Optimized quality detection system for rubber production

Abstract

The invention relates to an optimized quality detection system for rubber production, which comprises a multi-source parameter synchronous acquisition module, the multi-source parameter synchronous acquisition module is in data transmission connection with a dynamic weighting quality evaluation module, and the dynamic weighting quality evaluation module is in data transmission connection with a two-dimensional abnormity traceability module; the two-dimensional anomaly traceability module is in data transmission connection with the quantitative process regulation and control module; the multi-source parameter synchronous acquisition module acquires three types of core parameters of the whole rubber production process in real time through a distributed sensing array; and the dynamic weighting quality evaluation module constructs a comprehensive scoring model fused with time sequence characteristics based on the collected parameters. According to the structure provided by the embodiment of the invention, detection is comprehensively innovated, physical, chemical and technological parameters and time sequence characteristics are fused for the first time to construct the scoring model, and compared with traditional single parameter detection, the quality evaluation accuracy is improved from 65% to 98% or above.

Description

Technical Field

[0001] This application relates to the field of rubber production technology, and in particular to an optimized quality inspection system for rubber production. Background Technology

[0002] Existing rubber production quality inspection methods suffer from three major pain points: First, the inspection dimensions are isolated, mostly focusing on physical indicators such as tensile strength and Mooney viscosity of the vulcanized finished product, without considering process parameters such as mixing temperature and vulcanization pressure, or chemical parameters such as crosslinking density and ash content. This leads to a one-sided quality assessment and an inability to identify hidden defects such as "parameter compliance but finished product non-compliance" in advance. Second, anomaly tracing lacks quantitative basis. When non-conforming products are detected, manual experience is used to check parameter deviations one by one, which is not only inefficient (each tracing takes more than 30 minutes) but also prone to misjudging core anomaly factors. Third, process adjustment lacks precise guidance. Manual parameter adjustment often relies on trial and error, which easily leads to over-adjustment or under-adjustment, resulting in a high non-conforming product rate. Therefore, we propose an optimized quality inspection system for rubber production. Summary of the Invention

[0003] This application provides an optimized quality inspection system for rubber production to solve the problems mentioned above.

[0004] This application provides an optimized quality inspection system for rubber production, comprising:

[0005] A multi-source parameter synchronous acquisition module, wherein the data transmission of the multi-source parameter synchronous acquisition module is connected to a dynamic weighted quality assessment module, and the data transmission of the dynamic weighted quality assessment module is connected to a two-dimensional anomaly tracing module, wherein the data transmission of the two-dimensional anomaly tracing module is connected to a quantitative process control module;

[0006] The multi-source parameter synchronous acquisition module collects three types of core parameters in real time throughout the entire rubber production process through a distributed sensor array.

[0007] The dynamic weighted quality assessment module constructs a comprehensive scoring model that integrates time-series features based on the collected parameters;

[0008] The dual-dimensional anomaly tracing module includes calculating the sensitivity of parameter impact, the quality contribution rate of calculated parameters, and screening core anomaly parameters.

[0009] The quantitative process control module generates adjustment instructions for core abnormal parameters.

[0010] Preferably, the three types of core parameters include physical parameters, chemical parameters, and process parameters;

[0011] The physical parameters include the Mooney viscosity (M) of raw rubber and the tensile strength (σ) of vulcanized rubber. t and elongation at break εβ The chemical parameters include the original crosslinking density ρ0 and the ash content A, wherein the original crosslinking density ρ0 is calibrated to obtain the calibrated crosslinking density ρ after relaxation time calibration. The process parameters include the mixing temperature T, the mixing time t, the vulcanization pressure P, and the vulcanization temperature T. s .

[0012] Preferably, the formula for the calibrated crosslinking density ρ is:

[0013] ρ=ρ0*(T 20 / T2)*[1+0.003*(T m -25)]

[0014] In the formula: T2 represents the actual detected transverse relaxation time, T 20 T represents the transverse relaxation time corresponding to the standard crosslinking density. m This indicates the ambient temperature being detected.

