Production quality real-time monitoring and optimization management method based on AI technology

Abstract

The application discloses a production quality real-time monitoring and optimization management method based on an AI technology, and particularly relates to the field of artificial intelligence, and comprises S1: interval division, S2: production management feature extraction, S3: production planning, S4: production management rationality analysis, S5: correlation analysis, S6: production management prediction, and S7: production management optimization. The application carries out production management in the product production process, carries out production planning based on the production management features, further extracts production quality features, thereby carries out production management rationality analysis, and obtains a production management relationship formula by planning a production frequency, planning production equipment and planning a production speed, thereby providing data support for subsequent actual production frequency, actual production equipment and actual production speed adjustment, simultaneously generates a production quality optimization scheme, reduces product optimization management and adjustment time length in the production process, and evaluates the optimization effect according to the quality improvement amplitude, thereby ensuring that quality improvement is combined with production efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, more particularly, the present application relates to a production quality real-time monitoring and optimization management method based on AI technology. BACKGROUND

[0002] With the global manufacturing industry transforming into intelligentization, manufacturing industry needs to maintain competitiveness by improving production efficiency and reducing costs, at the same time, customers' requirements for product quality are increasingly stringent, especially in precision manufacturing, automobile, electronics and other industries, any small defect may cause serious consequences, enterprises need to realize whole-process quality control to ensure product consistency.

[0003] Traditional production mode relies on manual monitoring and post-detection, which is difficult to respond to quality problems in the production process in real time; through basic automatic detection tools, basic sensors such as temperature, pressure and flow are used to monitor key parameters in the production process in real time; by collecting data such as size, weight and temperature in the production process, statistical tools such as control charts are used to analyze process stability and identify abnormal fluctuations.

[0004] However, in actual use, there are still some shortcomings, first, traditional manual monitoring may lead to high scrap rate, increased rework cost, slow detection speed, and is easily affected by fatigue and subjective factors, and is difficult to cover high-speed production lines or complex products; traditional management method relies on single parameter adjustment to deal with quality risks, lacks the ability to optimize multiple variables, and in the process of manufacturing process conversion, factors affecting quality may include equipment pressure, material ratio, running temperature and other multiple dimensions, which will limit production efficiency, delay the construction period, and may also cause new quality problems due to unbalanced parameter matching;

[0005] Secondly, in the traditional monitoring system, equipment operation data, process parameters, environmental impact and other information are often scattered in independent modules, which is difficult to form a global perspective, resulting in adjustment measures focusing on local parameters, while ignoring the comprehensive evaluation of surrounding environmental carrying capacity and long-term equipment wear and tear, which will eventually lead to reduced adjustment effect and even greater hidden dangers due to insufficient system coordination. SUMMARY

[0006] Therefore, the embodiments of the present application provide a production quality real-time monitoring and optimization management method based on AI technology, which extracts production management features and production quality features in the production process, and makes fine and flexible adjustments, thereby maximizing the production efficiency and production quality, and predicting and optimizing the production management for the production quality, effectively solving the problems in the background art.

[0007] To achieve the above purpose, the present application provides the following technical scheme:

[0008] S1: interval division: according to the planning of the product production process, the product production process is divided into interval management, and production management is carried out before each interval management production;

[0009] S2: extraction of production management features: based on the interval management production management result, the interval management is divided into each process section, and the production management features of each process section are obtained;

[0010] S3: production planning: based on the production management features of the process section, production planning is carried out, specifically including planning production frequency, planning production equipment and planning production speed;

[0011] S4: production management rationality analysis: according to the production planning of each process section corresponding to each interval management, real-time production quality monitoring is carried out, thereby obtaining production quality features, and then based on the production quality features, production management rationality analysis is carried out;

[0012] S5: correlation analysis: correlation analysis is carried out on the planning production frequency, planning production equipment and planning production speed corresponding to each interval management and production management rationality, and a production management mathematical model is constructed, thereby obtaining a production management relationship;

[0013] S6: production management prediction: after the product production is completed, the actual production frequency, actual production equipment and actual production speed corresponding to each process section of each interval management are uploaded, and the actual production process rationality is predicted based on the production management relationship, thereby generating a production quality optimization scheme;

[0014] S7: production management optimization: based on the production quality optimization scheme, the production management terminal is issued to the production workshop, and the operating personnel adjusts the production frequency, production equipment and production speed according to the scheme, and the production quality is re-monitored after adjustment, and the optimization effect is evaluated.

