Intelligent identification and state monitoring method and system for central axis of complex wire harness structure

Abstract

According to the intelligent recognition and state monitoring method and system for the central axis of the complex wire harness structure, the axis feature database is established in advance, the axis in the complex wire harness is scanned and compared through the machine vision technology, identity recognition is achieved, working data collected in real time are combined, the state change trend of the axis is judged through the time sequence analysis algorithm, and the state of the complex wire harness structure is monitored. When the trend is abnormal, an embedded sensor is used for obtaining temperature vibration data, the position and the type of an abnormal axis are determined through a data fusion technology, historical production data are further combined, clustering analysis is adopted for judging the relevance between abnormity and a production link, and the signal quality is optimized based on real-time monitoring data. According to the invention, the block chain technology is fused to carry out anti-counterfeiting verification, and the long-term operation stability is analyzed through the prediction model. Finally, the deep learning technology is utilized to carry out iterative training on identification and monitoring precision, intelligent management of the axis in a high-density connection scene is realized, and the reliability and production efficiency of a complex wire harness structure are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to a method and system for intelligently identifying and monitoring the state of an axis in a complex wiring harness structure. Background Art

[0002] As a key component in the interconnection of modern electronic devices, the production process of USB wiring harnesses directly determines data transmission efficiency, device compatibility, and operational safety. With the popularization of smart devices and high-speed communication technologies, demand for USB wiring harnesses is growing in consumer electronics, industrial control, automotive electronics, and other fields. The advancement of USB wiring harness technology has become a focus of industry competition.

[0003] However, traditional production processes have gradually revealed their shortcomings in meeting diverse needs and improving reliability, which has promoted the urgency of research on intelligent and precise processes. In the current production of USB harnesses, axis identification mainly relies on manual inspection or simple mechanical marking. This method is inefficient and prone to errors, especially in complex harness structures and high-density connection scenarios. It is difficult to accurately distinguish the type and status of the axis. In addition, the existing technology lacks the ability to monitor the dynamic performance of the harness in real time, making it difficult to detect connection quality problems in a timely manner. These limitations not only increase production costs, but also limit the reliability of harnesses in high-end applications. Against this background, the core challenges of axis identification technology have gradually become prominent.

[0004] First, accurate identification of the axis identity poses a challenge, as traditional physical markings struggle to adapt to changing production requirements and equipment types. Second, real-time monitoring of axis status presents significant technical bottlenecks, particularly when data transmission rates and power requirements change dynamically, due to the lack of an effective feedback mechanism. Finally, insufficient integration of anti-counterfeiting verification and quality control results in substandard wiring harnesses entering the market, increasing the risk of equipment damage. Unresolved technical issues such as these directly lead to unique challenges, including a low level of intelligent production processes and a poor user experience.

[0005] Therefore, how to achieve effective integration of intelligent identification of axis identity, real-time monitoring of status, and anti-counterfeiting verification in the USB harness production process has become a key issue in improving the quality and reliability of the harness.

[0006] Solving this problem requires not only breaking through the limitations of existing identification technology, but also incorporating innovative monitoring and verification methods into process design to meet the needs of increasingly complex usage scenarios. Summary of the Invention

[0007] In order to solve the technical problems raised in the above background technology, the first aspect of the present invention provides a method for intelligently identifying and monitoring the state of the central axis of a complex wiring harness structure, the method comprising: S1, obtains the physical parameters and signal characteristics of each axis through the pre-established axis feature database, uses machine vision technology to scan and compare the axes in the complex wiring harness structure, and obtains the axis identity recognition result; S2, based on the axis identification results, extracts the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis status change trend. S3: If the axis state change trend exceeds the preset change threshold, the temperature and vibration data of the wiring harness are acquired through the embedded sensor. Combined with the state change trend, data fusion technology is used to determine the specific location and type of the abnormal axis. S4, based on the position and type of the abnormal axis, extract the corresponding process parameters and batch information from the historical production data, use the cluster analysis algorithm to determine whether the abnormality is related to the production link, and obtain the distribution characteristics of the process defects; S5, based on the distribution characteristics of process defects, obtains the signal attenuation and delay indicators of the abnormal axis from the real-time status monitoring data. In response to the fluctuation of data transmission efficiency, it uses adaptive filtering technology to optimize the signal quality and output the adjusted transmission parameters; S6, through the adjusted transmission parameters, extracts the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module, uses blockchain technology to compare the digital signature, determines the authenticity of the wiring harness and generates a verification result; S7, based on the verification results, obtains the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data. To meet the needs of enhanced reliability performance, a predictive model is used to analyze the stability trend of long-term operation; S8, if the stability trend shows an abnormality, the production process intelligent system will adjust the process parameters and equipment configuration, combine the real-time status monitoring data, use the feedback control algorithm to optimize the production process and output the improved process plan; S9, through the improved process plan, extracts the axis distribution pattern in high-density connection scenarios from complex wiring harness structures, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

[0008] Optionally, step S1, obtaining the physical parameters and signal characteristics of each axis through a pre-established axis feature database, and using machine vision technology to scan and compare the axes in the complex wiring harness structure to obtain axis identity recognition results, includes: Step S11, obtaining the physical parameters and signal characteristics of each axis through a preset axis characteristic database; Step S12, using the image processing function in the OpenCV library to perform a preliminary scan of the axis in the complex wiring harness structure to obtain initial data of the axis; Step S13: Based on the initial data, the SIFT algorithm is used to compare the axis features with the signal features in the database to determine the preliminary classification results of the axis; Step S14, using a high-precision laser scanner to perform in-depth analysis of the wiring harness structure in the complex structure to obtain supplementary parameters of the unidentified axis; Step S15: if the supplementary parameters match the physical parameters in the database, the identity recognition result of the axis is updated by using the cosine similarity calculation method; Step S16, verifying the updated identity recognition result through the image recognition function in the OpenCV library to determine whether there is any abnormal classification; Step S17: Based on the verification results, the SIFT algorithm is used to re-compare the signal features of the abnormally classified axes to obtain the final identity recognition result; In step S18, for the final result, the physical parameters and identification data of all axes are stored in a MySQL database to complete the complete analysis of the wiring harness structure.

[0009] Optionally, step S2 extracts the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data based on the axis identification result, and uses a timing analysis algorithm to determine the axis state change trend based on dynamic changes in high-density connection scenarios, including: Step S21 , obtaining real-time acquisition results from the harness data through axis identity recognition, extracting corresponding current data, voltage data, and signal waveform data to obtain a preliminary acquisition data set; Step S22, using a pre-established threshold segmentation method, separating the dynamically changing portion in the high-density connection scenario from the preliminary collected data set to determine a dynamically changing data set; Step S23: For the dynamically changing data set, a time series analysis algorithm is used to analyze the time series characteristics of the signal waveform and determine the time distribution characteristics of the dynamic changes; Step S24: if the time distribution characteristics of the dynamic change exceed the pre-established fluctuation range, then extract the local change segments from the dynamic change data set by using a sliding window method to obtain a local change feature set; Step S25, based on the local change feature set, using a support vector machine algorithm to classify abnormal states in the local change feature set and determine an abnormal state set; Step S26, using the mapping relationship between the abnormal state set and the change trend, a linear regression algorithm is used to fit the change trend of the axis state to obtain a state change trend result; Step S27, obtaining the slope change of the trend curve from the state change trend result, and determining the future evolution direction of the axis state.

[0010] Optionally, in step S3, if the axis state change trend exceeds a preset change threshold, the temperature and vibration data of the wiring harness are acquired through an embedded sensor, and the specific location and type of the abnormal axis are determined using data fusion technology in combination with the state change trend, including: Step S31 , if the change trend of the axis state exceeds the change threshold, collect harness temperature and vibration data through the embedded sensor to obtain the original collected data set; Step S32: using a low-pass filter to remove high-frequency noise from the original collected data set to obtain a clean data set; Step S33, extracting the mean and variance of the harness temperature and vibration data from the clean data set as feature parameters to form a feature data set; Step S34: Process the feature data set using Kalman filtering, and determine the preliminary judgment result of the abnormal axis based on the change trend; Step S35: Based on the preliminary judgment result, the deviation value of the specific position is calculated by interpolation to obtain the positioning data of the abnormal axis; Step S36: If the positioning data matches the pre-established type feature library, the support vector machine algorithm is used to classify the abnormal axis and determine the abnormal type; Step S37: Generate complete description data of the abnormal axis according to the abnormality type and positioning data.

[0011] Optionally, step S4 extracts corresponding process parameters and batch information from historical production data based on the position and type of the abnormal axis, uses a cluster analysis algorithm to determine whether the abnormality is related to the production process, and obtains process defect distribution characteristics, including: Step S41, obtaining the location and type of the abnormal axis from historical production data, and extracting the corresponding process parameters and batch information; Step S42, using database query technology to obtain parameter data and obtain preliminary extraction results; Step S43: Based on the preliminary extraction results, the K-means clustering algorithm is used to group the process parameters and abnormal axes; Step S44: Based on the grouping results, determine the correlation between the anomalies in each group and the production process, and obtain the anomaly grouping characteristics; Step S45: According to the abnormal grouping characteristics, the matching relationship between the batch information and the process parameters is obtained, the distribution of the abnormal axis in the production link is determined, and the batch association data is obtained; Step S46: Using the batch correlation data, a regression analysis method is used to analyze the corresponding pattern between the location type and the process parameters to determine whether the anomaly is caused by a specific production link and obtain the link impact result; Step S47: extract the correlation between defect distribution and process parameters based on the link impact results, determine the influence of the abnormal axis on the distribution characteristics, and obtain the defect distribution law; Step S48: According to the defect distribution law, the adjustment direction of the process parameters in the production process is obtained, the control point of the abnormal axis is determined, and parameter optimization suggestions are obtained; In step S49, the dynamic changes of location type and defect distribution are analyzed through parameter optimization suggestions to determine the distribution characteristics of process defects and obtain the final analysis conclusion.

