A method and system for identifying a matched tobacco appliance based on a metal heating induction body

Abstract

The application provides a metal heating induction body-based matching tobacco appliance identification method and system, belonging to the technical field of data identification, multiple temperature points are set during heating of the metal heating induction body, real-time temperature parameters and electromagnetic parameters of the metal heating induction body are acquired, and a temperature-electromagnetic mapping table is established; a three-dimensional parameter grid is constructed based on the electromagnetic parameters, the three-dimensional parameter grid is equally divided, an average grid is acquired, a coupling change rate matrix is constructed according to the average grid; the parameter coupling change rate matrix is subjected to eigenvalue decomposition, a deformation feature vector is acquired, and the deformation feature vector is compared with a pre-stored appliance mode identification vector to acquire a mode similarity score; an appliance judgment mode is acquired based on the mode similarity score, and a mode mechanism is executed according to the appliance judgment mode; the temperature drift interference on the identification result is reduced, the problem that a traditional identification mode relies on voltage, current and resistance value detection is solved, and the misjudgment probability is reduced.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of data recognition, and more particularly relates to a matching tobacco appliance recognition method and system based on a metal heating induction body. BACKGROUND

[0002] With the development of tobacco product technology, heat-not-burn tobacco appliances gradually become one of the mainstream products in the market because they can reduce harmful substances generated by traditional combustion. Such appliances heat a metal heating induction body through a heating module, so that the metal heating induction body releases aerosol in a non-combustion state. The use experience and safety of the metal heating induction body and the tobacco appliance are highly dependent on the matching of the metal heating induction body and the tobacco appliance.

[0003] Currently, there are various types of heat-not-burn tobacco appliances and matching metal heating induction bodies on the market. Different brands and specifications of appliances and metal heating induction bodies have significant differences in material, structure, and thermal response characteristics. If a non-matching metal heating induction body is combined with a tobacco appliance, it may lead to uncontrolled heating temperature. For example, a non-matching combination may cause the metal heating induction body to have a local temperature that is too high, causing carbonization of the metal heating induction body or an increase in the release of harmful substances. Or the heating temperature is insufficient, resulting in insufficient aerosol generation, affecting the use experience. In the prior art, the recognition of matching tobacco appliances is performed by detecting the voltage, current, and resistance value through an intelligent chip. However, the voltage, current, and resistance value are greatly affected by environmental temperature and aging degree, resulting in low recognition accuracy and easy misjudgment. SUMMARY

[0004] To solve the above technical problems, the present application provides a matching tobacco appliance recognition method and system based on a metal heating induction body to solve the technical problem in the prior art that the recognition of traditional matching tobacco appliances is performed by detecting the voltage, current, and resistance value through an intelligent chip. However, the voltage, current, and resistance value are greatly affected by environmental temperature and aging degree, resulting in low recognition accuracy and easy misjudgment.

[0005] The purpose and effect of the present application, a matching tobacco appliance recognition method and system based on a metal heating induction body, are achieved by the following specific technical means:

[0006] A matching tobacco appliance recognition method based on a metal heating induction body, the method comprising:

[0007] Establishing a temperature-electromagnetic mapping table by setting multiple temperature points and obtaining real-time temperature parameters and electromagnetic parameters of the metal heating induction body during the heating process of the metal heating induction body;

[0008] constructing a three-dimensional parameter grid based on the electromagnetic parameters, performing mean value division on the three-dimensional parameter grid to obtain a mean value grid, and constructing a coupling change rate matrix based on the mean value grid;

[0009] performing eigenvalue decomposition on the parameter coupling change rate matrix to obtain a deformation feature vector, and comparing the deformation feature vector with a pre-stored instrument pattern recognition vector to obtain a pattern similarity score;

[0010] obtaining an instrument determination mode based on the pattern similarity score, and performing a pattern mechanism according to the instrument determination mode.

[0011] According to a preferred embodiment, the multiple temperature points are set during the heating of the metal heating induction body, and the real-time temperature parameters and electromagnetic parameters of the metal heating induction body are obtained, and a temperature-electromagnetic mapping table is established, which includes:

[0012] obtaining the initial temperature parameters of the metal heating induction body, and heating the metal heating induction body to 400 degrees Celsius, and obtaining the real-time temperature parameters during the heating process;

[0013] Based on the initial temperature parameters and the real-time temperature parameters, multiple temperature points are set, including an initial temperature point and multiple temperature rise points, the initial temperature point represents the temperature point at the initial temperature parameter, and the temperature rise point represents the temperature point of each 5 degrees Celsius rise of the real-time temperature parameter compared with the initial temperature parameter;

[0014] Based on multiple temperature points, corresponding electromagnetic parameters are obtained, and the temperature-electromagnetic mapping table is generated according to multiple temperature points and corresponding electromagnetic parameters.

[0015] According to a preferred embodiment, the three-dimensional parameter grid is constructed based on the electromagnetic parameters, the mean value grid is obtained by performing mean value division on the three-dimensional parameter grid, and the coupling change rate matrix is constructed based on the mean value grid, which includes:

[0016] Based on the electromagnetic parameters and the temperature-electromagnetic mapping table, normalized temperature parameters, impedance phase product parameters and frequency offset parameters are obtained, a three-dimensional parameter grid is constructed based on the normalized temperature parameters, the impedance phase product parameters and the frequency offset parameters, the X-axis of the three-dimensional parameter grid is the normalized temperature parameter, the Y-axis is the impedance phase product parameter, and the Z-axis is the frequency offset parameter, the mean value grid is obtained by performing mean value division on the three-dimensional parameter grid;

[0017] Based on the six-neighborhood center difference, the mean value grid is subjected to partial derivative operation and voxel partial derivative data is obtained, and the parameter coupling change rate matrix is constructed based on the voxel partial derivative data.

[0018] According to a preferred embodiment, the obtaining the mean grid comprises:

[0019] The electromagnetic parameters include impedance amplitude parameters, phase angle parameters, and resonance frequency parameters.

[0020] The temperature points contained in the temperature-electromagnetic parameter mapping table are extracted, the set temperature parameters of the temperature points are linearly mapped to the interval [0, 1], and the normalized temperature parameters are obtained.

[0021] The impedance amplitude parameters and the phase angle parameters corresponding to each temperature point in the temperature-electromagnetic parameter mapping table are extracted, and the impedance phase product parameters are obtained based on the impedance amplitude parameters and the phase angle parameters.

[0022] The resonance frequency parameters corresponding to each temperature point in the temperature-electromagnetic parameter mapping table are extracted, and a reference resonance frequency parameter is obtained, the reference resonance frequency parameter is represented as a resonance frequency parameter obtained based on an initial temperature parameter, and a frequency deviation parameter is obtained according to the resonance frequency parameter and the reference resonance frequency parameter.

[0023] A three-dimensional parameter grid is constructed based on the frequency deviation parameter, the impedance phase product parameter, and the normalized temperature parameter.

[0024] The three-dimensional parameter grid is equally divided along the X / Y / Z axes into a 30x30x30 voxel grid, a single grid in the voxel grid is a voxel, and the normalized temperature parameter, the impedance phase product parameter, and the frequency deviation parameter contained in each voxel are averaged to obtain a mean grid.

[0025] According to a preferred embodiment, the mean grid is subjected to partial derivative operation based on six-neighborhood center difference, and voxel partial derivative data are obtained, and a parameter coupling change rate matrix is constructed based on the voxel partial derivative data, which comprises:

[0026] The neighboring voxels in the six directions of the grid of the non-boundary voxel contained in the mean grid are selected for center difference, and nine voxel partial derivative data are obtained, and the six directions of the grid represent six directions of the X axis, the Y axis, and the Z axis of the non-boundary voxel in the three-dimensional parameter grid.

