Multi-stage over-temperature grading protection atomizing heating control system

Abstract

The present application relates to the technical field of heating control of electronic atomization device, and particularly relates to an atomization heating control system with multi-stage over-temperature grading protection, which comprises a constant current sampling unit, a liquid supply margin calculation unit, an aging decoupling unit, a grading decision unit and a driving rollback unit. The constant current sampling unit injects a constant sampling current into a heating wire and collects the terminal voltage of the heating wire within a driving duty cycle off window, outputs a wire resistance sequence, splits the wire resistance sequence into a heating section resistance value trajectory and an intermittent section resistance value trajectory, and converts a wire temperature sequence. The liquid supply margin calculation unit calculates a single pulse resistance value rising rate from the heating section resistance value trajectory and compares the single pulse resistance value rising rate with a dry state reference rate, and outputs a liquid supply margin factor. The aging decoupling unit fits a falling time constant from the intermittent section resistance value trajectory, outputs a core aging degree, and corrects the dry state reference rate. The grading decision unit outputs a protection gear. The driving rollback unit proportionally rolls back the driving duty cycle or cuts off the heating wire power supply circuit according to an excess amount.

Description

Technical Field

[0001] This invention relates to the field of heating control technology for electronic atomizing devices, and more specifically to an atomizing heating control system with multi-stage over-temperature protection. Background Technology

[0002] Electronic atomizing devices rely on a heating wire to heat and vaporize the atomizing liquid adsorbed by the capillary wick, generating an aerosol. When the atomizing liquid level is insufficient, the liquid supply rate to the capillary wick cannot keep up with the vaporization rate, or the user continuously and intensively inhales, local sections of the heating wire will lose their liquid cover. The temperature in these sections will rise rapidly in a very short time, causing the atomizing liquid to decompose and produce a burnt odor and harmful components. To avoid this situation, the industry generally incorporates over-temperature protection mechanisms in the atomizing heating control.

[0003] Chinese patent application CN106102488A discloses a device and method for measuring and controlling the temperature of an atomizer heating coil, and an electronic cigarette. The device connects a first wire and a second wire, made of different conductor materials, to both ends of the heating coil to form a thermocouple circuit. The electromotive force signal output from the cold ends of the two wires is amplified and processed to obtain the current temperature value of the heating coil. When the current temperature value is greater than or equal to a preset value, the device controls the reduction of the atomization power of the heating coil or shuts off the power supply circuit of the heating coil. However, the preset value used for this decision is a fixed amount determined and stored at the time of manufacture and is not updated after the atomizer coil is put into use, regardless of the coil's own condition.

[0004] International patent application WO2022088081A1 discloses a method for preventing dry burning in electronic cigarettes and an electronic cigarette with this method. After the power supply device detects and acquires a dry burning signal from the atomizing device, the method lowers the heating temperature of the heating element from a first constant temperature to a second constant temperature and continues heating. When the number of times the dry burning signal is acquired within a set time reaches a set number, the heating element is controlled to stop heating. The dry burning signal is obtained based on whether the resistance value of the resistance wire reaches a dry burning resistance threshold or whether the temperature reaches a dry burning temperature threshold. However, the dry burning resistance threshold and dry burning temperature threshold in this solution are also preset fixed values, and the two constant temperatures are also preset constants.

[0005] US Patent Application Publication No. US20160374398A1 discloses a scheme for predicting and communicating the burning of an electronic cigarette. This scheme uses a low-resistance sampling resistor connected in series in the power supply circuit to measure the current flowing through the heating coil and the voltage across the coil, thereby calculating the coil resistance. It then uses the temperature-dependent resistance of the metal material to calculate the coil temperature, eliminating the need for a dedicated temperature sensing element. The scheme reduces the power supplied by the battery to the coil as the coil temperature rises, increases the power as the coil temperature falls, and cuts off the power supply when the temperature approaches the burning point. However, the reference resistance and temperature coefficient used to map the coil resistance to the coil temperature for this decision are still fixed values ​​calibrated during the manufacturing stage.

[0006] While the three existing technologies mentioned above each have their own trade-offs in temperature measurement methods, judgment levels, and execution actions, they are all constrained by the same bottleneck: the benchmark amount used to determine whether the atomizer is insufficient is assumed to be an inherent attribute of the atomizer coil at the factory, calibrated once during the manufacturing stage and effective for its entire lifespan. However, this is not the case in reality. During each over-temperature protection intervention, the capillary coil is subjected to a baking condition with moderate temperature and prolonged periods of poor liquid supply, causing its surface to continuously carbonize. This carbonized layer increases the thermal resistance between the coil and the heating wire, causing the wire temperature to rise faster at the same injection power; on the other hand, it blocks the capillary pores, reducing the capillary liquid supply rate. The combined result is that the rate of increase in wire resistance in a dry state for the same atomizer coil monotonically increases with the number of uses. In other words, the judgment benchmark itself continuously drifts on a scale of dozens of puffs, and the protection action itself is the accelerator of this drift.

[0007] The direct consequence of this is that carbonization aging and insufficient instantaneous liquid supply have the same effect on the rate of increase in coil resistance in the heating section, making them structurally indistinguishable in the heating section. A fixed baseline that is systematically low in the mid-to-late stages of the coil's lifespan causes the increase in coil resistance caused by aging to be misinterpreted as insufficient liquid supply, leading to premature intervention of the protection mechanism. Users subjectively perceive this as the coil becoming less vapor-producing with continued use. To avoid this phenomenon, manufacturers can only relax the baseline overall, which in turn prevents genuine localized liquid shortages from being intercepted before burnt formation. Therefore, within the framework of the fixed baseline, premature protection at the expense of flavor and delayed protection resulting in burnt vapor are two sides of the same coin, mutually exclusive.

[0008] Therefore, there is an urgent need in the field for an atomization heating control scheme that can separate the drift of the decision benchmark due to carbonization aging and the effect of insufficient instantaneous liquid supply on the decision quantity online without the need for external temperature sensing elements. Summary of the Invention

[0009] To address the core bottleneck in existing electronic atomizing devices where the over-temperature protection criterion is assumed to be an inherent factory attribute and cannot be calibrated online as the capillary core ages, making it indistinguishable from insufficient instantaneous liquid supply during heating, this invention provides a multi-stage over-temperature graded protection atomizing heating control system. By extracting the fallback time constant, determined solely by the core's thermal resistance and heat capacity, during the non-phase-change cooling zone between draw cycles, the system online isolates the core's aging degree and reversely corrects the dry-state reference rate. Without adding any dedicated temperature sensing elements, and using only the heating wire itself as the sensor, the system achieves consistency in the intervention time of graded over-temperature thresholds throughout the entire lifespan of the atomizing core, based on the principle of the physical orthogonality between the heating and cooling domains.