[0015] Preferably, the scoring algorithm formula of the comprehensive scoring model is:

[0016] S=α*(k1M+k2σ t +k3ε β )+β*(k4ρ-k5A)+γ*(k6 / T+k7 / t+k8P+k9 / T s )+δ*S t-1

[0017] In the formula: S is the current overall quality score, and the overall quality score is 0-100 points, with a score greater than or equal to 85 points considered qualified. t-1 The score for the previous acquisition time is given by α, β, γ, and δ, which are the weight coefficients of physical parameters, chemical parameters, process parameters, and time-series correlation terms, respectively, satisfying α+β+γ+δ=1. By improving the analytic hierarchy process and introducing entropy weight correction, the initial values ​​are determined as α=0.4, β=0.25, γ=0.25, and δ=0.1. k1-k9 are parameter correction coefficients with initial values ​​of k1=0.09, k2=2.2, k3=0.11, k4=1600, k5=1.9, k6=380, k7=1.3, k8=0.85, and k9=420.

[0018] Preferably, the sensitivity of the calculated parameters, the contribution rate of the calculated parameters to the quality of the parameters, and the screening of core abnormal parameters are activated when the overall quality score of the comprehensive scoring model is less than 85 points.

[0019] The calculation parameters affect the sensitivity Sens(X). i ):

[0020]

[0021] In the formula: For S with respect to parameter X i The partial derivatives;

[0022] The calculated parameter is the quality contribution rate Cont(X). i ):

[0023] Cont(X i ) = (calculated values ​​of corresponding terms α / β / γ / δ) / S

[0024] The specific method for filtering core anomaly parameters is to use Sens(X) i )>0.18 and Cont(X) i Parameters greater than 0.05 are used as the core anomaly tracing objects.

[0025] Preferably, the adjustment amount algorithm formula of the quantitative process control module is:

[0026]

[0027] In the formula: ΔS tar X represents the difference between the target score and the current score. i0 This represents the standard value of the parameter, [1-0.02*|(X i -X i0 ) / X i0 |] represents the deviation correction factor.

[0028] Preferably, the ash content A is collected by the ignition-weighing method, and then corrected for humidity using the following formula:

[0029] A = A0 * (1 - 0.01 * H)

[0030] In the formula: A0 is the original ash content detection value, and H is the sample moisture content.

[0031] Preferably, the entropy weight correction algorithm of the improved analytic hierarchy process is as follows:

[0032] ω' i =ω i *(1-e i ) / ∑[ω i *(1-e i )]

[0033] In the formula: ω i e represents the initial weights of the analytic hierarchy process. i This represents the information entropy of the parameter.

[0034] Preferably, the adjustment instruction, after being generated, needs to be verified by parameter constraints, and the verification formula is:

[0035] X ine =X i+ΔX i

[0036] If X i x e If the parameters exceed the preset reasonable range, automatic correction will be triggered.

[0037] ΔX i =(X im -X i )*(ΔS tar / ΔS max )

[0038] In the formula: X im For parameter X i The upper limit of ΔS max To maximize the potential score improvement (100-S), ΔS tar This represents the difference between the target score and the current score.

[0039] Preferably, the optimized quality inspection system for rubber production further includes a quality prediction module, which predicts the next score based on the collected S-value using an ARIMA model. The prediction algorithm formula is as follows:

[0040]

[0041] In the formula: St+1 is the next predicted score, φ1 to φ p These are the autoregressive coefficients, θ1 to θ q ε is the moving average coefficient. t Let be the error term at the current time, and p and q be the autoregression order and the moving average order, respectively.

[0042] The technical solutions provided in this application have the following advantages compared with the prior art:

[0043] The structure provided in this application offers comprehensive innovation in detection: It is the first to integrate physical, chemical, and process parameters with time-series features to construct a scoring model. Compared to traditional single-parameter detection, the accuracy of quality assessment is increased from 65% to over 98%, enabling the identification of hidden quality defects. It also improves traceability efficiency: through a dual-dimensional traceability algorithm, the anomaly traceability time is shortened from over 30 minutes to within 1 minute, with a 95% accuracy rate in identifying core anomaly parameters, avoiding human error. Furthermore, it optimizes control precision: the constrained quantitative control formula increases the first-pass yield of process adjustments from 60% to 92%, reducing single-batch production costs by 15%-20%. Finally, it provides early risk warnings: the quality prediction step can provide early warnings of quality risks 1-2 collection cycles in advance, allowing operators sufficient adjustment time and further reducing the generation of non-conforming products. Attached Figure Description

[0044] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of the overall principle and structure of the present invention. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Various embodiments of this application may exist in the form of a range. It should be understood that the description in the form of a range is merely for convenience and brevity and should not be construed as a rigid limitation on the scope of this application. Therefore, it should be considered that the range description has specifically disclosed all possible sub-ranges and single numerical values ​​within that range. For example, it should be considered that the range description from 1 to 6 has specifically disclosed sub-ranges, such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6, etc., and single numbers within the range, such as 1, 2, 3, 4, 5, and 6, regardless of the range. In addition, whenever a numerical range is indicated in this application, it means including any referenced number (fraction or integer) within the indicated range. Unless otherwise specified, all raw materials, reagents, instruments, and equipment used in this application can be purchased commercially or prepared using existing equipment.