[0012] Optionally, step S5, based on the distribution characteristics of process defects, obtains the signal attenuation and delay indicators of the abnormal axis from the real-time status monitoring data, optimizes the signal quality using adaptive filtering technology to address fluctuations in data transmission efficiency, and outputs adjusted transmission parameters, including: Step S51, collecting status monitoring information through real-time data collection, extracting the signal attenuation and delay indicators of the abnormal axis, and extracting the spectrum characteristics of the signal using fast Fourier transform; Step S52, separating abnormal frequency components from the spectrum characteristics and establishing a mapping relationship between abnormal axis and signal attenuation; Step S53, fitting a mathematical model of signal attenuation and delay indicators using the least squares method according to the mapping relationship; Step S54, using numerical differentiation to calculate the rate of change of the delay indicator based on the mathematical model; Step S55, using Kalman filtering to calculate the fluctuation range of the transmission efficiency according to the change rate; Step S56: using spectrum analysis to determine the source of data fluctuations based on the fluctuation amplitude; Step S57, using an adaptive filter to process data fluctuations, setting the filter cutoff frequency to the center frequency of the abnormal frequency component; Step S58, outputting the optimized signal characteristics through the filter; Step S59, updating the transmission parameters using the gradient descent method according to the optimized signal characteristics; Step S510, recalculating the transmission efficiency based on the updated parameters; Step S511: if the transmission efficiency is lower than the efficiency threshold, re-extract the delay index and adjust the filter parameters; Step S512: Acquire final transmission parameters and output a combination of signal attenuation and delay indicators in a stable state.

[0013] Optionally, step S6, extracting the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module using the adjusted transmission parameters, comparing the digital signature using blockchain technology, determining the authenticity of the wiring harness and generating a verification result, includes: Step S61: Adjust the parameters of the transmission protocol, obtain the axis signature and production identification data from the anti-counterfeiting verification module, and generate an initial data set; Step S62: encrypt the axis signature using the SHA-256 hash algorithm, extract the encrypted digital signature from the initial data set, and calculate the signature feature value; Step S63: Accessing the distributed ledger through the Ethereum blockchain, obtaining a reference signature corresponding to the production identifier from the ledger, and generating a comparison benchmark value; Step S64: If the signature feature value is consistent with the comparison reference value, the authenticity of the wiring harness is determined by digital comparison and a preliminary verification status is generated; Step S65: Perform secondary confirmation on the initial verification status using preset verification rules to determine whether the status is stable and generate a confirmation result; Step S66: Based on the confirmation result, use SQL query to filter abnormal data and generate the final verification result; Step S67: Record the final verification result to the Ethereum blockchain account book through the RESTful API to complete the verification process.

[0014] Optionally, in step S7, based on the verification results, the response time and power fluctuation of the wiring harness under different loads are obtained from the dynamic performance monitoring data, and a prediction model is used to analyze the stability trend of long-term operation to meet the demand for enhanced reliability performance, including: Step S71, extracting response time and power fluctuation data of the wiring harness under different loads through dynamic performance monitoring to obtain an initial data set; Step S72, using Python's Pandas library to process noise and outliers in the initial data set to obtain an optimized data set; Step S73, using the Scikit-learn library to calculate the characteristic values ​​of response time and power fluctuation under load changes from the optimized data set to determine the characteristic distribution; Step S74: If the feature distribution exceeds the distribution threshold, the reliability prediction model is trained using the support vector machine algorithm in the Scikit-learn library to determine the short-term stability; Step S75, based on the short-term stability results, use the ARIMA model in the Statsmodels library to predict the trend changes in the long-term operation and obtain a trend curve; Step S76, obtaining the reliability performance of the wiring harness under different loads by fitting the stability characteristics through trend curve; Step S77: extract key indicators from the reliability performance to determine the degree of realization of the enhancement requirements.

[0015] Optionally, in step S8, if the stability trend shows an abnormality, the production process intelligent system adjusts the process parameters and equipment configuration, combines the real-time status monitoring data, uses a feedback control algorithm to optimize the production process, and outputs an improved process plan, including: Step S81: if the monitoring data exceeds the monitoring threshold, extract key variables from the monitoring data as real-time status data; Step S82, using statistical analysis methods to determine abnormal indications based on real-time status data; Step S83, obtaining the corresponding process parameter adjustment range according to the abnormal indication from the pre-established process parameter mapping table; Step S84, calculating the adjusted parameter value within the adjustment range using interpolation method; Step S85, inputting the adjusted parameter values ​​into the device configuration management system to update the device configuration parameters; Step S86, extracting key indicators of the production process from the equipment configuration parameters, and using the proportional-integral-differential control algorithm to calculate the optimization direction of the production process; Step S87, adjusting key parameters in the production process according to the optimization direction to generate an improved process plan; Step S88: Input the improved process plan into the real-time monitoring system to update the real-time status data; Step S89, judging whether the production process has returned to normal based on the changing trend of the real-time status data; Step S810: If the production process returns to normal, the adjusted parameter values ​​and process plan are recorded as the final improvement plan.

[0016] A second aspect of the present invention provides an intelligent identification and status monitoring system for the central axis of a complex wiring harness structure, which uses the above-mentioned method to intelligently identify and monitor the central axis of the complex wiring harness structure. The system includes: The axis feature recognition module is used to obtain the physical parameters and signal characteristics of each axis through a pre-established axis feature database. It uses machine vision technology to scan and compare the axes in the complex wiring harness structure to obtain the axis identity recognition results. The abnormal state detection module is used to extract the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data based on the axis identity recognition results. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis state change trend; The data fusion analysis module is used to obtain the temperature and vibration data of the wiring harness through embedded sensors if the axis state change trend exceeds the preset change threshold. Combined with the state change trend, the module uses data fusion technology to determine the specific location and type of the abnormal axis. The process defect analysis module is used to extract the corresponding process parameters and batch information from historical production data based on the location and type of the abnormal axis, and use the cluster analysis algorithm to determine whether the abnormality is related to the production link to obtain the distribution characteristics of the process defects; The signal optimization module is used to obtain the signal attenuation and delay indicators of abnormal axes from real-time status monitoring data based on the distribution characteristics of process defects. In response to fluctuations in data transmission efficiency, it uses adaptive filtering technology to optimize signal quality and output adjusted transmission parameters; The anti-counterfeiting verification module is used to extract the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module through the adjusted transmission parameters, compare the digital signature using blockchain technology, determine the authenticity of the wiring harness, and generate a verification result; The performance prediction module is used to obtain the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data based on the verification results. To meet the needs of enhancing reliability performance, a prediction model is used to analyze the stability trend of long-term operation; The process optimization module is used to adjust process parameters and equipment configurations through the intelligent production process system if the stability trend shows abnormalities. In combination with real-time status monitoring data, it uses feedback control algorithms to optimize the production process and output an improved process plan; The model enhancement module is used to extract the axis distribution pattern in high-density connection scenarios from complex wiring harness structures through improved process solutions, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

[0017] The present invention discloses a method and system for intelligent identification and status monitoring of axes in complex wiring harness structures. By pre-establishing an axis feature database, machine vision technology is used to scan and compare the axes in the complex wiring harness to achieve identity recognition. In combination with real-time collected working data, a time series analysis algorithm is used to determine the trend of axis status changes. When the trend is abnormal, embedded sensors are used to obtain temperature and vibration data, and the position and type of the abnormal axis are determined through data fusion technology. Further combined with historical production data, cluster analysis is used to determine the correlation between the abnormality and the production link, and the signal quality is optimized based on real-time monitoring data. The present invention also integrates blockchain technology for anti-counterfeiting verification, and analyzes long-term operational stability through predictive models. Finally, deep learning technology is used to iteratively train the identification and monitoring accuracy, realize intelligent management of axes in high-density connection scenarios, and improve the reliability and production efficiency of complex wiring harness structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 The figure is a flow chart of the method for intelligently identifying and monitoring the status of the central axis of a complex wiring harness structure according to the present invention.

[0019] Figure 2It is a structural schematic diagram of the intelligent identification and status monitoring system of the central axis of the complex wiring harness structure of the present invention. DETAILED DESCRIPTION

[0020] To help those skilled in the art better understand the technical solutions in this specification, the following will provide a clear and complete description of the technical solutions in the embodiments of this specification, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this specification, not all of them. All other embodiments derived by those skilled in the art based on the embodiments in this specification without creative effort shall fall within the scope of protection of this specification.

[0021] like Figure 1 As shown, the first aspect of the present invention provides a method for intelligent identification and status monitoring of the central axis of a complex wiring harness structure, which may specifically include: S1, obtains the physical parameters and signal characteristics of each axis through the pre-established axis feature database, uses machine vision technology to scan and compare the axes in the complex wiring harness structure, and obtains the axis identity recognition result.

[0022] Optionally, this step also includes: Step S11: obtaining the physical parameters and signal characteristics of each axis through a preset axis characteristic database.

[0023] In step S12, the image processing function in the OpenCV library is used to perform a preliminary scan on the axis in the complex wiring harness structure to obtain initial data of the axis.

[0024] Step S13: Based on the initial data, the SIFT algorithm is used to compare the axis features with the signal features in the database to determine the preliminary classification results of the axis.

[0025] In step S14, a high-precision laser scanner is used to perform an in-depth analysis of the wiring harness structure in the complex structure to obtain supplementary parameters of the unidentified axis.

[0026] Step S15: If the supplementary parameters match the physical parameters in the database, the identity recognition result of the axis is updated by using the cosine similarity calculation method.

[0027] Optionally, use the following formula to calculate cosine similarity:

[0028] in, is the similarity, n represents the total number of feature dimensions, represents the i-th component of vector A, Represents the i-th component of vector B, cosine similarity The closer the result of is to 1, the more similar the two vectors are.

[0029] Step S16: Verify the updated identity recognition result through the image recognition function in the OpenCV library to determine whether there is any abnormal classification.

[0030] Step S17: Based on the verification result, the SIFT algorithm is used to re-compare the signal features of the axis of the abnormal classification to obtain the final identity recognition result.

[0031] In step S18, for the final result, the physical parameters and identification data of all axes are stored in a MySQL database to complete the complete analysis of the wiring harness structure.

[0032] Specifically, the physical parameters and signal characteristics of each axis are obtained through a preset axis feature database. The core of this process is to establish a reliable reference benchmark.