[0027] The nine voxel partial derivative data are filled according to a sequence rule to form a 3x3 parameter coupling change rate matrix.

[0028] The sequence filling means that the first row of the parameter coupling change rate matrix represents the variation characteristics of the normalized temperature parameter in three directions, the second row of the parameter coupling change rate matrix represents the response characteristics of the impedance phase product parameter in three directions, and the third row of the parameter coupling change rate matrix represents the variation characteristics of the frequency deviation parameter in three directions.

[0029] According to a preferred embodiment, the eigenvalue decomposition is performed on the parameter coupling change rate matrix, a deformation feature vector is obtained, and the deformation feature vector is compared with a pre-stored instrument pattern recognition vector to obtain a pattern similarity score, including:

[0030] Based on the plurality of 3x3 parameter coupling change rate matrices, a plurality of maximum change intensity parameters and a plurality of minimum change intensity parameters are extracted, the plurality of maximum change intensity parameters and the plurality of minimum change intensity parameters are respectively averaged to obtain an average maximum change intensity parameter and an average minimum change intensity parameter, and a change intensity ratio is obtained according to the maximum change intensity parameter and the average minimum change intensity parameter;

[0031] The plurality of 3x3 parameter coupling change rate matrices are arranged according to the corresponding positions of the plurality of 3x3 parameter coupling change rate matrices in the average grid to construct a 3xN parameter coupling change rate matrix, and a change direction is extracted based on the 3xN parameter coupling change rate matrix;

[0032] Based on the average maximum change intensity parameter, the average minimum change intensity parameter, the change direction, and the change intensity ratio, a deformation feature vector is constructed;

[0033] The deformation feature vector is compared with the pre-stored instrument pattern recognition vector item by item to obtain a pattern similarity score.

[0034] According to a preferred embodiment, the deformation feature vector is compared with the pre-stored instrument pattern recognition vector item by item to obtain a pattern similarity score, including:

[0035] It is checked whether the change direction is parallel to the standard direction in the pre-stored instrument pattern recognition vector;

[0036] It is compared whether the average maximum change intensity parameter, the average minimum change intensity parameter, and the standard maximum change intensity parameter, the standard minimum change intensity parameter in the pre-stored instrument pattern recognition vector are in the same order of magnitude;

[0037] The change intensity ratio is compared with the standard intensity ratio in the pre-stored instrument pattern recognition vector;

[0038] The similarity score is represented as the matching degree of each feature in the item-by-item comparison. The feature that is completely matched gets full score, and the feature that has deviation is deducted according to the deviation degree. The mapping of each matching degree to the [0, 100] interval obtains the similarity score.

[0039] According to a preferred embodiment, the instrument determination mode is obtained based on the pattern similarity score, and a pattern mechanism is performed according to the instrument determination mode, including:

[0040] When the similarity score is greater than or equal to 95, the mode is determined as a matching mode, in which a preset optimal heating curve is loaded based on the tobacco appliance to obtain a heating control parameter, the metal heating induction body is heated according to the heating control parameter, and the LED light on the tobacco appliance is started to display in green color constantly;

[0041] When the similarity score is in a closed interval of 85 to 94, the mode is determined as a suspected mode, in which the identification method is executed again, when the similarity score is greater than or equal to 95, the mode is determined as a matching mode, otherwise, the mode is determined as a non-matching mode;

[0042] When the similarity score is less than 85, the mode is determined as a non-matching mode, in which the heating of the metal heating induction body is stopped based on the tobacco appliance, a preset device protection instruction is obtained, the warning content is obtained based on the device protection instruction and is broadcasted through the sound emitting device, and the LED light on the tobacco appliance is started to display in red color constantly based on the device protection instruction.

[0043] According to a preferred embodiment, the metal heating induction body is porous vacuum silicon, which is made of nano silicon atoms.

[0044] A matching tobacco appliance identification system based on a metal heating induction body, comprising:

[0045] A collection module, which obtains real-time temperature parameters and electromagnetic parameters of the metal heating induction body during the heating of the metal heating induction body;

[0046] A storage unit, which is used for storing a pre-stored appliance mode identification vector;

[0047] An execution unit, which executes a mode mechanism according to an appliance determination mode;

[0048] An intelligent chip, which is electrically connected with the collection module, the storage unit and the execution unit, respectively, and is used for constructing a three-dimensional parameter grid, a parameter coupling change rate matrix, performing eigenvalue decomposition, comparing a deformation feature vector with the pre-stored appliance mode identification vector to obtain a mode similarity score, obtaining the appliance determination mode based on the mode similarity score, and sending a control instruction to the execution unit according to the appliance determination mode.

[0049] Compared with the prior art, the present application has the following beneficial effects:

[0050] Firstly, by collecting the real-time temperature parameters and electromagnetic parameters of the metal heating inductor at multiple temperature points, a temperature-electromagnetic mapping table is constructed, and a three-dimensional parameter grid is constructed according to the temperature-electromagnetic mapping table, so as to convert the environmental temperature variable into a controlled dimension of the three-dimensional parameter grid, eliminate the interference of temperature drift on the identification result, solve the limitation of the traditional identification method which only relies on single electrical parameters such as voltage, current and resistance, and more dimensional parameters can more comprehensively reflect the matching characteristics of the metal heating inductor and the tobacco appliance, reduce the feature distortion caused by single parameter affected by environmental temperature and aging, make the identification basis more sufficient, and thus reduce the misjudgment probability.

[0051] Secondly, by constructing the three-dimensional parameter grid and performing mean value division, combining with the six-neighbor center difference to obtain the voxel partial derivative data, constructing the parameter coupling change rate matrix, and then extracting the deformation feature vector through eigenvalue decomposition, the interference caused by environmental temperature fluctuation and aging can be effectively filtered, compared with the voltage, current and resistance value directly detected in the traditional method, the deformation feature vector extracted by the method more stably reflects the matching essential characteristics, can effectively distinguish the material structure difference, solves the problem that the counterfeit material cheats the traditional resistance detection, reduces the interference of external factors on the identification result, improves the stability of identification, and enhances the anti-interference ability.

[0052] Finally, a clear judgment mechanism is set based on the pattern similarity score, including matching, suspected and non-matching three modes, wherein the suspected mode re-executes the identification process to verify the result, and the non-matching mode triggers the device protection mechanism, so as to ensure the rigor of the identification result, and the hierarchical judgment logic can adapt to metal heating inductors in different states, avoids the misjudgment caused by single threshold judgment in the traditional method, makes the identification result more reliable in actual use scene, and ensures the safety of the equipment and use. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 It is a step flow chart of a kind of tobacco appliance identification method based on metal heating inductor of the present application.

[0054] Figure 2 It is a step flow chart of generating temperature-electromagnetic mapping table in a kind of tobacco appliance identification method based on metal heating inductor of the present application.

[0055] Figure 3 It is a step flow chart of obtaining mean grid in a kind of tobacco appliance identification method based on metal heating inductor of the present application.

[0056] Figure 4 It is the schematic diagram of a kind of tobacco appliance identification system based on metal heating inductor of the present application. DETAILED DESCRIPTION

[0057] For a further understanding of the present application, preferred embodiments thereof will be described in conjunction with the accompanying drawings and examples, it being understood that the description and examples are only by way of further illustrating the features and advantages of the present application, but are not intended to limit the scope of the claims of the present application.

[0058] Embodiment:

[0059] Referring to FIG. 1, the present application provides a metal heating induction body-based matching tobacco appliance identification method, comprising the following steps: Figures 1 to 3

[0060] S10: setting multiple temperature points during heating of the metal heating induction body and acquiring real-time temperature parameters and electromagnetic parameters of the metal heating induction body, and establishing a temperature-electromagnetic mapping table.