[0010] The technical solution of this invention is as follows: a multi-segment over-temperature graded protection atomization heating control system, comprising: a constant current sampling unit, used to inject a constant sampling current into the heating wire and collect the heating wire terminal voltage within the drive duty cycle off-window, outputting a wire resistance sequence, splitting the wire resistance sequence into a heating segment resistance trajectory and an intermittent segment resistance trajectory according to the suction sequence, and converting the wire resistance sequence into a wire temperature sequence; a liquid supply margin calculation unit, used to obtain the single-pulse resistance rise rate from the heating segment resistance trajectory, compare the single-pulse resistance rise rate with the dry-state reference rate, and output a liquid supply margin factor; and an aging decoupling unit, used to fit the intermittent segment resistance trajectory and output a fallback time constant, and convert the fallback time... A constant output core aging degree is provided, and the dry-state reference rate is corrected based on the core aging degree. A graded decision unit is provided, which includes a warning threshold, a power limit threshold, and a power-off threshold. The graded decision unit is used to adjust the graded thresholds based on the liquid supply margin factor and the core aging degree, compare the wire temperature sequence with the adjusted graded thresholds, and output a protection level, which includes a warning level, a power limit level, and a power-off level. A drive retraction unit is used to retract the drive duty cycle proportionally to the amount by which the wire temperature sequence exceeds the power limit threshold when the protection level is the power limit level, and to cut off the heating wire power supply circuit when the protection level is the power-off level.

[0011] Based on the above technical solution, the present invention further specifies: the liquid supply margin calculation unit uses the increment of the resistance rise rate in the later stage of conduction relative to the resistance rise rate in the earlier stage of conduction to characterize the degree of release of the vaporization latent heat clamp; the aging decoupling unit takes the section in the intermittent resistance trajectory where the driving duty cycle is zero and the wire temperature sequence is lower than the boiling point of the e-liquid as the non-phase change cooling section, and is equipped with a dual time scale recursive estimator, setting the core aging degree as a slow-changing state across the suction cycle and the liquid supply margin factor as a fast-changing state within the suction cycle; the grading decision unit is equipped with a hysteresis band and obtains the aging growth rate from the core aging degree, and uses the aging growth rate to adjust the bandwidth and retracement slope of the hysteresis band; the grading decision unit is also equipped with a gray box constraint model, which solves for the maximum allowable power by simultaneously solving for the capillary liquid supply rate constraint and the carbonization rate constraint.

[0012] The beneficial effects of this invention are as follows.

[0013] Firstly, this invention ensures that the overheating decision remains consistent throughout the entire lifespan of the atomizer core. The mechanism lies in the fact that the non-phase-change cooling section operates under conditions where the drive duty cycle is zero and the coil temperature sequence is below the boiling point of the e-liquid. There is neither electrical power injection nor latent heat of phase change involved. The fallback time constant is determined solely by the product of the core's thermal resistance and its heat capacity, and is independent of the instantaneous liquid content of the capillary core. The growth of the carbonized layer is precisely reflected in the monotonic increase of the core's thermal resistance. Based on this, the online core aging degree is used to correct the dry-state reference rate.

[0014] Secondly, this invention ensures that the protective intervention occurs before charring occurs. The mechanism is that the liquid supply margin calculation unit does not wait for the average wire temperature to exceed a threshold, but compares the rate of resistance increase before and after conduction within a single drive pulse conduction period. When the liquid coverage is sufficient, the latent heat of vaporization significantly amplifies the effective heat capacity of the heating wire, thus clamping the temperature rise in the latter part. When a local section loses coverage, this clamping is immediately released. The degree of release of the latent heat of vaporization clamping occurs abruptly before the average wire temperature exceeds the limit. Wire resistance is a spatial integral along the entire length of the heating wire. The contribution of the local dry area to the average wire temperature is diluted by the wet area still on the boiling platform. Therefore, any criterion based on the average wire temperature or its single threshold will inevitably result in a delayed action.

[0015] Third, there is a synergistic effect between the two effects mentioned above, exceeding the sum of their individual effects. The mechanism lies in the fact that the normalized denominator of the latent heat of vaporization clamp release is precisely the dry-state reference rate. If this denominator remains a fixed quantity, the preceding quantity will be systematically amplified with aging, losing its comparability. Furthermore, without the latent heat of vaporization clamp release as a rapidly changing observable measure, it is impossible to determine whether the current core is short of liquid. Both are derived from the heating and cooling domains, two physically orthogonal observation channels, respectively. Their merging produces a new capability that neither can possess alone. None of the three existing technologies mentioned above involve the cooling domain observation channel.

[0016] Fourth, this invention suppresses the positive feedback between the protection action and the aging of the capillary core, while maintaining a continuous suction experience within the power limit range. The mechanism is that as the aging growth rate increases, the hysteresis band bandwidth widens and the backoff slope slows down, causing the protection setting to recur less frequently, thereby shortening the cumulative time the capillary core is under medium-temperature baking conditions. Simultaneously, the drive backoff unit proportionally backs off the drive duty cycle according to the excess amount instead of directly cutting it off, and uses the maximum allowable power obtained from the gray box constraint model as the lower limit of the backoff. Attached Figure Description

[0017] Figure 1 This is the overall system architecture diagram of the present invention.

[0018] Figure 2 This is a structural diagram of the constant current sampling unit of the present invention.

[0019] Figure 3 This is a structural diagram of the liquid supply margin calculation unit of the present invention.

[0020] Figure 4 This is a structural diagram of the old decoupling unit of the present invention.

[0021] Figure 5 This is a structural diagram of the hierarchical decision unit of the present invention.

[0022] Figure 6 This is a structural diagram of the drive rollback unit of the present invention. Detailed Implementation

[0023] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. The following examples are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0024] like Figure 1As shown, the multi-stage over-temperature graded protection atomization heating control system of this embodiment is installed inside the power rod of the electronic atomization device and is electrically connected to the heating wire 6 on the atomization core. It includes a constant current sampling unit 1, a liquid supply margin calculation unit 2, an aging decoupling unit 3, a graded decision unit 4, and a drive retraction unit 5. The heating wire 6 is made of nickel-chromium alloy wire or stainless steel wire, with a room temperature resistance of 0.6Ω to 1.8Ω. It is wound around the cotton capillary core, and both ends of the capillary core are immersed in the liquid storage chamber.

[0025] The five units are interconnected in the data stream, forming multiple loops. Constant current sampling unit 1 injects a constant sampling current into the heating wire 6 within the drive duty cycle shutdown window and collects the heating wire terminal voltage, outputting a wire resistance sequence. This sequence is then divided into heating segment resistance trajectories and intermittent segment resistance trajectories according to the suction sequence, and the wire temperature sequence is calculated from the resistance sequence. The heating segment resistance trajectory is sent to the liquid supply margin calculation unit 2 to calculate the single-pulse resistance rise rate and compare it with the dry-state reference rate, outputting the liquid supply margin factor. The intermittent segment resistance trajectory is sent to the aging decoupling unit 3 to fit the output fallback time constant, thus outputting the core aging degree. The wire temperature sequence, liquid supply margin factor, and core aging degree are all sent to the grading decision unit 4, where the latter two are used to set the grading threshold. The wire temperature sequence is then compared with the set grading threshold to output the protection level. The protection level is sent to the drive retraction unit 5 to retract the drive duty cycle or cut off the heating wire power supply circuit.

[0026] The multiple loops in this system are manifested in four places. The first loop is that the core aging degree output by the aging decoupling unit 3 is fed back to the liquid supply margin calculation unit 2 to correct its dry-state reference rate, causing the output of the subsequent unit to rewrite the calculation parameters of the preceding unit. The second loop is that the core aging degree is simultaneously sent to the grading decision unit 4 to participate in the tuning of the grading threshold, causing the decision criterion to shift with the core state. The third loop is that the aging growth rate obtained by the grading decision unit 4 from the core aging degree is sent to the drive backoff unit 5 to tune its backoff slope, while simultaneously acting on the hysteresis bandwidth of the grading decision unit 4 itself. The fourth loop is that the drive duty cycle output by the drive backoff unit 5 determines the position and length of the drive duty cycle shutdown window for the next cycle, thereby changing the sampling timing of the constant current sampling unit 1, making the entire data stream a closed loop.

