Cosmetic filling control method based on large model mapping knowledge domain

Abstract

The invention relates to the technical field of filling automation control, in particular to a cosmetic filling control method based on a large model knowledge graph, which comprises the following steps: acquiring an actual filling rate to determine a dynamic filling efficiency index, and determining whether the filling efficiency is stable based on the dynamic filling efficiency index. Determining the abnormal level of the filling efficiency in combination with the nitrogen filling response delay; determining that the filling efficiency is reduced due to the fact that moisture absorption raw material components are influenced by environmental factors to change the viscosity of a material body based on the combination of the viscosity-temperature sensitivity coefficient and the actual viscosity of a product and the environmental humidity, and adjusting a temperature set value of a compression area; the actual response time for adjusting the nitrogen flow is obtained to determine a response efficiency index, whether the comprehensive efficiency of nitrogen charging meets the expectation or not is determined based on the response efficiency index, and the calibration delay for determining the abnormal level of the filling efficiency is adjusted. The filling process is regulated and controlled on the basis of the product characteristics of cosmetics needing to be filled and the characteristics of filling equipment.

Description

Cosmetic filling control method based on large model knowledge graph Technical Field

[0001] This invention relates to the field of automated filling control technology, and in particular to a cosmetic filling control method based on a large model knowledge graph. Background Technology

[0002] Cosmetics, especially skincare and high-end beauty products, have extremely stringent requirements for the precision, hygiene, and product stability of the filling process. Filling control not only directly affects whether the product's net content complies with regulations, but is also closely related to the product's sensory characteristics (such as texture and uniformity) and chemical stability.

[0003] The most common application method currently is for engineers to preset a set of fixed filling parameters (such as pump speed, time, and pressure) based on product formulas (e.g., viscosity, density) and experience before production. This is simple, but for categories like cosmetics with complex raw material characteristics that are easily affected by the environment (temperature, humidity), there are significant drawbacks: poor adaptability—when there are fluctuations between batches of raw materials, changes in ambient temperature, or slight wear and tear on the equipment, the fixed parameters cannot be automatically adjusted, easily leading to insufficient filling or overflow, resulting in raw material waste and product defects. Low changeover efficiency—when switching to different products, it is necessary to stop the machine and manually readjust the parameters, a time-consuming and material-intensive process that cannot meet the trend of flexible production of "multiple varieties and small batches." Reliance on manual experience—parameter setting heavily depends on the experience of operators, making it difficult to solidify and pass on knowledge, and posing production risks due to personnel changes.

[0004] While closed-loop feedback control based on classic PID controllers is used in some high-end equipment for real-time feedback via flow meters or weighing sensors, and dynamic adjustment is achieved through PID algorithms, limitations remain: weak ability to handle nonlinearity and hysteresis; cosmetic raw materials are mostly non-Newtonian fluids, with viscosity and shear force related, exhibiting nonlinear characteristics; and there is a hysteresis between parameter adjustment and flow rate stabilization. Classic PID controllers are ineffective for such complex controlled objects, easily producing overshoot or oscillation.

[0005] Lacking predictive and cognitive capabilities, PID can only compensate for deviations that have already occurred. It cannot make predictive feedforward adjustments based on raw material characteristics (such as viscosity changes due to temperature) or handle high-level changes such as "reformulation changes".

[0006] In summary, existing cosmetic filling control technologies, whether simple program control or classic feedback control, have failed to effectively solve the core technical problem of "how to achieve adaptive, predictable, and knowledge-reasoning-capable intelligent control when facing complex and ever-changing production factors." This has led to a series of pressing technical challenges during production, such as large fluctuations in filling accuracy, high quality control costs, and insufficient production flexibility.

[0007] Chinese Patent Publication No. CN117970894A discloses a quality management and control method and system based on the cosmetic production process. The method includes the following steps: performing raw material feeding and mixing, emulsification and homogenization mixing, and filling and packaging quality control processes on the raw material feeding and mixing production process, the cosmetic emulsification production process, and the cosmetic filling and packaging production process, respectively, to generate raw material feeding and mixing production control strategies, cosmetic emulsification production control strategies, and cosmetic filling and packaging production control strategies; and performing intelligent response regulation and management on the raw material feeding and mixing production control strategies, cosmetic emulsification production control strategies, and cosmetic filling and packaging production control strategies to obtain a cosmetic production process quality management and control report, thereby executing the corresponding cosmetic production process quality control work. It is evident that the aforementioned quality management and control method and system based on the cosmetic production process has the following problems: it cannot achieve adaptive and predictable intelligent control by incorporating batch-to-batch fluctuations in raw materials, changes in environmental temperature and humidity, and minor equipment wear. Summary of the Invention

[0008] To address this, the present invention provides a cosmetic filling control method based on a large model knowledge graph, which overcomes the problem in existing technologies that cannot achieve adaptive and predictable intelligent control by incorporating batch-to-batch fluctuations in raw materials, changes in environmental temperature and humidity, and minor wear and tear on equipment.

[0009] To achieve the above objectives, this invention provides a cosmetic filling control method based on a large-scale model knowledge graph, comprising: obtaining the actual filling rate to determine a dynamic filling efficiency index; determining whether the filling efficiency is stable based on the dynamic filling efficiency index; and determining the anomaly level of the filling efficiency by combining the nitrogen filling response delay; adjusting the filling set speed in response to the anomaly level or determining a viscosity-temperature sensitivity coefficient based on the actual temperature of each heating zone to determine the sensitivity of the material viscosity to temperature; and determining, based on the viscosity-temperature sensitivity coefficient combined with the actual viscosity of the product and the ambient humidity, that the decrease in filling efficiency is due to the hygroscopic raw material components changing the material viscosity due to environmental factors, and adjusting the compression zone... Temperature setpoint; Obtain the actual temperature of the metering zone and feeding zone, and determine the shear heat effect index by combining it with the actual screw speed. Determine that the decrease in filling efficiency is due to shear heating, and adjust the screw speed and the temperature setpoints of the feeding zone and metering zone. Based on the current real-time oxygen content, determine the potential conflict between adjusting the temperature setpoint and oxidation risk, and conduct a quantitative risk index assessment to determine the increase in oxidation risk and filling quality risk, so as to determine whether to adjust the temperature setpoint or increase the nitrogen flow rate. Obtain the actual response time of adjusting the nitrogen flow rate to determine the response efficiency index. Based on the response efficiency index, determine whether the overall efficiency of nitrogen filling meets expectations, and adjust the calibration delay to determine the abnormal level of filling efficiency.