[0061] It can be understood that multiple temperature points are set during heating of the metal heating induction body, and real-time temperature parameters and electromagnetic parameters of the metal heating induction body are acquired, and a temperature-electromagnetic mapping table is established; this step provides a data basis for subsequent identification by synchronously collecting electromagnetic parameters during heating, avoiding feature loss caused by single parameters, wherein the temperature parameters reflect the thermal response characteristics of the metal heating induction body, and the electromagnetic parameters reflect the electromagnetic characteristics of the metal heating induction body at different temperatures, and the combination of the two can more accurately reflect the matching state of the metal heating induction body and the tobacco appliance.

[0062] In this embodiment, step S10 comprises:

[0063] S100: acquiring an initial temperature parameter of the metal heating induction body, and simultaneously heating the metal heating induction body to 400 degrees Celsius, and acquiring the real-time temperature parameter during heating;

[0064] Specifically, after the heating module of the tobacco appliance receives a start signal, the temperature acquisition module is first triggered to collect the temperature value of the metal heating induction body, and the metal heating induction body is in an unheated state, and the temperature value is the initial temperature parameter, which reflects the temperature of the metal heating induction body at the ambient temperature; after the initial temperature parameter is acquired, the heating module starts to run, continuously applies heat to the metal heating induction body, so that the temperature of the metal heating induction body gradually rises from the initial temperature, and finally reaches 400 degrees Celsius, which is the working temperature of the metal heating induction body; during the entire process of heat application by the heating module, the temperature acquisition module continuously detects the temperature parameter of the metal heating induction body to acquire the real-time temperature parameter, which is used to record the temperature dynamic change of the metal heating induction body during heating;

[0065] ​Step S101, a plurality of temperature points are set based on the initial temperature parameter and the real-time temperature parameter, the temperature points including an initial temperature point and a plurality of temperature rise points, the initial temperature point representing a temperature point when the initial temperature parameter, and the temperature rise points representing temperature points each rising by 5 degrees Celsius compared with the initial temperature parameter;

[0066] Specifically, a plurality of temperature points are set based on the initial temperature parameter and the real-time temperature parameter obtained in step S100, the temperature points including an initial temperature point and a plurality of temperature rise points, wherein the initial temperature point corresponds to the initial temperature parameter and is used to mark the state at the beginning of heating, and the temperature rise points are determined according to the difference between the real-time temperature parameter and the initial temperature parameter, and a temperature rise point is recorded each time the real-time temperature parameter rises by 5 degrees Celsius compared with the initial temperature parameter. The purpose of setting a plurality of temperature points is to collect parameters at different stages of heating to avoid missing the influence of temperature change on the characteristics of the metal heat-generating inductor when collecting parameters at a single temperature, so that the parameter collection covers the whole heating process, thereby capturing the matching characteristics at different temperatures.

[0067] Step S102, the corresponding electromagnetic parameters are obtained based on the plurality of temperature points, and the temperature-electromagnetic mapping table is generated according to the plurality of temperature points and the corresponding electromagnetic parameters.

[0068] Specifically, when the real-time temperature parameter of the metal heat-generating inductor reaches a certain temperature point, the electromagnetic parameter collection module starts to apply an electromagnetic signal of a specific frequency to the metal heat-generating inductor, and the electromagnetic parameter corresponding to the temperature point is obtained by detecting the reflected signal. Then, the temperature parameter of each temperature point is corresponded to the corresponding electromagnetic parameter one by one to generate a temperature-electromagnetic mapping table. The temperature-electromagnetic mapping table has the effect of associating temperature change with electromagnetic characteristic change to form a structured data set, which is convenient for subsequent calling.

[0069] For example, when the initial temperature parameter of the metal heat-generating inductor is 25 degrees Celsius (initial temperature point), during the heating process, when the real-time temperature parameter reaches 30 degrees Celsius, which rises by 5 degrees Celsius compared with the initial temperature, the first temperature rise point is recorded, and the impedance amplitude parameter, the phase angle parameter and the resonance frequency parameter at this time are collected by the electromagnetic parameter collection module. When the temperature continues to rise to 35 degrees Celsius, 40 degrees Celsius, etc., the corresponding temperature rise points are recorded in turn and the electromagnetic parameters are collected. Finally, in the temperature-electromagnetic mapping table formed, each temperature value corresponds to a complete set of electromagnetic parameters, which clearly presents the characteristic change of the metal heat-generating inductor at different temperatures, and provides an analysis basis for subsequent identification.

[0070] S11: constructing a three-dimensional parameter grid based on the electromagnetic parameters, performing mean value division on the three-dimensional parameter grid to obtain a mean value grid, and constructing a coupling change rate matrix according to the mean value grid;

[0071] Specifically, normalized temperature parameters, impedance phase product parameters, and frequency offset parameters are obtained based on electromagnetic parameters and temperature-electromagnetic mapping tables. A three-dimensional parameter grid is constructed based on the normalized temperature parameters, impedance phase product parameters, and frequency offset parameters. The X-axis of the three-dimensional parameter grid is the normalized temperature parameter, the Y-axis is the impedance phase product parameter, and the Z-axis is the frequency offset parameter. The three-dimensional parameter grid is then divided into mean grids.

[0072] The mean grid is partially derived based on the six-neighbor central difference and voxel partial derivative data is obtained. A parameter coupling rate of change matrix is ​​constructed based on the voxel partial derivative data.

[0073] Furthermore, the electromagnetic parameters include impedance amplitude parameters, phase angle parameters, and resonant frequency parameters;

[0074] Impedance amplitude parameter refers to the magnitude of the total resistance of a metal heating sensor to an AC current when a specific frequency AC signal is applied to it. It reflects the degree to which the metal heating sensor impedes the AC current at the current temperature. During the heating process, as the temperature of the metal heating sensor changes, its internal material structure changes, which leads to a corresponding change in the impedance amplitude parameter. Different types of matching appliances have different impedance amplitude parameters for their corresponding metal heating sensors at the same temperature.

[0075] The phase angle parameter refers to the phase difference between the AC voltage across the metal heating sensor and the AC current passing through it when an AC signal is applied. When the temperature of the metal heating sensor changes, its inductance and capacitance characteristics will change, which will lead to a change in the phase relationship between voltage and current. The change in the value of the phase angle parameter can reflect the difference in this phase relationship. The phase angle parameters of metal heating sensors of different appliances are usually different at the same temperature.

[0076] The resonant frequency parameter refers to the AC signal frequency at which the impedance value reaches its minimum when a continuously varying AC signal is applied to a metal heating sensor. Due to differences in material composition, structure, and other factors, metal heating sensors have specific inductive and capacitive equivalent characteristics. Under the action of an AC signal, resonance will occur when the signal frequency matches the inherent frequency of the metal heating sensor itself. This frequency is the resonant frequency parameter. The resonant frequency parameter of metal heating sensors in different devices will exhibit different values ​​due to their own characteristics.

[0077] The mean grid is obtained by following these steps.

