielding layer characteristic parameters (shielding layer coverage, shielding layer resistance value, etc.). The parameters include shielding layer continuity, outer sheath characteristics (outer sheath thickness, outer sheath hardness, outer sheath abrasion resistance), and environmental parameters such as ambient temperature, ambient humidity, and ambient cleanliness. Each parameter is grouped independently without overlap or confusion. Based on the categorized parameter data, each type of parameter is sorted in ascending order according to the time of collection. A unique timestamp is assigned to each independent parameter data point and synchronized in real-time with the sensor's data acquisition action. The timestamp is generated and written to the data the instant the sensor triggers data acquisition, avoiding time-series deviations caused by time delays and ensuring accurate and traceable time dimensions for each parameter data point.

[0048] After timestamp allocation, spatial location information is added to each parameter data. This spatial location information is determined through a fixed coordinate system and a correlation with processing progress. The coordinate representation method for each parameter is tailored to its processing technology and testing requirements. Conductor core parameters use axial coordinates, with the center of the conductor core drawing equipment's outlet as the origin and the processing direction as the positive direction. The coordinate value equals the actual length of the cable drawn. Insulation layer parameters use a combination of radial and axial coordinates. The axial coordinates are consistent with the conductor core (ensuring spatiotemporal correspondence within the same cable segment), while the radial coordinates are measured along the insulation layer thickness direction, with the conductor core surface as the origin. Shielding layer parameters use segmented axial coordinates, starting from the shielding layer processing equipment's inlet and divided into pre-processing, forming, and testing segments. The coordinate values ​​are determined by segment numbering. The parameters are composed of the number and the axial distance within the segment; the parameters related to the outer sheath use a combination of circumferential and axial coordinates, with the axial coordinates linked to the insulation layer, and the circumferential coordinates divided into 360-degree fixed angles, with the coordinate values ​​composed of circumferential angles and axial distances; the environmental parameters use the workshop plane rectangular coordinate system, measuring the three-dimensional coordinates of the sensor installation position in meters (x-axis along the production line, y-axis perpendicular to the production line, z-axis consistent with the height of the processing surface). Finally, each parameter data is bound one-to-one with its corresponding timestamp and spatial location information to form a single time-series data point containing parameter values, timestamps, and spatial locations. All time-series data points are then arranged in ascending order by timestamp and integrated into a complete time-series data point set, ensuring that each data point in the set can be clearly traced back to its specific collection time, specific processing location, and corresponding parameter type.

[0049] Step 201 involves calculating the statistical characteristics of the distribution of each key parameter based on the time-series data point set, including parameter mean, parameter variance, and parameter distribution density. Specifically, this includes: based on the time-series data point set constructed in step 200, conducting statistical characteristic analysis according to the principle of independent calculation of each parameter, extracting all data point values ​​for each type of key parameter, such as extracting data point values ​​for all conductor core resistance values ​​and all insulation layer thicknesses, etc., and calculating the three core statistical characteristics of parameter mean, parameter variance, and parameter distribution density for each type of parameter to reflect the overall level, fluctuation degree, and spatiotemporal distribution pattern of the parameter; when calculating the parameter mean, first summing the values ​​of all data points under this type of parameter, retaining six decimal places of precision during the summation process to avoid numerical truncation errors, and obtaining the total value of this type of parameter; Next, count the total number of data points for this type of parameter. Finally, divide the sum of the values ​​by the total number of data points. The formula is: Parameter mean = Sum of all data point values ​​for this type of parameter ÷ Total number of data points for this type of parameter. The result is kept to four decimal places to ensure the accuracy of the mean. After obtaining the parameter mean, further calculate the parameter variance by subtracting the mean from the value of each data point to obtain the deviation value, retaining the positive and negative signs. Square each deviation value to eliminate the positive and negative effects, obtaining the square value of the deviation value. Sum all the square values ​​to obtain the sum of squares. Finally, divide the sum of squares by the total number of data points for this type of parameter. The formula is: Parameter variance = Sum of squares of (each data point value - parameter mean) ÷ Total number of data points for this type of parameter. The result is kept to eight decimal places to accurately quantify the degree of parameter fluctuation.

[0050] To accurately quantify the spatiotemporal density of parameters, it is also necessary to calculate the parameter distribution density. First, determine the time span by converting the millisecond-level timestamp difference between the first and last data points of this type of parameter into second-level units (time span = (last data point timestamp - first data point timestamp) ÷ 1000), ensuring consistent time units. Then, determine the spatial range based on the parameter type. The spatial range of structural layer characteristic parameters = axial length of the corresponding processing station × detection coverage area of ​​the parameter. For example, the detection coverage area of ​​the conductor core resistance value is the radial cross-sectional area of ​​the conductor core, and the spatial range of environmental parameters... = The volume of the rectangular block of the sensor deployment area (x-axis coverage length × y-axis coverage width × z-axis coverage height); Finally, divide the total number of data points for this type of parameter by the product of the time span and the spatial range to obtain the number of data points distributed per unit time and per unit space. The calculation formula is: parameter distribution density = total number of data points for this type of parameter ÷ (time span × spatial range). The calculation result is retained to two decimal places. After completing the calculation of the three statistical characteristics of all key parameters in the above manner, a statistical characteristic dataset of parameter name, parameter mean, parameter variance, and parameter distribution density is established for each parameter.

[0051] Step 202: Based on the statistical characteristics of the distribution of each key parameter, the cable material is divided into conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment. The conductor core segment corresponds to the conductor core resistance distribution characteristics, the insulation layer segment to the insulation layer resistance distribution characteristics, the shielding layer segment to the shielding layer resistance distribution characteristics, and the outer sheath segment to the outer sheath hardness distribution characteristics. Specifically, based on the differences in the core characteristics of each structural layer, the segmentation is logically designed to distinguish structural layers and avoid data omissions or overlaps. First, combining the material physical properties and manufacturing process requirements of each structural layer (conductor core, insulation layer, shielding layer, and outer sheath), as well as the parameter statistics of at least 100 qualified batches of products over the past year, a statistical characteristic threshold range is set for the core key parameters corresponding to each segment. The threshold range adopts... The 95% confidence interval principle is set to balance rigor with practicality. For the conductor core segment, the parameter mean threshold is ±5% of the mean of qualified products, the parameter variance threshold is 0 to the maximum variance of qualified products, and the parameter distribution density threshold is the minimum to maximum distribution density of qualified products. For the insulation layer segment, the parameter mean threshold is ±8% of the mean of qualified products (adapting to a reasonable fluctuation range in insulation resistance), and the variance and distribution density threshold logic is consistent with the conductor core. For the shielding layer segment, the parameter mean threshold is ±6% of the mean of qualified products, with the remaining threshold logic the same as above. For the outer sheath segment, the parameter mean threshold is ±7% of the mean of qualified products, with the remaining threshold logic remaining consistent to ensure that the thresholds effectively distinguish the parameter distribution differences between different structural layers.

[0052] After the threshold range is determined, the time-series data point set constructed in step 200 is traversed, and the core parameters (corresponding items in conductor core resistance, insulation layer resistance, shielding layer resistance, and outer sheath hardness) and their three statistical features corresponding to each data point are extracted one by one. The three statistical features of the data point are compared with the threshold range of the corresponding segment one by one. Only when the three statistical features simultaneously meet the threshold requirements of the corresponding segment is the classification determined. That is, if the conductor core resistance feature meets the standard, it belongs to the conductor core segment; if the insulation layer resistance feature meets the standard, it belongs to the insulation layer segment; if the shielding layer resistance feature meets the standard, it belongs to the shielding layer segment; and if the outer sheath hardness feature meets the standard, it belongs to the outer sheath segment. This avoids misclassification caused by a single feature.

[0053] After determining the segment affiliation of a single data point, a correlation check is performed on consecutive data points. For time correlation check, the timestamp difference between consecutive data points within the same segment must be controlled within ±10 milliseconds of the preset acquisition interval. For example, if the conductor core resistance value is acquired every 0.5 seconds, the difference must be between 0.49 seconds and 0.51 seconds. If it exceeds this range, it is considered a data breakpoint, requiring verification and correction of sensor or equipment records. For spatial correlation check, the spatial coordinate difference between consecutive data points must match the processing equipment's operating speed and acquisition interval. For example, if the wire drawing equipment speed is 1 meter / second and the acquisition interval is 0.5 seconds, the axial coordinate difference must be between 0.49 meters and 0.51 meters. To ensure cable segments are continuous and unbroken, after correlation verification, a final integrity and uniqueness verification is performed to ensure that all time-series data points belong to a specific segment without omission. Each data point belongs to only one segment without duplication. If three features simultaneously meet the thresholds of multiple segments, the matching degree is calculated (number of statistical features meeting the threshold ÷ 3), and the segment with the highest matching degree is selected. If the matching degrees are the same, spatial location information is used to determine the segment, prioritizing its belonging to the structural layer segment corresponding to the spatial location. This results in four independent segment datasets, each containing all time-series data points within the corresponding segment, achieving precise and continuous segmentation of cable materials according to structural layers.

[0054] This embodiment achieves precise segmentation of each structural layer of the cable material through time-series data integration and statistical feature analysis. The time-series data point set binds scattered key parameters with spatiotemporal information, ensuring that each parameter data can be traced back to the specific collection time and processing location. This solves the problems of data fragmentation and unclear spatiotemporal correlation in traceability, providing complete basic data support for full life-cycle traceability. The calculation of the statistical characteristics of key parameter distribution, by quantifying the mean, variance, and distribution density of parameters, objectively reflects the overall level, fluctuation degree, and spatial distribution law of each parameter, avoiding the segmentation deviation caused by subjective judgment and improving the accuracy and consistency of the segmentation results. Segmentation according to the core parameter distribution characteristics corresponding to the structural layer ensures that each segment can accurately correspond to the conductor core, insulation layer, shielding layer, or outer sheath of the cable, clarifying the parameter distribution boundaries of each structural layer. This avoids the pain points of ambiguous segments and difficulty in defining responsibility in traceability, providing a clear scope basis for accurate analysis of quality problems and determination of responsibility.