[0033] For example, when analyzing complex wire harness structures, the database may store information such as the length, diameter, color coding, and electrical signal frequency of the axis.

[0034] For example, the physical parameters of an axis may be a length of 50 cm, a diameter of 2 mm, and a signal characteristic of a frequency of 60 Hz. These data provide a basis for subsequent comparison.

[0035] In one possible implementation, the image processing function in the OpenCV library is used to perform a preliminary scan of the harness structure, and the initial contour of the axis can be identified through edge detection and contour extraction technology.

[0036] Specifically, after the camera captures the wiring harness image, OpenCV's grayscale processing and binarization functions separate the axis from the background, obtaining initial data such as the approximate position and shape of the axis. This method offers the advantage of quickly identifying key areas, easing the burden of subsequent precise analysis. Using this initial data, feature matching using the SIFT algorithm is a key step. SIFT excels at extracting key points and their descriptors from an image.

[0037] For example, the texture or end shape of an axis can generate a set of feature vectors that can be compared one by one with the signal features in the database.

[0038] In one embodiment, if the feature vector of an axis in the database contains 128 components and the scanned axis features match it at least 85%, it can be initially classified. This improves the robustness of classification, especially when the axes in complex harnesses are densely packed.

[0039] Preferably, a high-precision laser scanner is used for in-depth analysis, which can supplement the detailed parameters of the unidentified axis.

[0040] For example, laser scanning can accurately measure the surface roughness or slight curvature of an axis, obtaining supplementary parameters such as a curvature radius of 0.5 mm. These parameters are then compared with a database to further confirm the axis identity, thus compensating for the shortcomings of image scanning.

[0041] It should be noted that updating the identity recognition result through cosine similarity calculation depends on vectorized feature comparison.

[0042] For example, the supplementary parameter vector A and the database vector B each have 10 components, and the calculated cosine value is 0.95, which is close to 1, indicating that the two are highly similar. The advantage of this method is that it quantifies the degree of similarity and improves the scientific nature of the recognition.

[0043] In one embodiment, the image recognition function of OpenCV is used to verify the update results, and anomalies can be detected through template matching.

[0044] For example, if the color of a certain axis does not match the classification result, the system will mark it as an anomaly. This verification ensures the reliability of the results and reduces false positives.

[0045] It is understandable that re-using the SIFT algorithm to compare the abnormal classification axis can enable fine-tuning.

[0046] For example, by re-extracting key points, the matching degree of a misidentified axis increased from 60% to 90%, ultimately confirming its correct identity. This step improves overall recognition accuracy.

[0047] For example, a MySQL database is used to store the final results, with each axis's length, diameter, and classification result recorded as a row in a table. In one possible implementation, the database also supports real-time queries, for example, by entering an axis ID to retrieve all its parameters. This not only facilitates management but also provides data support for subsequent maintenance and optimization.

[0048] Specifically, the entire process, from database extraction to final storage, forms a closed-loop analysis system. The technology selected at each step focuses on improving accuracy and efficiency. For example, the SIFT algorithm enhances the robustness of feature matching, while laser scanning improves parameter accuracy, ultimately achieving comprehensive analysis of complex wiring harness structures. This approach is particularly practical in industrial scenarios, significantly reducing manual intervention costs and improving production consistency.

[0049] S2, based on the axis identity recognition results, extracts the corresponding current, voltage and signal waveforms from the real-time collected harness working data. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis status change trend.

[0050] Optionally, this step also includes: Step S21 , obtaining real-time acquisition results from the harness data through axis identity recognition, extracting corresponding current data, voltage data and signal waveform data, and obtaining a preliminary acquisition data set.

[0051] In step S22 , a pre-established threshold segmentation method is used to separate the dynamically changing portion in the high-density connection scenario from the preliminary collected data set to determine a dynamically changing data set.

[0052] Step S23 : For the dynamically changing data set, a time series analysis algorithm is used to analyze the time series characteristics of the signal waveform and determine the time distribution characteristics of the dynamic changes.

[0053] Step S24: If the time distribution characteristics of the dynamic changes exceed the pre-established fluctuation range, a local change segment is extracted from the dynamic change data set by a sliding window method to obtain a local change feature set.

[0054] Step S25 , based on the local change feature set, a support vector machine algorithm is used to classify abnormal states in the local change feature set to determine an abnormal state set.

[0055] Step S26, using the mapping relationship between the abnormal state set and the change trend, a linear regression algorithm is used to fit the change trend of the axis state to obtain a state change trend result.

[0056] Step S27, obtaining the slope change of the trend curve from the state change trend result, and determining the future evolution direction of the axis state.

[0057] Specifically, obtaining real-time acquisition results from the wiring harness data through axis identity recognition and extracting the corresponding current data, voltage data and signal waveform data is the starting point of the entire analysis process.

[0058] For example, on an identified axis, the acquisition device might record a current value of 2 amps and a voltage value of 24 volts, with a signal waveform showing a periodic square wave. This data forms the basis of the preliminary acquisition data set and provides the raw input for subsequent analysis.

[0059] In one possible implementation, the data collection device can use sensors to monitor the electrical signals of each axis in the wiring harness in real time. The data is stored with timestamps for easy tracking. A pre-established threshold segmentation method is used to isolate the dynamically changing components of high-density connection scenarios from the initial data set, focusing on key information.

[0060] Specifically, assuming the normal current threshold is set between 1 and 3 amps, when the current on a particular axis jumps to 5 amps, this portion of data is separated to form a dynamically changing data set. This method can quickly locate areas of abnormal fluctuations.

[0061] It should be noted that the threshold selection is usually based on historical data statistics to ensure coverage of common scenarios. For dynamically changing data sets, using a time series analysis algorithm to analyze the time series characteristics of the signal waveform can reveal the changing patterns.

[0062] For example, a waveform data segment might show that the current increases linearly from 2 amps to 5 amps within 10 seconds. Time series analysis can be used to determine whether this dynamic change is periodic or sudden.

[0063] In one embodiment, if the waveform jumps repeatedly within a short period of time, it may indicate an unstable connection, which provides a basis for subsequent judgment. If the time distribution characteristics of the dynamic change exceed the pre-established fluctuation range, the local change segment is extracted using a sliding window method.

[0064] Preferably, the sliding window can be set to 5 seconds in length with a step size of 1 second to capture the specific process of the current increasing from 2 amps to 5 amps, and obtain a local change feature set. The advantage of this method is that it refines the analysis granularity.

[0065] It's understandable that if the fluctuation range is set to ±20% and the actual change reaches 50%, window extraction can more accurately locate the problem period. Using the support vector machine algorithm to classify abnormal states based on the local change feature set is the core of anomaly detection.

[0066] For example, the feature set may include feature vectors such as current surge and voltage drop, which the algorithm classifies into two categories: "normal" and "abnormal".

[0067] In one embodiment, a shaft axis is flagged as abnormal due to a voltage drop to 20 volts. A support vector machine (SVM) is trained on historical data to achieve efficient classification. This step improves the accuracy of anomaly identification. By mapping the set of abnormal states to the changing trends, a linear regression algorithm is used to fit the changing trends of the shaft axis states, providing a direct reflection of the state evolution.

[0068] Specifically, anomaly clustering might show a continuous increase in current on a particular axis, with regression analysis yielding a trend curve that increases over time. This fitting can predict potential failures and enhance preventative measures.

[0069] For example, if the trend shows a current increase of 0.5 amperes per month, this may indicate an aging problem. The final step in the analysis is to determine the future direction of the axis state by analyzing the slope of the trend curve from the state change trend results.

[0070] For example, a change in slope from 0.1 to 0.5 may indicate an accelerating deterioration of the condition.

[0071] In one possible implementation, slope changes combined with historical data comparison can be used to infer whether the axis is nearing failure. This judgment provides data support for maintenance decisions and helps extend the service life of the wiring harness.

[0072] Optionally, the step S26, using a linear regression algorithm to fit the change trend of the axis state through the mapping relationship between the abnormal state set and the change trend to obtain the state change trend result, also includes: Step S261 : Obtain a mapping relationship between the abnormal state set and the change trend to obtain a mapping feature set.

[0073] Step S262: According to the mapping feature set, a linear regression algorithm is used to fit the change trend of the axis state to obtain trend fitting data.

[0074] Step S263 , for the trend fitting data, obtain local features of the time distribution to obtain a local feature set.

[0075] Step S264 , using a clustering algorithm based on the local feature set, determines the abnormal distribution in the trend result to obtain an abnormal distribution set.

[0076] Step S265: If the abnormal distribution set exceeds the preset abnormal threshold, the dynamic change segments are extracted by a sliding window method to obtain a dynamic segment set.

[0077] Step S266: Obtain the short-term fluctuation characteristics of the axis state according to the dynamic segment set to obtain a fluctuation characteristic set.

[0078] Step S267, by mapping the fluctuation feature set with the trend fitting data, the trend stability of the axis state is determined to obtain a stability result.

[0079] Specifically, the mapping relationship between the abnormal state set and the change trend is obtained to obtain the mapping feature set. The core of this process is to establish the correspondence between data.

[0080] For example, in a wiring harness operation scenario, an abnormal state may manifest as a sudden increase in the current of a certain axis.

[0081] It is understandable that the construction of this mapping relationship needs to rely on historical data, such as the current records of a certain axis under different loads in the past week, and then form a feature set to reflect the correlation between anomalies and trends.

[0082] In one possible implementation, assume that the normal current of an axis is 2 amps, but abnormally it reaches 5 amps. Mapping can initially determine whether this trend deviates from the norm. Based on the mapped feature set, a linear regression algorithm is used to fit the axis state's changing trend, generating trend fitting data. This step aims to quantify the pattern of change.

[0083] For example, assuming that the mapping feature set includes current values ​​at 10 time points, such as 2, 2.1, 2.3, 2.5, 3, 3.5, 4, 4.5, 5, and 5.2 amperes, linear regression can generate a trend line reflecting the characteristic that the current gradually increases over time.

[0084] Specifically, this fitting can help identify whether changes are stable, thereby providing basic data for subsequent analysis. The local features of the temporal distribution of trend fitting data are obtained, resulting in a local feature set. This process focuses on extracting detailed changes.