[0078] In this implementation, obtaining the mean grid includes the following steps:

[0079] S110, extract the multiple temperature points contained in the temperature-electromagnetic parameter mapping table, linearly map the set temperature parameters of the multiple temperature points to the interval [0, 1] to obtain the normalized temperature parameters;

[0080] Specifically, the set temperature parameters of all temperature points are extracted from the temperature-electromagnetic parameter mapping table, including the set temperature parameters of the initial temperature point and each temperature rise point. The set temperature parameters of the initial temperature point and each temperature rise point are linearly mapped to the interval [0, 1] to obtain the normalized temperature parameters. The specific operation is as follows: first, determine the minimum value and the maximum value in all set temperature parameters, wherein the set temperature parameter of the initial temperature point is the minimum value, and the highest temperature rise temperature point, such as the temperature parameter when heated to 400 degrees Celsius, is the maximum value; then calculate the total difference value of the temperature range, that is, the maximum value minus the minimum value; then for each set temperature parameter, calculate the difference value thereof from the minimum value, which reflects the relative position of the temperature in the entire range; then divide this difference value by the total difference value to obtain the normalized temperature parameter corresponding to the set temperature parameter, which falls within the interval [0, 1];

[0081] For example, if the initial temperature point is 25 degrees Celsius (minimum value) and the highest temperature rise temperature point is 400 degrees Celsius (maximum value), the total difference value is 375 degrees Celsius; when a temperature rise temperature point is 50 degrees Celsius, the difference value thereof from the minimum value is 25 degrees Celsius, and the normalized temperature parameter obtained by dividing the total difference value 375 degrees Celsius is 0.067. This value only reflects the relative position of the temperature in the entire heating range, and is irrelevant to whether the initial temperature is 25 degrees Celsius, so that the temperature parameters under different initial temperatures can be directly used for subsequent three-dimensional parameter grid construction and feature comparison. This operation unifies the absolute temperature values of different metal heating induction bodies to a relative interval, reducing the influence caused by environmental temperature differences;

[0082] S111, extract the impedance amplitude parameter and the phase angle parameter corresponding to each temperature point in the temperature-electromagnetic parameter mapping table, and obtain the impedance phase product parameter based on the impedance amplitude parameter and the phase angle parameter;

[0083] Specifically, in the temperature-electromagnetic parameter mapping table, each temperature point corresponds to a set of electromagnetic parameters, from which the impedance amplitude parameter and the phase angle parameter are extracted. The impedance phase product parameter is obtained by multiplying the impedance amplitude parameter and the phase angle parameter. This parameter integrates the information of the impedance amplitude parameter and the phase angle parameter, and can reflect the change of the transmission characteristics of the electromagnetic signal in the metal heating induction body. Compared with separately analyzing the impedance amplitude parameter or the phase angle parameter, the impedance phase product parameter is more sensitive to the material characteristics of the metal heating induction body;

[0084] S112, extract the resonance frequency parameter corresponding to each temperature point in the temperature-electromagnetic parameter mapping table, and obtain a reference resonance frequency parameter, which is represented as the resonance frequency parameter obtained based on the initial temperature parameter, and obtain a frequency deviation parameter according to the resonance frequency parameter and the reference resonance frequency parameter;

[0085] Specifically, the resonance frequency parameter corresponding to each temperature point is extracted from the temperature-electromagnetic parameter mapping table, the resonance frequency parameter corresponding to the initial temperature point is taken as the reference resonance frequency parameter, and the frequency deviation parameter is obtained by subtracting the reference resonance frequency parameter from the resonance frequency parameter and then dividing the result by the reference resonance frequency parameter. The frequency deviation parameter reflects the degree of deviation of the resonance frequency parameter at different temperatures relative to the initial state, and the change of the frequency deviation parameter is related to the physical structure change of the metal heat induction body in the heating process.

[0086] S113, constructing a three-dimensional parameter grid based on the frequency deviation parameter, the impedance phase product parameter and the normalized temperature parameter;

[0087] Specifically, the three-dimensional parameter grid is constructed based on the frequency deviation parameter, the impedance phase product parameter and the normalized temperature parameter; the normalized temperature parameter is taken as the X-axis, the impedance phase product parameter is taken as the Y-axis, and the frequency deviation parameter is taken as the Z-axis. The three parameter values corresponding to each temperature point are mapped into a three-dimensional space to form a discrete data point set. The distribution pattern of these data points in the three-dimensional space reflects the internal relationship between the temperature, electromagnetic characteristics and structural change of the metal heat induction body in the heating process. By constructing the three-dimensional parameter grid, the originally independent three parameters are integrated into a whole.

[0088] S114, dividing the three-dimensional parameter grid into a 30x30x30 voxel grid along the X / Y / Z axes at equal intervals, and taking a single grid in the voxel grid as a voxel. The normalized temperature parameter, the impedance phase product parameter and the frequency deviation parameter contained in each voxel are averaged to obtain a mean grid;

[0089] Specifically, on the basis of the three-dimensional parameter grid, 30 intervals are equally spaced along the X-axis (normalized temperature parameter), Y-axis (impedance phase product parameter), and Z-axis (frequency deviation parameter) directions to form a 30x30x30 cubic voxel grid, each voxel grid containing a plurality of original data points, the average values of the normalized temperature parameter, impedance phase product parameter, and frequency deviation parameter in each voxel are calculated to obtain a mean grid. This operation reduces random noise in the original data and smooths the parameter distribution through spatial division and parameter averaging. By dividing the three-dimensional parameter grid into a 30x30x30 cubic voxel grid, standardization is achieved, so that regardless of the number of grids in the three-dimensional parameter grid, it will eventually be divided into a standard 30x30x30 cubic voxel grid, which can adapt to the electromagnetic parameters obtained from different metal heat-inducing bodies, improving the applicability.

[0090] For example, when the metal heat-inducing body is at 25 degrees Celsius, the impedance amplitude parameter is 50Ω, the phase angle parameter is 30°, and the resonant frequency parameter is 10MHz, the reference resonant frequency parameter is 10MHz, the impedance phase product parameter is 50x30=1500, and the normalized temperature parameter is 0. When the temperature rises to 50 degrees Celsius, the impedance amplitude parameter becomes 60Ω, the phase angle parameter becomes 35°, and the resonant frequency parameter becomes 10.2MHz, the frequency deviation parameter is (10.2-10) / 10=0.02, the impedance phase product parameter is 60x35=2100, and the normalized temperature parameter is (50-25) / (400-25)≈0.067. After mapping the three parameters of the metal heat-inducing body at each temperature point to a three-dimensional space and performing voxel division, a voxel may contain 5 temperature points. The average of the normalized temperature parameter, impedance phase product parameter, and frequency deviation parameter corresponding to the 5 temperature points is taken to obtain the mean parameter of the voxel, and the mean parameters of all voxels form a mean grid. This grid reflects the overall parameter variation of the metal heat-inducing body during the heating process and can be used as a basic feature for identifying a matching appliance.

[0091] In this embodiment, the neighboring voxels in the six directions of the grid of the non-boundary voxel contained in the mean grid are selected for central difference to obtain 9 voxel partial derivative data.

[0092] Specifically, the six directions of the grid refer to the six directions of the non-boundary voxel in the X-axis, Y-axis, and Z-axis of the three-dimensional parameter grid, i.e., the positive direction of the X-axis, the negative direction of the X-axis, the positive direction of the Y-axis, the negative direction of the Y-axis, the positive direction of the Z-axis, and the negative direction of the Z-axis. These six directions cover all adjacent positions of the voxel in the three-dimensional space.

[0093] The central difference calculates the rate of change of the impedance phase product parameter with respect to the normalized temperature parameter in the X-axis direction. Specifically, the adjacent voxels in the positive and negative X-axis directions of the target voxel are taken, and the impedance phase product parameters of the two adjacent voxels are extracted. The difference between the two is calculated and divided by the distance between the two voxels in the X-axis direction. The result reflects the degree of change of the impedance phase product parameter corresponding to a unit change in the normalized temperature parameter. The greater the rate value, the more significant the influence of temperature change on the impedance phase product parameter.