[0055] In a preferred embodiment of the present invention, three mass points are set in each segment to construct a spatial topology network, and a multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units, including:

[0056] Step 300: Based on the statistical distribution characteristics of the conductor core segment, three mass points are set at the starting, middle, and ending positions of the conductor core segment; based on the statistical distribution characteristics of the insulation layer segment, three mass points are set at the starting, middle, and ending positions of the insulation layer segment; based on the statistical distribution characteristics of the shielding layer segment, three mass points are set at the starting, middle, and ending positions of the shielding layer segment; based on the statistical distribution characteristics of the outer sheath segment, three mass points are set at the starting, middle, and ending positions of the outer sheath segment. Specifically, this includes:

[0057] Based on the conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment divided in step 202, and combined with the statistical distribution characteristics of each segment (parameter mean, parameter variance, parameter distribution density), three mass points are set for each segment. The three mass points correspond to the starting position, middle position, and ending position of the segment, respectively, to ensure that the mass points can cover the entire range of the segment and reflect the distribution law of the core parameters. The determination of the mass point position is based on the spatial position data of each segment. The spatial position of the conductor core segment is represented by axial coordinates. First, the axial coordinates of all time series data points in the segment are extracted, and the minimum value (i.e., the axial coordinate of the starting end of the segment) and the maximum value (i.e., the axial coordinate of the ending end of the segment) of the axial coordinates are determined. The axial coordinate of the mass point at the starting position is set to the minimum value, and the axial coordinate of the mass point at the ending position is set to the maximum value. The axial coordinate of the mass point at the middle position is calculated. The calculation formula is: axial coordinate of the middle position = (maximum value of the axial coordinate of the segment - minimum value of the axial coordinate of the segment) ÷ 2 + minimum value of the axial coordinate of the segment. The spatial position of the three mass points only retains the axial coordinates (consistent with the spatial representation of the conductor core parameters).

[0058] The spatial position of the insulation layer segment is characterized by a combination of radial and axial coordinates. First, the axial coordinates of all data points within the segment are extracted, and the minimum and maximum values ​​of the axial coordinates are determined. The axial coordinate of the mass point at the starting position is set to the minimum value, and the axial coordinate of the ending position is set to the maximum value. The axial coordinate of the intermediate position is calculated as (maximum value - minimum value) ÷ 2 + minimum value. At the same time, for each axial position, the radial coordinates of all corresponding data points are extracted and the average value is calculated (average radial coordinate = sum of all radial coordinates of the corresponding axial position of the segment ÷ number of corresponding radial coordinates). This average value is used as the radial coordinate of the corresponding mass point, and finally three mass points containing radial and axial coordinates are formed.

[0059] The spatial position of the shielding layer section is represented by axial segmented coordinates. First, determine the corresponding processing steps (pre-processing section, forming section, and inspection section) and the range of axial distances within the section. Extract the axial distances within the section for all data points within the section. Determine the minimum and maximum values ​​of the axial distances within the section. Set the axial distance within the section for the starting mass point to the minimum value and the ending mass point to the maximum value. The axial distance within the section for the middle mass point is calculated as (maximum value - minimum value) ÷ 2 + minimum value. The spatial positions of the three mass points retain the section number and the calculated axial distance within the section (consistent with the spatial coordinate representation of the shielding layer).

[0060] The spatial position of the outer sheath section is represented by a combination of circumferential and axial coordinates. First, the axial coordinates of all data points within the section are extracted, and the minimum and maximum values ​​of the axial coordinates are determined. The calculation method for the axial coordinates of the starting, middle, and ending positions is consistent with that of the insulation layer. The circumferential coordinates are selected from the fixed angles with the highest frequency of use within the section, such as 90 degrees, to ensure consistency with the sensor detection angle. These are used as the circumferential coordinates of the three mass points, ultimately forming three mass points containing both circumferential and axial coordinates. Each mass point is bound to the statistical distribution characteristic data (parameter mean, parameter variance, parameter distribution density) of the corresponding section to ensure that the mass point has both spatial positioning attributes and reflects the parameter distribution state of the corresponding position. Step 301: Construct a spatial topology network based on the spatial location of the mass points, and calculate the spatial distance between the mass points and the correlation of the distribution of key parameters. Specifically, this includes: using all the mass points set in step 300 as network nodes, constructing a spatial topology network; establishing direct connection edges between three mass points in the same segment in the order of start position, middle position, and end position; establishing indirect connection edges between the end position mass point of adjacent segments and the start position mass point of the next segment, such as connecting the end position mass point of the conductor core segment to the start position mass point of the insulation layer segment, forming a spatial topology network that covers all segments and is logically coherent; the initial weight of the network edges is set to 1; after the network is constructed, calculate two core parameters, namely the spatial distance between mass points and the correlation of the distribution of key parameters; when calculating the spatial distance between mass points, adopt the corresponding calculation method according to the spatial coordinate representation type of the mass points, that is, the spatial distance of a mass point containing only axial coordinates (conductor core segment) = the absolute value of the difference between the axial coordinates of two mass points, and the calculation formula is: spatial distance = |mass point A axial coordinate ... The axial coordinate of mass point B is used to calculate the spatial distance for mass points containing both radial and axial coordinates (insulation layer section). First, calculate the difference between the axial and radial coordinates of the two mass points, then square the differences, sum them, and take the square root to obtain the spatial distance. For mass points containing segment numbers and axial distances within the same processing segment (shielding layer section), the spatial distance is equal to the absolute value of the difference in axial distances between the two mass points within the segment. For mass points in different processing segments, the spatial distance is equal to the remaining length of the previous segment plus the used length of the next segment. For example, from the end of the pre-processing segment to the middle of the forming segment, the distance is equal to the total length of the pre-processing segment minus the distance within the segment at the end of the pre-processing segment plus the distance within the segment at the middle of the forming segment. For mass points containing circumferential and axial coordinates (outer sheath section), the spatial distance is calculated by first converting the circumferential angle to a horizontal coordinate component (the horizontal value corresponding to the circumferential coordinate = cosθ × unit length, where θ is the circumferential angle), then calculating the difference between the axial coordinates and the difference between the horizontal coordinate components, squared them, summed, and taking the square root to obtain the spatial distance.

[0061] When calculating the correlation of key parameter distributions, the statistical characteristics bound to three quality points within the same segment are used as the analysis object. For each type of key parameter, such as the conductor core resistance value and the insulation layer resistance value, the linear correlation between different statistical characteristics is calculated, such as the correlation between the parameter mean and the parameter variance. The Pearson correlation coefficient method is used for calculation. First, the means of the two statistical characteristics are calculated. Let the mean of characteristic X be... The mean of feature Y is The calculation method is as follows =Sum of all mass point values ​​for this feature ÷Number of mass points Similarly; the covariance is calculated using the formula: Covariance = (X value of each mass point - ... ) × (Y value of each mass point - Sum of (x, y) ÷ Number of mass points; Calculate the standard deviation of the two features, the standard deviation of feature X = Similarly, the standard deviation of feature Y is calculated. The correlation coefficient is calculated using the formula: correlation coefficient = covariance ÷ (standard deviation of feature X × standard deviation of feature Y). The correlation coefficient ranges from -1 to 1. The closer the absolute value is to 1, the stronger the correlation, and the closer it is to 0, the weaker the correlation.

[0062] For the four sections—conductor core, insulation layer, shielding layer, and outer sheath—defined in step 202, a core point is first selected from the three mass points set in step 300, located at the center of the section's spatial range. This core point is both at the center of the section's spatial range and bound to the statistical distribution characteristics (such as parameter mean, variance, and distribution density) of the section's central position, reflecting the core state of the overall parameter distribution of the section. To further quantify the spatial correlation characteristics of the axial and radial parameter distributions of each section, a first spatial vector and a second spatial vector are specifically defined based on the spatial coordinates of the core point (determined by combining the coordinate representation type of each section) and the structural extension direction of each section. Specifically, the first spatial vector extends along the section's axial direction, maintaining consistency with the cable processing direction. Its parameter settings fully adapt to the structural representation characteristics of each section. For the conductor core section, the axial coordinate of the core point is denoted as Z, representing the position of the core point along the cable processing direction. The linear distance, with units consistent with the axial coordinates (e.g., millimeters), is used. Since the mass parameters of the conductor core are mainly distributed along the axial direction, there is no need to additionally characterize the radial characteristics. Therefore, the corresponding first spatial vector is set as a unit vector (1, 0), where 1 represents a unit increment along the processing direction, i.e., for every unit increase in axial distance, the vector extends synchronously by 1 unit along that direction; 0 indicates no radial component, ensuring that the vector only focuses on the axial extension characteristics. For the insulation layer segment, the spatial coordinates of the core point are (Z1, R1), where Z1 represents the axial coordinate of the core point (linear distance along the processing direction, in millimeters), and R1 represents the radial coordinate of the core point (linear distance perpendicular to the processing direction and pointing towards the outside of the insulation layer, in millimeters). The corresponding first spatial vector is set as (ΔZ, 0), where ΔZ represents the extension length of the vector along the axial direction, i.e., the axial increment, and its value is set to the total axial length of the insulation layer segment itself (denoted as ΔZ). 1 / 10 of the total axial distance from the beginning to the end of the section (ΔZ= ×1 / 10), this setting ensures that the vector length is reasonable and effectively covers the distribution range of the core parameters along the axial direction of the segment, while also preventing the vector from exceeding the segment boundary due to excessive length; for the shielding layer segment, the axial distance of the core point in the corresponding processing segment, such as the pre-processing segment or the forming segment, is denoted as L, where L represents the linear distance of the core point along the processing direction within the processing segment (in millimeters). Considering that the shielding layer adopts a segmented processing method, the corresponding first spatial vector is set as (ΔL, 0), where ΔL represents the axial increment within the processing segment, and its value is the total length of the shielding layer segment within the current processing segment (denoted as ΔL, 0). 1 / 10 of the total axial distance of the shielding layer within the processing section (ΔL = ×1 / 10), fully adapting to the segmented structure characteristics of the shielding layer; for the outer sheath section, the spatial coordinates of the core point are (Z2, θ), where Z2 represents the axial coordinate of the core point (linear distance along the processing direction, in millimeters), and θ represents the circumferential angle of the core point, that is, the rotation angle of the point around the cable axis, in degrees, consistent with the high-frequency detection angle of the sensor, such as 90 degrees); the corresponding first spatial vector is set as (ΔZ', 0), where ΔZ' represents the axial increment of the outer sheath section, and its value is the total axial length of the outer sheath section itself (denoted as ). 1 / 10 of ) (ΔZ'= ×1 / 10), ensuring precise matching between the vector and the axial distribution range of the outer sheath.

