Multi-scenario adaptive perfume dosing method and system

Abstract

This invention relates to the field of filter rod fragrance control, specifically to a multi-scenario adaptive fragrance metering and addition method and system. The method constructs a viscosity prediction model for the fragrance under test based on viscosity data at different ambient temperatures and humidity levels. It predicts the real-time viscosity of the fragrance in the conveying pipeline during filter rod production. For each viscosity of the fragrance under test, it performs linear fitting on the flow rate data at different speeds of the metering pump to obtain the fitting linear parameters for each viscosity. Combining the real-time predicted viscosity and the target set flow rate, it obtains the target feedforward speed of the metering pump. Based on the difference between the real-time flow rate of the fragrance under test in the conveying pipeline collected during filter rod production and the target set flow rate, it obtains the feedback compensation speed of the metering pump. Combining the target feedforward speed and the feedback compensation speed, it controls the speed of the metering pump. This invention can eliminate the problem of fragrance viscosity changes caused by temperature and humidity variations, leading to unstable flow output.

Description

Technical Field

[0001] This invention relates to the field of filter rod fragrance control, specifically to a multi-scenario adaptive fragrance metering and addition method and system. Background Technology

[0002] Adding flavorings to cigarette filters can improve the taste, increase aroma richness, or give it a unique flavor (such as menthol). The taste of cigarettes is a key factor for consumers when choosing a brand. Fluctuations in the amount of flavoring added can lead to differences in taste between different batches of the same brand, or even within the same pack. Adding flavorings to filters (such as flavoring the core wire) is to compensate for the aroma loss due to increased filtration efficiency. Inaccurate measurements can result in insufficient aroma (e.g., a rough taste) or excessive aroma (e.g., masking the original tobacco flavor), ruining the consumer experience.

[0003] In related technologies, fixed metering parameters are usually set, such as a fixed fragrance delivery flow rate or metering pump speed, to control the flavoring of cigarette filters. However, in actual production, changes in temperature and humidity can cause changes in the viscosity of the fragrance, which in turn can lead to errors in the output flow rate of the metering pump. In this case, using a flavoring mechanism with fixed metering parameters (such as metering pump speed) will not be able to adapt to dynamic working conditions, resulting in unstable actual fragrance delivery flow rate and increased loss. Ultimately, this will lead to quality problems such as uneven fragrance content in the filter and batch taste fluctuations. Summary of the Invention

[0004] To address the technical problem that flavoring mechanisms using fixed metering parameters (such as metering pump speed) cannot adapt to dynamic operating conditions determined by temperature and humidity, resulting in unstable actual flavor delivery flow rates, the present invention aims to provide a multi-scenario adaptive flavoring metering and addition method and system. The specific technical solution adopted is as follows:

[0005] This invention proposes a multi-scenario adaptive flavor metering method, the method comprising:

[0006] Acquire viscosity data of the fragrance under test at different ambient temperatures and humidity, and acquire flow rate data of the fragrance under test with different viscosities at different speeds of the metering pump;

[0007] Based on the viscosity data of the fragrance under different ambient temperatures and humidity, a viscosity prediction model for the fragrance is constructed. Using the viscosity prediction model, the viscosity of the fragrance in the conveying pipeline during filter rod production is predicted to obtain the real-time predicted viscosity of the fragrance.

[0008] Linear fitting is performed on the flow data of the test fragrance at different speeds of the metering pump for each viscosity to obtain the fitting line parameters of the test fragrance at each viscosity; based on the fitting line parameters of the test fragrance at each viscosity, and combined with the real-time predicted viscosity and the target set flow rate of the delivery pipeline, the target feedforward speed of the metering pump is obtained.

[0009] Based on the difference between the real-time flow rate of the fragrance to be tested in the delivery pipeline collected during filter rod production and the target set flow rate, the feedback compensation speed of the metering pump is obtained; combined with the target feedforward speed and the feedback compensation speed, the speed of the metering pump is controlled.

[0010] Furthermore, the construction of the viscosity prediction model for the fragrance to be tested includes:

[0011] Subtract the preset standard ambient temperature from the independent variable used to indicate the ambient temperature to obtain the variable used to indicate the temperature deviation. Multiply the variable used to indicate the temperature deviation with the temperature coefficient to obtain the variable used to indicate the temperature influence.

[0012] The difference between the preset standard ambient humidity and the independent variable used to indicate the ambient humidity is used as the variable used to indicate the amount of humidity deviation. The variable used to indicate the amount of humidity deviation is multiplied by the humidity coefficient to obtain the variable used to indicate the amount of humidity influence.

[0013] By using an exponential function with the natural constant as the base, the sum of the variables used to indicate the influence of temperature and the variables used to indicate the influence of humidity is mapped to obtain the variable used to indicate the overall influence.

[0014] A viscosity prediction model for the spice to be tested is constructed. The expression on the left side of the viscosity prediction model is the dependent variable indicating the real-time predicted viscosity of the spice to be tested. The expression on the right side of the viscosity prediction model is the product of the standard viscosity of the spice to be tested at a preset standard ambient temperature and a preset standard ambient humidity and the variable indicating the comprehensive influence quantity.

[0015] Furthermore, the methods for calculating the temperature coefficient and humidity coefficient include:

[0016] Under the preset standard ambient humidity, the viscosity data of the fragrance to be tested at different ambient temperatures were selected as the first viscosity data to be fitted.