[0010] Furthermore, the process of determining whether the filling efficiency is stable includes: when the dynamic filling efficiency index is in the normal range, the filling efficiency is considered stable; when the dynamic filling efficiency index is not in the normal range, the filling efficiency is considered abnormal, and the abnormality level is determined by combining the nitrogen filling response delay.

[0011] Furthermore, the process of determining the level of abnormality in filling efficiency includes: when the dynamic filling efficiency index is in the marginal range and the nitrogen charging response delay is less than the calibration delay, the abnormality is determined to be a level one abnormality, and the filling process exhibits efficiency drift or fluctuation; when the dynamic filling efficiency index is lower than or higher than the marginal range, or the nitrogen charging response delay is greater than or equal to the calibration delay, the abnormality is determined to be a level two abnormality, and the filling setting speed of the filling screw pump is adjusted.

[0012] Furthermore, the process of determining the temperature sensitivity of the material viscosity includes: the filling efficiency anomaly level is level one, and the viscosity temperature sensitivity coefficient is determined based on the actual viscosity and average temperature of several sampling points; if the viscosity temperature sensitivity coefficient is greater than the coefficient threshold, it is determined that the material viscosity is highly sensitive to temperature, and heating can effectively reduce viscosity; if the absolute value of the difference between the viscosity temperature sensitivity coefficient and zero is less than the difference evaluation value, it is determined that the material viscosity is not sensitive to temperature, and the shear heat characteristic temperature difference is calculated.

[0013] Furthermore, the process of determining the cause of the decrease in filling efficiency includes: if the actual viscosity of the essence is greater than the upper limit of the process standard range, and the absolute value of the viscosity temperature sensitivity coefficient is less than the preset sensitivity threshold, and the real-time relative humidity value of the environment is greater than the critical humidity for material hygroscopicity, then it is determined that the main reason for the current decrease in filling efficiency is that the ambient humidity is too high, and the hygroscopic raw material components are affected by environmental factors, changing the viscosity of the material beyond the process range, and adjusting the temperature setting value of the compression zone to increase the material temperature.

[0014] Furthermore, when the dynamic filling efficiency index is lower than the normal range, if the shear heat effect index is greater than the preset safety threshold and the actual viscosity is within the standard range, then the current efficiency decline is determined to be caused by shear heating, and the screw speed is adjusted and the temperature settings of the feeding zone and metering zone are compensated and adjusted.

[0015] Furthermore, when the change in the temperature setpoint of each heating zone is greater than zero, the current real-time oxygen content in the core area of ​​the filling cavity is obtained; if the current real-time oxygen content is greater than the oxygen content warning value but less than the safety absolute threshold, it is determined that there is a potential conflict between process optimization and oxidation risk, and a quantitative risk indicator assessment is performed; if the current real-time oxygen content is greater than or equal to the safety absolute threshold, the highest level alarm is triggered to stop the filling process.

[0016] Furthermore, the quantitative risk indicator assessment includes calculating the increase in oxidation risk due to temperature increase and the filling quality risk of maintaining the status quo without temperature increase; if the increase in oxidation risk due to temperature increase is less than the multiple threshold of the filling quality risk of maintaining the status quo without temperature increase, the filling quality risk is judged to be too high, and the determined temperature setpoints for each heating zone are adjusted; if the increase in oxidation risk due to temperature increase is greater than or equal to the multiple threshold of the filling quality risk of maintaining the status quo without temperature increase, the oxidation risk due to temperature increase is judged to be too high, the adjustment of the compression zone temperature setpoint is stopped, the nitrogen flow rate is increased, and the increase in oxidation risk is recalculated.

[0017] Furthermore, when there is a potential conflict between process optimization and oxidation risk, the maximum allowable temperature rise is calculated, and the adjustment amount of the compression zone temperature setpoint is limited based on the maximum temperature rise.

[0018] Furthermore, after adjusting the nitrogen flow rate, the response efficiency index for adjusting the nitrogen flow rate is determined. If the arithmetic mean of the response efficiency index is greater than the first index threshold, it is determined that the overall efficiency of nitrogen purging is consistently lower than expected, and there is a risk of false alarm. The calibration delay is reduced according to the calibration coefficient. If the arithmetic mean of the response efficiency index is less than the second index threshold, it is determined that the overall efficiency of nitrogen purging is consistently higher than expected, and there is a risk of false alarm. The calibration delay is increased according to the calibration coefficient.

[0019] Compared with existing technologies, the beneficial effects of this invention are as follows: by fusing dynamic filling efficiency index and nitrogen charging response delay, it achieves accurate graded early warning of early abnormalities in the filling process; based on multi-sensor data and knowledge graph, it can intelligently distinguish whether the root cause of viscosity increase is environmental moisture absorption or shear heating, and generate precise temperature control and speed adjustment strategies; by quantitatively assessing the oxidation risk increment of heating operation and the filling quality risk of maintaining the status quo, it conducts optimal arbitration while ensuring the activity of core ingredients; in addition, through closed-loop evaluation and adaptive learning of nitrogen charging efficiency, it continuously optimizes the abnormality detection threshold, fundamentally improving filling control from passive response to active intelligent control with causal diagnosis, conflict resolution and continuous learning capabilities, significantly improving the accuracy, stability and quality safety assurance level of high-activity cosmetic filling.

[0020] Furthermore, this invention achieves early, accurate, and graded warnings for filling accuracy degradation and nitrogen protection system performance decline by combining real-time calculation of the dynamic filling efficiency index with nitrogen charging response delay for joint diagnosis. When a level one anomaly is identified, the viscosity-temperature sensitivity coefficient and shear heat characteristic temperature difference are determined. When a level two anomaly is identified, the speed is immediately forced to a safe value, creating stable conditions for subsequent diagnosis and adjustment, effectively preventing batch product defects and raw material waste caused by abnormal equipment or process parameters. This method transforms traditional single post-event control into proactive preventive control based on the fusion of multiple physical quantities, significantly improving the stability and quality and safety assurance level of the high-value cosmetic filling process, while avoiding production risks caused by response lag or misjudgment in traditional methods.