[0049] In this application, unless otherwise stated, directional terms such as "upper" and "lower" specifically refer to the drawing directions in the accompanying drawings. Furthermore, in this application, the terms "comprising," "including," etc., mean "including but not limited to." In this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. In this application, "and / or" describes the relationship between related objects, indicating that three relationships may exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. A and B can be singular or plural. In this application, "at least one" means one or more, and "more than one" means two or more. "At least one," "at least one of the following," or similar expressions refer to any combination of these items, including any combination of a single item or a plural item. For example, "at least one of a, b, or c", or "at least one of a, b, and c", can both mean: a, b, c, ab, i.e., a and b, ac, bc, or abc, where a, b, and c can be a single or multiple.

[0050] like Figure 1 As shown in the figure, this application provides an optimized quality inspection system for rubber production, comprising:

[0051] A multi-source parameter synchronous acquisition module, wherein the data transmission of the multi-source parameter synchronous acquisition module is connected to a dynamic weighted quality assessment module, and the data transmission of the dynamic weighted quality assessment module is connected to a two-dimensional anomaly tracing module, wherein the data transmission of the two-dimensional anomaly tracing module is connected to a quantitative process control module;

[0052] The multi-source parameter synchronous acquisition module collects three types of core parameters in real time throughout the entire rubber production process through a distributed sensor array.

[0053] The dynamic weighted quality assessment module constructs a comprehensive scoring model that integrates time-series features based on the collected parameters;

[0054] The dual-dimensional anomaly tracing module includes calculating the sensitivity of parameter impact, the quality contribution rate of calculated parameters, and screening core anomaly parameters.

[0055] The quantitative process control module generates adjustment instructions for core abnormal parameters.

[0056] The three types of core parameters include physical parameters, chemical parameters, and process parameters;

[0057] The physical parameters include the Mooney viscosity (M) of raw rubber and the tensile strength (σ) of vulcanized rubber. t and elongation at break ε βThe chemical parameters include the original crosslinking density ρ0 and the ash content A, wherein the original crosslinking density ρ0 is calibrated to obtain the calibrated crosslinking density ρ after relaxation time calibration. The process parameters include the mixing temperature T, the mixing time t, the vulcanization pressure P, and the vulcanization temperature T. s .

[0058] The formula for the calibrated crosslinking density ρ is:

[0059] ρ=ρ0*(T 20 / T2)*[1+0.003*(T m -25)]

[0060] In the formula: T2 represents the actual detected transverse relaxation time, T 20 T represents the transverse relaxation time corresponding to the standard crosslinking density. m This indicates the ambient temperature being detected.

[0061] The scoring algorithm formula for the comprehensive scoring model is as follows:

[0062] S=α*(k1M+k2σ t +k3ε β )+β*(k4ρ-k5A)+γ*(k6 / T+k7 / t+k8P+k9 / T s )+ε*S t-1

[0063] In the formula: S is the current overall quality score, and the overall quality score is 0-100 points, with a score greater than or equal to 85 points considered qualified. t-1 The score for the previous acquisition time is given by α, β, γ, and δ, which are the weight coefficients of physical parameters, chemical parameters, process parameters, and time-series correlation terms, respectively, satisfying α+β+γ+δ=1. By improving the analytic hierarchy process and introducing entropy weight correction, the initial values ​​are determined as α=0.4, β=0.25, γ=0.25, and δ=0.1. k1-k9 are parameter correction coefficients with initial values ​​of k1=0.09, k2=2.2, k3=0.11, k4=1600, k5=1.9, k6=380, k7=1.3, k8=0.85, and k9=420.

[0064] The calculation parameters affect sensitivity, the contribution rate of the calculation parameters to quality, and the screening of core abnormal parameters are activated when the overall quality score of the comprehensive scoring model is less than 85 points.

[0065] The calculation parameters affect the sensitivity Sens(X). i ):

[0066]

[0067] In the formula: For S with respect to parameter X iThe partial derivatives;

[0068] The calculated parameter is the quality contribution rate Cont(X). i ):

[0069] Cont(X i ) = (calculated values ​​of corresponding terms α / β / γ / δ) / S

[0070] The specific method for filtering core anomaly parameters is to use Sens(X) i )>0.18 and Cont(X) i Parameters greater than 0.05 are used as the core anomaly tracing objects.