[0027] The existence of these four loops distinguishes this system from a simple feedforward threshold comparison structure. In particular, the first loop transmits the calculation benchmark rather than the decision result, which is the structural premise for this system to maintain the comparability of decision quantities under the condition that carbonization and aging continue.

[0028] like Figure 2As shown, the constant current sampling unit 1 includes a constant current source, a sampling switch, a differential amplifier, an analog-to-digital converter, and a timing controller. The output of the constant current source is connected to the high end of the heating wire 6 via the sampling switch. The two inputs of the differential amplifier are connected to the two ends of the heating wire 6, and its output is connected to the input of the analog-to-digital converter. The timing controller is connected to the sampling switch, the analog-to-digital converter, and the drive rollback unit 5.

[0029] The timing controller determines the start and end times of the drive duty cycle shutdown window based on the current drive duty cycle output by the drive rollback unit 5. Within each drive duty cycle shutdown window, the timing controller first completely disconnects the power stage to eliminate the voltage drop generated by the large current on the lead, then closes the sampling switch to inject a constant sampling current into the heating wire 6 from the constant current source, and then starts the analog-to-digital converter to collect the voltage at the heating wire terminals. The constant sampling current ranges from 1mA to 20mA, and in this embodiment, it is 5mA. The lower limit of this value is limited by the input offset and common-mode rejection capability of the differential amplifier. Below 1mA, the voltage at the heating wire terminals falls into the millivolt range, resulting in insufficient signal-to-noise ratio. The upper limit is limited by the self-heating of the sampling process itself. Above 20mA, the power generated by the sampling current on the 1Ω-level heating wire 6 will be non-negligible, which in turn contaminates the measured wire temperature. The sampling duration within the drive duty cycle turn-off window ranges from 50μs to 500μs; in this embodiment, it is 200μs. This duration must be longer than the settling time of the differential amplifier but shorter than the turn-off window itself to avoid encroaching on the power output time. The sampling period of the output wire impedance sequence of the constant current sampling unit 1 ranges from 1ms to 10ms; in this embodiment, it is 2ms, synchronized with the drive pulse period.

[0030] The wire resistance is obtained by dividing the voltage across the heating wire by the constant sampling current. The wire temperature is calculated from the wire resistance based on the resistance-temperature characteristics of the heating wire material. ,in: For the first The wire temperature at each sampling moment is a scalar, representing the spatial average temperature of the heating wire 6 along its entire length. Its value ranges from 20℃ to 400℃, and the unit is ℃. It is calculated by this formula and is an element of the wire temperature sequence. The sampling sequence number is a positive integer scalar with a value range from 1 to the total number of sampling points in a single sampling. It has no unit and is generated by the timing controller according to the sampling period. It is used to identify the sampling time. The reference temperature is a scalar value ranging from 20℃ to 30℃, with the unit being ℃. It is measured from the ambient temperature when the device is powered on and is used as the reference point for the resistance temperature characteristics. For the first The wire resistance at each sampling moment is a scalar, ranging from 0.5Ω to 3.0Ω, with the unit Ω. It is obtained by dividing the heating wire terminal voltage by the constant sampling current and is an element of the wire resistance sequence. The resistance of the wire at the reference temperature is a scalar value, ranging from 0.6Ω to 1.8Ω, and the unit is Ω. It is measured once when the device is powered on and the heating wire 6 is not powered on, and is used as the reference resistance value for the resistance temperature characteristic. The temperature coefficient of resistance of heating wire 6 is a scalar value. For nickel-chromium alloy, it ranges from 0.00010℃⁻¹ to 0.00020℃⁻¹, and for stainless steel, it ranges from 0.00075℃⁻¹ to 0.00110℃⁻¹. The unit is ℃⁻¹, determined by referring to a table based on the material grade of heating wire 6. A higher value indicates a greater change in resistance due to the same temperature rise, resulting in higher temperature measurement resolution. The timing controller simultaneously divides the wire resistance sequence into two categories based on the on / off state of the airflow switch. The wire resistance sequence segment corresponding to the period when the airflow switch is closed and the drive duty cycle is greater than zero constitutes the resistance trajectory of the heating section; the wire resistance sequence segment corresponding to the period when the airflow switch is open or the drive duty cycle is zero constitutes the resistance trajectory of the intermittent section. Both types of trajectories are labeled with the suction cycle number and then sent downstream.

[0031] like Figure 3 As shown, the liquid supply margin calculation unit 2 includes a pulse segmenter, a rate solver, a latent heat of vaporization clamping release calculator, and a margin mapper. The input terminal of the pulse segmenter is connected to the heating section resistance trajectory output terminal of the constant current sampling unit 1, and its two output terminals are respectively connected to the rate solver. The output terminal of the rate solver is connected to the latent heat of vaporization clamping release calculator, and the output terminal of the latent heat of vaporization clamping release calculator is connected to the margin mapper. The output terminal of the margin mapper is connected to the graded decision unit 4. The latent heat of vaporization clamping release calculator also has a reference input terminal, which is connected to the dry state reference rate output terminal of the aging decoupling unit 3.

[0032] The pulse segmenter divides the conduction period of a single driving pulse into a pre-conduction segment and a post-conduction segment along the time axis. The division point is taken at 35% to 55% of the total conduction period; in this embodiment, 45% is used. The rate solver performs least-squares linear fitting on the heating section resistance trajectory sampling points falling into the pre-conduction and post-conduction segments, respectively, and takes the slope as the resistance rise rate of the pre-conduction and post-conduction segments.

[0033] The physical basis for the above segmented comparison is that when the capillary core is sufficiently covered by liquid, a considerable proportion of the heat injected into the heating wire 6 is absorbed by the latent heat of vaporization of the atomizing liquid, which is equivalent to significantly amplifying the effective heat capacity of the heating wire 6. The temperature rise during the conduction period is clamped near the boiling plateau of the atomizing liquid, and the resistance rise rate in the later stage of conduction is comparable to that in the earlier stage of conduction. Once a local section of the heating wire 6 loses liquid coverage, no latent heat participates in heat absorption in that section, and its temperature rise quickly breaks free from clamping in the latter half of the conduction period. This is reflected in the wire resistance obtained by integrating along the entire length as a positive increment in the resistance rise rate in the later stage of conduction relative to the resistance rise rate in the earlier stage of conduction.