[0085] Preferably, the current value within a certain 5-minute period in the trend line, such as the current value rising from 3 amperes to 4 amperes, can be analyzed to extract the rising rate or the fluctuation amplitude as the local feature.

[0086] It should be noted that this local feature can reflect the state changes in a short period of time, making it easier to discover hidden anomalies. Using the local feature set, a clustering algorithm is used to determine the abnormal distribution in the trend results, and an abnormal distribution set is obtained. This method aims to classify similar features.

[0087] In one embodiment, if a local feature set contains multiple segments of data, such as one segment with a faster rising rate of 0.5 amps / minute and another with only 0.1 amp / minute, the clustering algorithm can classify the former as an abnormal distribution. This classification helps quickly locate problem areas. If the abnormal distribution set exceeds the abnormal threshold, a sliding window method is used to extract dynamically changing segments to obtain a dynamic segment set. This step emphasizes dynamic tracking.

[0088] For example, assuming the anomaly threshold is 0.3 amps / minute, if it exceeds this threshold, a sliding window with a width of 2 minutes can be used to extract segments from 4 amps to 5 amps. This approach captures the critical period when anomalies occur and improves analysis accuracy. The short-term fluctuation characteristics of the axis state are obtained from the dynamic segment set to form a fluctuation feature set. This process focuses on short-term characteristics.

[0089] In one embodiment, a dynamic segment shows that the current jumps from 4.5 amps to 5 amps and then drops back to 4.7 amps within a minute. A feature with a fluctuation amplitude of 0.5 amps can be extracted. This feature set can reveal unstable points in the axis state and provide a basis for subsequent judgment. By mapping the fluctuation feature set with trend fitting data, the trend stability of the axis state is determined and a stability result is obtained. This step integrates global and local information.

[0090] Specifically, if the trend fit shows a long-term rise, but the fluctuation feature set indicates frequent jumps in the short term, such as multiple fluctuations of 0.5 amps, it can be judged that the stability is low.

[0091] For example, in high-density connection scenarios, this analysis can provide early warning of potential failures and improve system reliability.

[0092] S3: If the axis state change trend exceeds the preset change threshold, the temperature and vibration data of the wiring harness are obtained through the embedded sensor, and the specific location and type of the abnormal axis are determined by data fusion technology based on the state change trend.

[0093] Optionally, this step also includes: Step S31 : If the change trend of the axis state exceeds the change threshold, the temperature and vibration data of the wiring harness are collected through the embedded sensor to obtain the original collected data set.

[0094] In step S32 , a low-pass filter is used to remove high-frequency noise from the original collected data set to obtain a cleaned data set.

[0095] Step S33 : extracting the mean and variance of the harness temperature and vibration data from the clean data set as feature parameters to form a feature data set.

[0096] In step S34, Kalman filtering is used to process the characteristic data set, and a preliminary judgment result of the abnormal axis is determined based on the change trend.

[0097] Step S35: Based on the preliminary judgment result, the deviation value of the specific position is calculated by interpolation method to obtain the positioning data of the abnormal axis.

[0098] Step S36: If the positioning data matches the pre-established type feature library, the support vector machine algorithm is used to classify the abnormal axis to determine the abnormal type.

[0099] Step S37: Generate complete description data of the abnormal axis according to the abnormality type and positioning data.

[0100] Specifically, if the change trend of the axis status exceeds the preset change threshold, collecting the wiring harness temperature and vibration data through embedded sensors will be the basis for subsequent analysis.

[0101] For example, on an identified axis, the embedded sensor may record a temperature of 75 degrees Celsius and a vibration frequency of 50 Hz. These data constitute the raw acquired data set, reflecting the real-time physical state of the harness.

[0102] In one possible implementation, sensors are installed at key nodes in the wiring harness, and data is stored once per second to ensure that instantaneous changes are captured. Using a low-pass filter to remove high-frequency noise from the original acquired data set is a key step. High-frequency noise may come from equipment interference or environmental jitter. Specifically, the low-pass filter can be set with a cutoff frequency of 100 Hz to retain the main signal characteristics. For example, after filtering out the 200 Hz noise mixed in the vibration data, the clean data set shows that the vibration frequency is stable at 50 Hz. This method ensures data reliability. Extracting the mean and variance from the clean data set as feature parameters can effectively characterize the wiring harness status.

[0103] For example, the mean temperature is 70 degrees Celsius and the variance is 5 degrees Celsius; the mean vibration is 50 Hz and the variance is 2 Hz. These characteristic parameters form a feature data set.

[0104] It's important to note that the mean reflects the overall level, while the variance reveals the degree of fluctuation. The combination of the two provides a multidimensional perspective for subsequent analysis. Using a Kalman filter to process feature datasets and identify abnormal axes based on changing trends is a dynamic optimization process. The Kalman filter smoothes the data through prediction and update steps.

[0105] In one embodiment, if the predicted mean temperature is 72 degrees Celsius and the measured value is 75 degrees Celsius, filtering and adjustment can identify the source of the deviation. Combined with trend analysis, if the temperature continues to rise, a preliminary assessment of the axis's potential anomaly is made. This approach improves the stability of the assessment. Based on this preliminary assessment, the deviation value is calculated through interpolation to pinpoint the anomalous axis, enabling precise location.

[0106] Preferably, if the sensor spacing is 10 cm, and the temperature jumps from 70°C to 75°C in a certain period, the interpolation method estimates that the abnormal point is located at the 3rd centimeter. This positioning data provides a spatial basis for subsequent classification.

[0107] It is understandable that interpolation methods, which infer the status of unmeasured points using existing data, are suitable for high-density wiring harness scenarios. If the positioning data matches the type feature library, using a support vector machine to classify the anomaly type is the core of intelligent processing.

[0108] For example, the feature library defines an "overheat" type corresponding to temperatures exceeding 70 degrees Celsius and normal vibration. If an axis has a temperature of 75 degrees Celsius and a vibration of 50 Hz, the support vector machine classifies it as "overheat."

[0109] In one possible implementation, training data includes historical anomaly samples to ensure classification accuracy. A complete description is generated based on the anomaly type and location data, providing a comprehensive overview of the problem. Specifically, the description might include: axis number A3, location 3 cm, type: overheating, temperature 75°C, vibration 50 Hz. This complete description facilitates rapid response by maintenance personnel.

[0110] In one embodiment, the description data may also be accompanied by a timestamp to track the time period when the anomaly occurred and improve management efficiency.

[0111] S4, through the position and type of the abnormal axis, extract the corresponding process parameters and batch information from the historical production data, use the cluster analysis algorithm to determine whether the abnormality is related to the production link, and obtain the distribution characteristics of the process defects.

[0112] Optionally, this step also includes: Step S41 , obtaining the position and type of the abnormal axis from historical production data, and extracting the corresponding process parameters and batch information.

[0113] Step S42: Use database query technology to obtain parameter data and obtain preliminary extraction results.

[0114] In step S43 , based on the preliminary extraction results, the process parameters and abnormal axes are grouped using the K-means clustering algorithm.

[0115] Step S44: Based on the grouping results, the correlation between the anomalies in each group and the production process is determined to obtain anomaly grouping features.

[0116] Step S45 , based on the abnormal grouping characteristics, obtain the matching relationship between the batch information and the process parameters, determine the distribution of the abnormal axis in the production link, and obtain batch association data.

[0117] Step S46: Using the batch association data, a regression analysis method is used to analyze the corresponding pattern between the location type and the process parameters, to determine whether the anomaly is caused by a specific production link, and to obtain the link impact result.

[0118] Step S47: Based on the link impact results, the correlation between the defect distribution and the process parameters is extracted, the influence degree of the abnormal axis on the distribution characteristics is determined, and the defect distribution law is obtained.

[0119] Step S48: According to the defect distribution law, the adjustment direction of the process parameters in the production link is obtained, the control point of the abnormal axis is determined, and parameter optimization suggestions are obtained.

[0120] In step S49, the dynamic changes of location type and defect distribution are analyzed through parameter optimization suggestions to determine the distribution characteristics of process defects and obtain the final analysis conclusion.

[0121] Specifically, obtaining the location and type of abnormal axes from historical production data and extracting the corresponding process parameters and batch information are the basis of analysis.

[0122] For example, in a certain production line, historical records show that an axis had an abnormality at the 5th centimeter, the type was "fracture", and the corresponding process parameters included a tensile force of 200 Newtons, a processing speed of 2 meters per second, and a batch number of B20250301.

[0123] It is understandable that these data reflect the specific context in which the anomaly occurred during the production process and provide the original basis for subsequent analysis. When using database query technology to obtain parameter data, the integrity of the data must be ensured.

[0124] Specifically, the tensile force, processing speed and other parameters of all axes in a batch can be extracted from the database through SQL queries.

[0125] In one example, the query results may show that the tensile force for batch B20250301 ranges from 180 to 220 Newtons, and the processing speed is stable at 2 meters per second. This preliminary extraction result lays the foundation for subsequent grouping. Based on this preliminary extraction result, the K-means clustering algorithm is used to group the process parameters and abnormal axes, and the division can be based on parameter similarity.

[0126] For example, if the number of clusters is set to 3, the algorithm may classify the axes with a tensile force higher than 210 Newtons into one group, where the abnormality type is mostly "fracture"; and the axes with a tensile force lower than 190 Newtons into another group, where the abnormality type is mostly "wear".

[0127] Preferably, this grouping can highlight the potential relationship between parameters and anomalies. Based on the grouping results, when determining the correlation between anomalies within each group and the production process, the changing trends of process parameters can be analyzed.

[0128] In one possible implementation, if abnormalities in the high-tensile-force group are concentrated in the stretching process, this indicates that this process may be the source of the problem. This grouping characteristic is thus reflected as "excessive tensile force leading to fracture." Based on this abnormality grouping characteristic, by obtaining a matching relationship between batch information and process parameters, specific batches can be traced.

