An online monitoring system for wire harness production based on artificial intelligence

Abstract

The present invention discloses an on-line monitoring system for wire harness production based on artificial intelligence, specifically relating to the technical field of wire harness production, including a wire harness production monitoring data acquisition module, a wire harness production monitoring data analysis module, a wire harness production intelligent monitoring module, a wire harness production intelligent monitoring comprehensive analysis module, and a wire harness production intelligent monitoring human-computer interaction module; the wire harness production monitoring data acquisition module: is used to collect data during the wire harness production process, obtain various parameters of on-line monitoring of wire harness production, and transmit the various parameters to the wire harness production monitoring data analysis module; Through artificial intelligence technology, the present invention adopts advanced sensor technology and data processing algorithms, enabling the system to accurately measure and analyze various parameters, improving the detection efficiency, thereby being able to more comprehensively monitor each link of the wire harness production line, improving the reliability and accuracy of the detection results, and helping to ensure the production quality of the wire harness.

Description

Technical Field

[0001] The present invention relates to the technical field of wire harness production, and specifically to an on-line monitoring system for wire harness production based on artificial intelligence. Background Art

[0002] With the rapid development of industries such as automobiles and electronics, the demand for wire harnesses is also increasing continuously. Therefore, improving production efficiency and quality has become an urgent need for wire harness production enterprises. Artificial intelligence technology has powerful data processing and analysis capabilities, can process a large amount of production data, extract key information, provide strong support for production monitoring, and thus realize automated and intelligent production monitoring to improve production efficiency.

[0003] In modern industrial manufacturing, wire harnesses, as key components for electrical connection and signal transmission, cover multiple key processes such as stranding, crimping, injection molding, and assembly. In the stranding process, parameters such as the tension of the wire core and the uniformity of the lay length are crucial for the electrical conductivity and anti-interference ability of the wire harness. Traditional monitoring methods mostly rely on manual sampling inspection, making it difficult to comprehensively and timely detect subtle defects in the stranding process; in the crimping process, the firmness of the crimping and the size of the contact resistance are directly related to the reliability of the wire harness connection. However, existing monitoring means are difficult to accurately measure the mechanical properties and electrical parameters at the crimping point; the injection molding process plays a decisive role in the protection and insulation performance of the wire harness, but traditional monitoring is prone to missed inspections; the assembly process requires ensuring the accurate installation and good cooperation of each component. Manual inspection of the assembly quality is not only inefficient but also prone to errors caused by human negligence.

[0004] With the progress of technology and the increasing market demand for high-quality wire harness products, traditional monitoring methods have been difficult to meet the requirements of high precision, high efficiency, and high stability in modern wire harness production. Therefore, there is an urgent need for an on-line monitoring system for wire harness production based on artificial intelligence, which can monitor key processes such as stranding, crimping, injection molding, and assembly in real time and accurately, timely discover problems in the production process, improve the quality and efficiency of wire harness production, reduce production costs, and enhance the competitiveness of enterprises in the market. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an on-line monitoring system for wire harness production based on artificial intelligence to solve the problems raised in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solution: An on-line monitoring system for wire harness production based on artificial intelligence, comprising:

[0007] Wire harness production monitoring data acquisition module: used to collect data during the wire harness production process, obtain various parameters for on-line monitoring of wire harness production, and transmit various parameters to the wire harness production monitoring data analysis module;

[0008] Wire harness production monitoring data analysis module: used to analyze various parameters obtained by the data acquisition module, obtain various monitoring indicators for on-line monitoring of wire harness production, and transmit various monitoring indicators to the wire harness production intelligent monitoring module. The wire harness production monitoring data analysis module includes a stranding quality monitoring unit, a crimping quality monitoring unit, an injection molding quality monitoring unit, and an assembly quality monitoring unit;

[0009] Wire harness production intelligent monitoring module: used to compare various monitoring indicators obtained by the data analysis module with thresholds respectively, transmit good comparison results to the wire harness production intelligent monitoring comprehensive analysis module, and send warning information to the wire harness production intelligent monitoring human-computer interaction module according to abnormal comparison results;

[0010] Wire harness production intelligent monitoring comprehensive analysis module: used to comprehensively analyze the good comparison results obtained by the intelligent monitoring module, obtain the comprehensive index for on-line monitoring of wire harness production, and transmit it to the wire harness production intelligent monitoring human-computer interaction module;

[0011] Wire harness production intelligent monitoring human-computer interaction module: used to receive the warning information of the intelligent monitoring module and the comprehensive index for on-line monitoring of wire harness production of the comprehensive analysis module for human-computer interaction.