[0021] Furthermore, this invention, through quantitative analysis of the viscosity-temperature sensitivity coefficient, innovatively distinguishes the root causes of viscosity anomalies leading to decreased filling efficiency into two fundamentally different physical reasons: environmental moisture absorption and shear heating. For the viscosity increase caused by moisture absorption, the compression zone temperature is precisely calculated and adjusted based on a pre-calibrated material temperature-viscosity influence coefficient to address the root cause. For the shear heating problem, the screw speed is intelligently reduced to suppress overheating by calculating the temperature rise index per unit rotational speed and comparing it to a safety threshold. The temperature setpoints of the inlet and outlet zones are simultaneously adjusted using a compensation coefficient calibrated by thermal characteristics, achieving rapid decoupling and stabilization of rotational speed and temperature. This method abandons the traditional one-size-fits-all viscosity control mode, achieving precise diagnosis and targeted proactive adjustment of complex fluid anomalies, effectively ensuring product quality and process stability during the filling process of highly active, heat-sensitive cosmetics.

[0022] Furthermore, this invention creatively resolves the inherent conflict between optimizing process parameters for temperature rise and viscosity reduction and the core quality red line of preventing oxidation by introducing a quantitative comparison model of oxidation risk increment and filling quality risk. It quantifies the potential hazards of temperature rise into specific values ​​related to oxygen content and temperature rise amplitude, and compares them with the quality risks caused by efficiency reduction, thereby making an interpretable and verifiable optimal decision. When the oxidation risk is determined to be too high, the solution is not simply rejected, but a dynamic adjustment mechanism is activated, prioritizing the reduction of ambient oxygen content by increasing nitrogen flow rate, creating a safety window for subsequent process adjustments. The method achieves intelligent trade-offs under complex constraints, ensuring that the filling process of highly active cosmetics, while pursuing efficiency and precision, never breaches the absolute safety bottom line of preventing active ingredient deactivation.

[0023] Furthermore, this invention calculates the dynamically permissible maximum temperature rise, transforming the oxygen content safety threshold into a specific upper limit for temperature regulation. This physically constrains any optimized operation from exceeding the oxidation prevention safety boundary. Furthermore, it comprehensively evaluates the oxygen content decrease and response speed after nitrogen purging, and adaptively calibrates the anomaly detection threshold (calibration delay) online based on the efficiency index. The method achieves dual protection: firstly, it eliminates the oxidation risk caused by process parameter adjustments through rigid constraints; secondly, it ensures the nitrogen purging system's performance monitoring remains sensitive and accurate through a self-learning mechanism, effectively avoiding missed or false alarms due to threshold drift. This invention combines static safety standards with dynamic system performance evaluation to construct a robust adaptive safety protection system, significantly improving the long-term operational reliability of the high-activity cosmetic filling process under complex conditions. Attached Figure Description

[0024] Figure 1 is a flowchart illustrating the cosmetic filling control method based on a large model knowledge graph in an embodiment of the present invention; Figure 2 is a flowchart illustrating the process of determining whether the filling efficiency is stable in an embodiment of the present invention; Figure 3 is a flowchart illustrating the process of quantitative risk indicator assessment in an embodiment of the present invention; Figure 4 is a flowchart illustrating the process of adjusting the calibration delay for determining the abnormal level of filling efficiency in an embodiment of the present invention. Detailed Implementation

[0025] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0026] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0027] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0028] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0029] Please refer to Figures 1-4. Figure 1 is a flowchart of the cosmetic filling control method based on a large model knowledge graph in an embodiment of the present invention; Figure 2 is a flowchart of determining whether the filling efficiency is stable in an embodiment of the present invention; Figure 3 is a flowchart of quantitative risk indicator assessment in an embodiment of the present invention; and Figure 4 is a flowchart of adjusting the calibration delay for determining the abnormal level of filling efficiency in an embodiment of the present invention.

[0030] This invention provides a cosmetic filling control method based on a large-scale knowledge graph, comprising: Step S1, obtaining the actual filling rate to determine a dynamic filling efficiency index, determining whether the filling efficiency is stable based on the dynamic filling efficiency index, and determining the abnormal level of the filling efficiency in conjunction with the nitrogen filling response delay; Step S2, adjusting the filling set speed in response to the abnormal level or determining the viscosity-temperature sensitivity coefficient based on the actual temperature of each heating zone to determine the sensitivity of the material viscosity to temperature; Step S3, determining, based on the viscosity-temperature sensitivity coefficient and the actual viscosity of the product and the ambient humidity, that the decrease in filling efficiency is due to the hygroscopic raw material components changing the material viscosity due to environmental factors, and adjusting the compression zone temperature set speed. Step S4: Obtain the actual temperatures of the metering zone and the feeding zone, and determine the shear heat effect index by combining the actual screw speed. Determine that the decrease in filling efficiency is due to shear heating, and adjust the screw speed and the temperature setpoints of the feeding zone and the metering zone. Step S5: Based on the current real-time oxygen content, determine the potential conflict between adjusting the temperature setpoint and the oxidation risk, and conduct a quantitative risk index assessment to determine the increase in oxidation risk and filling quality risk, so as to determine whether to adjust the temperature setpoint or increase the nitrogen flow rate. Step S6: Obtain the actual response time of adjusting the nitrogen flow rate to determine the response efficiency index. Based on the response efficiency index, determine whether the overall efficiency of nitrogen filling meets expectations, and adjust the calibration delay to determine the abnormal level of filling efficiency.