[0015] The technical effects and advantages of the present application are:

[0016] 1、The present application carries out production management in the product production process, obtains production management features, carries out production planning based on the production management features, and then extracts production quality features, thereby carrying out production management rationality analysis, which is not limited to single parameter adjustment, on the one hand, it can maximize the production management rationality, on the other hand, it can avoid the production efficiency reduction and the delay of production period caused by single adjustment of production speed, which is beneficial to improve the production quality under the premise of ensuring normal production;

[0017] 2. After conducting a rationality analysis of production management, this invention performs a correlation analysis. By planning the production frequency, production equipment, and production speed, it obtains the production management relationship, which provides data support for subsequent adjustments to the actual production frequency, actual production equipment, and actual production speed. This helps improve the accuracy of rationality prediction and generates production quality optimization schemes, reducing the time required for product optimization management and adjustment during the production process, thereby improving production efficiency.

[0018] 3. This invention manages production quality based on a production quality optimization scheme and re-monitors production quality after adjustments, thereby ensuring production continuity to a certain extent. At the same time, it can better control the stability of product quality, reduce quality anomalies caused by production processes, and evaluate the optimization effect based on the improvement in quality, ensuring that production efficiency is taken into account while improving quality. Attached Figure Description

[0019] Fig. 1 This is a flowchart illustrating the implementation steps of the method of the present invention.

[0020] Fig. 2 This is a flowchart of the interval management and division process of the present invention.

[0021] Fig. 3 This is a flowchart of the correlation analysis of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] like Figs. 1-3 The AI-based real-time monitoring and optimization management method for production quality, as shown in this invention, includes the following specific steps:

[0024] S1: Interval Division: Based on the product production process planning, the product production process is divided into interval management, and production management is carried out before production in each interval management.

[0025] In this embodiment, it should be specifically explained that production management obtains production efficiency parameters, production resource parameters, and production equipment parameters by referring to the normal stages of product production and conducting product testing. Among them, production efficiency parameters include production time, process connection time, and completed quantity; production resource parameters include unit energy consumption and material loss rate; and production equipment parameters include equipment utilization rate, downtime, and maintenance cycle.

[0026] S2: Extracting production management features: dividing the interval management into each process segment based on the interval management result of production management, and obtaining the production management features of each process segment.

[0027] In this embodiment, it needs to be specifically pointed out that the division of the interval management into each process segment is as follows:

[0028] A1: Obtain the total duration of the interval management and the beginning and end processes, thereby evenly distributing the interval management to generate a number of monitoring points;

[0029] A2: Arrange the monitoring points in the order of the interval management from the starting process to the final process, and perform production detection in turn according to the arrangement order of the monitoring points, and record the production management features of each monitoring point, wherein the production management features include product performance parameters, appearance defect parameters and production equipment parameters;

[0030] A3: Calculate the similarity of the production management features of each adjacent monitoring point, and compare it with the preset similarity threshold value, if the similarity of the production management features of the adjacent monitoring points is greater than or equal to the similarity threshold value, then the two points are classified into the same process segment, and the latter monitoring point in the adjacent monitoring points and the next monitoring point form a new adjacent monitoring point to continue the similarity calculation, if the similarity of the production management features of the adjacent monitoring points is less than the similarity threshold value, then the former monitoring point in the adjacent monitoring points is taken as the cutoff point of the current process segment, and the latter monitoring point in the adjacent monitoring points and the next monitoring point form a new adjacent monitoring point to continue the similarity calculation;

[0031] A4: Calculate the similarity of all monitoring points, thereby dividing the interval management into each process segment.

[0032] It needs to be added that the similarity calculation can be standardized for each detection point corresponding to the production quality characteristics, eliminating the dimensional difference of different characteristic dimensions. Specifically, the maximum value and the minimum value of the original production quality characteristics in the interval management can be extracted respectively, and the difference between the original production quality characteristic value and the minimum value is compared with the difference between the maximum value and the minimum value, thereby obtaining the standardized production quality characteristic value. Each monitoring point forms a feature vector in the standardized production quality characteristic value, and then the production quality characteristic similarity of the feature vectors of adjacent monitoring points is calculated through the Euclidean distance. At this time, the smaller the Euclidean distance, the greater the similarity of the feature vectors, and the greater the Euclidean distance, the smaller the similarity of the two feature vectors. The similarity can be obtained by comparing the value 1 with the sum of 1 and the Euclidean distance, and then a similarity threshold is set. Exemplarily, the similarity threshold is 0.7. The production management feature similarity of adjacent monitoring points is compared with the set similarity threshold. If the production management feature similarity of adjacent monitoring points is greater than or equal to the similarity threshold, it means that the production management feature similarity meets the similarity threshold, otherwise it means that it does not meet the similarity threshold.