[0129] For example, in batch B20250301, axes with high tensile forces were abnormally distributed in the stretching process, while batches with normal processing speeds did not exhibit this problem. This batch-related data reveals the distribution patterns of anomalies. Using regression analysis to analyze the correspondence between position types and process parameters using this batch-related data, we can observe the impact of parameter changes on anomalies.

[0130] Specifically, if the fracture probability increases by 5% for every 10 Newton increase in tensile force, it can be inferred that the tensile process is the key influencing factor. This process impact result provides guidance for subsequent optimization. Based on this process impact result, the correlation between defect distribution and process parameters can be extracted to identify areas of defect concentration.

[0131] For example, when the tensile force exceeds 210 Newtons, defects are most likely distributed 5 cm from the front of the axis. This defect distribution pattern helps pinpoint the root cause of the problem. Based on this defect distribution pattern, adjustments to process parameters during production can be identified, leading to recommendations for reducing the tensile force.

[0132] For example, adjusting the tensile force from 220 Newtons to 200 Newtons can significantly reduce the incidence of defects. This parameter optimization directly targets the defect control point. By analyzing the dynamic changes in location type and defect distribution through parameter optimization recommendations, the effects of the adjustments can be observed.

[0133] In one example, after the tensile force was reduced, the fracture anomaly shifted from the front 5 cm to a more uniform distribution, indicating that the distribution characteristics of the process defects were becoming more stable. This final analysis conclusion provides a reliable basis for production optimization.

[0134] S5, based on the distribution characteristics of process defects, obtains the signal attenuation and delay indicators of the abnormal axis from the real-time status monitoring data. In response to the fluctuation of data transmission efficiency, adaptive filtering technology is used to optimize the signal quality and output the adjusted transmission parameters.

[0135] Optionally, this step also includes: In step S51, real-time data is collected to collect status monitoring information, and the signal attenuation and delay indicators of the abnormal axis are extracted. Fast Fourier transform is used to extract the signal's spectral characteristics. In step S52, the abnormal frequency component is separated from the spectral characteristics, and a mapping relationship between the abnormal axis and signal attenuation is established. In step S53, a mathematical model of the signal attenuation and delay indicators is fitted using the least squares method based on the mapping relationship. In step S54, the rate of change of the delay indicator is calculated using numerical differentiation within the mathematical model. In step S55, the fluctuation amplitude of the transmission efficiency is calculated using the Kalman filter based on the rate of change. In step S56, based on the fluctuation amplitude, spectral analysis is used to determine the source of the data fluctuation. In step S57, an adaptive filter is used to process the data fluctuation, setting the filter cutoff frequency to the center frequency of the abnormal frequency component. In step S58, the filter outputs the optimized signal characteristics. In step S59, the transmission parameters are updated using the gradient descent method based on the optimized signal characteristics. In step S510, the transmission efficiency is recalculated based on the updated parameters. In step S511, if the transmission efficiency falls below the efficiency threshold, the delay indicator is re-extracted and the filter parameters are adjusted. Step S512: Acquire final transmission parameters and output a combination of signal attenuation and delay indicators in a stable state.

[0136] Specifically, the process of collecting status monitoring information through real-time data collection, extracting the signal attenuation and delay indicators of the abnormal axis, and using fast Fourier transform to extract the spectral characteristics of the signal, the core of which is to convert the time domain signal into frequency domain analysis.

[0137] For example, on a production line, the monitoring system collects signal data 1000 times per second. A specific axis shows a 20dB attenuation of signal strength and a 0.5 millisecond delay. Fast Fourier transform (FFT) can decompose this signal into frequency components. For example, suppose a dominant frequency of 50 Hz is detected, accompanied by abnormal high-frequency noise at 500 Hz. To isolate the abnormal frequency components from the spectral characteristics and establish a mapping between the abnormal axis and signal attenuation, filtering techniques can be used to isolate the target frequency.

[0138] Specifically, if the signal attenuation corresponding to the 500 Hz frequency component is significantly higher than that of other frequencies, it can be inferred to be the source of the anomaly. A table is then established that maps attenuation values ​​to specific frequencies, such as 500 Hz corresponding to 25 dB of attenuation. Based on this mapping, the least squares method is used to fit a mathematical model of signal attenuation and delay metrics, with the goal of quantifying the relationship between the two.

[0139] For example, if after collecting multiple sets of data, it is found that every 0.1 millisecond increase in delay is associated with a 5dB increase in attenuation, a preliminary linear relationship can be fitted, providing a basis for subsequent analysis. For mathematical models, when using numerical differentiation to calculate the rate of change of the delay indicator, the focus is on the trend of delay over time.

[0140] In one possible implementation, if a signal delay suddenly increases from 0.5 milliseconds to 0.7 milliseconds, with a rate of change of 0.2 milliseconds per second, this indicates that signal transmission may be blocked. Based on this rate of change, a Kalman filter is used to calculate the fluctuation in transmission efficiency, aiming to smooth out the effects of noise.

[0141] Preferably, if the rate of change causes the efficiency fluctuation range to be between 85% and 90%, the Kalman filter can predict a more stable efficiency value, such as 88%. In terms of the fluctuation amplitude, when using spectrum analysis to determine the source of data fluctuations, the distribution of abnormal frequency components can be focused on. Assuming that the analysis shows that the fluctuation is mainly caused by a frequency of 500 Hz, it means that the interference source may be in the high-frequency response part of the transmission equipment. An adaptive filter is used to process data fluctuations, and the filter cutoff frequency is set to the center frequency of the abnormal frequency component, such as 500 Hz, which can effectively suppress the frequency interference. After the filter outputs the optimized signal characteristics, the signal-to-noise ratio may be increased from 10 decibels to 15 decibels. According to the optimized signal characteristics, when the gradient descent method is used to update the transmission parameters, the parameters are gradually adjusted to reduce attenuation.

[0142] For example, the initial transmission power is 10 milliwatts, and after adjustment it is 12 milliwatts, reducing attenuation by 3 decibels. Based on the updated parameters, the transmission efficiency is recalculated. If the efficiency increases from 88% to 92%, the adjustment is effective. If the efficiency falls below the preset efficiency threshold of 95%, the delay index must be re-extracted and the filter parameters adjusted, such as fine-tuning the cutoff frequency to 510 Hz. After obtaining the final transmission parameters, the signal attenuation and delay index combination in the stable state is output. If, for example, the attenuation is reduced to 15 decibels and the delay is stabilized at 0.4 milliseconds, system performance is optimized.

[0143] S6, through the adjusted transmission parameters, extracts the digital signature and production mark of the axis from the anti-counterfeiting verification integrated module, uses blockchain technology to compare the digital signature, determines the authenticity of the wiring harness and generates a verification result.

[0144] Optionally, this step also includes: Step S61, adjust the parameters of the transmission protocol, obtain the axis signature and production identification data from the anti-counterfeiting verification module, and generate an initial data set. Step S62, use the SHA-256 hash algorithm to encrypt the axis signature, extract the encrypted digital signature from the initial data set, and calculate the signature characteristic value. Step S63, access the distributed ledger through the Ethereum blockchain, obtain the reference signature corresponding to the production identification from the ledger, and generate a comparison benchmark value. Step S64, if the signature characteristic value is consistent with the comparison benchmark value, the authenticity of the wiring harness is judged by digital comparison, and a preliminary verification status is generated. Step S65, for the preliminary verification status, use the preset verification rules to perform a second confirmation, determine whether the status is stable, and generate a confirmation result. Step S66, based on the confirmation result, use SQL query to filter abnormal data and generate a final verification result. Step S67, record the final verification result to the Ethereum blockchain ledger through the RESTful API to complete the verification process.

[0145] Specifically, the process of adjusting the parameters of the transmission protocol, obtaining the axis signature and production identification data from the anti-counterfeiting verification module, and generating the initial data set is centered on ensuring the integrity and traceability of the data.

[0146] For example, on a wire harness production line, the anti-counterfeiting verification module generates 100 sets of axis signatures and production identification data per minute, each set of data contains a 32-bit axis signature and a 16-bit production identification.

[0147] For example, the axis signature may be a unique serial number generated when the device is running, while the production identification is associated with the batch number and production time.

[0148] It should be noted that this data is transmitted intact even when the network is unstable by adjusting the timeout and retry count of the transmission protocol. The axis signature is encrypted using the SHA-256 hash algorithm, and the encrypted digital signature is extracted from the initial data set and the signature eigenvalue is calculated to enhance data security.

[0149] Specifically, the 32-bit axis signature is processed by SHA-256 to generate a 256-bit hash value as the digital signature.

[0150] In one possible implementation, if the signature of an axis is "AX123456," encryption may yield a fixed-length characteristic value. This encryption method ensures that even if the original data is leaked, it is difficult to reverse engineer. By accessing the distributed ledger via the Ethereum blockchain, a reference signature corresponding to the production identifier is retrieved from the ledger to generate a comparison benchmark, demonstrating the advantages of decentralized verification.

[0151] For example, the production identification “P20250321” corresponds to a recorded reference signature on the blockchain, and the result is returned through the smart contract query when accessed.

[0152] Ideally, the immutability of blockchain ensures the reliability of the reference signature. If the signature characteristic value matches the comparison reference value, the authenticity of the wiring harness is determined through digital comparison, generating a preliminary verification status. This step is a preliminary screening of data consistency.

[0153] In one embodiment, if the encrypted feature value completely matches the reference signature, the status is marked as "passed".

[0154] Understandably, this comparison is fast and efficient, avoiding the tediousness of manual verification. Based on the initial verification status, a secondary confirmation is performed using pre-set verification rules to determine whether the status is stable and generate a confirmation result, further improving the accuracy of the verification.

[0155] For example, the verification rule may require that the comparison results must be consistent three times in a row before the status is confirmed to be stable.

[0156] For example, if a wiring harness passes three verifications, the verification result is considered "authentic." Based on the verification results, SQL queries are used to filter out abnormal data and generate the final verification result, aiming to eliminate potential interference.

[0157] Specifically, the records marked as "unstable" in the database are queried and excluded through SQL statements.

[0158] In one embodiment, if a batch of data fluctuates due to network latency, it is filtered out to ensure a pure result. The final verification results are recorded to the Ethereum blockchain via a RESTful API, completing the verification process and enabling permanent archiving of the results.