[0031] Specifically, this invention achieves precise graded early warning of early anomalies in the filling process by fusing dynamic filling efficiency index with nitrogen charging response delay; based on multi-sensor data and knowledge graph, it can intelligently distinguish whether the root cause of viscosity increase is environmental moisture absorption or shear heating, and generate precise temperature control and speed adjustment strategies; by quantitatively assessing the increase in oxidation risk of heating operation and the filling quality risk of maintaining the status quo, it conducts optimal arbitration while ensuring the activity of core ingredients; in addition, through closed-loop evaluation and adaptive learning of nitrogen charging efficiency, it continuously optimizes the anomaly detection threshold, fundamentally upgrading filling control from passive response to proactive intelligent control with causal diagnosis, conflict resolution, and continuous learning capabilities, significantly improving the accuracy, stability, and quality safety assurance level of high-activity cosmetic filling.

[0032] In this embodiment, a high-activity antioxidant essence is nitrogen-filled using a screw pump and a filling nozzle. The screw pump delivers the essence to the filling chamber, and the filling nozzle fills the essence from the filling chamber into the container. The ideal viscosity range for the filling process is 850-1050 cP; exceeding this range will lead to decreased filling accuracy or uneven product texture. Specifically, empty bottles enter the filling station via a conveyor belt, located directly below the filling nozzle. The nitrogen filling valve is opened, and nitrogen gas, after being pressurized by a pressure reducing valve, is injected into the bottle through the annular gas path outside the filling nozzle for 1.5-2.0 seconds, purging the original air (oxygen content approximately 2%) from the bottle. 1%) is replaced and discharged, reducing the headspace oxygen content in the bottle to below 10ppm; after the pre-nitrogen charge is completed, the filling nozzle descends at a constant speed under the drive of the servo motor. After the filling nozzle is in place, the screw pump starts, and the essence is drawn from the storage tank into the screw pump through the feed pipeline. Under the propulsion of the rotating screw, it is continuously and quantitatively delivered to the filling chamber. The screw pump speed is preset according to the product viscosity and the target filling volume, usually 20-60rpm; during the conveying process, the three heating zones of the screw (feeding zone, compression zone, and metering zone) are independently temperature controlled to keep the material temperature constant within the process window (20-25℃) to ensure viscosity stability.

[0033] After the essence enters the filling chamber, the filling valve opens instantly. Under the combined action of pumping pressure and gravity, the material is injected into the bottle through the nozzle. The filling process adopts a time-pressure metering mode. Based on the target filling volume and real-time flow feedback, the opening time of the filling valve is precisely controlled (usually 0.5-3.0 seconds). When the filling volume reaches the set target, the filling valve closes.

[0034] The serum is sensitive to oxygen; exposure to oxygen levels greater than 50 ppm will cause core antioxidant ingredients such as vitamins C and E to become inactive by ≥15% within 24 hours. The core thickener is hygroscopic; when the relative humidity of the environment is greater than H%, such as 65%, the thickener will absorb moisture, leading to an increase in the measured viscosity of the serum. It is also shear-sensitive; a screw shear rate >1200 s⁻¹ will damage the microstructure of the emulsion, resulting in decreased product stability. Furthermore, it is temperature-sensitive; the material temperature needs to be maintained at 20-25℃, and the viscosity increases non-linearly below 18℃. Temperatures above 28℃ will accelerate the degradation of active ingredients.

[0035] When the ambient humidity is greater than 68%RH, the polyols in the raw materials absorb moisture, and the viscosity increase rate can reach 1.2% / h within 48 hours.

[0036] Therefore, this embodiment is particularly suitable for preparation scenarios that are sensitive to oxygen, humidity, high and low temperature, and stirring and shearing environments and require precise filling.

[0037] The real-time filling weight is obtained using a checkweigher, which is the weight detected during filling minus the initial weight at the start of filling. The actual filling rate is calculated based on the real-time filling weight per unit time: Actual filling rate = Real-time filling weight / Unit time. The dynamic filling efficiency index is determined by combining this with the theoretical maximum filling rate of the current batch of essence, calibrated in experiments: Dynamic filling efficiency index = Actual filling rate / Theoretical maximum filling rate. It is understood that the theoretical maximum filling rate for the current batch of essence can be measured by adjusting the filling equipment parameters, which will not be elaborated here.

[0038] The nitrogen filling response delay was obtained through experiments. The nitrogen filling response delay reflects the dynamic response performance of the nitrogen filling circuit. In practice, the moment when the initial command to increase the nitrogen filling flow rate was issued was recorded, and the actual oxygen content value of the laser oxygen content sensor located in the core area of ​​the filling cavity was continuously read. When the actual oxygen content continuously decreased from the initial value before the command was issued and stabilized at the level of "initial value - 10ppm", this moment was recorded as the command response moment. The nitrogen filling response delay is the difference between the initial command moment and the command response moment.

[0039] When the dynamic filling efficiency index is within the normal range, the filling efficiency is considered stable. When the dynamic filling efficiency index is outside the normal range, the filling efficiency is considered abnormal, and the level of abnormality is determined by combining the nitrogen charging response delay. Specifically, when the dynamic filling efficiency index is in the edge range and the nitrogen charging response delay is less than the calibration delay, the abnormality is classified as Level 1, indicating a slight, early efficiency drift or fluctuation in the filling process. When the dynamic filling efficiency index is below or above the edge range, or the nitrogen charging response delay is greater than or equal to the calibration delay, the abnormality is classified as Level 2. Specifically, an abnormal dynamic filling efficiency index indicates a significant deviation in filling efficiency, seriously affecting accuracy, while an abnormal nitrogen charging response delay indicates performance degradation or malfunction in the nitrogen protection system itself.

[0040] The normal range is [0.985, 1.015], the edge range is [0.975, 0.985) or (1.015, 1.025], and the calibration delay is 1.3 times the average normal response time in historical statistics.

[0041] Specifically, the normal range and the edge range were calibrated experimentally. The equipment was selected to continuously fill 1000 bottles of the same batch of serum A under optimal conditions. The actual filling rate of each bottle was recorded, and the corresponding dynamic filling efficiency index η was calculated, resulting in 1000 η sample values. After passing the normality test, the sample mean μ and standard deviation σ were calculated. The measured data showed that the mean η of the equipment under stable operating conditions was μ = 1.000, and the standard deviation σ = 0.005.