[0033] S3: Production planning: production planning based on process segment production management characteristics, specifically including planning production frequency, planning production equipment and planning production speed.

[0034] In this embodiment, it needs to be specifically explained that production planning can be obtained according to the historical product production records of the factory. Specifically, production management characteristics, production frequency, production cycle and other information are extracted from historical product production records for production planning.

[0035] It needs to be explained that the planned production frequency is based on the production task quantity and production time length in the production management characteristics of the process segment, combined with the corresponding production frequency extracted from the historical product production records; the planned production equipment is selected according to the processing requirements of the process segment and the equipment performance in the production management characteristics; the planned production speed is the production quantity expected to be completed in a single working day.

[0036] S4: Production management rationality analysis: real-time production quality monitoring is performed according to the production planning of each interval management corresponding to each process segment, thereby obtaining production quality characteristics, and then production management rationality analysis is performed based on the production quality characteristics.

[0037] In this embodiment, it needs to be specifically explained that the production quality characteristics include product performance, appearance defect and process quality.

[0038] Among them, the product performance is obtained according to the following process:

[0039] The product performance data corresponding to each process section is monitored in each interval, wherein the product performance data is determined according to the product type, for example, when the product type is an electrical product, the rated voltage, rated power and response speed thereof are monitored; when the product type is a mechanical product, the rotating speed, torque and accuracy thereof are monitored; and when the product type is a fluid equipment, the flow and lift thereof are monitored, and then the product performance stability is evaluated according to the monitoring results, so as to obtain the performance fluctuation coefficient.

[0040] It should be explained that the product performance stability is evaluated as follows: the performance sensing terminal is used to perform performance sensing in the production plan, so as to obtain the product performance data corresponding to each process section; the maximum value and the minimum value are extracted according to the product performance data, and then the difference between the maximum value and the minimum value is obtained, so as to obtain the performance fluctuation value of each product performance data, and then the performance fluctuation values corresponding to each product performance data are weighted and averaged, so as to obtain the performance fluctuation coefficient.

[0041] It should be explained that the performance sensing terminal is used to perform real-time monitoring and collection of the product performance data of each process section in the production plan, and is a basic equipment for evaluating the product performance stability. Different sensors are set to sense the performance of different product types, wherein the sensors include but are not limited to voltage sensors, current sensors, torque sensors, pressure sensors and the like. For example, the voltage sensor and the current sensor can collect the current and voltage of the electrical product, so as to calculate the power thereof; the torque sensor is pasted on the surface of the transmission shaft, and can convert the torque change of the mechanical product into resistance change; and the pressure sensor can be installed at the outlet end of the fluid equipment to measure the static pressure value.

[0042] The appearance defect obtaining process is as follows: the product contour is obtained by using the image sensing terminal, and the product appearance image is generated according to the product contour;

[0043] The appearance defect is extracted from the product appearance image, so as to obtain the appearance defect position and the appearance defect area corresponding to the position, and then the appearance defect area corresponding to the appearance defect position is accumulated and compared with the product surface area, so as to obtain the appearance defect degree;

[0044] The center points of each appearance defect position are selected, and then the center points of each appearance defect position are compared to calculate the distance between the center points;

[0045] The maximum distance and the minimum distance between the center points are extracted, and then the minimum distance and the maximum distance are compared to obtain the appearance defect repetition degree, wherein the smaller the appearance defect repetition degree is, the more the defects are concentrated in a small range and the larger the overall distribution range is, and the closer the ratio is to 1, the more uniform the defect distribution is, and the lower the repetition degree is. When there is only one appearance defect position, the appearance defect repetition degree is set to 1.

[0046] The appearance defect degree is multiplied by the appearance defect repeatability to obtain an appearance defect coefficient.

[0047] It should be explained that the image perception terminal is used to extract product appearance defect information through image acquisition and analysis technology, specifically through a high-resolution industrial camera and image recognition algorithm, which can automatically detect defects and avoid subjective bias of manual detection.