[0063] Corresponding to the first spatial vector, the second spatial vector extends radially along the segment, with its direction consistent with the segment's thickness direction, i.e., perpendicular to the axial direction and pointing outwards from the segment. Its definition is also specifically set based on the structural characteristics of each segment. The conductor core segment is a single conductor structure with no difference in thickness parameters and no actual radial structure. However, to meet the calculation requirements of the spatial orientation angle, a virtual radial component vector (0, 1) is set, where 0 represents no axial component and 1 represents a virtual radial unit increment. This vector remains strictly perpendicular to the axial direction and will not affect the accuracy of the angle calculation. The second spatial vector corresponding to the insulation layer segment is set as (0, ΔR), where ΔR represents the radial extension length of the vector, i.e., the radial increment, and its value is the total radial thickness of the insulation layer segment itself (denoted as ΔR). 1 / 10 of the maximum radial distance from the inner surface to the outer surface of the insulation layer (ΔR = ×1 / 10), and the vector extends outward from the radial coordinate R1 of the core point to ensure that it can reflect the radial parameter distribution trend of the insulation layer; the second spatial vector corresponding to the shielding layer segment is set as (0, ΔR'), where ΔR' represents the radial increment of the shielding layer, and its value is the total radial thickness of the shielding layer segment itself (denoted as ×1 / 10). 1 / 10 of ') (ΔR'= ×1 / 10), conforming to the thickness distribution characteristics of the shielding layer; the second spatial vector corresponding to the outer sheath section is set as (0, ΔR''), where ΔR'' represents the radial increment of the outer sheath, and its value is the total radial thickness of the outer sheath section itself (denoted as ). 1 / 10 of '' (ΔR''= (×1 / 10), ensuring that the vector can fully cover the radial parameter distribution range of the outer sheath.

[0064] Once the first and second spatial vectors are defined, the spatial orientation angle between them can be calculated using the vector dot product formula (to avoid confusion with the circumferential angle θ of the outer sheath, the angle is denoted as θ). The specific calculation is as follows: First, calculate the vector dot product. The calculation logic for the dot product result is: axial component of the first spatial vector × axial component of the second spatial vector + radial component of the first spatial vector × radial component of the second spatial vector. Taking the insulation layer section as an example, its first vector has an axial component of ΔZ and a radial component of 0, and its second vector has an axial component of 0 and a radial component of ΔR. Therefore, the dot product result = ΔZ × 0 + 0 × ΔR = 0. Other sections can be calculated according to the corresponding vector components using the same logic. Second, calculate the vector magnitude. The magnitude is the spatial length of the vector. The calculation formula is uniformly: That is, the magnitude of the first spatial vector = The magnitude of the second spatial vector is calculated using the same formula. For example, the magnitude of the first vector (1, 0) in the conductor core segment is = =1, the magnitude of the second vector (0,1) = =1; Third step, calculate the cosine of the included angle. Divide the vector dot product obtained in the first step by the product of the magnitudes of the first and second vectors calculated in the second step to obtain the cosine of the included spatial azimuth angle; Fourth step, calculate the included spatial azimuth angle. Calculate the cosine value obtained in the third step using the inverse cosine function to finally obtain the included spatial azimuth angle. , The unit is degrees, and the value ranges from 0 degrees to 90 degrees. This included angle can directly quantify the spatial correlation characteristics of the axial and radial parameter distributions of a segment. The larger the included angle, the more significant the difference between the axial and radial parameter distributions; the smaller the included angle, the more consistent the axial and radial parameter distributions are.

[0065] Step 303: Divide the segment into multiple quality control sub-regions based on the spatial azimuth angle. The spatial azimuth angle is used to determine the distribution boundaries of the quality control sub-regions. Specifically, based on the spatial azimuth angle calculated in step 302, and combined with the spatial range and parameter distribution density of each segment, divide each segment into multiple quality control sub-regions. The size of the spatial azimuth angle directly determines the distribution boundaries and number of sub-regions. This is because a larger angle indicates a more significant difference in parameter distribution between the radial and axial directions of the segment, requiring more sub-regions for refined monitoring. The specific division rules are as follows: when the spatial azimuth angle is between 0 and 30 degrees, divide into 2 quality control sub-regions; between 31 and 60 degrees, divide into 3 quality control sub-regions; and between 61 and 90 degrees, divide into 4 quality control sub-regions. The distribution boundaries of the sub-regions are determined with the core point as the origin. The angle formed by a spatial vector and a second spatial vector is divided equally according to the number of sub-regions. Each angle bisector after division is the boundary of the sub-region. For example, when the angle is 60 degrees and divided into 3 sub-regions, the 60-degree angle is divided into 3 smaller angles of 20 degrees each. The corresponding angle bisectors (0 degrees, 20 degrees, 40 degrees, and 60 degrees) are the boundaries of the sub-regions. The spatial position of these boundaries needs to be calculated by converting the direction of the angle bisector and the spatial coordinates of the segment. For example, the axial and radial coordinates of the boundary of a sub-region of the insulation layer segment are calculated according to the slope of the angle bisector (tanα, where α is the angle between the angle bisector and the axis) to ensure that the boundary can accurately fit the structural shape of the segment. Each quality control sub-region contains the time series data points within that range, the corresponding statistical feature data, and the correlation information of the quality points. It can independently reflect the local parameter distribution state and maintain a close relationship with the whole segment. Step 304: Establish a multi-level grid structure based on the spatial topology network and quality control sub-regions. Dynamically optimize the multi-level grid structure using the statistical characteristics of the distribution of key parameters and the spatial orientation angle. This includes adjusting the grid density of the multi-level grid structure based on the correlation of the distribution of key parameters, and optimizing the grid connection relationship of the multi-level grid structure based on spatial distance and spatial orientation angle to form data units with spatial constraints. Specifically, based on the spatial topology network constructed in step 301 and the quality control sub-regions divided in step 303, a multi-level grid structure is further established. This structure consists of three layers: the first layer is the segment layer of the entire cable, covering four segments: conductor core, insulation layer, shielding layer, and outer sheath, forming the macro framework of the grid; the second layer is the quality control sub-region layer for each segment, which is the intermediate association layer of the grid, connecting the segments and data units; the third layer is the specific data unit layer within the sub-region, where each data unit contains a set of time-series data points and corresponding spatiotemporal information and statistical characteristics, serving as the basic data carrier of the grid.

[0066] After the grid structure is established, dynamic optimization is required, taking into account the statistical characteristics of the key parameter distribution, spatial orientation angles, and spatial distances. The optimization process consists of two core steps. The first step is to adjust the grid density of the multi-level grid structure, based on the correlation of the key parameter distribution calculated in step 301. If the absolute value of the correlation coefficient between two types of key parameters is greater than 0.7, it indicates that the parameter distribution in this area is strongly correlated, and the grid density needs to be increased to accurately capture parameter changes. Specifically, the number of data units in the corresponding sub-region is increased by 50% (number of data units = original number of data units × 1.5, rounded up). If the absolute value of the correlation coefficient is between 0.3 and 0.7, it indicates that the parameter distribution is moderately correlated, and the grid density remains unchanged. If the absolute value of the correlation coefficient is less than 0.3, it indicates that the parameter distribution is discrete, and the grid density can be reduced to improve computational efficiency. The adjustment method is to reduce the number of data units by 30% (number of data units = original number of data units × 0.7, rounded down). The second step is to optimize the grid connection relationship of the multi-level grid structure, based on the spatial distance between mass points calculated in step 301 and the distance calculated in step 302. Based on the spatial orientation angle, mass points with a spatial distance less than 1 / 5 of the average axial length of the segment are spatially close and have closely related parameters. Their mesh connection strength needs to be enhanced, and the weight of the mesh edge is adjusted from 1 to 1.5. Mass points with a spatial distance greater than 1 / 2 of the average axial length of the segment are spatially distant and have weaker parameter correlations. Their connection strength is weakened, and the weight is adjusted to 0.5. For sub-regions with a spatial orientation angle greater than 60 degrees, the radial and axial parameter distributions differ significantly. Direct connection edges between data units and core points within the sub-region need to be added to ensure that the core point can quickly connect all data within the sub-region. For sub-regions with an orientation angle less than 30 degrees, the radial and axial parameter distributions are relatively consistent. Connection relationships are simplified, retaining only connections between data units and adjacent data units to reduce redundant connections. After optimization, data units with spatial constraints are finally formed. Each data unit not only contains its own parameter data and spatiotemporal information but is also closely connected to adjacent data units, its sub-region, and corresponding mass points through mesh connections. This reflects both the spatial distribution pattern of parameters and clearly demonstrates the correlation and spatial constraints between parameters.

[0067] This embodiment achieves a refined and structured upgrade in cable quality control through quality point setting, spatial topology network construction, sub-region division, and grid structure optimization. The quality point setting anchors the key monitoring locations in each section, avoiding blind monitoring. The construction of the spatial topology network links scattered quality points and parameter data, clearly presenting the correspondence between parameters and spatial locations, providing a logical framework for quality analysis across locations and sections. The calculation of spatial vectors and azimuth angles quantifies the axial and radial spatial relationships of each structural layer of the cable, providing an objective basis for sub-region division, enabling quality control to be refined from overall sections to local sub-regions, improving the accuracy of quality problem location. The establishment and dynamic optimization of the multi-level grid structure, combined with factors such as parameter correlation and spatial distance, adjusts the grid density and connection relationships, ensuring monitoring accuracy in parameter-dense areas while also considering overall computational efficiency.