[0042] According to industry-standard quality requirements, the filling process capability index Cpk must be ≥1.33. Calculated using Cpk=(1-|μ-1.0|) / 3σ, when μ=1.000 and σ=0.005, Cpk=1.000 / (3×0.005)=66.7, which is significantly higher than 1.33. To balance control sensitivity and false alarm rate, the control limit is narrowed to ±3σ. 3σ=3×0.005=0.015, therefore the lower limit of the normal range = 1.000-0.015=0.985, and the upper limit = 1.000+0.015=1.015, i.e., [0.985, 1.015].

[0043] The edge interval is defined as the transition zone outside the normal range that has not yet triggered a level 2 abnormal alarm. It is defined as the range of 3σ to 4σ: 4σ = 4 × 0.005 = 0.020. The lower limit of the edge interval is 1.000 - 0.020 = 0.980, but to avoid overlap with the normal range boundary, a left-closed, right-open interval representation is used, actually taking [0.975, 0.985]. The physical meaning of this value is: when η drops to the 0.985-0.975 range, the filling efficiency has deviated from the normal level by 1-3 standard deviations, which is an early sign of abnormality. Similarly, the upper limit of the edge interval is (1.015, 1.025], corresponding to the range of +3σ to +5σ. The upper limit is relaxed to 1.025 (5σ) to consider that the tolerance for artificially high efficiency caused by excessively fast pump speed is usually slightly wider.

[0044] When the abnormality is classified as Level 2, reduce the filling set speed of the filling screw pump. In practice, the filling set speed should be reduced to 80%.

[0045] Specifically, this invention achieves early, accurate, and graded early warning of filling accuracy decay and nitrogen protection system performance degradation by combining real-time calculation of the dynamic filling efficiency index with nitrogen charging response delay for joint diagnosis. When a level 1 anomaly is identified, the viscosity-temperature sensitivity coefficient and shear heat characteristic temperature difference are determined. When a level 2 anomaly is identified, the speed is immediately forced to a safe value, creating stable conditions for subsequent diagnosis and adjustment, effectively preventing batch non-conforming products and raw material waste caused by abnormal equipment or process parameters. This method transforms traditional single post-event control into proactive preventive control based on the fusion of multiple physical quantities, significantly improving the stability and quality and safety assurance level of the high-value cosmetic filling process, while avoiding the production risks caused by response lag or misjudgment in traditional methods.

[0046] The filling efficiency anomaly level is Level 1. When the filling efficiency is abnormal, the actual viscosity of the essence in the filling screw pump is obtained through an online viscometer, and the actual temperature of the three heating zones of the filling screw pump, namely the feeding zone, the compression zone, and the metering zone, is obtained through a temperature sensor. The actual viscosity and the actual temperature of the feeding zone, the compression zone, and the metering zone are obtained at several sampling points within a 30-minute interval at unit time intervals. The viscosity-temperature slope of the viscosity change with temperature is calculated and recorded as the viscosity-temperature sensitivity coefficient. Specifically, by performing a univariate linear regression fitting using the least squares method on the actual viscosity and average temperature of several sampling points, the linear equation is obtained as: Actual viscosity = Viscosity-temperature sensitivity coefficient × Average temperature + Equation constant, where the average temperature is the average of the actual temperatures of the feeding zone, the compression zone, and the metering zone. It can be understood that the linear equation is a linear equation in one variable, and the independent variables of the equation are the actual viscosity and average temperature corresponding to several sampling points. The viscosity-temperature sensitivity coefficient and the equation constant can be calculated by substituting several sampling points into the linear equation, which will not be elaborated here.

[0047] If the viscosity-temperature sensitivity coefficient is greater than the coefficient threshold, it is determined that the viscosity of the essence is highly sensitive to temperature, and heating can effectively reduce viscosity; if the absolute value of the difference between the viscosity-temperature sensitivity coefficient and zero is less than the difference evaluation value, it is determined that the viscosity of the essence is not sensitive to temperature, and the shear heat characteristic temperature difference is calculated; wherein, the coefficient threshold can be selected in the range of 0.8-1.5 cP / ℃, and the difference evaluation value is 1.0 cP / ℃.

[0048] The real-time relative humidity value of the environment is obtained by measuring the environmental sensors in the filling room. When the dynamic filling efficiency index is lower than the normal range, if the actual viscosity of the essence is greater than the upper limit of the process standard range, and the absolute value of the viscosity temperature sensitivity coefficient is less than the preset sensitivity threshold, and the real-time relative humidity value of the environment is greater than the material's hygroscopic critical humidity, then it is determined that the main reason for the current decrease in filling efficiency is that the environmental humidity is too high, which causes the hygroscopic raw material components to absorb moisture, resulting in the overall material viscosity exceeding the process range.

[0049] Specifically, if the absolute value of the viscosity-temperature sensitivity coefficient is less than the preset sensitivity threshold, the possibility that viscosity changes are mainly caused by temperature fluctuations can be ruled out.

[0050] The upper limit of the process standard range is 1050 cP, the sensitivity threshold is 0.5 cP / ℃, and the critical humidity of the material for hygroscopic absorption is the critical humidity of the essence recorded in the knowledge graph, which is 68% in practice.

[0051] When ambient humidity causes the overall viscosity of the material to exceed the process range, the temperature setting of the compression zone is adjusted to increase the material temperature, thereby reducing the viscosity increase due to moisture absorption and bringing it back to the standard range. In practice, the compression zone temperature, which is most effective for viscosity control, is selected as the material temperature adjustment target. The adjustment amount of the compression zone temperature setting is = (actual viscosity - target viscosity) / temperature viscosity influence coefficient. Here, the target viscosity is taken as the median value of the standard range, 950 cP, and the temperature viscosity influence coefficient is a known constant obtained through previous rheological experiments, with units of cP / ℃ and stored in the knowledge graph. If the temperature viscosity influence coefficient = -50 cP / ℃, it means that for every 1℃ increase in temperature, the viscosity of the material decreases by approximately 50 cP.

[0052] Specifically, the rated speed of the screw is obtained and the actual speed of the screw is obtained through feedback from the servo motor encoder. The shear heat effect index is calculated as (actual temperature of the metering zone - actual temperature of the feeding zone) / (actual speed of the screw / rated speed of the screw). The shear heat effect index quantifies the temperature rise generated per unit speed. A value higher than the historical baseline means that the screw generates an abnormally large amount of heat when shearing the material, which is usually caused by poor material flowability or improper clearance between the screw and the sleeve.