[0048] The process quality is obtained by specifically referring to the following process: using the quality perception terminal to analyze the composition of the raw material, recording the composition values of the raw material, and comparing the composition values with the standard composition values to obtain the raw material quality difference, which is specifically expressed as:

[0049] ,

[0050] Wherein Cl represents the raw material quality difference, C i represents the i-th composition value, C ib represents the standard composition value corresponding to the i-th composition, n represents the number of components in the raw material, i=1, 2, …, n;

[0051] The number of qualified products in the production planning in the historical production is extracted, and compared with the total number of products to obtain the product qualification rate, and then the product qualification rate of each process segment is averaged to obtain the product average qualification rate;

[0052] Based on the product qualification rate and the product average qualification rate, the process qualification fluctuation is calculated by the standard deviation formula, which is specifically expressed as:

[0053] ,

[0054] Wherein Ql represents the process qualification fluctuation, Q j and Q a respectively represent the product qualification rate and the product average qualification rate of the j-th process segment, m represents the number of process segments, j=1, 2, …, m;

[0055] Based on the raw material quality difference and the process qualification fluctuation, the process quality coefficient is obtained, which is specifically expressed as:

[0056] ,

[0057] Wherein L represents the process quality coefficient, Cl and Ql respectively represent the raw material quality difference and the process qualification fluctuation, and the smaller the raw material quality difference and the process qualification fluctuation, the larger the process quality coefficient.

[0058] It needs to be explained that the quality perception terminal is used for monitoring the quality of raw materials in the production process of products, which can be a spectrum analyzer, a hardness tester, and a size measuring instrument. The spectrum analyzer is used for detecting the composition of raw materials, and the hardness tester and the size measuring instrument can accurately measure the appearance of the product, so as to determine whether the product is qualified and record the qualified quantity.

[0059] It needs to be further explained that the production management rationality analysis is as follows: the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient of the production quality characteristics are extracted, so as to analyze the production management rationality and obtain the production management index, which is specifically expressed as:

[0060]

[0061] Wherein DQ represents the production management index of each interval management corresponding to each process segment, F, W and L represent the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient respectively;

[0062] The production management rationality index is calculated by using the production management index of each interval management corresponding to each process segment, which is specifically expressed as: Wherein D represents the production management rationality index, DQ j represents the production management index of the jth process segment, T j represents the required duration of the jth process segment, and Tl represents the total duration of interval management.

[0063] It needs to be explained that the numerical value 1 is converted into a positive index by subtracting the performance fluctuation coefficient and the appearance defect coefficient respectively. At this time, the greater the difference, the higher the production management rationality index, and the production management rationality is greatly affected by the performance fluctuation and the appearance defect, and is also affected by the quality in the production process. The production management rationality index calculated by the geometric mean method reflects the coordinated change of the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient. In this case, the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient of different production cycles or different products may differ greatly, and it is difficult to evaluate the rationality by comparing the absolute value of the index alone. The comprehensive index standardizes the three to a unified scale, which can be used for comparison of management level of the same product in different periods, and when the geometric mean value of the three is greater, it means that the production management is more reasonable.

[0064] S5: correlation analysis: extracting the correlation between the planning production frequency, the planning production equipment and the planning production speed corresponding to each interval management and the production management rationality, and constructing a production management mathematical model, so as to obtain the production management relationship.

[0065] In this embodiment, it needs to be specifically explained that the correlation analysis is as follows:

[0066] ​B1: Extracting the planning production frequency, planning production equipment and planning production speed corresponding to each process segment for each interval management;

[0067] B2: Calculating the correlation coefficient of the planning production frequency, planning production equipment and planning production speed of all process segments and the production management rationality index, thereby obtaining the correlation degree of the planning production frequency, planning production equipment and planning production speed corresponding to each process segment for each interval management and the production management rationality index;

[0068] B3: Setting a correlation threshold, for example, the correlation threshold is 0.6, when the absolute value of the correlation degree is greater than or equal to the correlation threshold, it is judged that the correlation between the planning production frequency and the production management rationality index is strong, otherwise it is judged that the correlation is weak, and similarly, the correlation between the planning production equipment and the planning production speed and the production management rationality index is judged strong or weak;

[0069] B4: Obtaining the production management relationship according to the correlation between the planning production frequency, planning production equipment and planning production speed of all process segments and the production management rationality index, which is specifically expressed as: D=f(Pt, Ft, Vt), where Pt, Ft and Vt represent the planning production frequency, planning production equipment and planning production speed respectively, and the planning production equipment can be quantified as production equipment performance data for fitting.