[0017] The independent variable in the viscosity prediction model that indicates the ambient humidity is set to a preset standard ambient humidity, so that the viscosity prediction model degenerates into a first viscosity prediction model with respect to temperature;

[0018] Take the natural logarithm on both sides of the first viscosity prediction model, substitute different ambient temperatures into the variable used to indicate the temperature deviation, and perform a straight line fitting on the two-dimensional data points formed by the temperature deviation of each ambient temperature as the abscissa and the natural logarithm of the first viscosity data to be fitted as the ordinate, and use the slope of the fitted straight line as the temperature coefficient.

[0019] At a preset standard ambient temperature, the viscosity data of the fragrance to be tested under different ambient humidity were selected as the second viscosity data to be fitted.

[0020] The independent variable in the viscosity prediction model that indicates the ambient temperature is set to a preset standard ambient temperature, so that the viscosity prediction model degenerates into a second viscosity prediction model with respect to humidity;

[0021] Take the natural logarithm on both sides of the equal sign of the second viscosity prediction model, substitute different ambient humidity into the variable used to indicate the amount of humidity deviation, and perform linear fitting on the two-dimensional data points formed by the amount of humidity deviation of each ambient humidity as the abscissa and the natural logarithm of the second viscosity data to be fitted as the ordinate, and use the slope of the fitted line as the temperature coefficient.

[0022] Furthermore, obtaining the real-time predicted viscosity of the flavoring agent to be tested includes:

[0023] The real-time ambient temperature and humidity collected during the spice delivery process are substituted into the independent variables indicating the ambient temperature and the independent variables indicating the ambient humidity in the viscosity prediction model, respectively, and the viscosity prediction model outputs the real-time predicted viscosity of the spice to be tested.

[0024] Furthermore, obtaining the fitting linear parameters of the fragrance under test at each viscosity includes:

[0025] Using any viscosity as the target viscosity, a straight line is fitted to the two-dimensional data points formed by the different rotation speeds of the metering pump as the abscissa and the flow data of the fragrance under test at different rotation speeds of the metering pump as the ordinate, to obtain the fitting straight line parameters of the fragrance under test at the target viscosity. The fitting straight line parameters include the slope and intercept of the fitted straight line.

[0026] Furthermore, obtaining the target feedforward speed of the metering pump includes:

[0027] If any viscosity exists that is the same as the real-time predicted viscosity of the fragrance to be tested, then the corresponding linear equation is constructed using the fitting linear parameters of the fragrance to be tested under the real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump.

[0028] Otherwise, linear interpolation is performed on the fitting straight line parameters of the fragrance to be tested at various viscosities to obtain the target feedforward speed of the metering pump.

[0029] Furthermore, the process of linearly interpolating the fitting straight line parameters of the fragrance under test at various viscosities to obtain the target feedforward speed of the metering pump includes:

[0030] From all viscosities, the maximum value of the viscosity that is less than the real-time predicted viscosity is selected as the first viscosity, and the minimum value of the viscosity that is greater than the real-time predicted viscosity is selected as the second viscosity.

[0031] Linear interpolation is performed on the fitting line parameters of the first viscosity and the fitting line parameters of the second viscosity to obtain the interpolation fitting line parameters of the fragrance to be tested under the real-time predicted viscosity. The interpolation fitting line parameters include the slope and intercept of the line.

[0032] The corresponding linear equation is constructed using the interpolation fitting linear parameters of the fragrance under real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump.

[0033] Furthermore, obtaining the feedback compensation speed of the metering pump includes:

[0034] The difference between the target set flow rate and the real-time flow rate of the spice to be tested is taken as the real-time flow rate deviation of the spice to be tested in the delivery pipeline.

[0035] The real-time flow deviation is input to the PID controller, which then outputs the feedback compensation speed of the metering pump.

[0036] Furthermore, controlling the rotational speed of the metering pump includes:

[0037] The target feedforward speed and the feedback compensation speed and value of the metering pump are taken as the actual operating speed of the metering pump, and the metering pump operates at the actual operating speed to stabilize the spice flow rate.

[0038] The present invention also proposes a multi-scenario adaptive spice metering and addition system, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any of the steps of a multi-scenario adaptive spice metering and addition method.

[0039] The present invention has the following beneficial effects:

[0040] This invention addresses the issue that existing flavoring mechanisms using fixed metering parameters are ill-suited to dynamic operating conditions determined by temperature and humidity, leading to unstable actual flavor delivery flow. Therefore, this invention first measures the viscosity of the flavoring under different ambient temperatures and humidity levels, as well as the flow rate of flavorings with different viscosities at different metering pump speeds. Considering that ambient temperature and humidity directly affect the viscosity of the flavoring, and that real-time measurement of flavoring viscosity in the delivery pipeline is more difficult and costly than monitoring ambient temperature and humidity during filter rod production, this invention first constructs a viscosity prediction model that can predict the viscosity of the flavoring based on ambient temperature and humidity. This accurately predicts the real-time viscosity of the flavoring determined by the current environmental conditions, providing accurate flavoring viscosity data for subsequent metering pump speed control. This avoids the problem of unstable flavoring flow caused by viscosity changes due to variations in ambient temperature and humidity when using a fixed speed. The metering pump is an actuator whose output flow rate is affected by both rotational speed and medium viscosity. Traditional methods assume constant viscosity and obtain the rotational speed by finding a single mapping curve based on the flow rate, without considering the influence of changes in spice viscosity, which is the root cause of error. This invention uses the estimated real-time viscosity as input and dynamically selects or calculates pump characteristics suitable for the current viscosity, i.e., fitting linear parameters, thereby directly compensating for the significant impact of viscosity changes on the flow rate and obtaining the target feedforward rotational speed of the metering pump. At the same time, considering that the above feedforward control is based on a viscosity prediction model, but cannot completely eliminate all errors (such as model deviation, pump wear, small pressure fluctuations, etc.), this invention further uses the difference between the real-time flow rate and the target set flow rate, and through closed-loop feedback of a PID controller, can continuously correct these residual, unmodeled small disturbances, thereby achieving control of the metering pump's rotational speed, ensuring that the flow rate is accurately locked at the set value, and eliminating the problem of unstable flow output caused by changes in spice viscosity due to temperature and humidity variations. Attached Figure Description