[0159] For example, the verification result "real" is uploaded in JSON format through the API, and the record timestamp is 2025-03-2114:30:00.

[0160] In one embodiment, the API call returns a success status code 200, indicating that the data has been uploaded to the chain.

[0161] It should be noted that this method facilitates subsequent tracing and auditing.

[0162] S7, based on the verification results, obtains the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data. To meet the demand for enhanced reliability performance, a predictive model is used to analyze the stability trend of long-term operation.

[0163] Optionally, this step also includes: Step S71: Extract the response time and power fluctuation data of the wiring harness under different loads through dynamic performance monitoring to obtain an initial data set. Step S72: Use Python's Pandas library to process the noise and outliers in the initial data set to obtain an optimized data set. Step S73: Use the Scikit-learn library to calculate the characteristic values ​​of the response time and power fluctuation under load changes from the optimized data set to determine the characteristic distribution. Step S74: If the characteristic distribution exceeds the preset distribution threshold, use the support vector machine algorithm in the Scikit-learn library to train the reliability prediction model to determine the short-term stability. Step S75: Based on the short-term stability results, use the ARIMA model in the Statsmodels library to predict trend changes in long-term operation and obtain a trend curve. Step S76: Fit the stability characteristics through the trend curve to obtain the reliability performance of the wiring harness under different loads. Step S77: Extract key indicators from the reliability performance to determine the degree of realization of the enhancement requirements.

[0164] Specifically, the response time and power fluctuation data of the wiring harness under different loads are extracted through dynamic performance monitoring to obtain the initial data set.

[0165] It can be understood that the dynamic performance monitoring here refers to the real-time collection of performance data of the wiring harness during operation.

[0166] For example, on a wire harness production line, equipment may operate at low load, medium load, and high load, with response time and power fluctuations recorded separately.

[0167] For example, under low load, the response time might be 20 milliseconds with a power fluctuation of ±5 watts, while under high load, the response time might be extended to 50 milliseconds with a power fluctuation of ±15 watts. This acquisition method comprehensively reflects the performance of the wiring harness under different operating conditions. The Python Pandas library was used to process noise and outliers in the initial dataset to obtain an optimized dataset.

[0168] Specifically, Pandas can identify outliers through statistical methods.

[0169] For example, if the response time collected at a certain time suddenly jumps to 200 milliseconds, which is significantly different from the average, Pandas can mark it as noise and remove it.

[0170] In one possible implementation, the power fluctuation data might be capped at ±20 watts, with data outside this range considered abnormal. This process allows the optimized dataset to more accurately reflect the actual performance of the wiring harness. The Scikit-learn library is used to calculate the characteristic values ​​of response time and power fluctuation under load variations from the optimized dataset to determine the characteristic distribution.

[0171] It should be noted that the characteristic value may be the mean of the response time or the standard deviation of the power fluctuation.

[0172] For example, the standard deviation of power fluctuations under low load is 3 watts, but it may rise to 8 watts under high load. This characteristic distribution can intuitively show the behavior of the wiring harness under load changes.

[0173] Preferably, the degree of data dispersion can be quickly determined by visualizing the distribution trend. If the feature distribution exceeds the preset distribution threshold, the reliability prediction model is trained using the support vector machine algorithm in the Scikit-learn library to determine short-term stability.

[0174] In one embodiment, assuming that the preset response time threshold is 60 milliseconds, if the characteristic value reaches 70 milliseconds under high load, model training is triggered.

[0175] For example, a support vector machine can use historical data to learn the boundaries between normal and abnormal conditions and predict the stability of the harness over the next several hours. This approach allows for the timely detection of potential problems. Based on the short-term stability results, an ARIMA model from the Statsmodels library is used to predict long-term trends and generate a trend curve.

[0176] Specifically, the ARIMA model can predict the changing trend of the next month based on the response time data of the past week.

[0177] For example, if response times have been increasing over the recent period, the trend curve might show an upward slope.

[0178] In one possible implementation, the predicted curve for power fluctuation data may exhibit periodic oscillations. This prediction provides a basis for long-term planning. By fitting the stability characteristics of the trend curve, the reliability performance of the wiring harness under different loads can be obtained.

[0179] For example, after fitting, it may be found that the stability characteristic value under medium load is the highest, indicating that the wiring harness runs most smoothly under this working condition.

[0180] Understandably, this analysis helps optimize wiring harness design or usage scenarios.

[0181] In one embodiment, low reliability performance under high load may indicate a need to adjust materials or processes. Key indicators are extracted from the reliability performance to determine the degree to which enhancement requirements have been met.

[0182] For example, a key metric might be a stability score or failure rate. If the score is 90 under medium load but only 60 under high load, this indicates that the enhancement requirements for high-load scenarios have not been fully met.

[0183] By comparing the indicators of different batches of wiring harnesses, it is possible to select a production solution with better performance. This approach provides a clear direction for subsequent improvements.

[0184] S8: If the stability trend shows abnormality, the process parameters and equipment configuration are adjusted through the intelligent production process system. Combined with the real-time status monitoring data, the feedback control algorithm is used to optimize the production process and output the improved process plan.

[0185] Optionally, this step also includes: In step S81, if the monitoring data exceeds the monitoring threshold, key variables are extracted from the monitoring data as real-time status data. In step S82, based on the real-time status data, statistical analysis methods are used to determine abnormality indications. In step S83, based on the abnormality indications, the corresponding process parameter adjustment range is obtained from a pre-established process parameter mapping table. In step S84, the adjusted parameter values ​​are calculated within the adjustment range using interpolation. In step S85, the adjusted parameter values ​​are input into the equipment configuration management system to update the equipment configuration parameters. In step S86, key indicators of the production process are extracted from the equipment configuration parameters, and the optimization direction of the production process is calculated using a proportional-integral-differential control algorithm. In step S87, key parameters in the production process are adjusted based on the optimization direction to generate an improved process plan. In step S88, the improved process plan is input into the real-time monitoring system to update the real-time status data. In step S89, based on the changing trends of the real-time status data, it is determined whether the production process has returned to normal. In step S810, if the production process has returned to normal, the adjusted parameter values ​​and process plan are recorded as the final improvement plan.

[0186] Specifically, when the monitoring data exceeds the preset monitoring threshold, key variables are extracted from it as real-time status data.

[0187] For example, on a wire harness production line, assuming the preset power fluctuation threshold is ±20 watts, and a certain monitoring shows that the fluctuation reaches ±25 watts, the key variables extracted at this time may include power fluctuation value, load status and ambient temperature.

[0188] It is understood that these variables directly reflect the real-time operating conditions of the wiring harness.

[0189] Preferably, power fluctuation values ​​can indicate whether the equipment is overloaded, while ambient temperature may reveal the impact of external factors on performance. Based on real-time status data, statistical analysis methods are used to determine abnormal indications.

[0190] Specifically, anomalies can be determined by calculating the mean and variance of key variables.

[0191] For example, if the mean of power fluctuations is typically around ±10 watts, but the mean of the current data rises to ±22 watts and the variance increases significantly, this indicates an anomaly.

[0192] In one embodiment, time series analysis can be combined to observe whether the anomaly persists. This method can quickly locate the root cause of the problem. Based on the anomaly indication, the corresponding adjustment range is obtained from a pre-established process parameter mapping table.

[0193] For example, assume that the mapping table shows that when the power fluctuation exceeds ±20 watts, it is recommended to adjust the current input range to between 5-10 amperes.

[0194] It should be noted that the mapping table is based on historical data and experience, which can ensure that the adjustment range is reasonable.

[0195] In one possible implementation, if the anomaly is also associated with high temperature, the mapping table may additionally recommend reducing the operating speed. Within the adjustment range, the adjusted parameter value is calculated using interpolation.

[0196] For example, if the current adjustment range is 5-10 amps and the current value is 12 amps, the new value can be estimated to be 7 amps through linear interpolation.

[0197] Preferably, this method can achieve smooth transition and avoid sudden changes in parameters.

[0198] It is understandable that the interpolation method simplifies the calculation process while ensuring accuracy. The adjusted parameter values ​​are input into the device configuration management system to update the device configuration parameters.

[0199] Specifically, the new current value of 7 amps will be transmitted to the control system and take effect in real time.

[0200] In one embodiment, the system also records the update timestamp for easy traceability. This approach ensures timely and controllable adjustments. Key production process indicators are extracted from equipment configuration parameters, and the proportional-integral-derivative control algorithm is used to calculate the optimization direction.

[0201] For example, key metrics might be production efficiency and failure rate. If, after adjustments, efficiency improves by 5% and the failure rate drops to less than 1%, the algorithm will recommend further fine-tuning of load distribution.

[0202] For example, this algorithm can balance short-term response and long-term stability. According to the optimization direction, it adjusts key parameters in the production process and generates an improved process plan.

[0203] In one possible implementation, if the load distribution is adjusted to be mainly medium load, the solution may require reducing the high load operation time to 20%.

[0204] Specifically, this can reduce fatigue losses in the wiring harness.

[0205] It is understandable that the new solution is more suitable for the current status. The improved process solution is input into the real-time monitoring system to update the real-time status data.

[0206] For example, the monitoring system will show that power fluctuations are reduced to within ±15 watts.

[0207] The system can also generate a visual curve to show the adjusted stabilization trend. This real-time feedback helps verify the effectiveness of the solution. Based on the changing trend of the real-time status data, it can be judged whether the production process has returned to normal.

[0208] For example, if the power fluctuation is stable within ±15 watts for three consecutive days and the response time is restored to 30 milliseconds, it is considered normal.

[0209] In one embodiment, historical data can also be compared to confirm the absence of potential hazards. This determination provides a basis for process optimization. If the production process returns to normal, the adjusted parameter values ​​and process plan are recorded as the final improvement plan.

[0210] Specifically, the plan with a current of 7 amps and mainly medium load will be archived.

[0211] Preferably, the environmental conditions, such as a temperature of 25 degrees Celsius, can also be marked for future reuse.

[0212] It is understandable that such records provide a reference for subsequent production.