[0071] The algorithm formula for the adjustment amount of the quantitative process control module is as follows:

[0072]

[0073] In the formula: ΔS tar X represents the difference between the target score and the current score. i0 This represents the standard value of the parameter, [1-0.02*|(X i -X i0 ) / X i0 |] represents the deviation correction factor.

[0074] The ash content A was collected by the ignition-weighing method, and then corrected for humidity using the following formula:

[0075] A = A0 * (1 - 0.01 * H)

[0076] In the formula: A0 is the original ash content detection value, and H is the sample moisture content.

[0077] The entropy weight correction algorithm of the improved analytic hierarchy process is as follows:

[0078] ω' i =ω i *(1-e i ) / ∑[ω i *(1-e i )]

[0079] In the formula: ω i e represents the initial weights of the analytic hierarchy process. i This represents the information entropy of the parameter.

[0080] After the adjustment instruction is generated, it needs to be verified by parameter constraints. The verification formula is as follows:

[0081] X ine =X i +ΔX i

[0082] If Xine If the parameters exceed the preset reasonable range, automatic correction will be triggered.

[0083] ΔX i =(X im -X i )*(ΔS tar / ΔS max )

[0084] In the formula: X im For parameter X i The upper limit of ΔS max To maximize the potential score improvement (100-S), ΔS tar This represents the difference between the target score and the current score.

[0085] The optimized quality inspection system for rubber production also includes a quality prediction module, which predicts the next score based on the collected S-value using an ARIMA model. The prediction algorithm formula is as follows:

[0086]

[0087] In the formula: St+1 is the next predicted score, φ1 to φ p These are the autoregressive coefficients, θ1 to θ q ε is the moving average coefficient. t Let be the error term at the current time, and p and q be the autoregression order and the moving average order, respectively.

[0088] Example

[0089] Taking a batch of styrene-butadiene rubber (SBR 1502) produced by a tire factory as an example, the actual application process of this method is explained in detail:

[0090] 1. Multi-source parameter synchronous acquisition module

[0091] During this batch of production, the multi-source parameter synchronous acquisition module acquired parameters at a frequency of once every 20 seconds. The raw data and calibrated data acquired at a certain moment are as follows:

[0092] Physical parameters: M = 58, σ t =23.5MPa, ε β =560%;

[0093] Chemical parameter: ρ0 = 1.15 × 10 -4 mol / cm 3 (T2=86ms, T m =28℃), A0=0.9% (H=0.5%);

[0094] Crosslinking density calibration: ρ = 1.15 × 10 -4×(80 / 86)×[1+0.003×(28-25)]≈1.08×10 - 4 mol / cm 3 ;

[0095] Ash content correction: A = 0.9% × (1 - 0.01 × 0.5) = 0.8955%;

[0096] Process parameters: T = 112℃, t = 7.5min, P = 14MPa, T s =155℃;

[0097] Previous rating: S t-1 =86 points.

[0098] 2. Dynamic Weighted Quality Assessment Module

[0099] Substitute the above parameters into the scoring model to calculate the current quality score:

[0100] S=0.4×(0.09×58+2.2×23.5+0.11×560)+0.25×(1600×1.08×10 -4 -1.9×0.8955)+0.25×(380 / 112+1.3 / 7.5+0.85×14+420 / 155)+0.1×86

[0101] Physical parameter: 0.4 × (5.22 + 51.7 + 61.6) = 0.4 × 118.52 = 47.408;

[0102] Chemical parameter: 0.25×(0.1728-1.7014)=0.25×(-1.5286)=-0.3822;

[0103] Process parameter item: 0.25×(3.3929+0.1733+11.9+2.7097)=0.25×18.1759≈4.544;

[0104] Timing-related terms: 0.1 × 86 = 8.6;

[0105] The final score is S = 47.408 - 0.3822 + 4.544 + 8.6 ≈ 59.17 (far below the passing threshold of 85 points, so anomaly tracing is initiated).