[0041] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of a multi-scenario adaptive spice metering and adding method provided in one embodiment of the present invention;

[0043] Figure 2This is a schematic diagram illustrating the results of linear fitting of flow data of various viscosities of the tested fragrance at different speeds of the metering pump, according to an embodiment of the present invention.

[0044] Figure 3 This is a framework diagram of a multi-scenario adaptive spice metering and adding system provided in one embodiment of the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a multi-scenario adaptive flavor metering and addition method and system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] The following description, in conjunction with the accompanying drawings, details the specific solution of the multi-scenario adaptive flavor metering and addition method and system provided by the present invention.

[0048] Please see Figure 1 The diagram illustrates a flowchart of a multi-scenario adaptive spice metering method according to an embodiment of the present invention, the method comprising:

[0049] Step S1: Obtain the viscosity data of the fragrance to be tested under different ambient temperatures and humidity, and obtain the flow rate data of the fragrance to be tested with different viscosities at different speeds of the metering pump.

[0050] In the production of filter rods, flavorings are often added to enhance aroma richness or impart unique flavor (such as adding fragrance to the core wire). However, if the flavoring is not accurately measured, it can lead to quality problems such as uneven flavoring content and batch-to-batch taste fluctuations. Therefore, existing technologies typically set a fixed speed for the metering pump that delivers the flavoring to maintain a stable flow rate. However, in the actual production process of filter rods, changes in temperature and humidity can alter the viscosity of the flavoring, which in turn can cause errors in the output flow rate of the metering pump. At this point, using a flavoring mechanism with fixed metering parameters (such as the metering pump speed) will not be able to adapt to dynamic operating conditions, resulting in unstable actual flow rate of the flavoring and thus causing quality problems in filter rod production.

[0051] Considering that in the actual production process of filter rods, measuring the viscosity of fragrances in the delivery pipeline in real time is more difficult and costly than measuring the ambient temperature and humidity, this embodiment of the invention first uses a viscometer (e.g., a rotational viscometer) in the laboratory to measure the viscosity data of the fragrance under different ambient temperatures and humidity levels. The viscosity data is stored in the form of a matrix, where the rows of the matrix can be ambient temperature or ambient humidity, and the corresponding columns can be ambient humidity or ambient temperature. The elements in the matrix are the viscosity data under the corresponding ambient temperature and humidity. The ambient temperature and ambient humidity are discrete data, and both need to be determined based on the specific production environment. For example, in one embodiment of the invention, the expected temperature range of the production environment is 20℃~30℃, and the ambient temperature step size is 1℃, so the temperature is selected every 1℃. The expected humidity range of the production environment is 50%RH~70%RH, and the ambient humidity step size is 2%RH, so the humidity is selected every 2%RH. The expected temperature range, temperature step size, humidity range, and humidity step size of the production environment can all be set by the implementer according to the specific implementation scenario, and are not limited here.

[0052] At the same time, it is necessary to ensure that the preset standard ambient temperature and the preset standard ambient humidity are within the temperature range and humidity range, so as to facilitate the construction of the viscosity prediction model. In one embodiment of the present invention, the preset standard ambient temperature is set to 25°C and the preset standard ambient humidity is set to 60%RH. The preset standard ambient temperature and preset standard ambient humidity can also be set by the implementer according to the specific implementation scenario, and are not limited here.

[0053] Then, in this embodiment of the invention, multiple (e.g., 10, which can be set according to the specific scenario) test media with different viscosities are prepared for the same batch of fragrance using an external temperature control device (e.g., water bath / oil bath) and physical dilution / thickening methods. The viscosity range should cover the viscosity range that the fragrance to be tested may have under all possible production environments (e.g., 20℃~30℃, 50%RH~70%RH). The actual viscosity of each medium needs to be accurately measured and confirmed using a laboratory rotational viscometer. Then, for any test medium of any viscosity, the metering pump speed is set from the lowest operating speed to the highest operating speed, increasing in fixed steps (e.g., 100 rpm, which can be set according to the specific scenario). At each set speed, after the flow rate stabilizes, the flow rate data of the test medium of that viscosity at that speed is measured using a high-precision mass flow meter. The flow rate data is also stored in the form of a matrix, where the rows of the matrix can be viscosity or speed, and the corresponding columns of the matrix can be speed or viscosity. The elements in the matrix are the flow rate data at the corresponding viscosity and speed.