[0012] Preferably, the various parameters in the wire harness production monitoring data acquisition module include administrator identity verification information, stranding quality monitoring parameters, crimping quality monitoring parameters, injection molding quality monitoring parameters, and assembly quality monitoring parameters. Among them, the stranding quality monitoring parameters include the length, pitch, tension, and diameter of the stranded wire; the crimping quality monitoring parameters include the peak pressure of crimping, the height of the crimped terminal, the pull-out force of the terminal after crimping, the conductor resistance before crimping, and the conductor resistance after crimping; the injection molding quality monitoring parameters include the actual pressure of the injection molding cycle, the size of the injection molded part, the weight of the injection molded part, the number of surface defects of the injection molded part, and the number of measurement samples of the injection molded part; the assembly quality monitoring parameters include the assembly step data set, the actual execution order of each assembly step, the number of steps with the actual order meeting the requirements, the total number of assembly steps, the number of terminals with the insertion depth meeting the requirements, the total number of terminals, the actual assembly time of each step, and the corresponding standard assembly time.

[0013] Preferably, in the stranding quality monitoring unit of the wire harness production monitoring data analysis module: first, the length L of the stranded wire is monitored in real time through a high-precision length sensor act , and compared with the designed length L des to obtain the stranding length deviation index LDR, LDR = |L act-L des | / L des ; Then, the pitch of the stranded wire is monitored in real time through a pitch sensor, n pitch values are recorded, the average value μ(p) and the standard deviation σ(p) of the pitch are obtained, and the pitch difference index PCI is obtained, PCI = σ(p) / μ(p); Secondly, the tension of the stranded wire during the processing is monitored in real time through a tension sensor, m tension values are recorded, the average value μ(f) and the standard deviation σ(p) of the tension are obtained, and the tension fluctuation index TFR is obtained, TFR = σ(p) / μ(f); Then, through a laser rangefinder, the diameter values of the stranded wire at k time points are recorded, and the diameter change rate DCR of the stranded wire is obtained. , j ∈ k, j is at least 1 and j ≤ k, R j and R j+1 represent the diameter values at the j-th and the (j + 1)-th time points respectively, t j and t j+1 represent the time values at the j-th and the (j + 1)-th time points respectively, and max() represents taking the maximum value; Finally, the stranded wire quality monitoring index SQMI is obtained through big data analysis technology. , DCR0 represents the maximum allowable diameter change rate.

[0014] Preferably, in the crimping quality monitoring unit of the harness production monitoring data analysis module: First, the pressure during the crimping process is monitored in real time through a pressure sensor, and the pressure peak value P peak of each crimping is recorded, and the crimping pressure deviation index PDR is obtained, PDR = |P peak -P peak,0 | / P peak , P peak,0 represents the set crimping pressure peak value; Then, the height of the harness terminal after each crimping is measured through a measuring device, N height values are recorded, the average value μ(h) and the standard deviation σ(h) of the height of the harness crimped terminal are obtained, and the harness crimped terminal height difference index HCI is obtained, HCI = σ(h) / μ(h); Secondly, the pull-out force test is carried out on the terminal after each crimping, N test results are recorded, the average value μ(F) and the standard deviation σ(F) of the terminal pull-out force test are obtained, and the terminal pull-out force difference index SCI is obtained, SCI = σ(F) / μ(F); Finally, through big data analysis technology, the crimping quality monitoring index CQMI is obtained. , F hou and F pre represent the conductor resistance after crimping and the conductor resistance before crimping respectively.

[0015] Preferably, in the injection molding quality monitoring unit of the harness production monitoring data analysis module: First, the pressure data is recorded in real time through the pressure sensor on the injection molding machine, and the actual pressure of L injection molding cycles is obtained, and the injection molding pressure fluctuation index PFR is obtained. , l ∈ L, IP act,l represents the actual pressure of the l-th injection molding cycle, IP set represents the set pressure; then through a measuring tool, according to the dimensional production process parameters of the injection molded part, obtain the dimensional measurement value D of the injection molded part mea , obtain the dimensional deviation index DDC of the injection molded part , i ∈ M, D mea，i represents the dimensional measurement value of the i-th injection molded part, D des represents the design dimension, M is the number of injection molded part measurement samples; secondly, weigh the injection molded parts through a weight sensor, record the weight of each injection molded part, and obtain the weight difference index WCI of the injection molded parts , W act，i represents the actual weight of the i-th injection molded part, W des represents the design weight; thirdly, through visual recognition technology, identify the number n_d of injection molded parts with surface defects, and obtain the surface defect index SDR of the injection molded parts, SDR = n_d / M; finally, obtain the injection molding quality monitoring index IQMI through big data analysis technology , add 0.01 to avoid the denominator being 0