[0106] 3. Two-dimensional anomaly tracing module

[0107] Calculate the sensitivity of each parameter, Sens(X). i ) and contribution rate Cont(X i The calculation results for some key parameters are as follows:

[0108] Mixing temperature T(X) i =112℃):

[0109]

[0110] Sens(T) = |-0.0076| × (112 / 59.17) ≈ 0.0076 × 1.893 ≈ 0.0144 (< 0.18, not satisfied);

[0111] vulcanization temperature T s (X i =155℃):

[0112]

[0113] Sens(T s = |-0.0044|×(155 / 59.17)≈0.0044×2.62≈0.0115 (<0.18, not satisfied);

[0114] Tensile strength σ t (X i =23.5MPa):

[0115]

[0116] Sens(σ t = 0.88 × (23.5 / 59.17) ≈ 0.88 × 0.397 ≈ 0.349 (> 0.18, satisfies)

[0117] Cont(σ t = (0.4 × 2.2 × 23.5) / 59.17 = (20.68) / 59.17 ≈ 0.349 (> 0.05, satisfied);

[0118] Crosslinking density ρ(X) i =1.08×10 -4 mol / cm 3 ):

[0119]

[0120] Sens(ρ)=400×(1.08×10 -4 / 59.17)≈400×1.825×10 -6 ≈0.00073 (<0.18, not satisfied);

[0121] After screening, only the tensile strength σ t Simultaneously satisfying Sens(X) i )>0.18 and Cont(X) iThe value of t > 0.05 was identified as a core abnormal parameter. Further investigation revealed that the low tensile strength was caused by uneven dispersion of carbon black during the mixing process. The tensile strength needed to be indirectly improved by adjusting the mixing process. Therefore, the mixing time t was added as a secondary abnormal parameter.

[0122] 4. Quantitative process control module

[0123] Preset target score S tar =90 points, standard value of mixing time X i0 =8min, reasonable range 5-12min, calculate adjustment amount:

[0124] ΔS tar =90-59.17=30.83;

[0125]

[0126] Deviation correction factor = 1 - 0.02 × |(7.5 - 8) / 8| = 1 - 0.02 × 0.0625 = 0.99875;

[0127] The basic adjustment amount Δt = 30.83 × (1 / 0.0058) × 0.99875 ≈ 30.83 × 172.41 × 0.99875 ≈ 5300 min (significantly exceeds the reasonable range, triggering constraint correction);

[0128] Constraint Modification: X i =12min (maximum mixing time), ΔS max =100-59.17=40.83;

[0129] The adjusted amount Δt' after correction is (12-7.5)×(30.83 / 40.83)=4.5×0.755≈3.4min;

[0130] The final adjustment instruction is: "It is recommended to extend the mixing time from 7.5 min to 10.9 min (7.5 + 3.4), and monitor the changes in carbon black dispersion simultaneously."

[0131] 5. Quality Prediction Module

[0132] Based on the previous 5 scores (89, 86, 82, 75, and 59.17), the ARIMA model is used to predict the next score:

[0133] Assume the error term ε in the first two iterations t =59.17-68=-8.83, ε t-1 =75-80=-5;

[0134]

[0135] because The system will trigger an early warning, prompting operators to pay close attention to parameter changes after the mixing time is adjusted.

[0136] After the operator executed the adjustment instructions, the parameters were collected again to calculate the quality score, which improved to 87 points (qualified). The next predicted score improved to 84 points, the warning was lifted, and the rapid rectification of non-conforming products and quality stability were achieved.

[0137] The above description is merely a specific embodiment of this application, enabling those skilled in the art to understand or implement this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed in this application.

Claims

1. An optimized quality inspection system for rubber production, characterized in that, include: A multi-source parameter synchronous acquisition module, wherein the data transmission of the multi-source parameter synchronous acquisition module is connected to a dynamic weighted quality assessment module, and the data transmission of the dynamic weighted quality assessment module is connected to a two-dimensional anomaly tracing module, wherein the data transmission of the two-dimensional anomaly tracing module is connected to a quantitative process control module; The multi-source parameter synchronous acquisition module collects three types of core parameters in real time throughout the entire rubber production process through a distributed sensor array. The dynamic weighted quality assessment module constructs a comprehensive scoring model that integrates time-series features based on the collected parameters; The dual-dimensional anomaly tracing module includes calculating the sensitivity of parameter impact, the quality contribution rate of calculated parameters, and screening core anomaly parameters. The quantitative process control module generates adjustment instructions for core abnormal parameters.