[0094] The central difference calculates the sensitivity of the normalized temperature parameter with respect to the impedance phase product parameter in the Y-axis direction. Specifically, the adjacent voxels in the positive and negative Y-axis directions of the target voxel are taken, and the normalized temperature parameters of the two adjacent voxels are extracted. The difference between the two is calculated and divided by the distance between the two voxels in the Y-axis direction. The result reflects the amplitude of the change in the normalized temperature parameter caused by a unit change in the impedance phase product parameter. The sensitivity value can reflect the reverse influence of the change in the impedance phase product parameter on the temperature characteristic.

[0095] The central difference calculates the response of the impedance phase product parameter with respect to the frequency offset parameter in the Z-axis direction. Specifically, the adjacent voxels in the positive and negative Z-axis directions of the target voxel are taken, and the impedance phase product parameters of the two adjacent voxels are extracted. The difference between the two is calculated and divided by the distance between the two voxels in the Z-axis direction. The response value is used to measure the adjustment amplitude of the impedance phase product parameter when the frequency offset parameter changes, and can reflect the influence relationship of structural changes on electromagnetic characteristics.

[0096] Through the calculation of the above three axis directions, 2 voxel partial derivative data can be obtained for each axis direction. In combination with the cross change parameters in each axis direction, such as the rate of change of the normalized temperature parameter in the X-axis direction, the rate of change of the impedance phase product parameter in the Y-axis direction, and the rate of change of the frequency offset parameter in the Z-axis direction, a total of 9 voxel partial derivative data are obtained.

[0097] The 9 voxel partial derivative data are filled into a 3x3 parameter coupling rate matrix according to a sequence rule.

[0098] The sequence filling is represented as:

[0099] The first row of the parameter coupling rate matrix represents the change characteristics of the normalized temperature parameter in three directions, i.e., the rate of change of the normalized temperature parameter in the X-axis direction, the sensitivity of the normalized temperature parameter in the Y-axis direction to the change in the impedance phase product parameter, and the response of the normalized temperature parameter in the Z-axis direction to the change in the frequency offset parameter.

[0100] The second row represents the response characteristics of the impedance phase product parameter in three directions, i.e., the rate of change of the impedance phase product parameter in the X-axis direction with respect to the normalized temperature parameter, the rate of change of the impedance phase product parameter in the Y-axis direction with respect to itself, and the response of the impedance phase product parameter in the Z-axis direction to the change in the frequency offset parameter.

[0101] The third row represents the variation characteristics of the frequency offset parameter in three directions, i.e., the rate of change of the frequency offset parameter in the X-axis direction with the normalized temperature parameter, the sensitivity of the frequency offset parameter in the Y-axis direction to the impedance phase product parameter, and the rate of change of the frequency offset parameter in the Z-axis direction with itself.

[0102] This matrix structure can systematically integrate the coupling change relationship of the three parameters in different directions, and provide structured numerical basis for subsequent extraction of deformation feature vectors.

[0103] For example, the impedance phase product parameter of a non-boundary voxel adjacent to the left in the X-axis direction is 2000, the impedance phase product parameter of a non-boundary voxel adjacent to the right in the X-axis direction is 2200, and the X-axis direction voxel spacing is 0.05. Therefore, the rate of change of the impedance phase product parameter in the X-axis direction with the normalized temperature parameter is (2200-2000) / 0.05=4000, which reflects the response strength of the impedance phase product parameter to temperature change of this voxel. If the normalized temperature parameter of a non-boundary voxel adjacent to the front in the Y-axis direction is 0.3, the normalized temperature parameter of a non-boundary voxel adjacent to the rear in the Y-axis direction is 0.25, and the Y-axis spacing is 500, then the sensitivity of the normalized temperature parameter in the Y-axis direction to the impedance phase product parameter is (0.3-0.25) / 500=0.0001, which reflects the influence of the change of the impedance phase product parameter on the temperature. After these calculation results are filled into the matrix in order, the second row and the first column of the matrix is 4000, the first row and the second column of the matrix is 0.0001, and so on. Finally, the matrix formed presents the parameter coupling change rule corresponding to the voxel.

[0104] S12: performing eigenvalue decomposition on the parameter coupling change rate matrix to obtain a deformation feature vector, and comparing the deformation feature vector with a pre-stored instrument pattern recognition vector to obtain a pattern similarity score;

[0105] Specifically, the eigenvalue decomposition includes extracting a plurality of maximum change strength parameters and a plurality of minimum change strength parameters based on a plurality of 3x3 parameter coupling change rate matrices, calculating the average value of all maximum change strength parameters to obtain an average maximum change strength parameter, and calculating the average value of all minimum change strength parameters to obtain an average minimum change strength parameter. According to the ratio of the average maximum change strength parameter to the average minimum change strength parameter, a change strength ratio is obtained. This ratio can reflect the unevenness of the parameter coupling change. The change strength ratio of different matching states of the metal heating induction body and the instrument combination has significant differences.

[0106] Further, the eigenvalue decomposition further comprises arranging the plurality of 3x3 parameter coupling rate of change matrices according to the corresponding positions of the plurality of 3x3 parameter coupling rate of change matrices in the mean grid, that is, according to the distribution order of the voxels in the three-dimensional space, sequentially splicing the row vectors of each matrix to construct a 3xN parameter coupling rate of change matrix, where N is the total number of voxels, and obtaining the eigenvector as the change direction based on the 3xN parameter coupling rate of change matrix;

[0107] The specific splicing manner is that the 3 row vectors of the first 3x3 matrix are sequentially taken as the first to third rows of the new matrix, the row vectors of the second 3x3 matrix are sequentially taken as the fourth to sixth rows of the new matrix, and so on, until all the 3x3 matrices are processed.

[0108] Further, the specific steps of obtaining the eigenvector as the change direction based on the 3xN parameter coupling rate of change matrix are as follows:

[0109] First, the covariance matrix of the 3xN matrix is calculated, the covariance matrix is obtained by calculating the covariances between the 3 row vectors in the matrix, and reflects the overall correlation degree of the normalized temperature parameter, the impedance phase product parameter and the frequency offset parameter in all voxel regions. Eigenvalue decomposition is performed on the obtained covariance matrix to obtain 3 eigenvalues and 3 corresponding eigenvectors, wherein the eigenvector corresponding to the maximum eigenvalue is the change direction.

[0110] For example, the eigenvector corresponding to the maximum eigenvalue of the covariance matrix of a certain 3xN matrix after eigenvalue decomposition is (0.6, 0.3, 0.1), which indicates that in all voxel regions, the dominant trend of parameter coupling change is mainly driven by the normalized temperature parameter (X axis), followed by the impedance phase product parameter (Y axis), and the influence of the frequency offset parameter (Z axis) is weak. The closer the direction is to the direction of the pre-stored appliance pattern recognition vector, the higher the matching degree of the current metal heating induction body and the tobacco appliance.

[0111] The deformation feature vector is constructed based on the mean value of the maximum change intensity parameter, the mean value of the minimum change intensity parameter, the change direction and the change intensity ratio. The vector integrates the strength, trend and proportional relationship of parameter change, converts the complex coupling change in three-dimensional space into a structured feature that can be directly compared, and makes the characteristic differences of different metal heating induction bodies and appliance combinations quantifiable through vector differences;

[0112] The deformation feature vector and the pre-stored appliance pattern recognition vector are compared item by item to obtain a pattern similarity score. The pre-stored appliance pattern recognition vector is a standard feature vector extracted based on a large number of matched metal heating induction bodies and appliance combinations through the same process.