[0054] Step S2: Based on the viscosity data of the fragrance to be tested under different ambient temperatures and humidity, construct a viscosity prediction model for the fragrance to be tested. Use the viscosity prediction model to predict the viscosity of the fragrance to be tested in the conveying pipeline of the filter rod production, and obtain the real-time predicted viscosity of the fragrance to be tested.

[0055] Since the temperature and humidity of the surrounding environment can directly affect the viscosity of the fragrance to be tested, and in the filter rod production process, it is more difficult and costly to measure the viscosity of the fragrance in the delivery pipeline in real time than to detect the temperature and humidity of the surrounding environment, this embodiment of the invention first constructs a viscosity prediction model based on the viscosity data of the fragrance to be tested under different ambient temperatures and humidity, which can predict the viscosity of the fragrance to be tested based on the ambient temperature and humidity. Subsequently, the viscosity prediction model can be used, combined with the temperature and humidity of the surrounding environment during the filter rod production process, to accurately predict the real-time viscosity of the fragrance to be tested.

[0056] Preferably, in one embodiment of the present invention, the method for constructing a viscosity prediction model for the fragrance to be tested specifically includes:

[0057] First, subtract the preset standard ambient temperature from the independent variable used to indicate the ambient temperature to obtain the variable used to indicate the temperature deviation. Multiply the variable used to indicate the temperature deviation with the temperature coefficient to obtain the variable used to indicate the temperature influence. The temperature coefficient is used to reflect the intensity of the temperature influence on viscosity, representing the rate of change of the natural logarithm of viscosity for every 1°C deviation of the temperature from the preset standard ambient temperature. It can be obtained later through linear fitting. The temperature influence is used to reflect the influence of temperature change on the viscosity of the tested fragrance.

[0058] The difference between the preset standard ambient humidity and the independent variable used to indicate ambient humidity is used as the variable used to indicate the humidity deviation. The variable used to indicate the humidity deviation is multiplied by the humidity coefficient to obtain the variable used to indicate the humidity effect. The humidity coefficient is used to reflect the intensity of the influence of humidity on viscosity, representing the rate of change of the natural logarithm of viscosity for every 1%RH deviation of humidity from the preset standard ambient humidity. It can be obtained by linear fitting in the future. The humidity effect is used to reflect the influence of humidity change on the viscosity of the tested fragrance.

[0059] Then, considering that fluid viscosity often satisfies an approximate Arrhenius exponential relationship with temperature and humidity, an exponential function with the natural constant as the base can fit this nonlinear change well. Therefore, the sum of the variables used to indicate the influence of temperature and the variables used to indicate the influence of humidity can be mapped using an exponential function with the natural constant as the base to obtain a variable used to indicate the comprehensive influence. The comprehensive influence reflects the influence of the combined changes in temperature and humidity on the viscosity of the spice being tested.

[0060] Furthermore, a viscosity prediction model for the spice to be tested is constructed. The expression on the left side of the viscosity prediction model is the dependent variable used to indicate the real-time predicted viscosity of the spice to be tested. The expression on the right side of the viscosity prediction model is the product of the standard viscosity of the spice to be tested under standard temperature and standard humidity and the variable used to indicate the comprehensive influence quantity. The standard viscosity of the spice to be tested under standard temperature and standard humidity is a known value and can be measured in step S1.

[0061] As an example, in one embodiment of the present invention, the viscosity prediction model expression for the fragrance to be tested can be specifically as follows:

[0062]

[0063] in, This represents the dependent variable used to indicate the real-time predicted viscosity of the spice being tested; The independent variable indicating ambient temperature; This represents the independent variable used to indicate ambient humidity. This indicates the standard viscosity of the fragrance being tested at a preset standard ambient temperature and preset standard ambient humidity. Indicates the preset standard ambient temperature; Indicates the preset standard ambient humidity; Indicates the temperature coefficient. It is always a positive value because an increase in temperature always leads to a decrease in viscosity; Indicates the humidity coefficient. The symbol and size are determined by the specific physicochemical properties of the spice; Represented by natural constant An exponential function with base 0.

[0064] Among them, when When the ambient temperature is significantly higher than the preset standard ambient temperature, the difference between the two is negative, the exponent is negative, and exp(negative number) < 1. Therefore... This conforms to the rule that viscosity decreases as temperature increases; when When the ambient humidity is higher than the preset standard ambient humidity (the difference is negative), for most hygroscopic alcohols and alcohol ether fragrance solvents, increased humidity will cause them to absorb moisture, thereby reducing the solution viscosity. (Usually positive values), but for a few special formulations, humidity may affect viscosity through other means (such as crystallization, phase separation), in which case... The value may be negative or zero, and needs to be determined through subsequent steps.

[0065] This invention requires further determination of the temperature coefficient and humidity coefficient in the viscosity prediction model. Preferably, in one embodiment of this invention, the method for obtaining the temperature coefficient and humidity coefficient specifically includes:

[0066] First, the temperature coefficient needs to be determined. Before determining the temperature coefficient, the ambient humidity needs to be fixed to a preset standard ambient humidity using the controlled variable method. Under the preset standard ambient humidity, the viscosity data of the fragrance to be tested at different ambient temperatures are selected as the first viscosity data to be fitted.

[0067] By setting the independent variable in the viscosity prediction model that indicates ambient humidity to a preset standard ambient humidity, the viscosity prediction model degenerates into a first viscosity prediction model with respect to temperature.