[0034] The latent heat of vaporization clamp release calculator calculates the latent heat of vaporization clamp release degree based on this:

[0035] ,in: For the first The degree of release of the latent heat of vaporization clamping in one suction cycle is a scalar quantity, which represents the degree to which the latent heat of vaporization clamping on the heating wire 6 is released. The value ranges from 0 to 1, has no unit, and is calculated by this formula. The closer the value is to 1, the greater the proportion of the section that has lost liquid coverage. This is the vaping cycle number, a positive integer scalar, ranging from 1 to the total number of vaping cycles within the lifespan of the atomizer core. It has no unit and is incremented by one each time the airflow switch is closed. It is used to identify the vaping cycle. For the first The rate of increase in resistance during the latter part of each pumping cycle is a scalar quantity, ranging from 0 Ω / s to 20 Ω / s, with units of Ω / s. It is obtained by least-squares fitting of the sampling points during the latter part of the conduction period by the rate solver, and characterizes the rate of temperature rise in the second half of the conduction period. For the first The rate of increase of the resistance in the conduction front section of each pumping cycle is a scalar with a value ranging from 0Ω / s to 20Ω / s and a unit of Ω / s. It is obtained by the rate solver through least-squares fitting of the sampling points in the conduction front section and is used as a self-reference baseline within the same pulse. For the first The dry reference rate for each vaping cycle is a scalar with a value ranging from 0.5Ω / s to 20Ω / s, and the unit is Ω / s. It is given by the aging decoupling unit 3 and is used as a normalization scale to make the clamp release degree of different batches of atomizing cores comparable. and These represent the minimum and maximum value operations, respectively, used to truncate the results between 0 and 1 to avoid out-of-bounds errors caused by noise. The margin mapper maps the latent heat of vaporization clamp release degree to the liquid supply margin factor:

[0036] ,in: For the first The liquid supply margin factor for each suction cycle is a scalar that characterizes the degree of surplus of the current liquid supply capacity of the capillary core relative to the heat generation power. Its value ranges from 0 to 1 and has no unit. It is calculated by this formula. When its value is 1, it means that the liquid supply is sufficient, and when it is 0, it means that the liquid supply is completely unable to match the heat generation. The definition is the same as the aforementioned formula for the degree of release of latent heat of vaporization clamping; The set release degree is a scalar value ranging from 0.05 to 0.20. In this embodiment, it is 0.12. It has no unit and is determined by the 95th percentile value of the clamp release degree measured by the whole machine after 50 continuous suctions at standard liquid level and standard power. Its function is to set a dead zone to accommodate sampling noise and normal breathing fluctuations of the capillary core. If the value is too small, normal suction will be frequently misjudged as insufficient liquid supply. If the value is too large, it will erode the time margin gained by the prior decision. The margin decay index is a scalar quantity with a value ranging from 1.0 to 3.0. In this embodiment, it is set to 2.0. It has no unit and is determined by the overall feel adjustment. The larger the value, the more gently the liquid supply margin factor decreases in the initial stage of clamp release and the steeper the decrease in the later stage. It is used to control the softness or hardness of the protective intervention. The liquid supply margin factor decreases monotonically with the increase of the latent heat of vaporization clamp release degree, and takes a full value of 1 when the latent heat of vaporization clamp release degree is lower than the set release degree, so that it does not interfere with normal suction when the liquid supply is sufficient.

[0037] like Figure 4 As shown, the aging decoupling unit 3 includes a segment filter, a time constant fitter, an aging degree normalizer, a reference corrector, a dual-time-scale recursive estimator, and a life indicator. The input of the segment filter is connected to the intermittent segment resistance trajectory output and the wire temperature sequence output of the constant current sampling unit 1, and its output is connected to the time constant fitter and the aging degree normalizer in sequence. The output of the aging degree normalizer is connected to the reference corrector, the dual-time-scale recursive estimator, the life indicator, and the graded decision unit 4, respectively. The output of the reference corrector is connected to the reference input of the liquid supply margin calculation unit 2.

[0038] The segment filter extracts continuous segments from the intermittent resistance trajectory that simultaneously meet two conditions as non-phase change cooling segments: first, the drive duty cycle at the corresponding moment of the segment is zero, thus excluding any electrical power injection; second, the coil temperature sequence corresponding to the segment is lower than the e-liquid boiling point, thus excluding the participation of latent heat of phase change. The e-liquid boiling point is selected between 188°C and 290°C based on the propylene glycol to glycerol ratio. In this embodiment, 225°C is selected for the atomizer with a propylene glycol to glycerol mass ratio of 50:50.

[0039] The significance of the above two conditions lies in the fact that, in segments where both conditions are simultaneously met, the heat dissipation of the heating wire from the capillary core to the environment is a purely thermal conduction process. Its decay rate is determined solely by the product of the core's thermal resistance and its heat capacity, and is unrelated to the amount of atomizing liquid contained in the capillary core at that moment. The accumulation of the carbonized layer is precisely manifested as a monotonic increase in the core's thermal resistance; therefore, the decay rate in this segment becomes a clean observation of the degree of carbonization. This is the fundamental reason why this invention places aging observation in the cooling domain and liquid supply observation in the heating domain: in the heating domain, both have the same direction of influence on the rate of increase in wire resistance and are therefore indistinguishable, while in the cooling domain, only the former leaves a trace.

[0040] The time constant fitter fits a first-order thermal decay model to the cooling section without phase change:

[0041] ,

[0042] in: The definition is the same as the aforementioned silk temperature conversion formula; The ambient temperature is a scalar value ranging from 0°C to 45°C, and the unit is °C. It is read from the internal temperature reference at the end of the suction interval and is used as an asymptotic value for the decay process. The filament temperature at the beginning of the phase-change-free cooling section is a scalar value ranging from 100℃ to 290℃, with the unit being ℃. It is taken from the value of the first sampling point of the filament temperature sequence in this section and is used as the initial value for the decay process. For the first Each sampling time is a scalar, with a value range from 0s to the interval between single suctions, and the unit is s. It is obtained by multiplying the sampling number by the sampling period. The starting time of the non-phase-change cooling section is a scalar quantity in seconds, given by the section filter. For the first The fallback time constant for each suction cycle is a scalar that represents the product of the core thermal resistance and the core thermal capacity. Its value ranges from 0.8s to 6.0s and is expressed in seconds. It is obtained by log-linear least squares fitting of this model and its value increases with the thickness of the carbonized layer. This represents an exponential function with the natural constant as its base.

[0043] When the number of sampling points in the non-phase-change cooling section is insufficient to support the above fitting, the time constant fitter uses the initial slope method to give the fallback time constant: ,in: , , , The definition is the same as the aforementioned first-order thermal decay model; The time derivative of the wire temperature sequence at the beginning of the cooling section without phase change is a scalar with a value range of -200℃ / s to 0℃ / s and a unit of ℃ / s. It is obtained by the difference between the first two to the first four sampling points in this section and characterizes the rate of decay in the initial stage. This represents the absolute value operation, used to eliminate the sign caused by the negative derivative during the descent process. The aging degree normalizer normalizes the fallback time constant to the core aging degree: ,in: For the first The core aging degree of one vaping cycle is a scalar quantity that characterizes the cumulative degree of carbonization on the surface of the capillary core. The value ranges from 0 to 1 and has no unit. It is calculated by this formula. When the value is 0, it means that the atomizing core is in a brand new state, and when it is 1, it means that it has reached the end of its life. The definition is the same as the aforementioned first-order thermal decay model; The fallback time constant of the new core is a scalar value ranging from 0.8s to 2.0s, with the unit being seconds. It is measured and the median value is taken and solidified within the first three vaping cycles after the initial assembly of the atomizing core, and used as the lower endpoint of the normalization. The fallback time constant at the end of the lifespan is a scalar value ranging from 3.0s to 6.0s, with the unit being seconds. It is obtained from accelerated aging tests of the same type of atomizer core. The fallback time constant corresponding to when the atomization output decreases to 70% of the initial value is taken as the upper limit of normalization. If its value is too small, it will cause premature aging saturation and loss of resolution. If it is too large, the aging degree at the end of the lifespan will still be too low and the compensation will be insufficient. and The definition is the same as the aforementioned formula for the degree of release of the latent heat of vaporization clamp. The reference corrector uses this formula to correct the dry-state reference rate: ,in: The definition is the same as the aforementioned formula for the degree of release of latent heat of vaporization clamping, and this formula is the method for obtaining it; The initial dry reference rate is a scalar value ranging from 0.5Ω / s to 12Ω / s, with the unit being Ω / s. It is determined by the rate of increase of the single-pulse resistance measured by a brand-new atomizing core of the same model driven at rated power in a liquid-free state, and is used as the baseline before aging compensation. The base correction coefficient is a scalar with a value range of 0.15 to 0.60. In this embodiment, it is 0.35. It has no unit and is obtained by fitting the dry-burn test of the same type of atomizing core under different aging degrees. Its physical meaning is the relative increase in the dry-state heating rate after the carbonized layer increases the thermal resistance of the core. If the value is too small, the aging compensation will be insufficient and the protection will still intervene too early. If the value is too large, it will mask the real liquid supply shortage at the end of the life. The definition is the same as the aforementioned core aging formula. It should be noted that this modification and the aging term in the subsequent grading decision unit 4 perform two independent functions. This modification eliminates the bias introduced by carbonization aging in the observation, making the liquid supply margin factor unbiased for aging; the aging term in grading decision unit 4 addresses the downward shift in the tolerance temperature of the roughened core itself after the carbonization layer has formed, and the two do not constitute redundant limitations.