[0053] When the dynamic filling efficiency index is lower than the normal range, if the shear heat effect index is greater than the preset safety threshold and the actual viscosity is within the standard range, then the current efficiency decline is judged to be caused by shear heating. The screw shearing work is excessive, resulting in excessively high local temperature of the material in the metering zone, which may cause changes in heat-sensitive components or abnormal local viscosity.

[0054] The current efficiency decrease is due to the adjustment of the screw speed during shear heating, which compensates for the temperature setpoints in the feeding and metering zones. Specifically, the screw speed setpoint is reduced proportionally, and the adjusted screw speed setpoint = the original screw speed setpoint × (safety threshold / shear heat effect index).

[0055] For example, if the safety threshold is 3℃ and the shear heat effect index is 5℃, then the set value of the screw speed after adjustment is equal to the set value of the screw speed before adjustment × (3.0 / 4.5) = 0.67 × the set value of the screw speed before adjustment, thus suppressing the temperature difference of the material to near the safety threshold. To adapt to the new speed and maintain a stable discharge temperature, the temperature set values ​​of the feeding zone and the metering zone are adjusted compensatorily. The set value of the temperature of the feeding zone after adjustment is equal to the set value of the temperature of the feeding zone before adjustment + C1 × (the set value of the screw speed before adjustment - the set value of the screw speed after adjustment), and the set value of the temperature of the metering zone after adjustment is equal to the set value of the temperature of the metering zone before adjustment - C2 × (the set value of the screw speed before adjustment - the set value of the screw speed after adjustment), where C1 and C2 are compensation coefficients obtained through equipment thermal characteristic calibration experiments.

[0056] During implementation, the initial temperature settings of the three temperature control zones of the screw are consistent, such as 25℃. The screw speed is set and stabilized at the rated speed, such as 50 rpm. After filling and operation stabilize for more than 30 minutes, the stable temperature of the feeding zone and the stable temperature of the metering zone are recorded. The screw speed is then gradually reduced from the rated speed to a lower stable value, such as 0.6 × rated speed = 30 rpm. The temperature settings of all temperature control zones are kept unchanged, and continuous monitoring is performed until they re-enter a new thermal equilibrium steady state for about 20-30 minutes. The new steady-state temperature of the feeding zone and the new steady-state temperature of the metering zone are recorded. The temperature changes in the feeding zone and metering zone caused by the reduction in speed, as well as the speed change, are calculated. C1 = temperature change in the feeding zone / speed change, C2 = temperature change in the metering zone / speed change.

[0057] The safety threshold is 3°C, and the standard range is 850 cP-1050 cP.

[0058] Specifically, this invention innovatively distinguishes the root causes of viscosity anomalies leading to decreased filling efficiency into two fundamentally different physical reasons: environmental moisture absorption and shear heating, through quantitative analysis of the viscosity-temperature sensitivity coefficient. For the viscosity increase caused by moisture absorption, the compression zone temperature is precisely calculated and adjusted based on a pre-calibrated material temperature-viscosity influence coefficient to address the root cause. For the shear heating problem, the screw speed is intelligently reduced to suppress overheating by calculating the temperature rise index per unit rotational speed and comparing it to a safety threshold. Furthermore, the temperature setpoints of the inlet and outlet zones are simultaneously adjusted using a compensation coefficient calibrated by thermal characteristics, achieving rapid decoupling and stabilization of rotational speed and temperature. This method abandons the traditional one-size-fits-all viscosity control mode, achieving precise diagnosis and targeted proactive adjustment of complex fluid anomalies, effectively ensuring product quality and process stability in the filling process of highly active, heat-sensitive cosmetics.

[0059] Adjusting the temperature setpoints of the three heating zones of the filling screw pump introduces new risks under the current overall operating conditions. When the change in the temperature setpoint of each heating zone is greater than zero (i.e., the temperature setpoint is increased), the real-time oxygen content in the core area of ​​the filling chamber is obtained using an oxygen analyzer. If the current real-time oxygen content is greater than the oxygen content warning value but less than the safety absolute threshold, a potential conflict between process optimization and oxidation risk is identified, as heating kinetically accelerates the oxidation reaction, and a quantitative risk indicator assessment is performed. If the current real-time oxygen content is greater than or equal to the safety absolute threshold, the highest-level alarm is triggered to stop the filling process. Specifically, the quantitative risk indicator assessment includes calculating the oxidation risk increment due to heating and the filling quality risk of maintaining the status quo without heating. The oxidation risk increment = oxidation reaction risk coefficient × (current real-time oxygen content - reference oxygen content) × the adjustment amount of the compression zone temperature setpoint. In practice, the oxidation reaction risk coefficient is determined based on accelerated oxidation test data of the product, and the unit of the oxidation reaction risk coefficient is (ppm⋅℃). −1 The oxidation risk increment is intended to quantify the additional oxidation risk introduced by performing a specific temperature rise at the current oxygen content.

[0060] The oxygen content warning value is 45 ppm, the absolute safety threshold is 50 ppm, and the reference oxygen content is 20 ppm.

[0061] Specifically, the filling quality risk = |Dynamic filling efficiency index - 1.0| / the lower limit of the edge range. The filling quality risk aims to quantify the risk of non-compliance in filling accuracy caused by the continuous existence of efficiency deviation.

[0062] If the increase in oxidation risk due to temperature increase is less than the multiple threshold of the filling quality risk of maintaining the status quo without temperature increase, the filling quality risk is judged to be too high, and the determined temperature setpoints for each heating zone are adjusted. If the increase in oxidation risk due to temperature increase is greater than or equal to the multiple threshold of the filling quality risk of maintaining the status quo without temperature increase, the oxidation risk due to temperature increase is judged to be too high, the adjustment of the compression zone temperature setpoint is stopped, the nitrogen flow rate is increased, and the increase in oxidation risk is re-determined.

[0063] The multiple threshold can be selected from a value range of 1.2 to 2.0.