[0113] In comparison, first analyze the direction consistency, judge whether the change direction is parallel by calculating the cosine value of the included angle of two vectors, the smaller the included angle, the higher the direction consistency; then evaluate the intensity matching degree, compare the magnitude difference of the maximum mean change intensity, the minimum mean change intensity and the corresponding intensity in the pre-stored vector, if they are in the same order of magnitude, the matching degree is high; finally, compare the intensity ratio tolerance, calculate the deviation of the change intensity ratio and the standard intensity ratio in the pre-stored vector, the smaller the deviation, the better the tolerance matching. Integrate the comparison results, give full marks to the items that match completely, deduct points according to the deviation degree for the items with deviation, and finally map the total score to the interval [0, 100], which is the pattern similarity score. This score directly reflects the matching degree of the current metal heating induction body and the tobacco appliance.

[0114] For example, the parameter coupling change rate matrix of a certain metal heating induction body and the matching appliance is processed to obtain a maximum mean change intensity of 5000, a minimum mean change intensity of 500, a change intensity ratio of 10, and a change direction vector of (0.8, 0.1, 0.1); the corresponding intensity in the pre-stored appliance pattern recognition vector is 5200, 510, the intensity ratio is 10.2, and the direction vector is (0.79, 0.11, 0.1). In comparison, the direction cosine value is close to 1, the direction consistency is high; the intensity order is the same, the matching degree is high; the intensity ratio deviation is 0.2, which is within the tolerance range, the sum of the scores of each item is 98, and the mapped pattern similarity score is 98, indicating that the matching degree of this combination is high.

[0115] S13: Obtain the appliance judgment mode based on the pattern similarity score, and execute the mode mechanism according to the appliance judgment mode;

[0116] Specifically, the matching degree of the quantized features is determined to determine the adaptation state of the metal heating induction body and the tobacco appliance, and then the corresponding operation is triggered according to the state, so that only the matched combination can operate in the optimal way, and the non-matched combination is safely controlled to avoid the use risk caused by the adaptation problem.

[0117] Further, the item-by-item comparison includes:

[0118] Analyze the direction consistency: check whether the change direction is parallel to the standard direction in the pre-stored appliance pattern recognition vector. The change direction vector is the dominant trend of parameter coupling change, and the standard direction vector is the typical trend direction determined by a large number of tests on the matching combination. In actual operation, the parallel degree is judged by calculating the cosine value of the included angle of two vectors: the closer the cosine value is to 1, the more consistent the direction is; the more the cosine value deviates from 1, the greater the direction difference is. Its role is to verify whether the overall trend of the current parameter change conforms to the matching characteristics, and if the direction is inconsistent, it often means that the electromagnetic or thermal characteristics of the metal heating induction body and the appliance do not match;

[0119] For example, the change direction vector of the matched combination is (0.6, 0.2, 0.2), and the standard direction vector is (0.58, 0.22, 0.2), the included angle cosine value of which is close to 1, and the direction consistency is high; the change direction vector of the non-matched combination is (0.3, 0.5, 0.2), and the included angle cosine value with the standard vector is small, and the direction consistency is low.

[0120] Evaluate the intensity matching degree: compare whether the mean maximum change intensity parameter, the mean minimum change intensity parameter and the standard maximum change intensity parameter, the standard minimum change intensity parameter in the pre-stored appliance pattern recognition vector are in the same order of magnitude. The order of magnitude is based on the order of magnitude of the numerical value, such as 10 4 , etc., rather than the specific numerical value: if the order of magnitude of the current intensity and the standard intensity is the same, it is considered that they are in the same order of magnitude; otherwise, they are in different orders of magnitude. The role is to ensure that the strength degree of parameter change conforms to the typical range of matched combination, and to avoid the heating out of control caused by abnormal strength, such as too high or too low.

[0121] For example, the mean maximum change intensity parameter of the matched combination is 5x10³, and the standard maximum change intensity parameter is 6x10³, both of which are in the order of 10³, and the intensity matching degree is high; the mean maximum change intensity parameter of a certain non-matched combination is 8x10 4 , which is significantly different from the order of 10³ of the standard value, and the intensity matching degree is low.

[0122] Intensity ratio tolerance comparison: compare the change intensity ratio with the standard intensity ratio in the pre-stored appliance pattern recognition vector. By calculating the deviation rate of the current change intensity ratio and the standard intensity ratio, the deviation amount is divided by the standard intensity ratio, to judge whether the tolerance is within the allowable range: the smaller the deviation rate, the closer the proportion to the standard; if the deviation rate exceeds the preset range, it is considered that the proportion is not matched. The role is to verify whether the proportion relationship of the strongest and weakest degree in the parameter change conforms to the matched characteristics. The imbalance of the proportion often reflects the poor characteristic synergy of the metal heating induction body and the appliance.

[0123] For example, the standard intensity ratio is 12, and the current change intensity ratio is 11.5, the deviation rate is small, and the tolerance comparison is passed; the change intensity ratio of a certain non-matched combination is 8, and the deviation rate from the standard value is large, and the tolerance comparison is not passed.

[0124] Further, when the similarity score is greater than or equal to 95, the mode is determined to be a matching mode. In the matching mode, the intelligent chip of the tobacco device calls the optimal heating curve pre-stored in the storage unit, which contains the target temperature, heating power adjustment parameter and the like corresponding to different time nodes. The intelligent chip generates heating control parameters according to the curve and sends the heating control parameters to the heating module, so that the heating module heats the metal heating induction body according to the curve, ensuring that the temperature of the metal heating induction body at each stage meets the optimal aerosol generation requirement. At the same time, the intelligent chip controls the LED lamp on the tobacco device to display green light constantly, feeding back to the user that the current is a matched combination and can be normally used. The function is to ensure that the matched metal heating induction body and the device work in the best way, maintain stable aerosol generation and release characteristics, and improve the use experience.

[0125] When the similarity score is in the closed interval of 85 to 94, the mode is determined to be a suspected mode. In the suspected mode, the intelligent chip triggers the whole identification method to be executed again, from collecting temperature and electromagnetic parameters to calculating the similarity score, in order to exclude accidental interference in single detection, such as transient electromagnetic fluctuation and temperature collection error, and ensure the reliability of the determination result. If the similarity score of the second detection is greater than or equal to 95, it is determined to be a matching mode and enters the matched processing flow. If the score of the second detection is still in the interval of 85 to 94 or less than 85, it is determined to be a non-matching mode and enters the non-matched processing flow. The function is to reduce the misjudgment caused by single detection deviation through secondary verification, and balance the identification efficiency and accuracy.

[0126] For example, the metal heating induction body is initially detected with a score of 92, and the mode is determined to be a suspected mode. After the second detection, the score is increased to 96 due to the exclusion of electromagnetic interference, and the mode is determined to be a matching mode. Another metal heating induction body is detected with a score of 83 in the second detection, and the mode is determined to be a non-matching mode.

[0127] When the similarity score is less than 85, it is determined to be a non-matching mode. In the non-matching mode, the intelligent chip sends a stop instruction to the heating module to cut off the heating power supply, and at the same time, starts the device protection instruction to prevent the non-matched combination from causing the metal heating induction body to overheat or the device to be damaged due to the mismatch of the heating parameters. The intelligent chip calls the pre-set warning content from the storage unit and broadcasts it through the sound device to prompt the user that the current is a non-matched combination. At the same time, the LED lamp is controlled to display red light constantly to strengthen the warning effect. The function is to actively issue a warning to avoid the safety risk caused by the non-matched combination.

[0128] The metal heating induction body is porous vacuum silicon, which is a heat insulation material formed by cross-linking of nano-silicon atom clusters, has a thermal conductivity of less than 0.014 W / (K.m), is made of nano-silicon atoms, does not produce odor at 300 DEG C to 400 DEG C, does not produce fine pollutants, is soft, has good heat preservation effect, and is pollution-free.

[0129] As shown in the accompanying drawings Figure 4 The application also provides a complete tobacco appliance identification system based on a metal heating induction body, which comprises:

[0130] The acquisition module comprises a temperature acquisition module and an electromagnetic parameter acquisition module.