[0016] Preferably, in the assembly quality monitoring unit of the wire harness production monitoring data analysis module: first, obtain the assembly step data set ASD according to the wire harness production assembly design parameters , K represents the number of steps required for the assembly process, AS J represents the J-th assembly step. Set sensors at each assembly step, record the actual execution order, and compare it with the designed order to obtain the assembly sequence compliance index CSR, CSR = n_as / Tn_as, n_as represents the number of steps where the actual order meets the requirements, Tn_as represents the total number of assembly steps; then measure the insertion depth of the wire harness terminals through a depth sensor and compare it with the design parameters to obtain the wire harness terminal insertion depth compliance index TDR, TDR = n_t / Tn_t, n_t represents the number of terminals with the insertion depth meeting the requirements, Tn_t represents the total number of terminals; at the same time, set a timer at each assembly step to record the actual assembly time T of each step act,J , and compare it with the corresponding standard assembly time T set,J to obtain the assembly time fluctuation index TTR , finally, through big data analysis technology, obtain the assembly quality monitoring index AQMI .

[0017] Preferably, the intelligent monitoring module in the wire harness production intelligent monitoring module includes the following steps:

[0018] Step 1: Compare the stranded wire quality monitoring index SQMI with the threshold SQMI0 to obtain the early warning factor ζ(SQMI) for stranded wire quality monitoring. ζ(SQMI) = exp(SQMI - SQMI0). If ζ(SQMI) ≥ 1, it indicates that the stranded wire quality monitoring is good; otherwise, it indicates that the stranded wire quality monitoring is abnormal, and a stranded wire quality early warning message is sent.

[0019] Step 2: Compare the crimping quality monitoring index CQMI with the threshold CQMI0 to obtain the early warning factor ζ(CQMI) for crimping quality monitoring. ζ(CQMI) = exp(CQMI - CQMI0). If ζ(CQMI) ≥ 1, it indicates that the crimping quality monitoring is good; otherwise, it indicates that the crimping quality monitoring is abnormal, and a crimping quality early warning message is sent.

[0020] Step 3: Compare the injection molding quality monitoring index IQMI with the threshold IQMI0 to obtain the early warning factor ζ(IQMI) for injection molding quality monitoring. ζ(IQMI) = exp(IQMI - IQMI0). If ζ(IQMI) ≥ 1, it indicates that the injection molding quality monitoring is good; otherwise, it indicates that the injection molding quality monitoring is abnormal, and an injection molding quality early warning message is sent.

[0021] Step 4: Compare the assembly quality monitoring index AQMI with the threshold AQMI0 to obtain the early warning factor ζ(AQMI) for assembly quality monitoring. ζ(AQMI) = exp(AQMI - AQMI0). If ζ(AQMI) ≥ 1, it indicates that the assembly quality monitoring is good; otherwise, it indicates that the assembly quality monitoring is abnormal, and an assembly quality early warning message is sent.

[0022] Preferably, the comprehensive analysis model in the intelligent monitoring and comprehensive analysis module for harness production is: WPCI = w1×ζ(SQMI) + w2×ζ(CQMI) + w3×ζ(IQMI) + w4×ζ(AQMI), where WPCI represents the comprehensive online monitoring index for harness production, w1 represents the weight of the early warning factor ζ(SQMI) for stranded wire quality monitoring, w2 represents the weight of the early warning factor ζ(CQMI) for crimping quality monitoring, w3 represents the weight of the early warning factor ζ(IQMI) for injection molding quality monitoring, and w4 represents the weight of the early warning factor ζ(AQMI) for assembly quality monitoring.

[0023] The technical effects and advantages of the present invention:

[0024] 1. Through artificial intelligence technology, the present invention adopts advanced sensor technology and data processing algorithms. The system can accurately measure and analyze various parameters, can more comprehensively monitor each link of the harness production line, improves the reliability and accuracy of the detection results, thereby improving the detection efficiency and helping to ensure the production quality of the harness.

[0025] 2. Through artificial intelligence technology, the present invention can optimize and adjust the production process according to the real-time monitored data. For example, by adjusting parameters such as stranding tension, crimping pressure, and injection molding temperature, it can avoid the occurrence of production accidents and the decline of product quality, thereby ensuring the product quality, reducing manual intervention, and lowering labor costs.

[0026] 3. Through artificial intelligence technology, the online monitoring system of the present invention can achieve automated detection. When detecting quality problems, the system will automatically send an alarm signal, timely discover and handle abnormal situations in the production process, avoid production stagnation and waste, prompt the operator to take measures to correct in time, improve the accuracy and efficiency of detection, and reduce the error and omission rate of manual detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0029] Please refer to Figure 1 As shown, the present invention provides an online monitoring system for wire harness production based on artificial intelligence, including a wire harness production monitoring data acquisition module, a wire harness production monitoring data analysis module, a wire harness production intelligent monitoring module, a wire harness production intelligent monitoring comprehensive analysis module, and a wire harness production intelligent monitoring human-computer interaction module.