[0044] The life indicator compares the coil aging degree with the preset life end aging degree. When the coil aging degree is not lower than the preset life end aging degree for 20 consecutive vaping cycles, it outputs an atomizer coil life indicator signal. The preset life end aging degree is between 0.80 and 0.95, and is 0.85 in this embodiment.

[0045] In real-world use, the intervals between suction stops vary. During intensive suction, the interval between adjacent stops can be as short as 0.5 seconds. At this point, the number of sampling points in the non-phase-change cooling section is insufficient, and the atomized liquid replenishment of the capillary core is not yet complete. If the supply margin factor is updated normally under these circumstances, the residual heat from the previous stop and the incomplete liquid replenishment will cause the observed values ​​in the heating section to be systematically low, thus misjudging intensive suction as insufficient supply. On the other hand, the core aging degree evolves on a scale of dozens of suction stops. Even if a single observation contains significant noise, the noise is significantly averaged out after accumulating across cycles. Based on this asymmetry, this embodiment sets up a dual-time-scale recursive estimator within the aging decoupling unit 3.

[0046] The dual-timescale recursive estimator sets the core aging degree as a slowly varying state across the pumping cycle and the liquid supply margin factor as a rapidly varying state within the pumping cycle. The recursive update formulas for the two are as follows:

[0047] ,

[0048] ,

[0049] in: For the first The slow-changing state across each pumping cycle is a scalar, representing the estimated core aging degree after cross-cycle smoothing. Its value ranges from 0 to 1 and has no unit. It is calculated recursively from the above formula and is used by the graded decision unit 4 and the benchmark corrector. This refers to the slowly varying state across the previous pumping cycle, with the same value range and unit. It is obtained by saving the recursive result of the previous cycle; The cumulative gain across pumping cycles is a scalar value ranging from 0.002 to 0.020. In this embodiment, it is 0.008. It has no unit and is determined by dividing the target smoothing time constant by the reciprocal of the average pumping cycle duration. The smaller the value, the smoother the slow-change state, but the slower the follow-up to the actual aging. The definition is the same as the aforementioned core aging formula, and it is used as a single observation value in this formula; For the first The rapidly changing state within each suction cycle is a scalar quantity, representing the estimated value of the liquid supply margin factor after filtering within the cycle. The value ranges from 0 to 1 and has no unit. It is calculated recursively from the above formula and is used by the grading decision unit 4. This refers to the rapidly changing state within the previous suction cycle, with the same value range and unit. ; The gain during the pumping cycle is a scalar quantity, ranging from 0.20 to 0.60. In this embodiment, it is 0.40. It has no unit and is determined by the maximum allowable response hysteresis within a single port. The larger the value, the faster the response to liquid shortage but the more sensitive it is to noise. The definition is the same as the aforementioned formula for the liquid supply margin factor, and it is used as a single observation value in this formula. For the first The number of sampling points in the non-phase change cooling section within each pumping cycle is a positive integer scalar with a value range of 0 to 3000. It has no unit and is obtained by counting by the section filter. The number of points is a positive integer scalar with a value range of 20 to 200. In this embodiment, it is set to 60. It has no unit and is determined by dividing the shortest acceptable interval duration by the sampling period. Its function is to identify intensive suction conditions. If the value is too small, intensive suction cannot be identified. If the value is too large, normal suction will also be falsely frozen.

[0050] As can be seen from the above formula, when the number of sampling points in the non-phase-change cooling section is lower than the set number of points, the dual-time-scale recursive estimator still updates the slow-changing state across the pumping cycle with the cumulative gain across the pumping cycle, while keeping the fast-changing state within the pumping cycle at the value of the previous cycle. In this case, the slow-changing state is based on the fallback time constant given by the initial slope method. Although its single estimation has a deviation, its impact on the cumulative result is limited due to the extremely small gain; the fast-changing state is frozen, thus avoiding systematic misjudgment under intensive pumping conditions.

[0051] The aforementioned structure mathematically eliminates the original indiscernibility. Core aging and transient fluid shortage have the same effect on the rate of increase in wire resistance during heating. If described by a single state variable, the observation equation would be rank deficient. In this embodiment, the two are placed on two time scales that differ by a factor of 50, meaning that the ratio of the cumulative gain across the pumping cycle to the gain within the pumping cycle is no higher than 0.05. This separates their observable subspaces in the frequency domain, thus eliminating the rank deficiency structurally rather than in terms of data.

[0052] like Figure 5 As shown, the graded decision unit 4 includes a threshold tuner, a comparator array, a hysteresis manager, an aging acceleration solver, and a slope monitor. The two inputs of the threshold tuner are connected to the two outputs of the dual-timescale recursive estimator, and its three outputs are connected to the comparator array. The other input of the comparator array is connected to the wire temperature sequence output of the constant current sampling unit 1, and its output is connected to the hysteresis manager. The input of the aging acceleration solver is connected to the slow-changing state output of the dual-timescale recursive estimator, and its output is connected to the hysteresis manager and the drive back-off unit 5, respectively. The input of the slope monitor is connected to the wire temperature sequence output, and its output is connected to the hysteresis manager. The output of the hysteresis manager is connected to the drive back-off unit 5. The threshold tuner sets the graded threshold using the following formula: ,in: For the first The grading threshold after each suction cycle is a scalar value, ranging from 150℃ to 330℃, and is calculated using this formula. Pick , , These correspond to the warning threshold, power limit threshold, and power outage threshold, respectively. For threshold category identifier, is the symbol subscript, and the set of values ​​is: Unitless, specified by threshold category; The nominal value of the corresponding category grading threshold is a scalar quantity in °C. In this embodiment, the nominal value of the warning threshold is 235 °C, the nominal value of the power limit threshold is 255 °C, and the nominal value of the power failure threshold is 285 °C. It is determined by the pyrolysis initiation temperature of the atomizing liquid and the sensory evaluation. The margin sensitivity coefficient is a scalar value ranging from 10°C to 40°C. In this embodiment, the three values ​​are 18°C, 22°C, and 15°C respectively, with the unit being °C. It is calibrated by the whole machine in the test of artificially evacuating the liquid level. Its function is to make the graded threshold press down synchronously when the liquid supply margin factor decreases so as to intervene in advance. If the value is too small, the advance judgment cannot be reflected. If the value is too large, the slight margin fluctuation during normal pumping will trigger the protection. The aging sensitivity coefficient is the corresponding category. It is a scalar value with a range of 5℃ to 30℃. In this embodiment, the three values ​​are 10℃, 12℃ and 20℃ respectively. The unit is ℃. It is obtained by fitting the measured curve of the cracking initiation temperature of the atomizing core with different aging degrees. Its function is to compensate for the downward shift of the capillary core's tolerance temperature after the formation of the carbonized layer. and The definition is the same as the aforementioned dual-timescale recursive formula. The comparator array compares the wire temperature sequence with the calibrated graded thresholds point by point and outputs the protection level. When the wire temperature sequence is below the warning threshold, the protection level is in normal mode, and no intervention is performed; when the wire temperature sequence enters between the warning threshold and the power limiting threshold, the protection level is in warning mode; when it enters between the power limiting threshold and the power off threshold, the protection level is in power limiting mode; when it exceeds the power off threshold, the protection level is in power off mode.