[0068] In a preferred embodiment of the present invention, according to the spatial constraint relationship of data units, time-series data points are mapped to corresponding data units, spatial distribution characteristic values ​​of data points within each data unit are calculated, and parameter calibration coefficients are generated based on the spatial distribution characteristic values, including:

[0069] Step 400: Based on the spatial constraints of the data units, map the time-series data points to the corresponding data units, and calculate the spatial distribution characteristic values ​​of the data points within each data unit, including data point density and data point distribution uniformity. Data point density is calculated as the ratio of the number of data points to the volume of the data unit. Data point distribution uniformity is calculated by constructing a convex boundary polygon based on the data point set and calculating the minimum circumcircle of the data point set. The inclusion relationship between the minimum circumcircle and the convex boundary polygon is detected by radiating rays from the center of the minimum circumcircle to the boundary of the convex boundary polygon, counting the number of intersections between the rays and the polygon boundary, and determining whether the center of the ray is inside the convex boundary polygon based on the parity of the number of intersections. The area of ​​the overlapping region between the minimum circumcircle and the convex boundary polygon is also calculated. The data point distribution uniformity is determined based on the ratio of the overlapping region area to the area of ​​the minimum circumcircle, specifically including:

[0070] The core of time-series data point mapping is to assign the time-series data points collected from each section of the cable to the corresponding grid data units, ensuring that the data assignment is accurate and complete. The specific operation is as follows: The cable is categorized into four sections: conductor core, insulation layer, shielding layer, and outer sheath. All time-series data points for each section are extracted sequentially. Each data point must be fully collected to match the spatiotemporal information of its corresponding section. For the conductor core, only axial distribution is focused, and the axial position information extracted from the data points is the coordinates along the cable length. The insulation layer involves both axial and radial two-dimensional distribution; the data points need to extract both axial position and radial position (distance from the cable's central axis). The shielding layer is categorized according to processing steps. For segment acquisition, data points need to be extracted based on the unique segment number of the processing stage at the time of acquisition (e.g., S1 for preprocessing segment, S2 for forming segment) and the axial distance within that segment. For the outer sheath, which covers both axial and circumferential distribution, data points need to be extracted based on the axial position and circumferential angle (rotation angle around the cable axis, 0 to 360 degrees). During the extraction process, the preset coordinate dimension template needs to be associated with the segment identifier to automatically verify the number and type of dimensions of the data points. For example, if radial information appears in the conductor core data points, it is judged as a mismatch and directly marked as invalid data to ensure that the spatial coordinate dimension of each valid data point is completely matched with the segment to which it belongs, without any missing or mismatched dimensions.

[0071] After information extraction, the coordinate range matching process begins. The starting and ending values ​​(including angular boundary starting values) of all data units are derived from the previously completed multi-level grid structure dynamic optimization results. During optimization, based on the actual structural and process parameters such as the length, thickness, and processing range of each cable segment, combined with quality control accuracy requirements, each segment is divided into uniform or differentiated grids. This ultimately clarifies the boundary coordinates of each data unit, i.e., various starting and ending values. The boundaries of the divided data units do not overlap and completely cover the corresponding segments. For each valid time-series data point, the corresponding segment is first selected by its segment identifier. Then, spatial coordinate range matching is performed only against all grid data units within that segment. During matching, each data point is checked item by item according to its coordinate dimension. For conductor core data points, only the axial position is checked. The axial starting value of the data unit is the starting coordinate along the axial direction after grid optimization, and the axial ending value is the corresponding ending coordinate. If the axial position of the data point is ≥ the starting value and ≤ the ending value, then a preliminary match is determined. For insulation layer data points, both axial and radial positions need to be checked separately. The logic for determining the end value is the same as that for the conductor core. The radial starting value is the coordinate of the inner boundary of the cell adjacent to the conductor core after mesh optimization, and the radial ending value is the coordinate of the outer boundary towards the shielding layer. If both conditions are met, a preliminary match is determined. For the shielding layer data points, all data units corresponding to the processing segment are locked according to the segment number. The axial starting and ending values ​​within the segment of this unit are the boundary coordinates divided according to the length of the corresponding processing segment after mesh optimization. If the axial distance within the segment of the data point is ≥ the starting value and ≤ the ending value, a preliminary match is determined. For the outer sheath data points, the axial position and circumferential angle need to be checked separately. The logic for determining the axial starting and ending values ​​is the same as above. The starting value of the circumferential angle is the starting point of the uniform division boundary preset during mesh optimization, such as 0 degrees, 90 degrees, etc., and the corresponding ending value is the ending point of the division boundary, such as 90 degrees, 180 degrees, etc. The error is allowed to be ±0.1 degrees. If both conditions are met, a preliminary match is determined. In the unique attribution confirmation stage, since the spatial coordinate ranges of the data units do not overlap, each time series data point will only satisfy the complete matching conditions with one data unit. After confirming the matching relationship, the data point is formally mapped to the data unit, and the data point's affiliation is marked to avoid duplicate mapping (one data point belongs to multiple units) or omission mapping (a data point has no corresponding unit). Finally, a unique correspondence is formed between each data unit and its time series data point, providing a regular data foundation for subsequent feature value calculation.

[0072] Based on the mapping relationship between data units and time-series data points, two core spatial distribution characteristic values ​​of data points within each data unit are calculated: data point density and data point distribution uniformity. The specific calculation process is as follows: Data point density calculation (characterizing the density of data points): Data point density is calculated by the ratio of the number of data points within a data unit to the volume (or area, length, adapting to the spatial dimension of different segments). First, the number of data points is counted. The total number of successfully mapped time-series data points within each data unit is counted one by one. This number represents the number of valid data points actually participating in the calculation, excluding duplicate or invalid data points. Second, the volume of the data unit is calculated (adapting to the spatial dimension of different segments). For the conductor core segment, which is a one-dimensional space with only axial direction, the data unit volume = data unit axial end coordinate - data unit axial start coordinate (unit consistent with the axial coordinate, such as millimeters). Since only axial distribution needs to be characterized, linear length is used instead of volume. For the insulation layer segment, which is a two-dimensional space with axial and radial directions, the data unit volume = (data unit axial end coordinate - data unit axial start coordinate) × (Radial end coordinate of data unit - Radial start coordinate of data unit) (unit: square millimeters), which is the area of ​​the data unit in a two-dimensional plane; the shielding layer section is a one-dimensional space along the axial direction within the section, and the volume of the data unit = axial end coordinate of the data unit section - axial start coordinate of the data unit section (unit: millimeters), consistent with the conductor core, using the linear length within the section to replace the volume; the outer sheath section is a two-dimensional space in the axial and circumferential directions, first converting the circumferential angle range into arc length (arc length = (data unit circumferential angle end value - data unit circumferential angle start value)). () ÷ 360 × 2 × pi × radial radius of outer sheath), then calculate the volume = (axial end coordinate of data unit - axial start coordinate of data unit) × arc length (unit is square millimeter), that is, the two-dimensional area formed by the axial length and the circumferential arc length; the third step is to calculate the data point density, data point density = total number of time series data points in the data unit ÷ volume (or area, length) of the data unit, the unit of the calculation result is points / mm, points / square millimeter, etc. The larger the value, the denser the data points in the data unit; the smaller the value, the sparser the data points.

[0073] Data point distribution uniformity calculation (characterizing the regularity of data point distribution): Data point distribution uniformity is calculated by analyzing the overlap relationship between the convex boundary polygon of the data points and the smallest circumcircle. The specific steps are as follows: First, construct the convex boundary polygon. Select all time series data points within the data unit as vertices of the polygon. Connect all vertices sequentially according to the core order of axial coordinates from smallest to largest, combined with the order of radial (or circumferential) coordinates (radial from inside to outside, circumferential angle from smallest to largest). During the connection process, it is necessary to ensure that the resulting polygon is a convex polygon, that is, the interior angle formed by any three consecutive vertices is less than 180 degrees, and there are no concave parts. If a concave trend appears, the vertex connection order needs to be adjusted. The final convex boundary polygon should completely enclose the data unit. The first step involves identifying all time-series data points within the polygon, ensuring no data point is missed outside the polygon. The second step is to calculate the minimum circumcircle, which is the circle with the smallest area that completely encloses all time-series data points. The calculation process begins by determining the initial center and calculating the mean of each coordinate dimension for all data points. For example, the mean of axial coordinates = the sum of the axial coordinates of all data points ÷ the number of data points; the mean of radial coordinates = the sum of the radial coordinates of all data points ÷ the number of data points. This mean coordinate is used as the initial center. Next, the minimum radius is determined by calculating the straight-line distance from each data point (axial coordinate, radial / circumferential coordinate) to the initial center (axial coordinate, radial / circumferential coordinate). The maximum distance is recorded as the initial radius. Finally, the center and radius are fine-tuned. First, adjust the coordinates of the initial center (each adjustment should not exceed 1% of the data cell volume). Recalculate the maximum distance from each data point to the new center. If the new maximum distance is less than the initial radius, update the center and radius. Repeat this process until the maximum distance cannot be reduced further. The center and radius at this point are the final parameters of the minimum circumcircle. Second, check the inclusion relationship between the minimum circumcircle and the convex boundary polygon. Starting from the center of the minimum circumcircle, emit a ray horizontally to the right (choosing horizontally to the right avoids the special case of the ray coinciding with a polygon vertex or edge; if a special case occurs, adjust to horizontally to the left, vertically upward, etc.). Count the number of intersections between this ray and the boundary of the convex boundary polygon. If the number of intersections is odd, it indicates that the ray originates from... If the number of intersection points is odd, the center of the circle is determined to be inside the convex polygon. If the number of intersection points is even, it means the ray enters the polygon from the outside, crosses the boundary, and exits without any additional intersections. The fourth step is to calculate the area of ​​the overlapping region. The overlapping region area is the area of ​​the space jointly covered by the smallest circumcircle and the convex polygon. The calculation is based on the inclusion relationship. In case one, the center of the circle is inside the convex polygon (the polygon is completely inside the circle). In this case, the overlapping region area equals the area of ​​the convex polygon. The polygon area is calculated using the Shoelace formula, arranging the coordinates of the polygon vertices in order. , () , )...( , ), area = (in +1= , +1= ), It is an index. =1 to = Case 2: The center of the circle is located outside the convex boundary polygon (the circle and the polygon partially overlap). First, calculate the intersection point of each edge of the convex boundary polygon with the smallest circumcircle. Each edge corresponds to two consecutive vertices of the convex boundary polygon. Let their coordinates be (…). , )and( , The two-point form of the equation of a line segment is determined as follows: if ≠ and ≠ The equation is (x- ) / ( - )=(y- ) / ( - );like = The equation of the perpendicular line is x= ;like = The equation of the horizontal line is y= Compare the equation of the line segment with the equation of the circle ( After solving the equations simultaneously to obtain the coordinates of the intersection point, it is necessary to further determine whether the x-coordinate of the intersection point is within the range of... and Is the y-coordinate between...? and The process involves several steps: First, confirming whether the intersection point lies on a line segment. Then, based on the valid intersection point, divide the polygon's sides into line segments inside and outside the circle, dividing the overlapping area into several triangles and sectors. Calculate the area of ​​each triangle (base × height ÷ 2) and the area of ​​each sector (central angle ÷ 360 × pi × radius²). Finally, add the areas of all triangles and sectors to obtain the total area of ​​the overlapping area. The fifth step is to determine the uniformity of data point distribution. The uniformity of data point distribution is calculated as: Area of ​​overlapping area ÷ Area of ​​minimum circumscribed circle (Area of ​​minimum circumscribed circle = pi × radius²). This ratio ranges from 0 to 1. The closer the ratio is to 1, the more regular and uniform the data points are distributed in space. The closer the ratio is to 0, the more concentrated the data points are in a small area or scattered in multiple isolated areas, indicating poorer distribution uniformity.