[0030] Specifically in this embodiment, the wire harness production process studied by the present invention includes stranding, crimping, injection molding, and assembly. The production process of the wire harness requires multiple crimping processes and multiple injection molding processes. Among them, the stranding process is to twist multiple single wires together according to certain rules and directions to form a stranded wire or a bundle of wires; the crimping process is to press-fit the terminal or connector with the core wire in the wire harness through a crimping die to form a firm mechanical and electrical connection; the injection molding process is to inject thermoplastic plastic into a pre-designed mold cavity, and after cooling and solidifying, obtain plastic products that meet the required shapes. In wire harness production, the injection molding process is usually used to produce key components such as connectors, plugs, and sockets, as well as to achieve a tight combination of the cable with the connector or other components; the assembly process is to assemble components such as stranded wires, crimped parts, and injection molded parts together according to the design requirements to form a complete wire harness product.

[0031] The wire harness production monitoring data acquisition module is connected to the wire harness production monitoring data analysis module. The wire harness production intelligent monitoring module is respectively connected to the wire harness production monitoring data analysis module and the wire harness production intelligent monitoring comprehensive analysis module. The wire harness production intelligent monitoring human-computer interaction module is respectively connected to the wire harness production intelligent monitoring module and the wire harness production intelligent monitoring comprehensive analysis module.

[0032] Wire harness production monitoring data acquisition module: It is used to collect data during the wire harness production process, obtain various parameters of the wire harness production online monitoring, and transmit various parameters to the wire harness production monitoring data analysis module;

[0033] It should be specifically noted in this embodiment that the various parameters include administrator identity verification information, stranding quality monitoring parameters, crimping quality monitoring parameters, injection molding quality monitoring parameters, and assembly quality monitoring parameters. Among them, the stranding quality monitoring parameters include the length, pitch, tension, and diameter of the stranded wire; the crimping quality monitoring parameters include the peak pressure of crimping, the height of the crimped terminal, the pull-out force of the terminal after crimping, the conductor resistance before crimping, and the conductor resistance after crimping; the injection molding quality monitoring parameters include the actual pressure of the injection molding cycle, the size of the injection molded part, the weight of the injection molded part, the number of surface defects of the injection molded part, and the number of injection molded part measurement samples; the assembly quality monitoring parameters include the assembly step data set, the actual execution order of each assembly step, the number of steps with the actual order meeting the requirements, the total number of assembly steps, the number of terminals with the insertion depth meeting the requirements, the total number of terminals, the actual assembly time of each step, and the corresponding standard assembly time.

[0034] Wire harness production monitoring data analysis module: It is used to analyze various parameters obtained by the data acquisition module, obtain various monitoring indicators of the wire harness production online monitoring, and transmit various monitoring indicators to the wire harness production intelligent monitoring module. The wire harness production monitoring data analysis module includes a stranding quality monitoring unit, a crimping quality monitoring unit, an injection molding quality monitoring unit, and an assembly quality monitoring unit. The steps to obtain various monitoring indicators are as follows:

[0035] It should be specifically noted in this embodiment that the stranding quality monitoring unit is used to analyze the stranding quality monitoring parameters, the crimping quality monitoring unit is used to analyze the crimping quality monitoring parameters, the injection molding quality monitoring unit is used to analyze the injection molding quality monitoring parameters, and the assembly quality monitoring unit is used to analyze the assembly quality monitoring parameters.

[0036] Step 1: Stranding quality monitoring unit: First, the length L of the stranded wire is monitored in real time through a high-precision length sensor act , and compared with the designed length L des to obtain the stranding length deviation index LDR, LDR = |L act - L des | / Ldes ; Then, the pitch of the stranded wire is monitored in real time by a pitch sensor, n pitch values are recorded, the average value μ(p) and the standard deviation σ(p) of the pitch are obtained, and the pitch difference index PCI is obtained, PCI = σ(p) / μ(p); Secondly, the tension of the stranded wire during the processing is monitored in real time by a tension sensor, m tension values are recorded, the average value μ(f) and the standard deviation σ(p) of the tension are obtained, and the tension fluctuation index TFR is obtained, TFR = σ(p) / μ(f); Then, a laser rangefinder is used to record the diameter values of the stranded wire at k time points, and the diameter change rate DCR of the stranded wire is obtained. , j ∈ k, j is at least 1 and j ≤ k, R j and R j+1 represent the diameter values at the j-th and the (j + 1)-th time points respectively, t j and t j+1 represent the time values at the j-th and the (j + 1)-th time points respectively, and max() represents taking the maximum value; Finally, the stranded wire quality monitoring index SQMI is obtained through big data analysis technology. , DCR0 represents the maximum allowable diameter change rate;

[0037] It should be specifically noted in this embodiment that with the aid of advanced artificial intelligence algorithms, the system can perform high-precision analysis on the collected data. Based on the stranded wire quality monitoring index, the artificial intelligence system can automatically adjust the parameters of production equipment, such as tension control, pitch adjustment, etc., so as to optimize the production process and improve production efficiency.