[0131] Specifically, the temperature acquisition module is installed in the tobacco appliance close to the metal heating induction body, and senses the temperature of the metal heating induction body during heating of the metal heating induction body, so as to obtain temperature data including an initial temperature parameter when the metal heating induction body is not heated, real-time temperature parameters during heating, and temperature data corresponding to a plurality of temperature points set according to the difference between the initial temperature and the real-time temperature; the data is transmitted to the intelligent chip through a signal line, and serves as a basis for subsequent construction of a temperature-electromagnetic mapping table and calculation of a normalized temperature parameter, so as to ensure that the system can capture the thermal response characteristics of the metal heating induction body at different temperature stages.

[0132] Further, the electromagnetic parameter acquisition module is installed in the tobacco appliance at a position corresponding to the metal heating induction body, and is connected to the intelligent chip; when the intelligent chip determines that the real-time temperature data reaches a set temperature point, the electromagnetic parameter acquisition module is controlled to emit an electromagnetic signal of a specific frequency to the metal heating induction body, and the amplitude, phase shift and resonance frequency parameter change of the reflected signal are detected to obtain electromagnetic parameters corresponding to the temperature point; the electromagnetic parameters reflect the electromagnetic characteristics of the metal heating induction body at a specific temperature, and cooperate with the temperature data to form multi-dimensional characteristics, thereby providing data support for subsequent parameter coupling analysis.

[0133] The storage unit is used for storing pre-stored appliance mode identification vectors, pre-set optimal heating curves and device protection instructions, and functions to provide comparison standard data and execution instruction basis for the system, so as to ensure that the intelligent chip has a clear reference benchmark during analysis and determination.

[0134] The execution unit comprises a heating module, an LED lamp and a sound device, the heating module is connected with the power supply circuit of the tobacco appliance, and the heating module is heated by adjusting the current or voltage to the metal heating induction body after receiving the control instruction of the smart chip; the LED lamp is installed on the surface of the shell of the tobacco appliance, and presents green constant light or red constant light according to the instruction of the smart chip, so as to intuitively feedback the determination result; the sound device is connected with the smart chip, and the sound device plays the preset prompt or warning content according to the instruction; the unit converts the determination result of the smart chip into specific operation, realizes heating control and state feedback.

[0135] The smart chip, as the core processing unit of the system, is electrically connected with the temperature collection module, the electromagnetic parameter collection module, the storage unit and the execution unit through wires, receives the temperature data transmitted by the temperature collection module and the electromagnetic parameters transmitted by the electromagnetic parameter collection module, first constructs a temperature-electromagnetic mapping table based on the data, then converts the parameters into normalized temperature parameters, impedance phase product parameters and frequency offset parameters through an algorithm, and constructs a three-dimensional parameter grid; then the grid is averaged and divided to obtain an average grid, the partial derivative data of the voxel is calculated based on the center difference of the six-neighborhood, and a parameter coupling change rate matrix is constructed; then eigenvalue decomposition is performed on the matrix, and a deformation feature vector containing the average maximum change strength parameter, the average minimum change strength parameter, the change direction and the change strength ratio is extracted; then the vector is compared with the pre-stored appliance pattern recognition vector in the storage unit item by item, and a pattern similarity score in the interval [0, 100] is generated; finally, the appliance is determined to be in a matching, suspected or non-matching mode according to the score, and corresponding control instructions are sent to the execution unit, such as the instruction of loading the optimal heating curve in the matching mode and the instruction of stopping heating and starting warning in the non-matching mode, and the role is to coordinate the cooperative work of each module, realize the accurate identification and control of the matched appliance through data processing, feature extraction and comparison determination, and ensure that the system operates according to the set logic.

[0136] The specific use mode and role of the embodiment are described below:

[0137] Firstly, a plurality of temperature points are set during the heating of the metal heating induction body in step S10, and real-time temperature parameters and electromagnetic parameters are obtained, a temperature-electromagnetic mapping table is established, and the plurality of temperature point parameter collection covering the whole heating process provides basic data containing the correlation between temperature and electromagnetic characteristics for subsequent analysis, avoiding the limitations of single temperature parameter collection; then, according to step S11, a three-dimensional parameter grid is constructed based on the electromagnetic parameters, the mean grid is obtained through mean value division, and then the coupling change rate matrix is constructed, through parameter normalization, spatial gridding and mean value processing, the noise interference in the original data is reduced, and then the coupling change law between parameters is captured through partial derivative calculation, and the multi-dimensional parameters are converted into structured matrix features, laying a foundation for subsequent feature extraction; then, according to step S12, the parameter coupling change rate matrix is subjected to eigenvalue decomposition, the deformation feature vector is obtained and compared with the pre-stored vector, and the mode similarity score is obtained; the core features of parameter change are extracted through eigenvalue decomposition, the complex coupling change is converted into a vector that can be quantitatively compared, and the multi-dimensional comparison ensures that the score can accurately reflect the matching degree of the metal heating induction body and the tobacco appliance; then, according to step S13, the appliance judgment mode is obtained based on the mode similarity score, and the corresponding mode mechanism is executed, specifically: when the score is ≥95, it is determined as a matching mode, the preset optimal heating curve is loaded to obtain the heating control parameter to control the heating operation, and the LED green light is started to be always on; when the score is between 85 and 94, it is determined as a suspected mode, the identification method is re-executed, and if the second score is ≥95, it is converted into a matching mode, otherwise it is a non-matching mode; when the score is <85, it is determined as a non-matching mode, the heating is stopped and the device protection instruction is started, the alarm content is broadcast through the sound device, and the LED red light is started to be always on; through the layered judgment mechanism, the matching combination is ensured to operate in the optimal way, the suspected combination is verified twice to reduce the misjudgment, and the non-matching combination is safely controlled, which not only guarantees the use effect but also avoids the risk caused by the adaptation problem.

[0138] Through multi-temperature point parameter collection, multi-dimensional parameter coupling analysis, structured feature comparison and layered judgment mechanism, the influence of environmental temperature, aging and other factors on identification is considered in the comprehensive consideration, so that the identification result is more stable and reliable, so as to reduce the probability of misjudgment.

[0139] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for identifying tobacco devices based on a metal heating sensor, characterized in that, The method includes: During the heating process of the metal heating sensor, multiple temperature points are set and the real-time temperature and electromagnetic parameters of the metal heating sensor are obtained to establish a temperature-electromagnetic mapping table. A three-dimensional parameter mesh is constructed based on the electromagnetic parameters. The three-dimensional parameter mesh is then divided into mean-valued subdivisions to obtain a mean-valued mesh. A parameter coupling rate of change matrix is ​​then constructed based on the mean-valued mesh. The parameter coupling rate of change matrix is ​​subjected to eigenvalue decomposition to obtain deformation feature vector, and the deformation feature vector is compared with the pre-stored device pattern recognition vector to obtain pattern similarity score; The device determination pattern is obtained based on the pattern similarity score, and the pattern mechanism is executed according to the device determination pattern.

2. The method for identifying tobacco devices based on a metal heating sensor according to claim 1, characterized in that, The process of establishing multiple temperature points and acquiring real-time temperature and electromagnetic parameters of the metal heating element during heating, and establishing a temperature-electromagnetic mapping table, includes: The initial temperature parameters of the metal heating element are obtained, and the metal heating element is heated to 400 degrees Celsius. The real-time temperature parameters are obtained during the heating process. Multiple temperature points are established based on the initial temperature parameter and the real-time temperature parameter. The temperature points include the initial temperature point and multiple temperature rise points. The initial temperature point is the temperature point when the initial temperature parameter is set. The temperature rise points are the temperature points when the real-time temperature parameter increases by 5 degrees Celsius compared to the initial temperature parameter. The electromagnetic parameters are obtained based on multiple temperature points, and a temperature-electromagnetic mapping table is generated based on the multiple temperature points and the corresponding electromagnetic parameters.