[0068] The expression for the first viscosity prediction model is:

[0069]

[0070] in, This represents the dependent variable used to indicate the real-time predicted viscosity of the tested fragrance, without considering changes in ambient humidity.

[0071] Taking the natural logarithm on both sides of the equality of the first viscosity prediction model, the result is:

[0072]

[0073] As can be seen from the results of taking the natural logarithm above, the expression is about The equation of the straight line, It is the slope of the straight line, so different ambient temperatures can be substituted into the variable used to indicate the amount of temperature deviation. In the process, a straight line is fitted to the two-dimensional data points formed by the temperature deviation of each ambient temperature as the abscissa and the natural logarithm of the first viscosity data to be fitted at each ambient temperature as the ordinate, and the slope of the fitted straight line is used as the temperature coefficient.

[0074] Then, the humidity coefficient is determined using the same method described above. First, at a preset standard ambient temperature, the viscosity data of the fragrance to be tested under different ambient humidity are selected as the second viscosity data to be fitted.

[0075] By setting the independent variable in the viscosity prediction model that indicates ambient temperature to a preset standard ambient temperature, the viscosity prediction model degenerates into a second viscosity prediction model with respect to humidity.

[0076] The expression for the second viscosity prediction model is:

[0077]

[0078] in, This represents the dependent variable used to indicate the real-time predicted viscosity of the tested fragrance, without considering changes in ambient temperature.

[0079] Taking the natural logarithm on both sides of the equality of the second viscosity prediction model, the result is:

[0080]

[0081] Similarly, as can be seen from the results of taking the natural logarithm above, the expression is... about The equation of the straight line, It is the slope of the straight line, so different ambient humidity levels can be substituted into the variable used to indicate humidity deviation. In the process, a straight line is fitted to the two-dimensional data points formed by the humidity deviation of each ambient humidity as the abscissa and the natural logarithm of the second viscosity data to be fitted for each ambient humidity as the ordinate, and the slope of the fitted straight line is used as the humidity coefficient.

[0082] In one embodiment of the present invention, line fitting can be achieved using, for example, the least squares method or other line fitting methods, without limitation, and the line fitting method in subsequent steps can also be achieved using the least squares method.

[0083] Once the viscosity prediction model of the fragrance to be tested is obtained, the viscosity of the fragrance to be tested in the conveying pipeline of the filter rod production can be predicted using the viscosity prediction model, thereby obtaining the real-time predicted viscosity of the fragrance to be tested. This provides an accurate data basis for the fragrance viscosity of the metering pump for subsequent speed control, avoiding the problem of unstable fragrance flow due to fragrance viscosity changes caused by changes in ambient temperature and humidity, which would otherwise require a fixed speed.

[0084] Preferably, in one embodiment of the present invention, the method for obtaining the real-time predicted viscosity of the fragrance to be tested specifically includes:

[0085] The real-time ambient temperature and humidity collected during the fragrance transportation process are substituted into the independent variables for indicating ambient temperature and ambient humidity in the viscosity prediction model, respectively. The viscosity prediction model outputs the real-time predicted viscosity of the fragrance to be tested. In the actual production process of the filter rod, temperature and humidity sensors can be installed near the inlet of the fragrance transportation pipeline to collect the real-time ambient temperature and humidity during the fragrance transportation process.

[0086] Thus, the viscosity of the tested fragrance under current environmental conditions was predicted using the constructed model.

[0087] Step S3: Perform linear fitting on the flow data of the fragrance under test at different speeds of the metering pump for each viscosity to obtain the fitting linear parameters of the fragrance under test at each viscosity; based on the fitting linear parameters of the fragrance under test at each viscosity, and combined with the real-time predicted viscosity and the target set flow rate of the delivery pipeline, obtain the target feedforward speed of the metering pump.

[0088] Since metering pumps (such as gear pumps, plunger pumps, or peristaltic pumps) are actuators, their output flow rate is affected by both rotational speed and medium viscosity. Traditional methods assume constant viscosity and obtain rotational speed by finding a single mapping curve based on flow rate, without considering the influence of changes in fragrance viscosity, which is the root cause of error. Furthermore, considering that under ideal conditions where fragrance viscosity remains constant, the flow rate of the fragrance under the action of the metering pump exhibits a highly linear proportional relationship with the pump's rotational speed, this invention performs linear fitting on the flow rate data of each viscosity of the fragrance under test at different rotational speeds of the metering pump to obtain the fitting linear parameters for each viscosity. The fitting linear parameters for a certain viscosity determine a straight line that reflects the relationship between the fragrance flow rate and the metering pump rotational speed when the metering pump acts on the fragrance of that viscosity. Subsequently, the real-time predicted viscosity of the fragrance under test obtained above can be combined to dynamically select or calculate the pump characteristics suitable for the current viscosity, i.e., the fitting linear parameters, thereby directly compensating for the significant impact of viscosity changes on flow rate.

[0089] Preferably, in one embodiment of the present invention, the method for obtaining the fitting line parameters of the fragrance to be tested at each viscosity specifically includes:

[0090] Using any viscosity as the target viscosity, a straight line is fitted to the two-dimensional data points formed by the different rotation speeds of the metering pump as the abscissa and the flow data of the fragrance under test at different rotation speeds of the metering pump as the ordinate. The fitted straight line parameters of the fragrance under test at the target viscosity are obtained, including the slope and intercept of the fitted straight line.