[0038] Step 2: Crimping quality monitoring unit: First, the pressure during the crimping process is monitored in real time by a pressure sensor, and the pressure peak value P peak of each crimping is recorded, and the crimping pressure deviation index PDR is obtained, PDR = |P peak -P peak,0 | / P peak , P peak,0 represents the set crimping pressure peak value; Then, the height of the wire harness terminal after each crimping is measured by a measuring device, N height values are recorded, the average value μ(h) and the standard deviation σ(h) of the height of the wire harness crimping terminal are obtained, and the wire harness crimping terminal height difference index HCI is obtained, HCI = σ(h) / μ(h); Secondly, the pull-out force test is performed on the terminal after each crimping, N test results are recorded, the average value μ(F) and the standard deviation σ(F) of the terminal pull-out force test are obtained, and the terminal pull-out force difference index SCI is obtained, SCI = σ(F) / μ(F); Finally, through big data analysis technology, the crimping quality monitoring index CQMI is obtained. , F hou and F pre represent the conductor resistance after crimping and the conductor resistance before crimping respectively;

[0039] In this embodiment, it should be specifically noted that by real-time monitoring of key parameters such as the pressure peak deviation rate and height difference index during the crimping process, subtle changes in the crimping quality such as insufficient pressure and inconsistent height can be accurately captured, thereby avoiding the production of defective products; by monitoring the resistance values before and after crimping and calculating the resistance change rate, if the resistance change rate is too large, it may mean that the crimping is defective, directly affecting the electrical conductivity and stability of the electrical connection.

[0040] Step 3: Injection molding quality monitoring unit: First, the pressure sensor on the injection molding machine is used to record the pressure data in real time, obtain the actual pressure of L injection molding cycles, and obtain the injection molding pressure fluctuation index PFR, , l ∈ L, IP act,l represents the actual pressure of the l-th injection molding cycle, IP set represents the set pressure; then, through measuring tools (such as vernier calipers, coordinate measuring machines, etc.), according to the size production process parameters of the injection molded parts, the size measurement value D of the injection molded parts is obtained mea , and the injection molded part size deviation index DDC is obtained, , i ∈ M, D mea，i represents the size measurement value of the i-th injection molded part, D des represents the design size, and M is the number of injection molded part measurement samples; secondly, the injection molded parts are weighed by a weight sensor, the weight of each injection molded part is recorded, and the injection molded part weight difference index WCI is obtained, , W act，i represents the actual weight of the i-th injection molded part, W des represents the design weight; thirdly, through visual recognition technology, the number n_d of injection molded parts with surface defects is identified. Surface defects of injection molded parts include, for example, bubbles, cracks, shrinkage, etc., and the injection molded part surface defect index SDR is obtained, SDR = n_d / M; finally, the injection molding quality monitoring index IQMI is obtained through big data analysis technology, , add 0.01 to avoid the denominator being 0;

[0041] In this embodiment, it should be specifically noted that through artificial intelligence technology, the pressure fluctuation data is accurately analyzed, and then the potential faults of the injection molding machine are predicted; high-precision sensors and algorithms can be used to monitor the size deviation of injection molded parts in real time to ensure that the products meet the design requirements; by analyzing the weight difference data, the production process parameters can be optimized to improve the overall quality of the products; the AI vision detection system can automatically identify the defects on the surface of injection molded parts, such as scratches, pits, protrusions, foreign objects, etc., to ensure the consistency of product quality.

[0042] Step 4: Assembly quality monitoring unit: First, the assembly step data set ASD is obtained according to the wiring harness production and assembly design parameters, , K represents the number of steps required for the assembly process, AS JDenote the Jth assembly step. Sensors are set at each assembly step to record the actual execution sequence and compare it with the designed sequence to obtain the assembly sequence compliance index CSR. CSR = n_as / Tn_as, where n_as represents the number of steps with the actual sequence meeting the requirements, and Tn_as represents the total number of assembly steps. Then, the insertion depth of the wire harness terminals is measured by a depth sensor and compared with the designed parameters to obtain the wire harness terminal insertion depth compliance index TDR. TDR = n_t / Tn_t, where n_t represents the number of terminals with the insertion depth meeting the requirements, and Tn_t represents the total number of terminals. At the same time, a timer is set at each assembly step to record the actual assembly time T of each step act,J , and compare it with the corresponding standard assembly time T set,J to obtain the assembly time fluctuation index TTR, , and finally, through big data analysis technology, obtain the assembly quality monitoring index AQMI, ;

[0043] It should be specifically noted in this embodiment that the correct assembly sequence is crucial for ensuring the performance and safety of the wire harness. If the assembly sequence is incorrect, it may lead to internal short circuits, open circuits, or other faults in the wire harness; if the insertion depth of the terminals is insufficient, it may result in poor electrical connections, thereby affecting the performance of the entire wire harness; if the assembly time fluctuates greatly, it may mean that there are unstable factors during the assembly process, such as equipment failures and improper operations by employees.