[0053] The aging rate calculation unit obtains the aging rate by differentially calculating the slow-changing state across pumping cycles using a sliding window. The sliding window length ranges from 20 to 100 pumping cycles; in this embodiment, 50 pumping cycles are used. The value is the difference between the slow-changing states at both ends of the window divided by the window length, with the unit being the negative first power of the period. The hysteresis manager has a hysteresis band, the bandwidth of which is set by the following formula: ,in: For the first The hysteresis bandwidth of each suction cycle is a scalar, representing the width of the wire temperature drop range required to exit the current protection level. Its value ranges from 5℃ to 40℃, and the unit is ℃. It is calculated by this formula. The nominal value of the hysteresis bandwidth is a scalar value ranging from 5°C to 15°C. In this embodiment, it is taken as 8°C. The unit is °C. It is determined by three times the standard deviation of the measurement noise of the wire temperature sequence and is used to prevent jitter near the threshold. is the bandwidth setting coefficient, which is a scalar with a value range of 50 cycles to 400 cycles. In this embodiment, it is taken as 200 cycles. The unit is the cycle. It is calibrated by the lowest protection mode switching frequency in the accelerated aging test of the whole machine. The larger the value, the more significant the hysteresis widening during accelerated aging. For the first The aging growth rate per pumping cycle is a scalar value ranging from 0 cycles⁻¹ to 0.01 cycles⁻¹, with the unit being cycles⁻¹. It is given by the aging growth rate solver and characterizes the speed of carbonization propagation.

[0054] The above equation reveals a neglected causal relationship: each time the protection mode enters the power-limiting mode, the capillary core is subjected to a prolonged period of moderate temperature and poor liquid supply. This condition is precisely the one that accelerates carbonization the most, thus the protection action itself acts as an accelerator for aging. If the hysteresis bandwidth is fixed, the repeated switching of the protection mode after accelerated aging will become more frequent, further accelerating aging and forming positive feedback. In this embodiment, the hysteresis bandwidth is widened in reverse by increasing the aging growth rate, thus transforming this positive feedback into negative feedback.

[0055] The slope monitor calculates the temperature rise slope of the wire by the first-order difference of the wire temperature sequence. When the temperature rise slope exceeds the set slope threshold, it directly outputs a power-off signal. The set slope threshold ranges from 40℃ / s to 120℃ / s, and in this embodiment, it is set to 75℃ / s. This channel runs parallel to the aforementioned threshold comparison channel, serving as a redundant safety channel for extreme conditions such as when the heating wire 6 is completely detached from the liquid cover.

[0056] The hysteresis manager also manages the exit of the protection mode. When the protection mode exits from the power-off mode, the wire temperature sequence must be continuously lower than the difference between the power-off threshold and the hysteresis bandwidth, and the recovery time must be continuously set; the set recovery time is between 0.5s and 3s, and is 1.5s in this embodiment. The lower limit of this time is taken from the shortest time required for the capillary core to re-wet the heating wire 6, and the upper limit is limited by the user's tolerance for the waiting time for fog to reappear.

[0057] like Figure 6 As shown, the drive rollback unit 5 includes a proportional rollback unit, a gray box constraint solver, a duty cycle synthesizer, and a power stage. The two inputs of the proportional rollback unit are connected to the protection mode output of the hysteresis manager and the wire temperature sequence output of the constant current sampling unit 1, respectively. The input of the gray box constraint solver is connected to the core aging output of the aging normalizer. The outputs of the proportional rollback unit and the gray box constraint solver are connected to the duty cycle synthesizer, the output of the duty cycle synthesizer is connected to the power stage, and the output of the power stage is connected to the heating wire 6. At the same time, the current drive duty cycle is sent back to the timing controller of the constant current sampling unit 1.

[0058] When the proportional backoff device is in the power limit protection position, it proportionally backs off the drive duty cycle according to the amount by which the wire temperature exceeds the power limit threshold. ,in: For the first The midpoint of the drive duty cycle after the proportional back-off of each pumping cycle is a scalar, ranging from 0 to 1, and has no unit. It is calculated by this formula. The rated duty cycle is a scalar quantity with a value ranging from 0.6 to 1.0. In this embodiment, it is taken as 0.95. It has no unit and is determined by the rated power of the whole machine and the battery voltage. For the first The retrace slope of each pumping cycle is a scalar, ranging from 0.005℃⁻¹ to 0.080℃⁻¹, in ℃⁻¹, and is given by the tuning formula below, representing the percentage reduction in duty cycle for every 1℃ increase. The definition is the same as the aforementioned silk temperature conversion formula; For the first The power limit threshold after each suction cycle is tuned is defined in the aforementioned graded threshold tuning formula. Pick The situation; superscript This indicates that the value is an intermediate value that has not yet been combined with the lower limit of regression. The regression slope is adjusted according to the aging growth rate: ,in: The definition is the same as the aforementioned proportional regression formula; The nominal value of the back slope is a scalar, ranging from 0.010℃⁻¹ to 0.080℃⁻¹. In this embodiment, it is taken as 0.030℃⁻¹, and the unit is ℃⁻¹. It is determined by the requirement that the time required for the wire temperature to converge to the power limit threshold within the power limit range does not exceed 0.3s. is the slope tuning coefficient, a scalar, with a value range of 50 to 400 cycles. In this embodiment, it is taken as 150 cycles, with the unit being cycles. It is calibrated from the same source as the bandwidth tuning coefficient but is taken independently. The definition is the same as the aforementioned hysteresis bandwidth tuning formula. The gray box constraint solver solves for the maximum allowable power by simultaneously solving for the capillary supply rate constraint and the carbonization rate constraint. The capillary supply rate constraint characterizes the maximum evaporation power that the supply side can support after the carbonization layer blocks the capillary pores:

[0059] ,in: For the first The maximum evaporation power that the liquid supply side can support in one suction cycle is a scalar quantity, ranging from 0W to 30W, and is calculated by this formula. The upper limit of the capillary liquid supply rate of the new core is a scalar quantity, ranging from 0.0002 kg / s to 0.0012 kg / s. In this embodiment, it is taken as 0.0005 kg / s. The unit is kg / s. It is measured by the weight loss method of the new atomizing core and characterizes the maximum transport capacity of the capillary structure under saturated pressure difference. The definition is the same as the aforementioned core aging formula; The aging degradation index is a scalar value ranging from 2.0 to 4.0. In this embodiment, it is 3.0. It has no unit and is determined by the power law relationship between capillary permeability and porosity. The larger the value, the more rapidly the liquid supply capacity decays with carbonization. The latent heat of vaporization of the atomizing fluid is a scalar value ranging from 800,000 J / kg to 1,000,000 J / kg. In this embodiment, 914,000 J / kg is used for an atomizing fluid with a propylene glycol to glycerol mass ratio of 50:50. The unit is J / kg, determined by referring to a table based on the atomizing fluid formulation. The carbonization rate constraint is given by the carbonization reaction kinetics, which provides the upper limit of the wire temperature. The carbonization rate increases exponentially with the wire temperature. By ensuring that it does not exceed the set carbonization rate, the upper limit of the thermodynamic temperature of the wire temperature can be obtained by inverse solving.