[0070] It should be understood that the correlation coefficient can be specifically Pearson correlation coefficient, when calculating the correlation coefficient of the planning production frequency, planning production equipment and planning production speed of the process segment and the production management rationality index, the planning production frequency, planning production equipment and production management rationality index of all process segments can be respectively regarded as four variables, including Pt j , Ft j , Vt j , DQ j , where j represents the number of process segments, thereby dividing the four variables into four arrays, and then calculating the covariance and standard deviation of the four arrays, and finally calculating the Pearson correlation coefficient using the covariance and standard deviation, the correlation coefficient obtained by calculation can obtain the correlation strength of the planning parameter and the management rationality, provide data basis for variable selection and relationship construction of the production process mathematical model in S5, and improve the accuracy of model prediction.

[0071] It needs to be further explained that the production management relationship formula takes the planned production frequency, the planned production equipment, and the planned production speed as input variables, and takes the comprehensive rationality index as an output variable, constructs a production process mathematical model, processes the nonlinear relationship between the input variables and the output variables according to the correlation degree, and combines a neural network model to improve the fitting accuracy of the model, to obtain the production management relationship formula. Exemplarily, when the correlation degree of the planned production speed and the production management rationality index is high, the nonlinear relationship is retained, which is specifically represented as: D = ω1 × Pt + ω2 × Ft + ω3 × Vt + ω4 × Vt 2 wherein Pt, Ft, and Vt represent the planned production frequency, the planned production equipment, and the planned production speed, respectively, ω1, ω2, and ω3 represent the regression coefficients corresponding to the planned production frequency, the planned production equipment, and the planned production speed, respectively, and ω4 represents the quadratic term coefficient corresponding to the planned production speed. The regression coefficients are fitted by using the least square method, so that the error sum of squares of the predicted value and the actual production management rationality index is minimized.

[0072] S6: Production management prediction: After the product production is completed, the actual production frequency, the actual production equipment, and the actual production speed corresponding to each process section of each interval management are uploaded, and the actual production process rationality is predicted based on the production management relationship formula, thereby generating a production quality optimization scheme.

[0073] In this embodiment, it needs to be specifically explained that the production quality optimization scheme is generated by referring to the following process:

[0074] After the product production is completed, the actual production frequency, the actual production equipment, and the actual production speed corresponding to each process section of each interval management are uploaded through the production management terminal;

[0075] The actual production frequency, the actual production equipment, and the actual production speed are imported into the production management relationship formula, thereby obtaining the actual production management rationality index;

[0076] A rationality threshold is set. If the actual production management rationality index is greater than or equal to the rationality threshold, it is predicted that the actual production management is reasonable. If the actual production management rationality index is less than the rationality threshold, it is predicted that the actual production management is unreasonable, and the production management needs to be optimized to generate a production quality optimization scheme.

[0077] It needs to be explained that the production quality optimization scheme is based on the actual production management rationality prediction results. Specifically, the actual production management rationality index is compared with the rationality threshold value. The rationality threshold value can be screened from the historical product production records due to the production management irrationality leading to substandard product quality or decreased production efficiency. Then the specific rationality threshold value is extracted from the screened rationality index, including but not limited to the rationality index critical value corresponding to different product types and different process segments. If the actual production management rationality index is lower than the rationality threshold value, it is identified that the production management is not reasonable. When the production management is not reasonable or there is a quality risk, the production speed is reduced, and the production frequency is reduced. At the same time, the production qualified rate before adjustment is compared. If the production qualified rate is reduced, the production equipment is calibrated for precision and the maintenance cycle is shortened. If the production qualified rate is increased, the production speed is locked.

[0078] It needs to be added that reducing the production speed is not limited to the impact of unreasonable production management or quality risk. When the production equipment fails, the production speed also needs to be reduced.

[0079] S7: Production management optimization: based on the production quality optimization scheme, the production management terminal is issued to the production workshop, and the operating personnel adjusts the production frequency, production equipment and production speed according to the scheme, and re-monitors the production quality after adjustment to evaluate the optimization effect.