[0044] Wire harness production intelligent monitoring module: used to compare each monitoring index obtained by the data analysis module with the threshold respectively, transmit the good comparison results to the wire harness production intelligent monitoring comprehensive analysis module, and send warning messages to the wire harness production intelligent monitoring human-computer interaction module according to the abnormal comparison results. The intelligent monitoring module includes the following steps:

[0045] Step 1: Compare the stranded wire quality monitoring index SQMI with the threshold SQMI0 to obtain the warning factor ζ(SQMI) of the stranded wire quality monitoring. ζ(SQMI) = exp(SQMI - SQMI0). If ζ(SQMI) ≥ 1, it means that the stranded wire quality monitoring is good; otherwise, it means that the stranded wire quality monitoring is abnormal, and a stranded wire quality warning message is sent;

[0046] Step 2: Compare the crimping quality monitoring index CQMI with the threshold CQMI0 to obtain the warning factor ζ(CQMI) of the crimping quality monitoring. ζ(CQMI) = exp(CQMI - CQMI0). If ζ(CQMI) ≥ 1, it means that the crimping quality monitoring is good; otherwise, it means that the crimping quality monitoring is abnormal, and a crimping quality warning message is sent;

[0047] Step 3: Compare the injection molding quality monitoring index IQMI with the threshold IQMI0 to obtain the warning factor ζ(IQMI) for injection molding quality monitoring. ζ(IQMI) = exp(IQMI - IQMI0). If ζ(IQMI) ≥ 1, it indicates that the injection molding quality monitoring is good; otherwise, it indicates that the injection molding quality monitoring is abnormal, and an injection molding quality warning message is issued.

[0048] Step 4: Compare the assembly quality monitoring index AQMI with the threshold AQMI0 to obtain the warning factor ζ(AQMI) for assembly quality monitoring. ζ(AQMI) = exp(AQMI - AQMI0). If ζ(AQMI) ≥ 1, it indicates that the assembly quality monitoring is good; otherwise, it indicates that the assembly quality monitoring is abnormal, and an assembly quality warning message is issued.

[0049] Intelligent monitoring comprehensive analysis module for wire harness production: It is used to comprehensively analyze the good comparison results obtained by the intelligent monitoring module to obtain the wire harness production online monitoring comprehensive index WPCI, and transmit it to the intelligent monitoring human-computer interaction module for wire harness production. The comprehensive analysis model is: WPCI = w1×ζ(SQMI) + w2×ζ(CQMI) + w3×ζ(IQMI) + w4×ζ(AQMI), where w1 represents the weight of the warning factor ζ(SQMI) for stranding quality monitoring, w2 represents the weight of the warning factor ζ(CQMI) for crimping quality monitoring, w3 represents the weight of the warning factor ζ(IQMI) for injection molding quality monitoring, and w4 represents the weight of the warning factor ζ(AQMI) for assembly quality monitoring. For example, w1 = 0.25, w2 = 0.25, w3 = 0.25, and w4 = 0.25.

[0050] Intelligent monitoring human-computer interaction module for wire harness production: It is used to receive the warning information from the intelligent monitoring module and the wire harness production online monitoring comprehensive index from the comprehensive analysis module for human-computer interaction. Among them, according to different warning information, it prompts the administrator to take corresponding management measures; according to the wire harness production online monitoring comprehensive index, it judges whether it is within the set allowable range. If so, it indicates that the wire harness production online monitoring is good; otherwise, it prompts the administrator to take measures in time, such as tension control, pitch adjustment, etc.; optimize the crimping parameters, improve the mold design, etc.; find the optimal combination of process parameters through machine learning, so as to optimize the production process, improve product quality and production efficiency.

[0051] Secondly: In the accompanying drawings of the disclosed embodiments of the present invention, only the structures related to the disclosed embodiments are involved. Other structures can refer to the general design. Without conflict, the same embodiment and different embodiments of the present invention can be combined with each other.