[0091] The fitting linear parameters for the tested fragrance at each viscosity can be obtained using the same method described above. (See also...) Figure 2 The diagram illustrates the results of linear fitting of flow data of various viscosities of the tested fragrance at different speeds of the metering pump, according to an embodiment of the present invention. Each straight line represents the linear fitting result corresponding to a certain viscosity.

[0092] After obtaining the fitting linear parameters (i.e. pump characteristics) of the fragrance under test at each viscosity, the target feedforward speed of the metering pump can be obtained based on the fitting linear parameters of the fragrance under test at each viscosity, combined with the real-time predicted viscosity and the target set flow rate of the delivery pipeline. Thus, the target feedforward speed is used to actively offset the main flow deviation caused by viscosity changes. The target set flow rate is a known fixed value set for the system to control the fragrance flow rate in the delivery pipeline to remain at the target set flow rate.

[0093] Preferably, in one embodiment of the present invention, the method for obtaining the target feedforward speed of the metering pump specifically includes:

[0094] If there is a viscosity among all the viscosities that is the same as the real-time predicted viscosity of the fragrance to be tested, then the corresponding linear equation can be constructed directly using the fitting linear parameters of the fragrance to be tested under the real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump.

[0095] The specific process is as follows: First, the linear equation constructed using the fitting linear parameters of the fragrance to be tested at the real-time predicted viscosity is:

[0096]

[0097] in, This represents the dependent variable indicating the flow rate of the tested fragrance at the predicted viscosity in real time; The independent variable representing the rotational speed of the metering pump; and This represents the parameters of the fitted straight line for the tested fragrance at the real-time predicted viscosity, namely the slope and intercept.

[0098] Then substitute the target traffic setting. Solving by reverse engineering The target feedforward speed can then be calculated.

[0099] Otherwise, if there is no viscosity among all the viscosities that is the same as the real-time predicted viscosity of the fragrance to be tested, then linear interpolation can be performed on the fitting straight line parameters of the fragrance to be tested at each viscosity to obtain the target feedforward speed of the metering pump.

[0100] Preferably, in one embodiment of the present invention, the method for obtaining the target feedforward speed of the metering pump in this case specifically includes:

[0101] From all viscosities, the maximum value of the viscosity that is less than the real-time predicted viscosity is selected as the first viscosity, and the minimum value of the viscosity that is greater than the real-time predicted viscosity is selected as the second viscosity. At this time, the first viscosity and the second viscosity are the two viscosities that are closest to the real-time predicted viscosity, and the real-time predicted viscosity is between the first viscosity and the second viscosity.

[0102] Linear interpolation is performed on the fitting line parameters of the first viscosity and the fitting line parameters of the second viscosity to obtain the interpolated fitting line parameters of the fragrance under the real-time predicted viscosity. The interpolated fitting line parameters include the slope and intercept of the line.

[0103] The calculation method for the interpolation fitting line parameters is as follows:

[0104]

[0105]

[0106] in, and The parameters represent the interpolation fitting line parameters of the fragrance under test at the real-time predicted viscosity, which correspond to the slope and intercept of the line determined by the two parameters, respectively. The slope of the fitted line is represented by the parameters of the fitted line for the first viscosity. The slope of the fitted line is represented by the parameters of the fitted line for the second viscosity. This indicates the real-time predicted viscosity of the fragrance being tested; Indicates the first viscosity; Indicates the second viscosity; The intercept of the line fitted by the fitting line parameters for the first viscosity. The intercept of the line fitted by the parameters of the fitted line for the second viscosity is expressed as the intercept of the line fitted by the parameters of the fitted line.

[0107] in, As a scaling factor, it calculates the real-time predicted viscosity. Within the known viscosity range The relative position in the equation is a proportional value between 0 and 1. , The slope represents the slope over the entire known viscosity range. The total change within, This reflects from Starting from the beginning, the change that needs to be increased (or decreased) needs to be addressed; similarly, for... The analytical logic is the same as above.

[0108] Then, the corresponding linear equation is constructed by using the interpolation fitting linear parameters of the fragrance under real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump.

[0109] The specific process is as follows: First, the linear equation constructed by interpolating and fitting the linear parameters of the fragrance under test at the real-time predicted viscosity is:

[0110]

[0111] in, This represents the dependent variable indicating the flow rate of the tested fragrance at the predicted viscosity in real time; The independent variable representing the rotational speed of the metering pump; and This represents the parameters of the interpolation fitting line for the fragrance under test at the real-time predicted viscosity, namely the corresponding slope and intercept.

[0112] Then substitute the target traffic setting. Solving by reverse engineering The target feedforward speed can then be calculated.

[0113] It should be noted that during the interpolation process, if the real-time predicted viscosity is lower than the minimum of all known viscosities, the corresponding linear equation can be constructed directly using the fitting linear parameters corresponding to the minimum viscosity. The target set flow rate is then substituted into the linear equation, and the independent variable obtained by inverse calculation is used as the target feedforward speed of the metering pump. Similarly, if the real-time predicted viscosity is higher than the maximum of all known viscosities, the corresponding linear equation can be constructed directly using the fitting linear parameters corresponding to the maximum viscosity. The target set flow rate is then substituted into the linear equation, and the independent variable obtained by inverse calculation is used as the target feedforward speed of the metering pump to eliminate boundary problems.