[0074] Step 401: Generate parameter calibration coefficients based on spatial distribution characteristic values, including density calibration coefficients and distribution uniformity calibration coefficients. Specifically, the generation of density calibration coefficients proceeds as follows: First, determine the standard density benchmark for data points. Firstly, collect approximately 10 to 20 batches of inspected and qualified cables, classifying them by conductor core, insulation layer, and other sections. Extract the historical actual density of all data units in the corresponding section of each batch (these historical actual densities are calculated in past production using the method of dividing the total number of valid time-series data points within the data unit in step 400 by the volume (or area, length) of the corresponding data unit). Next, calculate the arithmetic mean of all extracted historical actual densities; for example, the statistical average is 5 units / square millimeter. Then, combine this with a preset ±10% process error range. The following steps are performed: First, the average value is verified. If the average value is within the allowable error range, it is directly set as the standard density reference value (i.e., 5 points / square millimeter). If the average value exceeds the error range, such as an average value of 5.6 points / square millimeter, which exceeds the upper limit of 5.5 points / square millimeter corresponding to +10%, the critical value within the error range (5.5 points / square millimeter) is taken as the standard density reference value to ensure that the reference meets the process requirements. Second, the actual data point density of the data unit is obtained. This value has been calculated in step 400 and can be directly called. Its calculation logic is the total number of valid time-series data points in the data unit ÷ the volume (or area, length) of the corresponding data unit. Third, the calibration coefficient is calculated according to the formula: density calibration coefficient = actual data point density of the data unit ÷ standard density reference value.

[0075] When applying, follow these logics: If the density calibration coefficient > 1.1, it indicates that the data points within the data unit are too dense and redundant. In this case, it is necessary to first obtain the original calculated values. These values ​​are the extracted time-series original sensor data (such as resistance measurements, radial position data, thickness detection values, etc.) from the data unit, substituted into the basic calculation formulas for the corresponding quality parameters, such as resistance = resistivity × length ÷ cross-sectional area, insulation thickness = maximum radial position value - minimum radial position value. These are the initial results of the quality parameters directly calculated without any calibration correction. In the resistance calculation, the length is the axial boundary span of the current data unit, i.e., the difference between the axial end coordinate and the start coordinate of the unit determined after mesh optimization; the cross-sectional area is the nominal cross-sectional area of ​​the cable conductor (preset according to the production material and specifications) or based on the radial dimensions within the data unit. The equivalent cross-sectional area is calculated by inch; the maximum and minimum radial position values ​​in the insulation layer thickness calculation are the maximum and minimum values ​​selected from the radial position information of all valid time-series data points in the data unit, and the radial positions of the data points all come from the valid sensor data verified in step 400; then the influence weight of redundant data is reduced by the quality parameter correction value = original calculated value ÷ calibration coefficient; if the calibration coefficient < 0.9, it means that the data points in the data unit are too sparse. The original calculated value is obtained in the same way as above, and then the weight of the valid data is increased by the correction value = original calculated value × reciprocal of calibration coefficient to compensate for the missing information; if the calibration coefficient is between 0.9 and 1.1, it means that the data point density is reasonable and no additional adjustment is needed. The original calculated value obtained by the above method can be used directly.

[0076] The generation and application of the distribution uniformity calibration coefficient is based on the actual distribution uniformity of the data units accurately calculated in step 400. By setting key parameters, quantifying the calculation coefficients, and specifically correcting the quality parameters, the evaluation bias caused by uneven data point distribution is ultimately eliminated, ensuring that the quality parameter results accurately reflect the actual production status of the cable. First, the setting logic of two key parameters is clarified. The standard distribution uniformity of data points is determined one by one according to the cable conductor core, insulation layer, shielding layer, and outer sheath. The specific acquisition process is as follows: First, focus on the target section and collect historical production data from nearly 10 to 20 batches of cables that have passed quality inspection, ensuring the representativeness and reliability of the data source. Second, extract the historical actual distribution uniformity of all data units in the corresponding section from each batch of historical data. This historical value must strictly follow the unified calculation logic set in step 400, i.e., data unit... The first step involves calculating the convex boundary polygon formed by the valid time-series data points, and dividing the area of ​​the overlap between this convex boundary and the smallest circumscribed circle by the area of ​​the smallest circumscribed circle. This ensures consistency between historical data and current calculation standards. The second step involves calculating the arithmetic mean of all extracted historical actual distribution uniformity data, using this as the basic reference value for standard setting, reflecting the actual distribution level of past qualified products. The third step involves optimizing and calibrating the basic reference value based on two core criteria: ideal processing state characteristics (i.e., data points are neither concentrated nor isolated, uniformly covering the entire data unit, with a theoretical distribution uniformity approaching 1), and industry-wide quality acceptance standards or enterprise-specific acceptance specifications (such as the recommended standard value of 0.9 in the conventional cable industry). Finally, by comprehensively considering the basic reference value, ideal state requirements, and industry / enterprise standards, a unique "standard distribution uniformity of data points" for this section is determined, which is both in line with actual production and meets the bottom line of quality control. The fixed adjustment coefficient is set according to the different sensitivities of production quality control. For high-precision production scenarios (such as aerospace cables), the value is 0.8-1.0 to strengthen the correction of slight uneven distribution. For conventional production scenarios (such as civilian communication cables), the value is 0.3-0.7 to balance the calibration effect with the authenticity of the original data. For extensive production scenarios, the value is 0.1-0.2 to only make slight corrections to data with severe uneven distribution, adapting to the quality control needs of different scenarios.

[0077] The generation and application of the distribution uniformity calibration coefficient is based on the actual distribution uniformity of the data units calculated in step 400. By setting key parameters, quantifying the calculated coefficients, and specifically correcting the quality parameters, the evaluation bias caused by uneven data point distribution is ultimately eliminated, ensuring that the quality parameter results accurately reflect the actual production status of the cable. First, the setting logic of two key parameters is clarified: the standard distribution uniformity of data points is determined according to the cable conductor core, insulation layer, and other sections. The specific acquisition process is as follows: first, collect historical production data from nearly 10 to 20 batches of cables that have passed quality inspection; extract the historical actual distribution uniformity of the corresponding data units (this historical value strictly follows the calculation logic of the overlapping area of ​​the convex boundary polygon and the minimum circumscribed circle ÷ the area of ​​the minimum circumscribed circle in step 400 to ensure data source consistency); then, these historical actual distribution... The uniformity is calculated using an arithmetic mean, which serves as the basic reference value for standard setting. If this basic reference value falls within the reasonable range defined by the ideal processing characteristics (data points uniformly cover data units, and the theoretical uniformity approaches 1) and the industry / enterprise quality acceptance standards, then the basic reference value is directly set as the standard distribution uniformity for that segment. If the basic reference value exceeds this reasonable range, then the industry / enterprise quality acceptance standard value is used as the final standard distribution uniformity (prioritizing the quality acceptance baseline requirements) to ensure that the standard is both aligned with actual production and complies with quality control specifications. The fixed adjustment coefficient is set according to the differences in production quality control sensitivity: 0.8 to 1.0 for high-precision production scenarios (such as aerospace cables), 0.3 to 0.7 for conventional production scenarios (such as civilian communication cables), and 0.1 to 0.2 for extensive production scenarios.

[0078] After the parameters are set, the calibration coefficient calculation and quality parameter correction are carried out according to the continuous process. First, the calibration coefficient for distribution uniformity is calculated using the formula: Distribution Uniformity Calibration Coefficient = 1 + (Standard Distribution Uniformity of Data Points - Actual Distribution Uniformity of Data Units) × Fixed Adjustment Coefficient. The actual distribution uniformity of data units directly uses the result calculated in step 400, eliminating the need for recalculation and ensuring process efficiency. The original calculated values ​​used for correction are completely consistent with the definitions in density calibration. They are obtained by extracting the time-series original sensor data within the data units and substituting it into the corresponding basic calculation formulas. Specifically, this includes: Outer Sheath Hardness = Arithmetic Mean of Effective Hardness Sensor Measurements within the Data Unit (effective data is the compliant data verified in step 400); Shielding Coverage = (Number of Effectively Covered Data Points ÷ Total Number of Effective Data Points) × 100% (effectively covered data points are the data points determined by the sensor to be covered). The initial calculated value has not undergone any calibration correction and may have deviations due to uneven distribution. In specific correction, different logic is executed based on the comparison results between the actual distribution uniformity and the standard value. If the actual distribution uniformity of the data unit is less than the standard distribution uniformity of the data point, the corresponding difference in the formula is positive and the calibration coefficient is greater than 1. At this time, the deviation caused by uneven distribution is compensated by the quality parameter correction value = original calculated value × calibration coefficient. If the actual distribution uniformity of the data unit is greater than or equal to the standard distribution uniformity of the data point, the corresponding difference is 0 or negative and the calibration coefficient is approximately 1. This indicates that the data point distribution is regular and the original calculated value is highly reliable. The original calculated value is directly used to ensure parameter stability. Finally, the generated distribution uniformity calibration coefficient and the density calibration coefficient mentioned above are respectively associated with the quality parameters (such as resistance, thickness, hardness, etc.) of the corresponding data unit to reflect the actual production status of the cable.