[0064] Specifically, this invention creatively resolves the inherent conflict between optimizing process parameters for temperature rise and viscosity reduction and the core quality red line of preventing oxidation by introducing a quantitative comparison model of oxidation risk increment and filling quality risk. It quantifies the potential hazards of temperature rise into specific values ​​related to oxygen content and temperature rise amplitude, and compares them with the quality risks caused by efficiency reduction, thereby making an interpretable and verifiable optimal decision. When the oxidation risk is determined to be too high, the solution is not simply rejected, but a dynamic adjustment mechanism is activated, prioritizing the reduction of ambient oxygen content by increasing nitrogen flow rate, creating a safety window for subsequent process adjustments. The method achieves intelligent trade-offs under complex constraints, ensuring that the filling process of highly active cosmetics, while pursuing efficiency and precision, never breaches the absolute safety bottom line of preventing active ingredient deactivation.

[0065] When there is a potential conflict between process optimization and oxidation risk, the maximum allowable temperature rise is calculated. The maximum temperature rise = (absolute safety threshold - current real-time oxygen content) × safety margin coefficient. In the formula, the unit of the safety margin coefficient is ℃ / ppm. For example, if the safety margin coefficient = 0.1℃ / ppm, it means that for every 1ppm that the current oxygen content approaches the safety threshold, the maximum allowable temperature rise decreases by 0.1℃. When the current oxygen content = 45ppm, the maximum temperature rise of the temperature setpoint = 0.5℃.

[0066] The adjustment amount of the compression zone temperature setpoint is compared with the maximum temperature rise of the temperature setpoint, and the final output adjustment amount is min(adjustment amount of compression zone temperature setpoint, maximum adjustment amount of temperature setpoint), ensuring that any temperature rise operation is strictly limited within the dynamic safety boundary.

[0067] After adjusting the nitrogen flow rate, the actual response delay, the final oxygen content, and the target oxygen content are obtained. The target oxygen content is the difference between the current real-time oxygen content and 10 ppm. Specifically, the actual response delay is the actual response time from the issuance of the nitrogen flow rate adjustment command to the oxygen content decreasing by 10 ppm, and the final oxygen content is the stable oxygen content value after the nitrogen flow rate adjustment command response is completed.

[0068] Calculate the response efficiency index for adjusting nitrogen flow rate: Response efficiency index = (target decrease / actual decrease) × (baseline response time / actual response time), where the target decrease is 10 ppm, and the actual decrease is (real-time oxygen content before adjusting nitrogen flow rate - final oxygen content upon response). In practice, at least 3-5 standard nitrogen filling tests are conducted before filling, and the actual time from issuing the command to the oxygen content decreasing by 10 ppm is recorded for each test. The average of several actual times is obtained as the baseline response time.

[0069] Specifically, the first item (target decrease / actual decrease) is used to assess the adequacy of the response. Ideally, the actual decrease equals the target of 10 ppm, and this item is 1. If the actual decrease is insufficient, such as only 8 ppm, this item is greater than 1, indicating insufficient effectiveness. If there is overshoot, such as a decrease of 12 ppm, this item is less than 1, indicating possible over-response. The second item (baseline response time / actual response time) is used to assess the agility of the response. The shorter the actual response time, the larger this item is, indicating high agility. The ideal value for the response effectiveness index is 1.0, indicating that the response is both sufficient and agile. A response effectiveness index greater than 1 indicates that the overall effectiveness is lower than expected, too slow or insufficient. A response effectiveness index less than 1 indicates that the overall effectiveness is higher than expected, too fast or excessive.

[0070] After acquiring N consecutive performance data for nitrogen purging operations (N=5 in practice), the arithmetic mean of the response performance index is calculated. If the arithmetic mean of the response performance index is greater than the first index threshold, it is determined that the overall performance of nitrogen purging is consistently lower than expected, indicating that the calibration delay may be too lenient, causing the first layer to be insufficiently sensitive to actual performance degradation, resulting in a risk of false alarms. If the arithmetic mean of the response performance index is less than the second index threshold, it is determined that the overall performance of nitrogen purging is consistently higher than expected, indicating that the calibration delay may be too strict, causing the first layer to misjudge normal fluctuations, resulting in a risk of false alarms. If the second index threshold is less than or equal to the arithmetic mean of the response performance index, which is less than or equal to the first index threshold, it is determined that the calibration delay threshold is appropriate and no adjustment is needed.

[0071] During implementation, if the overall efficiency of nitrogen purging is consistently lower than expected, the calibration delay is reduced according to the calibration coefficient; if the overall efficiency of nitrogen purging is consistently higher than expected, the calibration delay is increased according to the calibration coefficient.

[0072] The first index threshold is 1.1, the second index threshold is 0.9, and the calibration coefficient is a ratio to the arithmetic mean of the response performance index.

[0073] Specifically, this invention calculates the dynamically permissible maximum temperature rise, transforming the oxygen content safety threshold into a specific upper limit for temperature regulation. This physically constrains any optimized operation from exceeding the oxidation prevention safety boundary. Furthermore, it comprehensively evaluates the reduction in oxygen content and response speed after nitrogen purging, and adaptively calibrates the anomaly detection threshold (calibration delay) online based on the efficiency index. The method achieves dual protection: firstly, it eliminates the oxidation risk caused by process parameter adjustments through rigid constraints; secondly, it ensures the nitrogen purging system's performance monitoring remains sensitive and accurate through a self-learning mechanism, effectively avoiding missed or false alarms due to threshold drift. This invention combines static safety standards with dynamic system performance evaluation to construct a robust adaptive safety protection system, significantly improving the long-term operational reliability of the high-activity cosmetic filling process under complex conditions.