[0114] At this point, the target feedforward speed of the metering pump has been obtained, and the target feedforward speed can be used to actively offset the main flow deviation caused by viscosity changes.

[0115] Step S4: Based on the difference between the real-time flow rate of the fragrance to be tested in the conveying pipeline collected during filter rod production and the target set flow rate, obtain the feedback compensation speed of the metering pump; combine the target feedforward speed and the feedback compensation speed to control the speed of the metering pump.

[0116] The target feedforward speed of the metering pump, calculated above, can effectively offset the flow instability caused by changes in fragrance viscosity due to variations in ambient temperature and humidity. However, it cannot completely eliminate all errors (such as model deviations, pump wear, and minor pressure fluctuations). Therefore, in order to further control the stability of the fragrance flow rate in the conveying pipeline during filter rod production, this embodiment of the invention needs to obtain the feedback compensation speed of the metering pump based on the difference between the real-time flow rate of the fragrance to be tested in the conveying pipeline collected during filter rod production and the target set flow rate. Subsequently, the operating speed of the metering pump can be determined more accurately by combining the feedback compensation speed and the target feedforward speed to achieve high-precision flow stability. This can be achieved by installing a high-precision flow meter at the output end of the metering pump and using the flow meter to collect the real-time flow rate of the fragrance to be tested in the conveying pipeline collected during filter rod production.

[0117] Preferably, in one embodiment of the present invention, the method for obtaining the feedback compensation speed of the metering pump specifically includes:

[0118] Since the system expects the flow rate of the spice to be tested in the delivery pipeline to be stable at the target set flow rate, the difference between the target set flow rate and the real-time flow rate of the spice to be tested can be used as the real-time flow deviation of the spice to be tested in the delivery pipeline. The real-time flow deviation quantifies the gap between the current flow performance of the system and the ideal target. By working to eliminate this deviation, the system can achieve accurate tracking.

[0119] The PID controller can effectively compensate for the metering pump speed based on the system's flow error using the proportional + integral + derivative rule, thereby stabilizing the spice flow. Therefore, the real-time flow deviation can be input into the PID controller, which will output the feedback compensation speed of the metering pump.

[0120] The target feedforward speed is used to eliminate the flow deviation caused by the viscosity change of the spice being tested, while the feedback compensation speed is used to compensate for the flow deviation caused by factors such as model deviation, pump wear, and small pressure fluctuations. Therefore, the two can be combined to control the speed of the metering pump, so that the metering pump can maintain a stable flow output after acting on the spice being tested.

[0121] Preferably, in one embodiment of the present invention, the method for controlling the rotational speed of the metering pump specifically includes:

[0122] The target feedforward speed and feedback compensation speed and value of the metering pump are taken as the actual operating speed of the metering pump. The metering pump operates at the actual operating speed to stabilize the spice flow rate.

[0123] This invention also provides a multi-scenario adaptive spice metering and addition system. Please refer to [link / reference]. Figure 3 The diagram illustrates a framework of a multi-scenario adaptive spice metering and adding system according to an embodiment of the present invention. The system may include: a memory 201, a processor 202, and a computer program stored on the memory 201 and capable of running on the processor 202. When the processor 202 executes the program, it implements the multi-scenario adaptive spice metering and adding method provided in the above embodiment.

[0124] Furthermore, the system also includes a communication interface 203 for communication between the memory 201 and the processor 202.

[0125] The memory 201 may include high-speed RAM memory, and may also include nonvolatile memory, such as at least one disk storage.

[0126] If the memory 201, processor 202, and communication interface 203 are implemented independently, then the communication interface 203, memory 201, and processor 202 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 A bus is represented by a single line, but this does not mean that there is only one bus or one type of bus.

[0127] Optionally, in a specific implementation, if the memory 201, processor 202, and communication interface 203 are integrated on a single chip, then the memory 201, processor 202, and communication interface 203 can communicate with each other through an internal interface.

[0128] The processor 202 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0129] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0130] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A multi-scenario adaptive spice metering method, characterized in that, The method includes: Acquire viscosity data of the fragrance under test at different ambient temperatures and humidity, and acquire flow rate data of the fragrance under test with different viscosities at different speeds of the metering pump; Based on the viscosity data of the fragrance under different ambient temperatures and humidity, a viscosity prediction model for the fragrance is constructed. Using the viscosity prediction model, the viscosity of the fragrance in the conveying pipeline during filter rod production is predicted to obtain the real-time predicted viscosity of the fragrance. Linear fitting is performed on the flow data of the test fragrance at different speeds of the metering pump for each viscosity to obtain the fitting line parameters of the test fragrance at each viscosity; based on the fitting line parameters of the test fragrance at each viscosity, and combined with the real-time predicted viscosity and the target set flow rate of the delivery pipeline, the target feedforward speed of the metering pump is obtained. Based on the difference between the real-time flow rate of the fragrance to be tested in the delivery pipeline collected during filter rod production and the target set flow rate, the feedback compensation speed of the metering pump is obtained; combined with the target feedforward speed and the feedback compensation speed, the speed of the metering pump is controlled.

2. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The viscosity prediction model for the spice to be tested includes: Subtract the preset standard ambient temperature from the independent variable used to indicate the ambient temperature to obtain the variable used to indicate the temperature deviation. Multiply the variable used to indicate the temperature deviation with the temperature coefficient to obtain the variable used to indicate the temperature influence. The difference between the preset standard ambient humidity and the independent variable used to indicate the ambient humidity is used as the variable used to indicate the amount of humidity deviation. The variable used to indicate the amount of humidity deviation is multiplied by the humidity coefficient to obtain the variable used to indicate the amount of humidity influence. By using an exponential function with the natural constant as the base, the sum of the variables used to indicate the influence of temperature and the variables used to indicate the influence of humidity is mapped to obtain the variable used to indicate the overall influence. A viscosity prediction model for the spice to be tested is constructed. The expression on the left side of the viscosity prediction model is the dependent variable indicating the real-time predicted viscosity of the spice to be tested. The expression on the right side of the viscosity prediction model is the product of the standard viscosity of the spice to be tested at a preset standard ambient temperature and a preset standard ambient humidity and the variable indicating the comprehensive influence quantity.

3. The multi-scenario adaptive flavor metering method according to claim 2, characterized in that, The methods for calculating the temperature coefficient and humidity coefficient include: Under the preset standard ambient humidity, the viscosity data of the fragrance to be tested at different ambient temperatures were selected as the first viscosity data to be fitted. The independent variable in the viscosity prediction model that indicates the ambient humidity is set to a preset standard ambient humidity, so that the viscosity prediction model degenerates into a first viscosity prediction model with respect to temperature; Take the natural logarithm on both sides of the first viscosity prediction model, substitute different ambient temperatures into the variable used to indicate the temperature deviation, and perform a straight line fitting on the two-dimensional data points formed by the temperature deviation of each ambient temperature as the abscissa and the natural logarithm of the first viscosity data to be fitted as the ordinate, and use the slope of the fitted straight line as the temperature coefficient. At a preset standard ambient temperature, the viscosity data of the fragrance to be tested under different ambient humidity were selected as the second viscosity data to be fitted. The independent variable in the viscosity prediction model that indicates the ambient temperature is set to a preset standard ambient temperature, so that the viscosity prediction model degenerates into a second viscosity prediction model with respect to humidity; Take the natural logarithm on both sides of the equal sign of the second viscosity prediction model, substitute different ambient humidity into the variable used to indicate the amount of humidity deviation, and perform linear fitting on the two-dimensional data points formed by the amount of humidity deviation of each ambient humidity as the abscissa and the natural logarithm of the second viscosity data to be fitted as the ordinate, and use the slope of the fitted line as the temperature coefficient.

4. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The real-time predicted viscosity of the spice to be tested includes: The real-time ambient temperature and humidity collected during the spice delivery process are substituted into the independent variables indicating the ambient temperature and the independent variables indicating the ambient humidity in the viscosity prediction model, respectively, and the viscosity prediction model outputs the real-time predicted viscosity of the spice to be tested.

5. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The parameters for obtaining the fitted linear parameters of the fragrance under test at each viscosity include: Using any viscosity as the target viscosity, a straight line is fitted to the two-dimensional data points formed by the different rotation speeds of the metering pump as the abscissa and the flow data of the fragrance under test at different rotation speeds of the metering pump as the ordinate, to obtain the fitting straight line parameters of the fragrance under test at the target viscosity. The fitting straight line parameters include the slope and intercept of the fitted straight line.

6. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The target feedforward speed of the metering pump includes: If any viscosity is found that matches the real-time predicted viscosity of the fragrance to be tested, then the corresponding linear equation is constructed using the fitting linear parameters of the fragrance to be tested at the real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump. Otherwise, the fitting linear parameters of the fragrance to be tested at each viscosity are linearly interpolated to obtain the target feedforward speed of the metering pump.

7. The multi-scenario adaptive flavor metering method according to claim 6, characterized in that, The process of linearly interpolating the fitting straight line parameters of the fragrance under test at various viscosities to obtain the target feedforward speed of the metering pump includes: From all viscosities, the maximum value of the viscosity that is less than the real-time predicted viscosity is selected as the first viscosity, and the minimum value of the viscosity that is greater than the real-time predicted viscosity is selected as the second viscosity. Linear interpolation is performed on the fitting line parameters of the first viscosity and the fitting line parameters of the second viscosity to obtain the interpolation fitting line parameters of the fragrance to be tested under the real-time predicted viscosity. The interpolation fitting line parameters include the slope and intercept of the line. The corresponding linear equation is constructed using the interpolation fitting linear parameters of the fragrance under real-time predicted viscosity. The target set flow rate is substituted into the linear equation, and the independent variable obtained by inverse solution is used as the target feedforward speed of the metering pump.

8. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The feedback compensation speed of the metering pump includes: The difference between the target set flow rate and the real-time flow rate of the spice to be tested is taken as the real-time flow rate deviation of the spice to be tested in the delivery pipeline. The real-time flow deviation is input to the PID controller, which then outputs the feedback compensation speed of the metering pump.

9. The multi-scenario adaptive flavor metering method according to claim 1, characterized in that, The control of the metering pump speed includes: The target feedforward speed and the feedback compensation speed and value of the metering pump are taken as the actual operating speed of the metering pump, and the metering pump operates at the actual operating speed to stabilize the spice flow rate.

10. A multi-scenario adaptive spice metering and adding system, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 9.

Images