2. The optimized quality inspection system for rubber production according to claim 1, characterized in that: The three types of core parameters include physical parameters, chemical parameters, and process parameters; The physical parameters include the Mooney viscosity (M) of raw rubber and the tensile strength (σ) of vulcanized rubber. t and elongation at break ε β The chemical parameters include the original crosslinking density ρ0 and the ash content A, wherein the original crosslinking density ρ0 is calibrated to obtain the calibrated crosslinking density ρ after relaxation time calibration. The process parameters include the mixing temperature T, the mixing time t, the vulcanization pressure P, and the vulcanization temperature T. s .

3. The optimized quality inspection system for rubber production according to claim 3, characterized in that: The formula for the calibrated crosslinking density ρ is: ρ=ρ0*(T 20 / T2)*[1+0.003*(T m -25)] In the formula: T2 represents the actual detected transverse relaxation time, T 20 T represents the transverse relaxation time corresponding to the standard crosslinking density. m This indicates the ambient temperature being detected.

4. The optimized quality inspection system for rubber production according to claim 1, characterized in that: The scoring algorithm formula for the comprehensive scoring model is as follows: S=α*(k1M+k2σ t +k3e β )+β*(k4ρ-k5A)+γ*(k6 / T+k7 / t+k8P+k9 / T s )+δ*S t-1 In the formula: S is the current overall quality score, and the overall quality score is 0-100 points, with a score greater than or equal to 85 points considered qualified. t-1 The score for the previous acquisition time is given by α, β, γ, and δ, which are the weight coefficients of physical parameters, chemical parameters, process parameters, and time-series correlation terms, respectively, satisfying α+β+γ+δ=1. By improving the analytic hierarchy process and introducing entropy weight correction, the initial values ​​are determined as α=0.4, β=0.25, γ=0.25, and δ=0.

1. k1-k9 are parameter correction coefficients with initial values ​​of k1=0.09, k2=2.2, k3=0.11, k4=1600, k5=1.9, k6=380, k7=1.3, k8=0.85, and k9=420.

5. The optimized quality inspection system for rubber production according to claim 1, characterized in that: The calculation parameters affect sensitivity, the contribution rate of the calculation parameters to quality, and the screening of core abnormal parameters are activated when the overall quality score of the comprehensive scoring model is less than 85 points. The calculation parameters affect the sensitivity Sens(X). i ): In the formula: For S with respect to parameter X i The partial derivatives; The calculated parameter is the quality contribution rate Cont(X). i ): Cont(X i ) = (calculated values ​​of corresponding terms α / β / γ / δ) / S The specific method for filtering core anomaly parameters is to use Sens(X) i )>0.18 and Cont(X) i Parameters greater than 0.05 are used as the core anomaly tracing objects.

6. The optimized quality inspection system for rubber production according to claim 1, characterized in that: The algorithm formula for the adjustment amount of the quantitative process control module is as follows: In the formula: ΔS tar X represents the difference between the target score and the current score. i0 This represents the standard value of the parameter, [1-0.02*|(X i -X i0 ) / X i0 |] represents the deviation correction factor.

7. The optimized quality inspection system for rubber production according to claim 3, characterized in that: The ash content A was collected by the ignition-weighing method, and then corrected for humidity using the following formula: A = A0 * (1 - 0.01 * H) In the formula: A0 is the original ash content detection value, and H is the sample moisture content.

8. The optimized quality inspection system for rubber production according to claim 4, characterized in that: The entropy weight correction algorithm of the improved analytic hierarchy process is as follows: oh i =ω i *(1-e i ) / ∑[ω i *(1-e i )] In the formula: ω i e represents the initial weights of the analytic hierarchy process. i This represents the information entropy of the parameter.

9. The optimized quality inspection system for rubber production according to claim 1, characterized in that: After the adjustment instruction is generated, it needs to be verified by parameter constraints. The verification formula is as follows: X ine =X i +ΔX i If X ine If the parameters exceed the preset reasonable range, automatic correction will be triggered. ΔX i =(X im -X i )*(ΔS tar / ΔS max ) In the formula: X im For parameter X i The upper limit of ΔS max To maximize the potential score improvement (100-S), ΔS tar This represents the difference between the target score and the current score.

10. The optimized quality inspection system for rubber production according to claim 1, characterized in that: The optimized quality inspection system for rubber production also includes a quality prediction module, which predicts the next score based on the collected S-value using an ARIMA model. The prediction algorithm formula is as follows: In the formula: St+1 is the next predicted score, φ1 to φ p These are the autoregressive coefficients, θ1 to θ q ε is the moving average coefficient. t Let be the error term at the current time, and p and q be the autoregression order and the moving average order, respectively.

Images