[0060] ,in: The upper limit of the thermodynamic temperature of the wire temperature allowed by the carbonization rate constraint is a scalar quantity with a value range of 450K to 620K and a unit of K. It is calculated by this formula. The carbonization activation energy is a scalar quantity, ranging from 80,000 J / mol to 140,000 J / mol. In this example, it is taken as 110,000 J / mol. The unit is J / mol, which is determined by thermogravimetric analysis of cotton capillary core. is the molar gas constant, a scalar quantity, with a value of 8.314 J / (mol·K), and the unit is J / (mol·K); Here, is the pre-exponential factor for carbonization, a scalar with a value range of 1 × 10⁻⁶. 8 s⁻¹ to 1×10¹²s⁻¹, in this embodiment we take 1×10¹. 0 s⁻¹, in units of s⁻¹, is obtained by synchronous fitting with the same set of thermogravimetric analysis data as the carbonization activation energy; To define the carbonization rate, is a scalar value with a range of 1 × 10⁻ 7 s⁻¹ to 1×10⁻ 5 s⁻¹, in this embodiment, 2×10⁻ 6 s⁻¹, the unit is s⁻¹, which is determined by back-calculation based on the cumulative carbonization conversion rate not exceeding 30% within the target life. If its value is too large, it will lose the meaning of limiting carbonization, and if it is too small, the upper limit of the wire temperature will be too low and the atomization output will be insufficient. This represents the natural logarithm operation. The maximum allowable power is determined jointly by the liquid supply side and the heat transfer side.

[0061] ,in: For the first The maximum allowable power per pumping cycle is a scalar quantity, ranging from 0W to 40W, and is calculated using this formula. The rated power of the whole machine is a scalar quantity with a value range of 5W to 40W. In this embodiment, it is taken as 18W. The unit is W. The definition is the same as the aforementioned formula for the maximum evaporation power on the liquid supply side; The definition is the same as the aforementioned formula for the upper limit of silk temperature; Thermodynamic temperature is the ambient temperature, a scalar value ranging from 273K to 318K, in K, and is obtained by adding 273.15 to the ambient temperature. The core thermal resistance of the new core is a scalar value ranging from 3K / W to 12K / W. In this embodiment, it is taken as 6K / W. The unit is K / W, which is obtained by steady-state thermal testing of the new atomizing core. The thermal resistance aging coefficient is a scalar value ranging from 0.3 to 1.5. In this embodiment, it is 0.8. It has no unit and is obtained by fitting steady-state thermal tests of atomizing cores with different aging degrees. Its physical meaning is the extent to which the carbonization layer relatively increases the thermal resistance of the core. The definition is the same as the aforementioned core aging formula; The definition is the same as the aforementioned formula for the release degree of latent heat of vaporization clamping. The duty cycle synthesizer converts the maximum allowable power into the lower limit of the duty cycle and takes the larger of this value and the intermediate value given by the proportional backoff unit as the final output: ,in: For the first The final output drive duty cycle of each pumping cycle is a scalar with a value range of 0 to 1 and no unit. It is calculated by this formula, sent to the power stage and fed back to the timing controller. The definition is the same as the aforementioned maximum allowable power formula; The definition is the same as the aforementioned silk temperature conversion formula; The driving voltage output from the power stage to the heating wire 6 is a scalar value ranging from 2.8V to 8.4V, measured in volts (V) from the battery voltage after boost conversion. The definition is the same as the aforementioned proportional regression formula; The definition is the same as the aforementioned formula for the degree of release of the latent heat of vaporization clamp.

[0062] When the protection setting is the power off setting, the duty cycle synthesizer directly outputs zero and causes the power stage to cut off the power supply circuit of the heating wire. At this time, the above synthesis will no longer take effect.

[0063] The prompting unit 7 includes an indicator light and a linear vibration motor, with its input connected to the protection level output of the graded decision unit 4. When the protection level is the warning level, the prompting unit 7 outputs a prompt signal, which includes an indicator light flashing signal and a vibration signal. The indicator light flashing frequency is 1Hz to 3Hz, and the vibration duration is 100ms to 300ms. At the same time, the drive duty cycle remains unchanged, meaning that the warning level only provides a prompt without reducing the output, thus strictly limiting the intervention on the suction experience to the necessary range. When the protection level is the power limit level, the indicator light turns on continuously; when the protection level is the power off level, the indicator light flashes rapidly at a frequency of 5Hz to 8Hz.

[0064] Example 1 uses a brand new atomizing coil with a measured fallback time constant of 1.2s and a coil aging degree of 0. After continuous vaping with the reservoir level dropping to 10%, the latent heat of vaporization clamp release rate jumped from 0.05 to 0.31 on the third puff, the liquid supply margin factor decreased from 1 to 0.42, and the power limit threshold was reduced from 255℃ to 242.2℃ after adjustment. The protection mode entered the power limit mode when the coil temperature reached 243℃, and the drive duty cycle decreased from 0.95 to 0.61. The atomization output was visibly reduced but not interrupted on this puff, and no burnt smell was observed in the sensory evaluation.

[0065] Example 2 uses an aged atomizing core with a cumulative 1200 puffs. Its measured fallback time constant is 3.6s, and the core aging degree is 0.62. Under normal atomization conditions with sufficient liquid level in the reservoir, the measured latent heat clamp release degree is 0.14. Since the dry reference rate has been adjusted by the reference corrector to 1.217 times the factory initial value according to the core aging degree, the calculated liquid supply margin factor is still 0.96. The protection setting remains normal, and no false intervention occurs.

[0066] Comparative Example 1 uses a fixed reference scheme based on existing technology. Under the same conditions as Example 2, because the dry reference rate is still taken from the factory initial value, the latent heat of vaporization clamp release degree is overestimated to more than 0.17, the liquid supply margin factor is reduced to 0.79, and the power limit threshold after setting is reduced to below 259.6°C. The protection mode enters the power limit mode during normal suction, and the atomization output is unnecessarily reduced.

[0067] Comparative Example 2 is a two-stage constant temperature scheme according to WO2022088081A1. Under the same conditions as Example 1, since the second constant temperature heating temperature is a preset constant and is not related to the current core state, its cooling range is more conservative on new cores and more aggressive on aged cores. Moreover, after the set number of times is reached within the set time, the heating stops directly, and the continuous suction is interrupted.

[0068] The above embodiments and comparative examples show that the protection intervention time of the present invention remains consistent throughout the entire life cycle of the atomizing core, neither prematurely intervening due to aging nor lagging behind the formation of char due to relaxed benchmarks.