3. The method for identifying tobacco devices based on a metal heating sensor according to claim 1, characterized in that, The process of constructing a three-dimensional parameter mesh based on the electromagnetic parameters, dividing the three-dimensional parameter mesh into mean-valued subdivisions to obtain a mean-valued mesh, and constructing a parameter coupling rate of change matrix based on the mean-valued mesh includes: Based on the electromagnetic parameters and the temperature-electromagnetic mapping table, normalized temperature parameters, impedance phase product parameters, and frequency offset parameters are obtained. A three-dimensional parameter grid is constructed based on the normalized temperature parameters, the impedance phase product parameters, and the frequency offset parameters. The X-axis of the three-dimensional parameter grid is the normalized temperature parameters, the Y-axis is the impedance phase product parameters, and the Z-axis is the frequency offset parameters. The three-dimensional parameter grid is then divided into mean grids. The mean grid is subjected to partial derivative operation based on the six-neighbor central difference to obtain voxel partial derivative data, and a parameter coupling rate of change matrix is ​​constructed based on the voxel partial derivative data.

4. The method for identifying tobacco devices based on a metal heating sensor according to claim 3, characterized in that, The process of obtaining the mean grid includes: The electromagnetic parameters include impedance amplitude parameters, phase angle parameters, and resonant frequency parameters; Extract multiple temperature points contained in the temperature-electromagnetic parameter mapping table, linearly map the set temperature parameters of the multiple temperature points to the [0,1] interval, and obtain the normalized temperature parameters. Extract the impedance amplitude parameter and phase angle parameter corresponding to each temperature point in the temperature-electromagnetic parameter mapping table, and obtain the impedance phase product parameter based on the impedance amplitude parameter and the phase angle parameter; Extract the resonant frequency parameter corresponding to each temperature point in the temperature-electromagnetic parameter mapping table, and obtain the reference resonant frequency parameter. The reference resonant frequency parameter is represented as the resonant frequency parameter obtained based on the initial temperature parameter. Obtain the frequency offset parameter based on the resonant frequency parameter and the reference resonant frequency parameter. A three-dimensional parametric mesh is constructed based on the frequency offset parameter, the impedance phase product parameter, and the normalized temperature parameter. The three-dimensional parametric mesh is divided into a 30×30×30 voxel mesh at equal intervals along the X / Y / Z axes. Each voxel mesh is a voxel. The normalized temperature parameter, impedance phase product parameter and frequency offset parameter contained in each voxel are averaged to obtain the mean mesh.

5. The method for identifying tobacco devices based on a metal heating sensor according to claim 3, characterized in that, The step of performing a partial derivative operation on the mean grid based on the six-neighbor central difference and obtaining voxel partial derivative data, and constructing a parameter coupling rate of change matrix based on the voxel partial derivative data, includes: Central difference was performed on adjacent voxels in the six directions of the non-boundary voxels contained in the mean grid to obtain the partial derivative data of 9 voxels. The six directions of the grid represent the 6 directions of the non-boundary voxels on the X-axis, Y-axis and Z-axis in the three-dimensional parametric grid. The nine voxel partial derivative data are filled into a 3×3 parameter coupling rate of change matrix according to the order rules; The sequential rule filling is represented by the first row of the parameter coupling rate of change matrix representing the variation characteristics of the normalized temperature parameter in three directions, the second row representing the response characteristics of the impedance phase product parameter in three directions, and the third row representing the variation characteristics of the frequency offset parameter in three directions.

6. The method for identifying tobacco devices based on a metal heating sensor according to claim 1, characterized in that, The step of performing eigenvalue decomposition on the parameter coupling rate of change matrix to obtain deformation feature vectors, and comparing the deformation feature vectors with pre-stored device pattern recognition vectors to obtain a pattern similarity score includes: Based on several 3×3 parameter coupling rate of change matrices, extract several maximum change intensity parameters and several minimum change intensity parameters. Then, average the maximum change intensity parameters and several minimum change intensity parameters respectively to obtain the averaged maximum change intensity parameters and averaged minimum change intensity parameters. Finally, obtain the change intensity ratio based on the maximum change intensity parameters and the averaged minimum change intensity parameters. Several 3×3 parameter coupling rate of change matrices are arranged according to their corresponding positions in the mean grid to construct a 3×N parameter coupling rate of change matrix, and the direction of change is extracted based on the 3×N parameter coupling rate of change matrix. Deformation feature vectors are constructed based on the mean-based maximum change intensity parameter, mean-based minimum change intensity parameter, change direction, and change intensity ratio. The deformation feature vector is compared with the pre-stored device pattern recognition vector item by item to obtain the pattern similarity score.

7. The method for identifying tobacco devices based on a metal heating sensor according to claim 6, characterized in that, The step of comparing the deformation feature vector with the pre-stored device pattern recognition vector item by item to obtain the pattern similarity score includes: Check whether the direction of change is parallel to the standard direction in the pre-stored device pattern recognition vector; Compare whether the mean-based maximum change intensity parameter, the mean-based minimum change intensity parameter, and the standard maximum change intensity parameter and standard minimum change intensity parameter in the pre-stored device pattern recognition vector are of the same order of magnitude; The variation intensity ratio is compared with the standard intensity ratio in the pre-stored device pattern recognition vector; The similarity score represents the degree of matching of each feature in the item-by-item comparison. Features that match perfectly receive full marks, while features with deviations are deducted points according to the degree of deviation. Each degree of matching is mapped to the [0,100] interval to obtain the similarity score.

8. The method for identifying tobacco devices based on a metal heating sensor according to claim 1, characterized in that, The method for obtaining the device determination pattern based on pattern similarity scoring, and executing the pattern mechanism according to the device determination pattern, includes: When the similarity score is greater than or equal to 95, the judgment mode is the matching mode. In the matching mode, the heating control parameters are obtained based on the preset optimal heating curve of the tobacco device. The metal heating sensor is heated according to the heating control parameters, and the LED light on the tobacco device is turned on to display a solid green light. When the similarity score is in the closed range of 85 to 94, the mode is determined to be a suspected mode. The recognition method is executed again in the suspected mode. When the similarity score is greater than or equal to 95, the mode is determined to be a matching mode; otherwise, it is a non-matching mode. When the similarity score is less than 85, it is determined to be a non-matching mode. In the non-matching mode, the tobacco device stops heating the metal heating element, obtains a preset device protection command, obtains warning content based on the device protection command and broadcasts it through a sound device, and activates the LED light on the tobacco device to display a constant red light based on the device protection command.

9. The method for identifying tobacco devices based on a metal heating sensor according to claim 1, characterized in that: The metal heating element is porous vacuum silicon, which is made of nano-silicon atom material.

10. A matching tobacco device identification system based on a metal heating sensor, characterized in that, Including: The acquisition module acquires the real-time temperature and electromagnetic parameters of the metal heating element during the heating process. A storage unit, wherein the storage unit is used to store pre-stored device pattern recognition vectors; The execution unit executes the mode mechanism according to the device determination mode; The intelligent chip is electrically connected to the acquisition module, the storage unit and the execution unit respectively. It is used to construct a three-dimensional parameter grid and a parameter coupling rate of change matrix, perform eigenvalue decomposition, and compare the deformation feature vector with the pre-stored device pattern recognition vector to obtain a pattern similarity score. The device determination pattern is obtained based on the pattern similarity score, and control commands are sent to the execution unit according to the device determination pattern.

Images