[0074] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0075] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A cosmetic filling control method based on a large-scale knowledge graph, characterized in that, include: The actual filling rate of the filling process is obtained to determine the dynamic filling efficiency index. Based on the dynamic filling efficiency index, it is determined whether the filling efficiency is stable. Combined with the nitrogen filling response delay, the abnormal level of the filling efficiency is determined. In response to the abnormality level, the filling setting speed is adjusted or the viscosity temperature sensitivity coefficient is determined based on the actual temperature of each heating zone to determine the sensitivity of the material viscosity to temperature. Based on the viscosity-temperature sensitivity coefficient combined with the actual viscosity of the product and the ambient humidity, the reason for the decrease in filling efficiency was determined to be that the viscosity of the hygroscopic raw material component was affected by environmental factors, thus adjusting the temperature setpoint of the compression zone in the heating zone. The actual temperatures of the metering and feeding zones were obtained, and the actual screw speed was used to determine the shear heat effect index, confirming that the decrease in filling efficiency was due to shear heating. The screw speed and the temperature setpoints of the feeding and metering zones were adjusted accordingly. Based on the current real-time oxygen content in the filling cavity, the potential conflict between adjusting the temperature setpoint and oxidation risk was determined. A quantitative risk index assessment was conducted to determine the increase in oxidation risk and filling quality risk, thus determining whether to adjust the compression zone temperature setpoint or increase the nitrogen flow rate. The actual response time of adjusting the nitrogen flow rate was obtained to determine the response efficiency index. Based on the response efficiency index, it was determined whether the overall efficiency of nitrogen filling met expectations, and the calibration delay used to determine the level of abnormal filling efficiency was adjusted.

2. The cosmetic filling control method based on a large-scale knowledge graph according to claim 1, characterized in that, The process of determining whether the filling efficiency is stable includes: when the dynamic filling efficiency index is in the normal range, the filling efficiency is considered stable; when the dynamic filling efficiency index is not in the normal range, the filling efficiency is considered abnormal, and the level of abnormality is determined by combining the nitrogen filling response delay.

3. The cosmetic filling control method based on a large-scale knowledge graph according to claim 2, characterized in that, The process of determining the level of abnormality in filling efficiency includes: when the dynamic filling efficiency index is in the edge range and the nitrogen charging response delay is less than the calibration delay, the abnormality is determined to be a level one abnormality, and the filling process exhibits efficiency drift or fluctuation; when the dynamic filling efficiency index is lower than or higher than the edge range, or the nitrogen charging response delay is greater than or equal to the calibration delay, the abnormality is determined to be a level two abnormality, and the filling setting speed of the filling screw pump is adjusted; wherein, the edge range is a preset index range adjacent to the normal range.

4. The cosmetic filling control method based on a large-scale knowledge graph according to claim 3, characterized in that, The process of determining the temperature sensitivity of material viscosity includes: the filling efficiency is classified as Level 1 abnormal; the viscosity temperature sensitivity coefficient is determined based on the actual viscosity and average temperature of several sampling points; if the viscosity temperature sensitivity coefficient is greater than the coefficient threshold, it is determined that the material viscosity is highly sensitive to temperature, and heating can effectively reduce viscosity; if the absolute value of the difference between the viscosity temperature sensitivity coefficient and zero is less than the difference evaluation value, it is determined that the material viscosity is not sensitive to temperature, and the shear heat characteristic temperature difference is calculated.

5. The cosmetic filling control method based on a large-scale knowledge graph according to claim 4, characterized in that, The process of determining the cause of the decrease in filling efficiency includes: if the actual viscosity of the essence is greater than the upper limit of the process standard range, and the absolute value of the viscosity temperature sensitivity coefficient is less than the preset sensitivity threshold, and the real-time relative humidity value of the environment is greater than the material's critical humidity for hygroscopic absorption, then it is determined that the current decrease in filling efficiency is due to excessively high ambient humidity, and the hygroscopic raw material components are affected by environmental factors, causing the material viscosity to exceed the process range. The temperature setting value of the compression zone is adjusted to increase the material temperature.

6. The cosmetic filling control method based on a large-scale knowledge graph according to claim 5, characterized in that, When the dynamic filling efficiency index is lower than the normal range, if the shear heat effect index is greater than the preset safety threshold and the actual viscosity is within the standard range, the current efficiency decline is determined to be caused by shear heating. The screw speed is adjusted and the temperature settings of the feeding zone and metering zone are compensated and adjusted.

7. The cosmetic filling control method based on a large-scale knowledge graph according to claim 6, characterized in that, When the change in the temperature setpoint of each heating zone is greater than zero, the current real-time oxygen content of the core area of ​​the filling chamber is obtained; if the current real-time oxygen content is greater than the oxygen content warning value but less than the safety absolute threshold, it is determined that there is a potential conflict between process optimization and oxidation risk, and a quantitative risk indicator assessment is performed; if the current real-time oxygen content is greater than or equal to the safety absolute threshold, the highest level alarm is triggered to stop the filling process.

8. The cosmetic filling control method based on a large-scale knowledge graph according to claim 7, characterized in that, The quantitative risk assessment includes calculating the increase in oxidation risk due to heating and the filling quality risk of maintaining the status quo without heating; If the increase in oxidation risk due to heating is less than the multiple threshold of the filling quality risk of maintaining the status quo without heating, then the filling quality risk is judged to be too high, and the determined temperature setpoints for each heating zone are adjusted. If the increase in oxidation risk due to temperature increase is greater than or equal to the multiple threshold of the filling quality risk of maintaining the status quo without temperature increase, then the oxidation risk caused by temperature increase is judged to be too high. The adjustment of the compression zone temperature setpoint is stopped, and the nitrogen flow rate is increased to redetermine the increase in oxidation risk.

9. The cosmetic filling control method based on a large-scale knowledge graph according to claim 8, characterized in that, When there is a potential conflict between process optimization and oxidation risk, calculate the maximum allowable temperature rise and limit the adjustment of the compression zone temperature setpoint based on the maximum temperature rise.

10. The cosmetic filling control method based on a large-scale knowledge graph according to claim 1, characterized in that, After adjusting the nitrogen flow rate, determine the response efficiency index of the nitrogen flow rate adjustment; if the arithmetic mean of the response efficiency index is greater than the first index threshold, it is judged that the overall efficiency of nitrogen filling is consistently lower than expected, and there is a risk of false alarm. The calibration delay is reduced according to the calibration coefficient. If the arithmetic mean of the response performance index is less than the second index threshold, it is determined that the overall performance of nitrogen filling is consistently higher than expected, and there is a risk of false alarm. The calibration delay is then increased based on the calibration coefficient.

Images