[0213] S9, through the improved process plan, extracts the axis distribution pattern in high-density connection scenarios from complex wiring harness structures, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

[0214] Optionally, this step also includes: Step S91: Process the wire harness structure using the optimized process plan to obtain an optimized wire harness sample. Step S92: Perform quality inspection on the optimized wire harness sample to ensure that the sample meets the preset quality standards. Step S93: Extract the axis distribution data in the high-density connection scenario from the wire harness sample that meets the quality standards. Step S94: Clean and standardize the extracted axis distribution data to obtain a preprocessed axis distribution dataset. Step S95: Use a convolutional neural network to train the preprocessed axis distribution dataset to obtain preliminary recognition accuracy results. Step S96: Compare the preliminary recognition accuracy results with the preset accuracy threshold. Step S97: If the recognition accuracy is lower than the preset accuracy threshold, iterative training is performed by adjusting the learning rate and batch size parameters until the recognition accuracy meets the preset requirements. Step S98: Construct an initial axis management model based on the model parameters after iterative training. Step S99: Use a test dataset to evaluate the initial axis management model and calculate the model's accuracy and recall indicators. Step S910: Determine the axis distribution pattern based on the evaluation results. Step S911: Apply the axis management model to a high-density connection scenario analysis to obtain the scenario analysis results. Step S912: Based on the scenario analysis results, optimize the model parameters to obtain an enhanced management model. Step S913: Calculate the prediction accuracy of the enhanced management model in the high-density connection scenario. Step S914: If the prediction accuracy is lower than the preset requirement, optimize the model structure by increasing the number of network layers or adjusting the activation function. Step S915: Finally, obtain an axis management model that meets the accuracy requirements.

[0215] Specifically, when processing the wire harness structure through the optimized process plan, the material distribution of the wire harness can be replanned first.

[0216] For example, if the original design had uneven insulation thickness on a wire harness, potentially leading to localized overheating, the optimized solution could adjust the extruder's pressure distribution to keep the thickness deviation within ±0.05 mm. Next, a multi-dimensional inspection method could be used to test the quality of the optimized wire harness samples.

[0217] For example, a high-precision resistance tester is used to measure the conductivity performance to ensure that the resistance value is stable within 0.01 ohms. At the same time, a tensile testing machine is used to test the mechanical strength of the wiring harness to confirm that the tensile force reaches more than 200 Newtons. When extracting axis distribution data in high-density connection scenarios from samples that meet the standards, attention should be paid to the spatial coordinates of the connection points. In one possible implementation method, the coordinate data of each connection point can be obtained through a 3D scanner to form a preliminary data set. When cleaning and standardizing data, outliers need to be removed and the format needs to be unified.

[0218] Specifically, if the coordinates of a connection point deviate from the mean by more than 3 standard deviations, it is considered a noise point and removed, and the coordinate data is normalized to the range of 0 to 1. When using a convolutional neural network to train a preprocessed dataset, the initial network structure can be set to three convolutional layers plus two fully connected layers.

[0219] Preferably, during the training process, the input axis distribution data is represented as an image, with each pixel corresponding to the density value of a connection point. The initial recognition accuracy result may be shown as 85%. If the preset accuracy threshold is 90%, the parameters need to be adjusted.

[0220] In one embodiment, the learning rate can be reduced from 0.01 to 0.005, the batch size can be increased from 32 to 64, and the accuracy can be improved to 91% after iterative training. After the initial axis management model is constructed, when the test data set is used for evaluation, the accuracy can be calculated to be 93% and the recall rate can be 88%. According to the evaluation results, the axis distribution pattern is determined. For example, it is found that the connection points in high-density areas are mostly distributed in a grid shape. When it is applied to the analysis of high-density connection scenarios, a density distribution map of the connection points can be output. The model parameters are optimized according to the analysis results, such as increasing the convolution kernel size, to obtain an enhanced management model. If the calculated prediction accuracy is 87%, which is lower than the preset 90%, the structure can be optimized by adding a convolution layer or changing the activation function from ReLU to LeakyReLU. The prediction accuracy of the final model reached 92%, which met the requirements.

[0221] See also Figure 2 In a second aspect, the present invention provides an intelligent identification and status monitoring system for the central axis of a complex wiring harness structure, which uses the method described in the above claims to intelligently identify and monitor the central axis of the complex wiring harness structure, and mainly includes: The axis feature recognition module is used to obtain the physical parameters and signal characteristics of each axis through a pre-established axis feature database. It uses machine vision technology to scan and compare the axes in the complex wiring harness structure to obtain the axis identity recognition results. The abnormal state detection module is used to extract the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data based on the axis identity recognition results. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis state change trend; The data fusion analysis module is used to obtain the temperature and vibration data of the wiring harness through embedded sensors if the axis state change trend exceeds the preset change threshold. Combined with the state change trend, the module uses data fusion technology to determine the specific location and type of the abnormal axis. The process defect analysis module is used to extract the corresponding process parameters and batch information from historical production data based on the location and type of the abnormal axis, and use the cluster analysis algorithm to determine whether the abnormality is related to the production link to obtain the distribution characteristics of the process defects; The signal optimization module is used to obtain the signal attenuation and delay indicators of abnormal axes from real-time status monitoring data based on the distribution characteristics of process defects. In response to fluctuations in data transmission efficiency, it uses adaptive filtering technology to optimize signal quality and output adjusted transmission parameters; The anti-counterfeiting verification module is used to extract the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module through the adjusted transmission parameters, compare the digital signature using blockchain technology, determine the authenticity of the wiring harness, and generate a verification result; The performance prediction module is used to obtain the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data based on the verification results. To meet the needs of enhancing reliability performance, a prediction model is used to analyze the stability trend of long-term operation; The process optimization module is used to adjust process parameters and equipment configurations through the intelligent production process system if the stability trend shows abnormalities. In combination with real-time status monitoring data, it uses feedback control algorithms to optimize the production process and output an improved process plan; The model enhancement module is used to extract the axis distribution pattern in high-density connection scenarios from complex wiring harness structures through improved process solutions, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

[0222] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. Intelligent identification and status monitoring method of the central axis of complex wiring harness structure, characterized by: The method comprises: S1, obtains the physical parameters and signal characteristics of each axis through the pre-established axis feature database, uses machine vision technology to scan and compare the axes in the complex wiring harness structure, and obtains the axis identity recognition result; S2, based on the axis identification results, extracts the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis status change trend. S3: If the axis state change trend exceeds the preset change threshold, the temperature and vibration data of the wiring harness are acquired through the embedded sensor. Combined with the state change trend, data fusion technology is used to determine the specific location and type of the abnormal axis. S4, based on the position and type of the abnormal axis, extract the corresponding process parameters and batch information from the historical production data, use the cluster analysis algorithm to determine whether the abnormality is related to the production link, and obtain the distribution characteristics of the process defects; S5, based on the distribution characteristics of process defects, obtains the signal attenuation and delay indicators of the abnormal axis from the real-time status monitoring data. In response to the fluctuation of data transmission efficiency, it uses adaptive filtering technology to optimize the signal quality and output the adjusted transmission parameters; S6, through the adjusted transmission parameters, extracts the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module, uses blockchain technology to compare the digital signature, determines the authenticity of the wiring harness and generates a verification result; S7, based on the verification results, obtains the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data. To meet the needs of enhanced reliability performance, a predictive model is used to analyze the stability trend of long-term operation; S8, if the stability trend shows an abnormality, the production process intelligent system will adjust the process parameters and equipment configuration, combine the real-time status monitoring data, use the feedback control algorithm to optimize the production process and output the improved process plan; S9, through the improved process plan, extracts the axis distribution pattern in high-density connection scenarios from complex wiring harness structures, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

2. The method according to claim 1, characterized in that The step S1, obtaining the physical parameters and signal characteristics of each axis through a pre-established axis feature database, and using machine vision technology to scan and compare the axes in the complex wiring harness structure to obtain the axis identity recognition results, includes: Step S11, obtaining the physical parameters and signal characteristics of each axis through a preset axis characteristic database; Step S12, using the image processing function in the OpenCV library to perform a preliminary scan of the axis in the complex wiring harness structure to obtain initial data of the axis; Step S13: Based on the initial data, the SIFT algorithm is used to compare the axis features with the signal features in the database to determine the preliminary classification results of the axis; Step S14, using a high-precision laser scanner to perform in-depth analysis of the wiring harness structure in the complex structure to obtain supplementary parameters of the unidentified axis; Step S15: if the supplementary parameters match the physical parameters in the database, the identity recognition result of the axis is updated by using the cosine similarity calculation method; Step S16, verifying the updated identity recognition result through the image recognition function in the OpenCV library to determine whether there is any abnormal classification; Step S17: Based on the verification results, the SIFT algorithm is used to re-compare the signal features of the abnormally classified axes to obtain the final identity recognition result; In step S18, for the final result, the physical parameters and identification data of all axes are stored in a MySQL database to complete the complete analysis of the wiring harness structure.

3. The method according to claim 1, characterized in that Step S2, based on the axis identification result, extracts the corresponding current, voltage, and signal waveform from the real-time collected harness working data, and uses a timing analysis algorithm to determine the axis state change trend based on the dynamic changes in the high-density connection scenario, including: Step S21 , obtaining real-time acquisition results from the harness data through axis identity recognition, extracting corresponding current data, voltage data, and signal waveform data to obtain a preliminary acquisition data set; Step S22, using a pre-established threshold segmentation method, separating the dynamically changing portion in the high-density connection scenario from the preliminary collected data set to determine a dynamically changing data set; Step S23: For the dynamically changing data set, a time series analysis algorithm is used to analyze the time series characteristics of the signal waveform and determine the time distribution characteristics of the dynamic changes; Step S24: if the time distribution characteristics of the dynamic change exceed the pre-established fluctuation range, then extract the local change segments from the dynamic change data set by using a sliding window method to obtain a local change feature set; Step S25, based on the local change feature set, using a support vector machine algorithm to classify abnormal states in the local change feature set and determine an abnormal state set; Step S26, using the mapping relationship between the abnormal state set and the change trend, a linear regression algorithm is used to fit the change trend of the axis state to obtain a state change trend result; Step S27, obtaining the slope change of the trend curve from the state change trend result, and determining the future evolution direction of the axis state.