[0052] Finally, the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An artificial intelligence-based online monitoring system for wire harness production, characterized by: include: Wire harness production monitoring data acquisition module: used to collect data during the wire harness production process, obtain various parameters for online monitoring of wire harness production, and transmit various parameters to the wire harness production monitoring data analysis module; Wire harness production monitoring data analysis module: used to analyze the various parameters obtained by the data acquisition module, obtain the various monitoring indicators of the online monitoring of the wire harness production, and transmit the various monitoring indicators to the wire harness production intelligent monitoring module. The wire harness production monitoring data analysis module includes a stranding quality monitoring unit, a crimping quality monitoring unit, an injection molding quality monitoring unit, and an assembly quality monitoring unit; The strand quality monitoring unit in the harness production monitoring data analysis module first monitors the length L of the strand in real time through a high-precision length sensor. act and with the design length L des By comparison, the strand length deviation index LDR is obtained, LDR=|L act -L des | / L des ; Then, the pitch of the stranded wire is monitored in real time by the pitch sensor, n pitch values ​​are recorded, the average value μ(p) and standard deviation σ(p) of the pitch are obtained, and the pitch difference index PCI is obtained, PCI=σ(p) / μ(p); Secondly, the tension of the stranded wire during processing is monitored in real time by the tension sensor, m tension values ​​are recorded, the average value μ(f) and standard deviation σ(p) of the tension are obtained, and the tension fluctuation index TFR is obtained, TFR=σ(p) / μ(f); Then, the diameter values ​​of the stranded wire at k time points are recorded by the laser rangefinder, and the diameter change rate DCR of the stranded wire is obtained, , j∈k, j is at least 1 and j≤k, R j and R j+1 Represent the diameter values ​​at the jth and j+1th time points, t j and t j+1 They represent the time values ​​of the jth and j+1th time points respectively, and max() represents the maximum value. Finally, the strand quality monitoring index SQMI is obtained through big data analysis technology. , DCR0 represents the maximum allowed diameter change rate; The crimping quality monitoring unit in the harness production monitoring data analysis module first monitors the pressure during the crimping process in real time through a pressure sensor, and records the peak pressure P of each crimping. peak , get the crimping pressure deviation index PDR, PDR=|P peak -P peak,0 | / P peak , P peak,0 Indicates the set crimping pressure peak value; then the height of the wire harness terminal after each crimping is measured by measuring equipment, N height values ​​are recorded, the average value μ(h) and standard deviation σ(h) of the height of the wire harness crimped terminal are obtained, and the height difference index HCI of the wire harness crimped terminal is obtained, HCI=σ(h) / μ(h); secondly, the pull-out force test is performed on the terminal after each crimping, N test results are recorded, the average value μ(F) and standard deviation σ(F) of the terminal pull-out force test are obtained, and the terminal pull-out force difference index SCI is obtained, SCI=σ(F) / μ(F); finally, the crimping quality monitoring index CQMI is obtained through big data analysis technology, , F hou and F pre They represent the conductor resistance after crimping and the conductor resistance before crimping respectively; The injection molding quality monitoring unit in the wiring harness production monitoring data analysis module first records the pressure data in real time through the pressure sensor on the injection molding machine, obtains the actual pressure of L injection molding cycles, and obtains the injection molding pressure fluctuation index PFR. , l∈L, IP act,l Indicates the actual pressure of the first injection cycle, IP set Indicates the set pressure; then, through the measuring tool, according to the size production process parameters of the injection molded part, the size measurement value D of the injection molded part is obtained. mea , get the injection molded part dimensional deviation index DDC, , i∈M,D mea，i represents the dimensional measurement value of the i-th injection molded part, D des represents the design size, M is the number of injection molded parts measurement samples; secondly, the injection molded parts are weighed by a weight sensor, the weight of each injection molded part is recorded, and the injection molded part weight difference index WCI is obtained. , W act，i represents the actual weight of the i-th injection molded part, W des Indicates the design weight; again, the number of injection molded parts with surface defects n_d is identified through visual recognition technology, and the surface defect index SDR of injection molded parts is obtained, SDR=n_d / M; finally, the injection molding quality monitoring index IQMI is obtained through big data analysis technology, , add 0.01 to avoid the denominator being 0; The assembly quality monitoring unit in the wire harness production monitoring data analysis module first obtains the assembly step data set ASD according to the wire harness production assembly design parameters, , K represents the number of steps required for the assembly process, AS J represents the Jth assembly step, a sensor is set in each assembly step, the actual execution order is recorded, and compared with the design order, the assembly order qualification index CSR is obtained, CSR=n_as / Tn_as, n_as represents the number of steps that meet the requirements of the actual order, Tn_as represents the total number of assembly steps; then the wiring harness terminal insertion depth is measured by the depth sensor, and compared with the design parameters, the wiring harness terminal insertion depth qualification index TDR is obtained, TDR=n_t / Tn_t, n_t represents the number of terminals whose insertion depth meets the requirements, Tn_t represents the total number of terminals; at the same time, a timer is set in each assembly step to record the actual assembly time T of each step act,J , and the corresponding standard assembly time T set,J By comparison, we can get the assembly time fluctuation index TTR. Finally, the assembly quality monitoring index AQMI is obtained through big data analysis technology. ; Wire harness production intelligent monitoring module: used to compare the various monitoring indicators obtained by the data analysis module with the threshold value, transmit the good comparison results to the wire harness production intelligent monitoring comprehensive analysis module, and send warning information to the wire harness production intelligent monitoring human-computer interaction module according to the abnormal comparison results; Wire harness production intelligent monitoring comprehensive analysis module: used to comprehensively analyze the good comparison results obtained by the intelligent monitoring module, obtain the comprehensive indicators of online monitoring of wire harness production, and transmit them to the wire harness production intelligent monitoring human-computer interaction module; Wire harness production intelligent monitoring human-computer interaction module: used to receive the early warning information of the intelligent monitoring module and the comprehensive indicators of wire harness production online monitoring of the comprehensive analysis module for human-computer interaction.