[0080] In this embodiment, it needs to be specifically explained that the evaluation of the optimization effect is as follows:

[0081] According to the production quality optimization scheme, the specific adjustment instructions of each process segment are disassembled, including the target value of the production frequency, the replacement model or maintenance requirement of the production equipment, and the lock value of the production speed. Through the production management terminal, the corresponding operating personnel of the process segment is synchronously pushed, and the operating parameters of the production equipment are updated in turn according to the adjustment requirements. The device running state and parameter change record are recorded in real time during the adjustment process;

[0082] After the adjustment is completed, the production quality is re-monitored, and the adjusted production management characteristics, production quality characteristics and production quality rationality index are collected;

[0083] The adjusted production quality rationality index is compared with the production quality rationality index before adjustment, the quality improvement amplitude is calculated, if the improvement amplitude reaches the preset target value, it is determined that the optimization effect is effective; if it does not reach the target value, the abnormal data is fed back to S5, the production process mathematical model is recalibrated, a new production quality optimization scheme is generated, and the next round of adjustment cycle is entered. Exemplarily, the preset target value is 5%.

[0084] It needs to be explained that the production management terminal is used for receiving the production quality optimization scheme, and analyzing the monitoring data and controlling the related parameters in the production process, ensuring that the optimization scheme is controllable from the formulation to the landing, and improving the efficiency and accuracy of the production quality adjustment.

[0085] In the effect evaluation process, the production management characteristics and production quality characteristic changes in the adjustment process are recorded, thereby obtaining the production quality rationality index, and then the optimization effect is evaluated according to the quality improvement amplitude, so that the quality improvement is ensured while the production efficiency is considered.

[0086] Secondly: in the embodiment of the present application, only the structures involved in the embodiment of the present application are involved, other structures can refer to the usual design, and in the case of no conflict, the same embodiment and different embodiments of the present application can be combined with each other;