4. The method according to claim 1, wherein In step S3, if the axis state change trend exceeds a preset change threshold, the temperature and vibration data of the wiring harness are acquired through the embedded sensor, and the specific location and type of the abnormal axis are determined by using data fusion technology in combination with the state change trend, including: Step S31 , if the change trend of the axis state exceeds the change threshold, collect harness temperature and vibration data through the embedded sensor to obtain the original collected data set; Step S32: using a low-pass filter to remove high-frequency noise from the original collected data set to obtain a clean data set; Step S33, extracting the mean and variance of the harness temperature and vibration data from the clean data set as feature parameters to form a feature data set; Step S34: Process the feature data set using Kalman filtering, and determine the preliminary judgment result of the abnormal axis based on the change trend; Step S35: Based on the preliminary judgment result, the deviation value of the specific position is calculated by interpolation to obtain the positioning data of the abnormal axis; Step S36: If the positioning data matches the pre-established type feature library, the support vector machine algorithm is used to classify the abnormal axis and determine the abnormal type; Step S37: Generate complete description data of the abnormal axis according to the abnormality type and positioning data.

5. The method according to claim 1, wherein In step S4, the corresponding process parameters and batch information are extracted from the historical production data based on the position and type of the abnormal axis, and a cluster analysis algorithm is used to determine whether the abnormality is related to the production process, thereby obtaining the distribution characteristics of the process defects, including: Step S41, obtaining the location and type of the abnormal axis from historical production data, and extracting the corresponding process parameters and batch information; Step S42, using database query technology to obtain parameter data and obtain preliminary extraction results; Step S43: Based on the preliminary extraction results, the K-means clustering algorithm is used to group the process parameters and abnormal axes; Step S44: Based on the grouping results, determine the correlation between the anomalies in each group and the production process, and obtain the anomaly grouping characteristics; Step S45: According to the abnormal grouping characteristics, the matching relationship between the batch information and the process parameters is obtained, the distribution of the abnormal axis in the production link is determined, and the batch association data is obtained; Step S46: Using the batch correlation data, a regression analysis method is used to analyze the corresponding pattern between the location type and the process parameters to determine whether the anomaly is caused by a specific production link and obtain the link impact result; Step S47: extract the correlation between defect distribution and process parameters based on the link impact results, determine the influence of the abnormal axis on the distribution characteristics, and obtain the defect distribution law; Step S48: According to the defect distribution law, the adjustment direction of the process parameters in the production process is obtained, the control point of the abnormal axis is determined, and parameter optimization suggestions are obtained; In step S49, the dynamic changes of location type and defect distribution are analyzed through parameter optimization suggestions to determine the distribution characteristics of process defects and obtain the final analysis conclusion.

6. The method according to claim 1, characterized in that Step S5, based on the distribution characteristics of process defects, obtains the signal attenuation and delay indicators of the abnormal axis from the real-time status monitoring data, optimizes the signal quality using adaptive filtering technology to address fluctuations in data transmission efficiency, and outputs adjusted transmission parameters, including: Step S51, collecting status monitoring information through real-time data collection, extracting the signal attenuation and delay indicators of the abnormal axis, and extracting the spectrum characteristics of the signal using fast Fourier transform; Step S52, separating abnormal frequency components from the spectrum characteristics and establishing a mapping relationship between abnormal axis and signal attenuation; Step S53, fitting a mathematical model of signal attenuation and delay indicators using the least squares method according to the mapping relationship; Step S54, using numerical differentiation to calculate the rate of change of the delay indicator based on the mathematical model; Step S55, using Kalman filtering to calculate the fluctuation range of the transmission efficiency according to the change rate; Step S56: using spectrum analysis to determine the source of data fluctuations based on the fluctuation amplitude; Step S57, using an adaptive filter to process data fluctuations, setting the filter cutoff frequency to the center frequency of the abnormal frequency component; Step S58, outputting the optimized signal characteristics through the filter; Step S59, updating the transmission parameters using the gradient descent method according to the optimized signal characteristics; Step S510, recalculating the transmission efficiency based on the updated parameters; Step S511: if the transmission efficiency is lower than the efficiency threshold, re-extract the delay index and adjust the filter parameters; Step S512: Acquire final transmission parameters and output a combination of signal attenuation and delay indicators in a stable state.

7. The method according to claim 1, characterized in that The step S6, extracting the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module using the adjusted transmission parameters, comparing the digital signature using blockchain technology, determining the authenticity of the wiring harness and generating a verification result, includes: Step S61: Adjust the parameters of the transmission protocol, obtain the axis signature and production identification data from the anti-counterfeiting verification module, and generate an initial data set; Step S62: encrypt the axis signature using the SHA-256 hash algorithm, extract the encrypted digital signature from the initial data set, and calculate the signature feature value; Step S63: Accessing the distributed ledger through the Ethereum blockchain, obtaining a reference signature corresponding to the production identifier from the ledger, and generating a comparison benchmark value; Step S64: If the signature feature value is consistent with the comparison reference value, the authenticity of the wiring harness is determined by digital comparison and a preliminary verification status is generated; Step S65: Perform secondary confirmation on the initial verification status using preset verification rules to determine whether the status is stable and generate a confirmation result; Step S66: Based on the confirmation result, use SQL query to filter abnormal data and generate the final verification result; Step S67: Record the final verification result to the Ethereum blockchain account book through the RESTful API to complete the verification process.

8. The method according to claim 1, characterized in that In step S7, based on the verification results, the response time and power fluctuation of the wiring harness under different loads are obtained from the dynamic performance monitoring data. To meet the demand for enhanced reliability performance, a prediction model is used to analyze the stability trend of long-term operation, including: Step S71, extracting response time and power fluctuation data of the wiring harness under different loads through dynamic performance monitoring to obtain an initial data set; Step S72, using Python's Pandas library to process noise and outliers in the initial data set to obtain an optimized data set; Step S73, using the Scikit-learn library to calculate the characteristic values ​​of response time and power fluctuation under load changes from the optimized data set to determine the characteristic distribution; Step S74: If the feature distribution exceeds the distribution threshold, the reliability prediction model is trained using the support vector machine algorithm in the Scikit-learn library to determine the short-term stability; Step S75, based on the short-term stability results, use the ARIMA model in the Statsmodels library to predict the trend changes in the long-term operation and obtain a trend curve; Step S76, obtaining the reliability performance of the wiring harness under different loads by fitting the stability characteristics through trend curve; Step S77: extract key indicators from the reliability performance to determine the degree of realization of the enhancement requirements.

9. The method according to claim 1, characterized in that In step S8, if the stability trend shows an abnormality, the production process intelligent system adjusts the process parameters and equipment configuration, combines the real-time status monitoring data, uses a feedback control algorithm to optimize the production process, and outputs an improved process plan, including: Step S81: if the monitoring data exceeds the monitoring threshold, extract key variables from the monitoring data as real-time status data; Step S82, using statistical analysis methods to determine abnormal indications based on real-time status data; Step S83, obtaining the corresponding process parameter adjustment range according to the abnormal indication from the pre-established process parameter mapping table; Step S84, calculating the adjusted parameter value within the adjustment range using interpolation method; Step S85, inputting the adjusted parameter values ​​into the device configuration management system to update the device configuration parameters; Step S86, extracting key indicators of the production process from the equipment configuration parameters, and using the proportional-integral-differential control algorithm to calculate the optimization direction of the production process; Step S87, adjusting key parameters in the production process according to the optimization direction to generate an improved process plan; Step S88: Input the improved process plan into the real-time monitoring system to update the real-time status data; Step S89, judging whether the production process has returned to normal based on the changing trend of the real-time status data; Step S810: If the production process returns to normal, the adjusted parameter values ​​and process plan are recorded as the final improvement plan.

10. Intelligent identification and status monitoring system for the central axis of complex wiring harness structure, characterized by: The method according to any one of claims 1 to 9 is used to perform intelligent identification and status monitoring of the central axis of a complex wiring harness structure, the system comprising: The axis feature recognition module is used to obtain the physical parameters and signal characteristics of each axis through a pre-established axis feature database. It uses machine vision technology to scan and compare the axes in the complex wiring harness structure to obtain the axis identity recognition results. The abnormal state detection module is used to extract the corresponding current, voltage, and signal waveforms from the real-time collected harness operating data based on the axis identity recognition results. Based on the dynamic changes in high-density connection scenarios, a timing analysis algorithm is used to determine the axis state change trend; The data fusion analysis module is used to obtain the temperature and vibration data of the wiring harness through embedded sensors if the axis state change trend exceeds the preset change threshold. Combined with the state change trend, the module uses data fusion technology to determine the specific location and type of the abnormal axis. The process defect analysis module is used to extract the corresponding process parameters and batch information from historical production data based on the location and type of the abnormal axis, and use the cluster analysis algorithm to determine whether the abnormality is related to the production link to obtain the distribution characteristics of the process defects; The signal optimization module is used to obtain the signal attenuation and delay indicators of abnormal axes from real-time status monitoring data based on the distribution characteristics of process defects. In response to fluctuations in data transmission efficiency, it uses adaptive filtering technology to optimize signal quality and output adjusted transmission parameters; The anti-counterfeiting verification module is used to extract the digital signature and production identification of the axis from the anti-counterfeiting verification integrated module through the adjusted transmission parameters, compare the digital signature using blockchain technology, determine the authenticity of the wiring harness, and generate a verification result; The performance prediction module is used to obtain the response time and power fluctuation of the wiring harness under different loads from the dynamic performance monitoring data based on the verification results. To meet the needs of enhancing reliability performance, a prediction model is used to analyze the stability trend of long-term operation; The process optimization module is used to adjust process parameters and equipment configurations through the intelligent production process system if the stability trend shows abnormalities. In combination with real-time status monitoring data, it uses feedback control algorithms to optimize the production process and output an improved process plan; The model enhancement module is used to extract the axis distribution pattern in high-density connection scenarios from complex wiring harness structures through improved process solutions, and uses deep learning technology to iteratively train the recognition and monitoring accuracy to obtain an enhanced axis management model.

Images