[0069] The value ranges of several key parameters and their selection criteria in this embodiment are given one by one in the symbol descriptions of each unit. Overall, the set resolution and margin decay index jointly determine the degree of protection intervention, and the two need to be adjusted in conjunction: a set resolution of 0.05 and a margin decay index of 1.0 provide the most aggressive protection, suitable for atomizing fluid formulations that are extremely sensitive to scorching; a set resolution of 0.20 and a margin decay index of 3.0 provide the most lenient protection, suitable for formulations with a high glycerol ratio and a slow fluid supply. The ratio of cumulative gain across aspiration cycles to gain within aspiration cycles is key to the discriminability of the dual-timescale recursive estimator. In this embodiment, it is set to 0.02, and actual measurements show stable separation within the range of 0.01 to 0.05; when this ratio is higher than 0.10, the slow-changing state will be contaminated by rapid-changing perturbations, resulting in a significant jump in the aging estimation.

[0070] Several implementation details of this invention can be replaced with equivalent substitutions. The heating wire 6 can be replaced with a thick film structure printed on a ceramic substrate. In this case, the temperature coefficient of resistance is taken according to the measured value of the thick film slurry, while the structure and parameter range of the other units remain unchanged. The fitting of the first-order thermal decay model can be replaced with a second-order model to further distinguish between the heat capacity of the wire and the heat capacity of the core. However, the second-order model requires a higher number of sampling points in the non-phase change cooling section, and the proportion of degradation to the initial slope method increases under intensive inhalation conditions. The carbonization pre-exponential factor and carbonization activation energy in the carbonization rate constraint can be uniformly written by the whole machine during the accelerated aging calibration before leaving the factory, or they can be grouped and stored according to the atomizing liquid formula and selected and called by the cartridge identification resistor.

[0071] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A multi-stage over-temperature graded protection atomization heating control system, characterized in that, include: The constant current sampling unit is used to inject a constant sampling current into the heating wire and collect the terminal voltage of the heating wire within the driving duty cycle off window, output the wire resistance sequence, split the wire resistance sequence into the resistance trajectory of the heating section and the resistance trajectory of the intermittent section according to the suction sequence, and convert the wire resistance sequence into the output wire temperature sequence. The liquid supply margin calculation unit is used to obtain the single-pulse resistance rise rate from the resistance trajectory of the heating section, compare the single-pulse resistance rise rate with the dry state reference rate, and output the liquid supply margin factor. The aging decoupling unit is used to output a fallback time constant by fitting the resistance trajectory of the intermittent segment, output the core aging degree by the fallback time constant, and correct the dry state reference rate by the core aging degree. The graded decision unit is equipped with graded thresholds, including a warning threshold, a power limit threshold, and a power failure threshold. The graded decision unit is used to adjust the graded thresholds with the liquid supply margin factor and the core aging degree, compare the wire temperature sequence with the adjusted graded thresholds, and output the protection level, including a warning level, a power limit level, and a power failure level. The drive retraction unit is used to retract the drive duty cycle proportionally to the amount by which the wire temperature sequence exceeds the power limit when the protection setting is the power limit setting, and to cut off the power supply circuit of the heating wire when the protection setting is the power off setting.

2. The atomization heating control system with multi-stage over-temperature graded protection according to claim 1, characterized in that, The liquid supply margin calculation unit divides the conduction period of a single driving pulse into a pre-conduction segment and a post-conduction segment, and calculates the resistance rise rate of the pre-conduction segment and the resistance rise rate of the post-conduction segment respectively. The increment of the resistance rise rate of the post-conduction segment relative to the resistance rise rate of the pre-conduction segment represents the degree of release of the latent heat clamping. The liquid supply margin factor decreases as the degree of release of the latent heat clamping increases, and the liquid supply margin factor takes full value when the degree of release of the latent heat clamping is lower than the set release degree.

3. The atomization heating control system with multi-stage over-temperature graded protection according to claim 2, characterized in that, The aging decoupling unit takes the section in the intermittent resistance trajectory where the drive duty cycle is zero and the coil temperature sequence is lower than the boiling point of the e-liquid as the non-phase change cooling section, and outputs the fallback time constant from the non-phase change cooling section. The aging degree of the coil increases as the fallback time constant increases. The aging decoupling unit also outputs an atomizer coil life indicator signal, which is obtained by comparing the coil aging degree with the preset life end aging degree.

4. The atomization heating control system with multi-stage over-temperature graded protection according to claim 3, characterized in that, The aging decoupling unit is equipped with a dual-time-scale recursive estimator. The dual-time-scale recursive estimator sets the core aging degree to a slow-changing state across the pumping cycle and sets the liquid supply margin factor to a fast-changing state within the pumping cycle. The slow-changing state across the pumping cycle is recursively updated with the cumulative gain across the pumping cycle, and the fast-changing state within the pumping cycle is recursively updated with the gain within the pumping cycle. When the number of sampling points in a single phase-change-free cooling section is lower than the set number of points, the dual-time-scale recursive estimator only updates the slow-changing state across the pumping cycle and maintains the fast-changing state within the pumping cycle.

5. The atomization heating control system with multi-stage over-temperature graded protection according to claim 4, characterized in that, The graded decision unit is equipped with a hysteresis band and the aging growth rate is obtained from the core aging degree. The hysteresis band is the temperature drop range required to exit the protection level. The graded decision unit adjusts the bandwidth of the hysteresis band according to the aging growth rate, and the drive back-off unit adjusts the back-off slope according to the aging growth rate. The larger the aging growth rate, the wider the bandwidth of the hysteresis band and the gentler the back-off slope.

6. The atomization heating control system with multi-stage over-temperature graded protection according to claim 5, characterized in that, The graded decision unit is equipped with a gray box constraint model, which is composed of capillary liquid supply rate constraint and carbonization rate constraint. The capillary liquid supply rate constraint is determined by the core aging degree to determine the upper limit of the capillary liquid supply rate, and the carbonization rate constraint is determined by the filament temperature sequence and carbonization activation energy to determine the carbonization rate. The graded decision unit solves the maximum allowable power that ensures the carbonization rate does not exceed the set carbonization rate using the gray box constraint model. When the protection mode is the power limit mode, the drive back-off unit uses the maximum allowable power as the lower limit of the drive duty cycle back-off.

7. The atomization heating control system with multi-stage over-temperature graded protection according to claim 1, characterized in that, When the protection mode exits from the power-off mode, the graded decision unit requires the wire temperature sequence to remain below the power-off threshold and continuously set a recovery time, wherein the set recovery time ranges from 0.5s to 3s.

8. The atomization heating control system with multi-stage over-temperature graded protection according to claim 1, characterized in that, The graded decision unit also calculates the temperature rise slope of the wire temperature sequence, and directly outputs the power-off setting when the temperature rise slope exceeds the set slope threshold. The set slope threshold ranges from 40℃ / s to 120℃ / s.

9. The atomization heating control system with multi-stage over-temperature graded protection according to claim 1, characterized in that, It also includes a prompting unit, which is connected to the graded decision unit. When the protection level is the warning level, the prompting unit outputs a prompting signal and maintains the drive duty cycle unchanged. The prompting signal includes an indicator light flashing signal and a vibration signal.

10. The atomization heating control system with multi-stage over-temperature graded protection according to claim 1, characterized in that, The constant sampling current ranges from 1mA to 20mA, the sampling duration within the drive duty cycle turn-off window ranges from 50μs to 500μs, and the sampling period of the wire resistance sequence output by the constant current sampling unit ranges from 1ms to 10ms.

Images