[0079] In this embodiment, the mapping of time-series data points ensures clear data ownership for each data unit, avoiding data confusion or omissions. The quantitative calculation of data point density and distribution uniformity enables a concrete representation of the spatial distribution of time-series data within a data unit, filling the gap in quality control that focuses only on parameter values ​​while ignoring data distribution characteristics. The parameter calibration coefficients generated based on spatial distribution characteristic values ​​can specifically correct quality parameter deviations caused by uneven data density and irregular distribution, improving the accuracy and reliability of cable quality assessment. The deep correlation between the calibration coefficients and the grid structure and quality point parameters strengthens the coupling relationship between spatial modeling and quality control, helping to achieve refined and intelligent management and control of cable quality.

[0080] In a preferred embodiment of the present invention, multi-level calibration of key parameter data is performed using parameter calibration coefficients to obtain calibrated key parameter data. Signal processing is then performed on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf optimization algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, resulting in verified feature data, including:

[0081] Step 500: Perform density calibration using density calibration coefficients and distribution uniformity calibration using distribution uniformity calibration coefficients to obtain calibrated key parameter data. Specifically, this includes: First, clarifying that the calibration object is the original calculated value defined in step 401, including the initial results of key quality parameters such as resistance, insulation layer thickness, outer sheath hardness, and shielding layer coverage. The calibration is performed in the order of density calibration first, followed by distribution uniformity calibration. During density calibration, for each key parameter's original calculated value, the corresponding density calibration coefficient is called for calculation. If the density calibration coefficient is greater than 1.1, it indicates that the data points in the data unit are too dense and redundant. The calibration calculation process is to divide the original calculated value by the density calibration coefficient. If the density calibration coefficient is less than 0.9, it indicates that the data points in the data unit are too sparse. The calibration calculation process is to multiply the original calculated value by the reciprocal of the density calibration coefficient. If the density calibration coefficient is between 0.9 and 1.1, it indicates that the data point density is reasonable, and the original calculated value is directly retained as the density-calibrated data.

[0082] The distribution uniformity calibration is performed by calling the corresponding distribution uniformity calibration coefficient based on the density-calibrated data. If the distribution uniformity calibration coefficient is greater than 1, it indicates that the data points are unevenly distributed. The calibration calculation process is to multiply the density-calibrated data by the distribution uniformity calibration coefficient. If the distribution uniformity calibration coefficient is approximately 1, it indicates that the data points are evenly distributed. The density-calibrated data is directly retained. After completing the two calibrations, the calibration results of all parameters are integrated to form the calibrated key parameter data, ensuring that the data has eliminated the deviations caused by abnormal data point density and uneven distribution.

[0083] Step 501: Perform signal processing on the calibrated key parameter data to eliminate signal interference components, obtaining processed data. Specifically, this includes: first, identifying that signal interference components mainly include measurement noise generated during sensing, environmental electromagnetic interference, random interference caused by equipment vibration, and other irrelevant signals; using a moving average filtering method for signal processing; and then setting the sliding window size based on the data sampling frequency and interference frequency characteristics, selecting 3 to 20 consecutive data points as a window to ensure effective filtering of high-frequency interference without losing the core data trend; iterating through the calibrated key parameter data, for each data point, selecting all consecutive data points within the sliding window centered on that data point, and calculating the arithmetic mean of these data points. The calculation process involves dividing the sum of all data point values ​​within the window by the number of data points within the window. The calculated arithmetic mean is then used to replace the original value of the central data point. This replacement process is repeated for all data points to obtain the processed data.

[0084] Step 502 involves using the gray wolf optimization algorithm to extract features from the signal-processed data. This includes optimizing the selection of feature parameters by simulating gray wolf hunting behavior, including initializing the gray wolf population, calculating fitness values, updating gray wolf positions, and determining the final combination of feature parameters. Specifically, step 502a involves initializing the gray wolf population by setting the population size and randomly generating initial feature parameter combinations. Fitness values ​​are calculated based on these initial feature parameter combinations to evaluate the ability of each combination to distinguish the signal-processed data. This includes: firstly, clarifying the core components of the feature parameters; all parameters are designed around the needs of distinguishing cable quality data, ensuring direct correlation with the data distinction targets (qualified and unqualified data, valid and invalid data). Specifically, this includes: one, classification thresholds corresponding to key quality parameters, covering resistance qualification thresholds, insulation layer thickness validity thresholds, and outer sheath hardness qualification thresholds; and two, time-series feature extraction. The window size refers to the length of the continuously acquired segment of time-series data after signal processing. Thirdly, there are the weighting coefficients for multi-source sensor data, including radial position data weight, hardness sensor data weight, and thickness sensor data weight. These three weighting coefficients work together to improve the accuracy of multi-source data fusion. Next, the number of gray wolf populations is set. The population size is dynamically adjusted based on the feature parameter dimensions and optimization precision. If the feature parameter dimensions are 3 to 5, for example, only including the resistance qualification threshold, the effective insulation layer thickness threshold, and the time-series feature extraction window size, the population size is set to 20. If the dimensions are 6 to 10, for example, including the resistance qualification threshold, the effective insulation layer thickness threshold, the outer sheath hardness qualification threshold, the time-series feature extraction window size, and the three types of multi-source sensor data weighting coefficients, the population size is set to 50. Each individual in the population uniquely corresponds to a complete set of feature parameters, avoiding parameter redundancy that leads to reduced optimization efficiency.

[0085] Then, initial feature parameter combinations are randomly generated. The parameter values ​​of each combination strictly follow physical meaning and industry standard constraints. The specific values ​​are as follows: In the classification threshold, the resistance qualification threshold is set to 0.02 to 0.08 Ω / m (based on the industry standard for copper conductor low-voltage cables), the effective threshold for insulation layer thickness is set to 0.9 to 1.1 mm (adapting to the design specifications of conventional low-voltage cable insulation layers), and the qualified threshold for outer sheath hardness is set to 65 to 75 HB (meeting the industry performance requirements for rubber outer sheaths). The size of the time-series feature extraction window needs to match the data sampling frequency. If the sampling frequency is 10 Hz, that is, sampling every second... With 10 data points collected, the window size is set to 3 to 20, corresponding to continuous time-series data of 0.3 to 2 seconds. This captures short-term data trends while avoiding trend lag caused by an excessively large window. The total weight coefficient of the multi-source sensor data is fixed at 1, and the individual coefficients are controlled within the range of 0.1 to 0.5. Specifically, the weights are 0.25 for radial position data, 0.35 for hardness sensor data, and 0.4 for thickness sensor data. This value ensures a balanced contribution from various sensor data types while highlighting the core role of thickness sensor data in determining the quality of cable insulation and outer sheath, avoiding a single data point dominating the differentiation results.

[0086] After the initial population is generated, the fitness value of each feature parameter combination is calculated for each group. The fitness value is used to directly quantify the accuracy of the combination in distinguishing the processed signal data. The calculation process is coherent and logically rigorous. The first step is to classify the processed signal data based on the current feature parameter combination. First, the corresponding sensor data is weighted and fused using multi-source sensor data weighting coefficients (radial position 0.25, hardness 0.35, thickness 0.4). Then, the data is divided into qualified and unqualified data based on the resistance qualification threshold of 0.02 to 0.08 Ω / m, and the data is divided into valid and invalid data based on the effective insulation layer thickness threshold of 0.9 to 1.1 mm. Time-series feature extraction is then performed. The first step involves determining the overall data validity after extracting the data trend by window size. The second step involves retrieving the actual offline quality inspection results corresponding to the batch of signal-processed data, including manually measured resistance values ​​and qualification status, measured insulation layer thickness, and outer sheath hardness test results. These measured results are used as the true labels for data classification. The third step involves counting the number of samples whose classification results of the current feature parameter combination match the true labels. The fitness value is calculated by dividing the number of correctly classified samples by the total number of samples participating in the classification. The value ranges from 0 to 1. The closer the value is to 1, the higher the discrimination accuracy of the feature parameter combination and the stronger its ability to distinguish cable quality data.

[0087] Step 502b involves determining the feature parameter combinations corresponding to the first fitness value, the second fitness value, and the third fitness value based on the fitness value ranking results. Specifically, this includes: sorting the fitness values ​​of all gray wolf individuals in descending order; if individuals with the same fitness value appear during the ranking process, prioritizing individuals with parameter combinations closer to industry standard parameters (e.g., combinations with resistance thresholds closer to industry recommended values); and identifying three core final combinations based on the ranking results. The first-ranked combination is the current globally optimal feature parameter combination (the combination corresponding to the first fitness value), representing the combination with the strongest discriminative ability in the current iteration round; the second and third-ranked combinations are locally optimal feature parameter combinations (the combinations corresponding to the second and third fitness values), representing combinations with suboptimal discriminative ability and different parameter distributions within the round. These three combinations together constitute the guiding benchmark for population updates, ensuring that the update direction has both global optimality and diversity.

[0088] Step 502c: Update the position of the gray wolf population based on the feature parameter combinations corresponding to the first fitness value, the second fitness value, and the third fitness value, generating new feature parameter combinations. Specifically, this includes: updating the position of each individual in the population according to the position update rules of the gray wolf algorithm based on the three selected optimal feature parameter combinations, generating new feature parameter combinations. The specific update process is as follows: First, calculate the distance between each individual in the population and the three optimal combinations. The distance is calculated using Euclidean distance. The calculation process is to take the square root of the sum of the squares of the differences of each feature parameter. That is, for an individual's feature parameter combination and the first optimal combination, the distance is equal to the sum of the squares of all corresponding feature parameters (individual parameter value - first optimal combination parameter value), and then take the square root of this sum. Similarly, calculate the distances with the second and third optimal combinations. Then, let... The coefficients a and c are fixed. The coefficient a decreases linearly with the number of iterations, with an initial value of 2. After each iteration, the coefficient a is the previous coefficient a minus (2 divided by the preset number of iterations), until the coefficient a drops to 0 at the end of the iteration. The coefficient c is a random number between 0 and 2, which is randomly generated in each iteration. Finally, the new position of each individual is calculated. The calculation process of the new position is as follows: the parameter value of the first optimal combination is multiplied by (1 minus the coefficient c multiplied by the distance between the individual and the first optimal combination divided by the sum of three distances) + the parameter value of the second optimal combination is multiplied by (1 minus the coefficient c multiplied by the distance between the individual and the second optimal combination divided by the sum of three distances) + the parameter value of the third optimal combination is multiplied by (1 minus the coefficient c multiplied by the distance between the individual and the third optimal combination divided by the sum of three distances). The sum is then multiplied by the coefficient a to obtain a new combination of feature parameters, ensuring that the new combination moves closer to the optimal combination.