2. According to claim 1, an artificial intelligence-based online monitoring system for wire harness production is characterized in that: The parameters in the wire harness production monitoring data acquisition module include administrator identity authentication information, stranded wire quality monitoring parameters, crimping quality monitoring parameters, injection molding quality monitoring parameters and assembly quality monitoring parameters, wherein the stranded wire quality monitoring parameters include the length, pitch, tension and diameter of the stranded wire; the crimping quality monitoring parameters include the crimping pressure peak, crimping terminal height, terminal pull-out force after crimping, conductor resistance before crimping and conductor resistance after crimping; the injection molding quality monitoring parameters include the actual pressure of the injection molding cycle, the size of the injection molded part, the weight of the injection molded part, the number of surface defects of the injection molded part and the number of injection molded part measurement samples; the assembly quality monitoring parameters include the assembly step data set, the actual execution order of each assembly step, the number of steps that meet the requirements of the actual order, the total number of assembly steps, the number of terminals whose insertion depth meets the requirements, the total number of terminals, the actual assembly time of each step and the corresponding standard assembly time.

3. The artificial intelligence-based online monitoring system for wire harness production according to claim 1 is characterized in that: The intelligent monitoring module in the wiring harness production intelligent monitoring module includes the following steps: Step 1: Compare the strand quality monitoring index SQMI with the threshold SQMI0 to obtain the early warning factor ζ(SQMI) for strand quality monitoring, ζ(SQMI)=exp(SQMI-SQMI0). If ζ(SQMI)≥1, it means that the strand quality monitoring is good; otherwise, it means that the strand quality monitoring is abnormal, and a strand quality early warning message is issued; Step 2: Compare the crimping quality monitoring index CQMI with the threshold CQMI0 to obtain the early warning factor ζ(CQMI) for crimping quality monitoring, ζ(CQMI)=exp(CQMI-CQMI0). If ζ(CQMI)≥1, it means that the crimping quality monitoring is good; otherwise, it means that the crimping quality monitoring is abnormal, and a crimping quality early warning information is issued; Step 3: Compare the injection molding quality monitoring index IQMI with the threshold value IQMI0 to obtain the early warning factor ζ(IQMI) for injection molding quality monitoring, ζ(IQMI)=exp(IQMI-IQMI0). If ζ(IQMI)≥1, it means that the injection molding quality monitoring is good; otherwise, it means that the injection molding quality monitoring is abnormal, and an injection molding quality early warning information is issued; Step 4: Compare the assembly quality monitoring index AQMI with the threshold AQMI0 to obtain the early warning factor ζ(AQMI) for assembly quality monitoring, ζ(AQMI)=exp(AQMI-AQMI0). If ζ(AQMI)≥1, it means that the assembly quality monitoring is good; otherwise, it means that the assembly quality monitoring is abnormal and an assembly quality early warning message is issued.

4. The artificial intelligence-based online monitoring system for wire harness production according to claim 1 is characterized in that: The comprehensive analysis model in the wire harness production intelligent monitoring comprehensive analysis module is: WPCI=w1×ζ(SQMI)+w2×ζ(CQMI)+w3×ζ(IQMI)+w4×ζ(AQMI), WPCI represents the comprehensive index of online monitoring of wire harness production, w1 represents the weight of the early warning factor ζ(SQMI) of stranded wire quality monitoring, w2 represents the weight of the early warning factor ζ(CQMI) of crimping quality monitoring, w3 represents the weight of the early warning factor ζ(IQMI) of injection molding quality monitoring, and w4 represents the weight of the early warning factor ζ(AQMI) of assembly quality monitoring.

Images