[0087] Finally: the above only describes the preferred embodiments of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for real-time monitoring and optimization management of production quality based on AI technology, characterized in that, Comprise: S1: interval division: according to the planning of product production process, the product production process is divided into interval management, and the production management is carried out before each interval management production; S2: extraction of production management characteristics: based on the interval management production management result, the interval management is divided into each process section, and the production management characteristics of each process section are obtained; S3: production planning: based on the production management characteristics of the process section, the production planning is carried out, which specifically includes planning production frequency, planning production equipment and planning production speed; S4: production management rationality analysis: according to the production planning of each process section corresponding to each interval management, real-time production quality monitoring is carried out, so as to obtain production quality characteristics, and then based on the production quality characteristics, production management rationality analysis is carried out; The production quality characteristics include product performance, appearance defect and process quality, wherein the product performance is obtained by referring to the following process: The product performance data of each interval management corresponding to each process section is monitored, wherein the product performance data is determined according to the product type; then the product performance stability is evaluated according to the monitoring result, so as to obtain the performance fluctuation coefficient; The process quality is obtained by referring to the following process: the composition of raw materials is analyzed by using quality sensing terminal, so as to record each component value of raw materials, and each component value is compared with standard component value to obtain raw material quality difference degree; The qualified product quantity in the production planning in the history production is extracted, and compared with the total product quantity to obtain the product qualified rate, and then the product qualified rate of each process section is averaged to obtain the product average qualified rate; Based on the product qualified rate and the product average qualified rate, the process qualified fluctuation degree is calculated by the standard deviation formula; Based on the raw material quality difference degree and the process qualified fluctuation degree, the process quality coefficient is obtained; S5: correlation analysis: the correlation between the planning production frequency, the planning production equipment and the planning production speed corresponding to each interval management and the production management rationality is analyzed, and a production management mathematical model is constructed, so as to obtain the production management relationship; S6: production management prediction: after the product production is completed, the actual production frequency, the actual production equipment and the actual production speed corresponding to each process section of each interval management are uploaded, and the actual production process rationality is predicted based on the production management relationship, so as to generate a production quality optimization scheme; S7: production management optimization: based on the production quality optimization scheme, the production management terminal is delivered to the production workshop, and the operation personnel adjusts the production frequency, the production equipment and the production speed according to the scheme, and the production quality is monitored again after adjustment to evaluate the optimization effect. 2.The AI technology-based production quality real-time monitoring and optimization management method of claim 1, wherein: The interval management is divided into process sections by referring to the following process: A1: the total time length and the beginning and end process of interval management are obtained, so as to generate a plurality of monitoring points by evenly distributing interval management; A2: the monitoring points are arranged in order from the starting process to the final process according to the interval management, and the production detection is carried out in turn according to the arrangement order of the monitoring points, and the production management characteristics of each monitoring point are recorded, wherein the production management characteristics include product performance parameters, appearance defect parameters and production equipment parameters; A3: Similarity calculation is performed on the production management features of each adjacent monitoring point, and compared with a preset similarity threshold. If the similarity of the production management features of the adjacent monitoring points is greater than or equal to the similarity threshold, the two points are classified into the same process segment, and the latter monitoring point in the adjacent monitoring points and the next monitoring point are combined to form a new adjacent monitoring point for similarity calculation. If the similarity of the production management features of the adjacent monitoring points is less than the similarity threshold, the former monitoring point in the adjacent monitoring points is taken as the end point of the current process segment, and the latter monitoring point in the adjacent monitoring points and the next monitoring point are combined to form a new adjacent monitoring point pair for similarity calculation; A4: Similarity calculation is performed on all monitoring points, thereby dividing the interval management into process segments. 3.The AI technology-based production quality real-time monitoring and optimization management method of claim 1, wherein: The appearance defect is obtained by the following process: obtaining the product contour by using the image sensing terminal, generating the product appearance image according to the product contour; The appearance defect is extracted from the product appearance image, thereby obtaining the appearance defect position and the appearance defect area corresponding to the position, and then the appearance defect area corresponding to the appearance defect position is accumulated and compared with the product surface area to obtain the appearance defect degree; The center points of each appearance defect position are selected, and then the center points of each appearance defect position are compared to calculate the distance between the center points; The maximum distance and the minimum distance between the center points are extracted, and then the minimum distance is compared with the maximum distance to obtain the appearance defect repetition degree; The appearance defect degree and the appearance defect repetition degree are multiplied to obtain the appearance defect coefficient. 4.The AI technology-based production quality real-time monitoring and optimization management method of claim 1, wherein: The production management rationality analysis is as follows: the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient of the production quality feature are extracted, thereby analyzing the production management rationality to obtain the production management index, which is specifically represented as: ， Wherein DQ represents the production management index of each interval management corresponding to each process segment, F, W and L represent the performance fluctuation coefficient, the appearance defect coefficient and the process quality coefficient respectively; The production management rationality index is calculated by using the production management index of each interval management corresponding to each process segment, and is specifically represented as: ， wherein D represents a production management rationality index, DQ j represents the production management index corresponding to the jth flow section, T j represents the required duration of the jth flow section, Tl represents the total duration of interval management, and m represents the number of flow sections. 5.The AI technology-based production quality real-time monitoring and optimization management method according to claim 4, characterized in that: The correlation analysis is as follows: B1: The planned production frequency, the planned production equipment and the planned production speed of each interval management corresponding to each process segment are extracted; B2: The planned production frequency, the planned production equipment and the planned production speed of all process segments and the production management rationality index are subjected to correlation coefficient calculation, thereby obtaining the correlation degree of the planned production frequency, the planned production equipment and the planned production speed of each interval management corresponding to each process segment and the production management rationality index; B3: A correlation threshold is set. When the absolute value of the correlation degree is greater than or equal to the correlation threshold, it is judged that the correlation between the planned production frequency and the production management rationality index is strong, otherwise it is judged that the correlation is weak. Similarly, the correlation between the planned production equipment, the planned production speed and the production management rationality index is judged. B4: The production management relationship is obtained according to the correlation between the planned production frequency, the planned production equipment, the planned production speed and the production management rationality index of all process sections, and is specifically expressed as: D=f(Pt, Ft, Vt), wherein Pt, Ft and Vt represent the planned production frequency, the planned production equipment and the planned production speed respectively. 6.The AI technology-based production quality real-time monitoring and optimization management method according to claim 1, characterized in that: The production quality optimization scheme is generated by referring to the following process: After the product production is completed, the actual production frequency, the actual production equipment and the actual production speed corresponding to each process section are uploaded by the production management terminal; The actual production frequency, the actual production equipment and the actual production speed are introduced into the production management relationship, and thus the actual production management rationality index is obtained; A rationality threshold is set, if the actual production management rationality index is greater than or equal to the rationality threshold, it is predicted that the actual production management is reasonable, if the actual production management rationality index is less than the rationality threshold, it is predicted that the actual production management is unreasonable, and then the production management needs to be optimized, and the production quality optimization scheme is generated.

Images