[0089] Step 502d involves repeating the fitness value calculation and position update process until the preset number of iterations is reached, determining the final feature parameter combination. Specifically, this includes: using the generated new feature parameter combination as the population individuals for the next iteration; repeating the fitness value calculation process of 502a (classifying data according to the new combination, comparing true labels, counting the number of correct results, and calculating fitness values); then re-selecting the new round of globally optimal and locally optimal combinations according to the rules of 502b; continuing to execute the position update of 502c to generate the next batch of new combinations. During the iteration process, the entire... If the improvement in the global optimal fitness value and its corresponding parameter combination is less than 0.001 over 10 consecutive rounds, i.e. the improvement in discrimination accuracy is less than 0.1%, the iteration can be terminated early (to avoid invalid loops). If the early termination condition is not triggered, the iteration continues until the preset number of iterations is reached (50 times for 3 to 5 parameter dimensions, and 100 times for 6 to 10 dimensions). After the iteration terminates, the combination with the largest fitness value and the best parameter combination stability (the smallest parameter fluctuation range in multiple iterations) is selected from the global optimal combinations recorded in all rounds as the final feature parameter combination.

[0090] Step 503 verifies the authenticity and timeliness of the processed data based on the final feature parameter combination, obtaining verified feature data. Specifically, this includes: based on the final feature parameter combination determined in step 502d, performing dual verification of the authenticity and timeliness of the signal-processed data, eliminating false and outdated data, and ensuring that the output data both accurately reflects the actual quality status of the cables and meets real-time control requirements. Authenticity verification (ensuring the data reflects the true quality of the cables) first establishes a true feature template, i.e., collecting data from nearly 100 batches of cables that have passed comprehensive offline quality inspection, and extracting the core features of this batch of data based on the final feature parameter combination. The template content includes: first, the numerical range of each quality parameter, such as resistance from 0.01 to 0.1 Ω / m, insulation layer thickness from 0.8 to 1.2 mm, and outer sheath hardness from 65 to 75 HB; second, the range of the slope of time-series data changes, such as the slope of resistance change over time ≤ 0.005 Ω / (m・min) and the slope of hardness change ≤ 0.2 HB / min; and third, the correlation between parameters, such as the difference between insulation layer thickness and the maximum value of radial position from 0.1 to 0.3 mm, and the positive correlation coefficient between shielding layer coverage and the proportion of effective coverage data points ≥ 0.85. The template needs to be updated regularly to adapt to fine-tuning of the production process.

[0091] Next, features of the data to be verified are extracted. Based on the final feature parameter combination, comprehensive feature extraction is performed on the signal-processed data. First, numerical features are extracted, including the actual measured value and range of each quality parameter, ensuring coverage of core parameters such as resistance, insulation thickness, and outer sheath hardness, and fully restoring the basic numerical attributes of the data. Second, trend features are extracted, calculating the slope of data change window by window according to the time-series feature extraction window size in the final feature parameter combination. This is achieved by subtracting the starting data value from the last data value within the window and then dividing by the window duration, capturing the dynamic change patterns of the time-series data. Third, correlation features are calculated, determining the correlation coefficients between different quality parameters. The calculation logic is based on the physical and technological correlations between parameters, selecting parameter groups that belong to the same cable structure or performance related. For example, the hardness of the outer sheath and the thickness of the insulation layer are both affected by the material formulation and extrusion process; the shielding layer coverage and radial position data both reflect the uniformity of the shielding layer wrapping, ensuring that the correlation coefficient calculation has practical significance. The specific calculation adopts the Pearson correlation coefficient method, and the steps are as follows: First, calculate the mean of the two sets of parameters respectively, then calculate the deviation of each data in each set of parameters from the corresponding mean, then divide the sum of the products of the deviations of the two sets of parameters by the number of data samples to obtain the covariance, and at the same time calculate the standard deviation of the two sets of parameters respectively (the square root of the sum of the squares of the deviations of each data from the mean divided by the number of samples), and finally divide the covariance by the product of the standard deviations of the two sets of parameters to obtain the correlation coefficient (the value range is -1 to 1, and a positive value indicates a positive correlation), ensuring that the extracted features correspond one-to-one with the real feature template.

[0092] Then, feature similarity is calculated, comparing the extracted features with the real feature template one by one. The criterion for each feature item is that the extracted feature falls within ±5% of the template range. The number of feature items that meet the template requirements is counted. Feature similarity is calculated by dividing the number of feature items that meet the requirements by the total number of feature items. If the feature similarity is ≥ the preset similarity threshold (set according to the strictness of quality control, ≥0.9 in normal production scenarios and ≥0.95 in high-precision scenarios), the data is considered to be authentic; otherwise, it is considered false data. If the value exceeds the template range, the trend of change is abnormal, or the parameter correlation is disordered, it will be marked and removed.

[0093] Timeliness verification (ensuring data meets real-time control requirements) first involves obtaining key time nodes. This includes retrieving the original sensor data acquisition time corresponding to the processed signal data from the metadata stored in the sensor device; and recording the system time of the current verification operation to ensure the accuracy of the time data. Then, the time difference is calculated and determined. The time difference is calculated by subtracting the original sensor data acquisition time from the current verification time. If the time difference is less than the preset timeliness threshold (set according to production rhythm: ≤5 minutes for online real-time detection scenarios, ≤1 hour for offline sampling inspection scenarios), it indicates that the data accurately reflects the current quality status of the cable and is deemed timely. If the time difference is greater than or equal to the timeliness threshold, it indicates that the data is outdated, possibly due to changes in production processes or cable storage environment leading to altered quality status; such data is marked and discarded.

[0094] After screening the processed signal data that simultaneously passes both authenticity and timeliness verification, all marked false and outdated data are thoroughly removed to ensure the basic reliability of the remaining data. For the screened valid data, core features are extracted again according to the final feature parameter combination, including resistance values ​​within the qualified range, stable hardness change trends, parameter groups that conform to correlation relationships, etc., to finally form verified feature data.

[0095] This embodiment effectively eliminates deviations caused by abnormal data point density and uneven distribution through a dual calibration mechanism of density calibration and distribution uniformity calibration, thereby improving the accuracy of key parameter data. The use of moving average filtering for signal processing successfully removes interference components from the data, making the data trend more stable and accurate, reducing the impact of interference signals on subsequent feature parameter optimization and data verification, and ensuring the effectiveness of data processing. The use of the Grey Wolf optimization algorithm to optimize feature parameter combinations, through iterative optimization, finds the feature parameter combinations with the strongest data discrimination ability, avoiding the subjectivity and limitations of manual parameter selection, and improving the scientific and rational nature of data classification and discrimination. The establishment of a dual verification mechanism for authenticity and timeliness ensures that the final output verified feature data is authentic, reliable, and meets real-time requirements, contributing to improved overall production quality control.

[0096] In a preferred embodiment of the present invention, the verified feature data is bound to an identifier to generate a digital fingerprint, and the digital fingerprint and key parameter data are uploaded to a blockchain network to complete the traceability record of the cable material throughout its entire lifecycle, including:

[0097] In this embodiment of the invention, a unique identifier for the cable is first determined. This identifier is generated by combining the production batch number, production date, production equipment number, and cable serial number, ensuring that each cable product corresponds to a unique identifier, with no possibility of duplication or confusion. Next, a data binding operation is performed. This unique identifier is used as the core index and integrated with verified feature data (including core features such as qualified resistance values, insulation layer thickness, and stable hardness change trends), raw data of key parameters for cable quality inspection (including raw resistance data, raw insulation layer thickness data, raw outer sheath hardness data, raw radial position data, and raw shielding layer coverage data directly collected by sensors), and calibration records (including data before and after density calibration, data before and after distribution uniformity calibration, and calibration coefficients) to form a structured data set, ensuring a one-to-one correspondence between the identifier and various types of data. Subsequently, a digital fingerprint is generated, and the SHA-256 hash algorithm is used to encrypt and calculate the integrated structured data set. First, the data set is converted into a binary data stream in the order of identifier, verified feature data, key parameter data, and calibration record. Then, the data stream is hashed using the SHA-256 algorithm to output a fixed-length hash value as a digital fingerprint. This fingerprint has uniqueness and immutability, and can characterize the data features of the corresponding cable. Finally, the blockchain upload process is executed. A secure connection is first established with a preset blockchain network (a consortium blockchain adapted to industrial traceability scenarios). The digital fingerprint, unique identifier, key parameter data, and verification record are packaged into a standard transaction data block through an encrypted transmission protocol. After the data block is submitted to the blockchain node, it waits for the node to complete the data verification and accounting through a consensus mechanism (such as the PBFT algorithm). After the transaction is confirmed, the data will be permanently stored in the blockchain distributed ledger. At this point, the key data of the cable material from production testing to subsequent circulation has been recorded on the chain. Relying on the immutability and traceability of the blockchain, traceability management of the entire life cycle of the cable material is realized.

[0098] like Figure 2 As shown, embodiments of the present invention also provide a blockchain-based cable material lifecycle traceability management system, including:

[0099] The data acquisition module is used to assign an identifier to each cable material and register the identifier to the blockchain network; during the processing of the cable material, it collects key parameter data in real time.

[0100] The segmentation module is used to convert key parameter data into a set of time-series data points to establish parameter distribution characteristics and to divide the cable material into conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment.

[0101] A module is established to set three mass points in each segment to construct a spatial topology network. A multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units.

[0102] The calculation module is used to map time series data points to corresponding data units according to the spatial constraints of the data units, calculate the spatial distribution characteristic value of the data points in each data unit, and generate parameter calibration coefficients based on the spatial distribution characteristic value.

[0103] The calibration module is used to perform multi-level calibration on key parameter data using parameter calibration coefficients to obtain calibrated key parameter data. Then, it performs signal processing on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf optimization algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, thus obtaining verified feature data.

[0104] The binding module is used to bind verified feature data with identifiers, generate digital fingerprints, and upload digital fingerprints and key parameter data to the blockchain network to complete the traceability record of the entire life cycle of cable materials.

[0105] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0106] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0107] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0108] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A blockchain-based method for full lifecycle traceability management of cable materials, characterized in that: The method includes: Assign an identifier to each cable material and register the identifier to the blockchain network; collect key parameter data in real time during the cable material processing. Key parameter data are converted into a set of time-series data points to establish parameter distribution characteristics, and the cable material is divided into conductor core section, insulation layer section, shielding layer section and outer sheath section; Three mass points are set in each segment to construct a spatial topology network, and a multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units. Based on the spatial constraints of the data units, time series data points are mapped to the corresponding data units, the spatial distribution characteristic values ​​of the data points in each data unit are calculated, and parameter calibration coefficients are generated based on the spatial distribution characteristic values. The key parameter data is calibrated at multiple levels using parameter calibration coefficients to obtain calibrated key parameter data. Signal processing is then performed on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf Optimization Algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, thus obtaining verified feature data. The verified feature data is bound to the identifier to generate a digital fingerprint, and the digital fingerprint and key parameter data are uploaded to the blockchain network to complete the traceability record of the entire life cycle of the cable material.

2. The blockchain-based cable material lifecycle traceability management method according to claim 1, characterized in that, Assign an identifier to each cable material and register the identifier to the blockchain network; During the cable material processing, key parameter data are collected in real time, including: Conductor core characteristic parameters, including conductor core resistance, conductor core diameter, and conductor core surface roughness, are collected by a first sensor array deployed on the conductor core processing device. The insulation layer characteristic parameters, including insulation layer thickness, insulation layer resistance value and insulation layer dielectric constant, are collected by a second sensor array deployed on the insulation layer processing equipment. The shielding layer characteristic parameters, including shielding layer coverage, shielding layer resistance value and shielding layer continuity, are collected by a third sensor array deployed on the shielding layer processing equipment. The outer sheath characteristic parameters, including outer sheath thickness, outer sheath hardness, and outer sheath wear resistance, are collected by a fourth sensor array deployed on the outer sheath processing equipment. Environmental parameters, including ambient temperature, ambient humidity, and ambient cleanliness, are collected through environmental monitoring equipment.

3. The blockchain-based cable material lifecycle traceability management method according to claim 2, characterized in that, Key parameter data are converted into a time-series data point set to establish parameter distribution characteristics, and the cable material is divided into conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment, including: The key parameter data is converted into a time series data point set in chronological order, with each data point containing a timestamp and corresponding spatial location information; Statistical characteristics of the distribution of each key parameter are calculated based on the time series data point set, including parameter mean, parameter variance, and parameter distribution density; Based on the statistical characteristics of the distribution of each key parameter, the cable material is divided into conductor core section, insulation layer section, shielding layer section and outer sheath section. The conductor core section corresponds to the conductor core resistance value distribution characteristics, the insulation layer section corresponds to the insulation layer resistance value distribution characteristics, the shielding layer section corresponds to the shielding layer resistance value distribution characteristics, and the outer sheath section corresponds to the outer sheath hardness distribution characteristics.

4. The blockchain-based cable material lifecycle traceability management method according to claim 3, characterized in that, Three mass points are set in each segment to construct a spatial topology network. A multi-level mesh structure is then established based on this network. The multi-level mesh structure is dynamically optimized using parameter distribution characteristics to form data units, including: Based on the statistical distribution characteristics of the conductor core segment, three mass points are set at the starting, middle, and ending positions of the conductor core segment; based on the statistical distribution characteristics of the insulation layer segment, three mass points are set at the starting, middle, and ending positions of the insulation layer segment; based on the statistical distribution characteristics of the shielding layer segment, three mass points are set at the starting, middle, and ending positions of the shielding layer segment; based on the statistical distribution characteristics of the outer sheath segment, three mass points are set at the starting, middle, and ending positions of the outer sheath segment. A spatial topology network is constructed based on the spatial location of mass points, and the spatial distance between mass points and the correlation of the distribution of key parameters are calculated. In the spatial topology network, for each segment, the mass point at the middle position is selected as the core point. A first spatial vector is defined by extending from the core point along the segment axis, and a second spatial vector is defined by extending from the core point along the segment radially. The spatial orientation angle between the first spatial vector and the second spatial vector is calculated. The section is divided into multiple quality control sub-regions based on the size of the spatial orientation angle, wherein the spatial orientation angle is used to determine the distribution boundary of the quality control sub-regions. A multi-level grid structure is established based on the spatial topology network and quality control sub-regions. The multi-level grid structure is dynamically optimized by utilizing the statistical characteristics of the distribution of each key parameter and the spatial orientation angle. This includes adjusting the grid density of the multi-level grid structure based on the correlation of the distribution of each key parameter, and optimizing the grid connection relationship of the multi-level grid structure based on the spatial distance and spatial orientation angle, thus forming data units with spatial constraints.

5. The blockchain-based cable material lifecycle traceability management method according to claim 4, characterized in that, Based on the spatial constraints of the data units, time-series data points are mapped to corresponding data units. The spatial distribution characteristic values ​​of the data points within each data unit are calculated, and parameter calibration coefficients are generated based on these spatial distribution characteristic values, including: Based on the spatial constraints of the data units, time-series data points are mapped to corresponding data units, and spatial distribution characteristic values ​​of data points within each data unit are calculated, including data point density and data point distribution uniformity. Data point density is calculated as the ratio of the number of data points to the volume of the data unit. Data point distribution uniformity is calculated by constructing a convex boundary polygon based on the data point set and calculating the minimum circumcircle of the data point set. The inclusion relationship between the minimum circumcircle and the convex boundary polygon is detected by radiating rays from the center of the minimum circumcircle to the boundary of the convex boundary polygon, counting the number of intersections between the rays and the polygon boundary, and determining whether the center of the ray is inside the convex boundary polygon based on the parity of the number of intersections. The area of ​​the overlapping region between the minimum circumcircle and the convex boundary polygon is also calculated. The data point distribution uniformity is determined based on the ratio of the overlapping region area to the area of ​​the minimum circumcircle. The parameter calibration coefficients are generated based on the spatial distribution characteristic values, including the density calibration coefficient and the distribution uniformity calibration coefficient.

6. The blockchain-based cable material lifecycle traceability management method according to claim 5, characterized in that, The key parameter data is calibrated at multiple levels using parameter calibration coefficients to obtain calibrated key parameter data. Then, the calibrated key parameter data is processed to obtain signal-processed data. The Grey Wolf optimization algorithm is used to extract features from the processed signal data to verify the authenticity and timeliness of the processed data, resulting in verified feature data, including: Density calibration is performed using density calibration coefficients, and distribution uniformity calibration is performed using distribution uniformity calibration coefficients to obtain calibrated key parameter data. Signal processing is performed on the calibrated key parameter data to eliminate signal interference components, resulting in signal-processed data. The gray wolf optimization algorithm is used to extract features from the signal-processed data. This includes optimizing the selection process of feature parameters by simulating the hunting behavior of gray wolves, including initializing the gray wolf population, calculating fitness values, updating gray wolf positions, and determining the final combination of feature parameters. The authenticity and timeliness of the processed data are verified based on the final combination of feature parameters, and the verified feature data is obtained.

7. The blockchain-based cable material lifecycle traceability management method according to claim 6, characterized in that, The gray wolf optimization algorithm is used to extract features from the signal-processed data. This includes optimizing the selection of feature parameters by simulating gray wolf hunting behavior, including initializing the gray wolf population, calculating fitness values, updating gray wolf positions, and determining the final combination of feature parameters. Initialize the gray wolf population, set the population size, and randomly generate initial feature parameter combinations; calculate fitness values ​​based on the initial feature parameter combinations, and evaluate the ability of each feature parameter combination to distinguish the data after signal processing. Based on the fitness value sorting results, determine the feature parameter combination corresponding to the first fitness value, the feature parameter combination corresponding to the second fitness value, and the feature parameter combination corresponding to the third fitness value; The position of the gray wolf population is updated based on the feature parameter combination corresponding to the first fitness value, the feature parameter combination corresponding to the second fitness value, and the feature parameter combination corresponding to the third fitness value, and a new feature parameter combination is generated. Repeat the fitness value calculation and position update process until the preset number of iterations is reached to determine the final combination of feature parameters.

8. A blockchain-based cable material lifecycle traceability management system, wherein the system implements the method as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to assign an identifier to each cable material and register the identifier to the blockchain network; During the processing of cable materials, key parameter data are collected in real time; The segmentation module is used to convert key parameter data into a set of time-series data points to establish parameter distribution characteristics and to divide the cable material into conductor core segment, insulation layer segment, shielding layer segment, and outer sheath segment. A module is established to set three mass points in each segment to construct a spatial topology network. A multi-level grid structure is established based on the spatial topology network. The multi-level grid structure is dynamically optimized using parameter distribution characteristics to form data units. The calculation module is used to map time series data points to corresponding data units according to the spatial constraints of the data units, calculate the spatial distribution characteristic value of the data points in each data unit, and generate parameter calibration coefficients based on the spatial distribution characteristic value. The calibration module is used to perform multi-level calibration on key parameter data using parameter calibration coefficients to obtain calibrated key parameter data. Then, it performs signal processing on the calibrated key parameter data to obtain signal-processed data. The Grey Wolf optimization algorithm is used to extract features from the signal-processed data to verify the authenticity and timeliness of the processed data, thus obtaining verified feature data. The binding module is used to bind verified feature data with identifiers, generate digital fingerprints, and upload digital fingerprints and key parameter data to the blockchain network to complete the traceability record of the entire life cycle of cable materials.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images