An aerosol generating device

Abstract

A battery-powered aerosol generating device comprises a monitoring circuit configured to output a parameter of the battery (16), such as its remaining capacity. A memory (22) stores information about the usage of the device in relation to one or more external conditions. A controller (20) calculates a number of heating sessions that the device can perform using the remaining charge stored in the battery (16), the calculation being based on the parameter of the battery (16), the stored usage information and the values of the one or more external conditions at the time of the calculation. It notifies the calculated remaining number via a user interface. The external conditions may include ambient conditions such as temperature or humidity. They may also include the time of day or the day of the week, which can be linked to typical patterns of usage of the device to improve the accuracy of the calculation.

Description

[0001]TITLE AN AEROSOL GENERATING DEVICE DESCRIPTION Technical field The invention relates to battery-powered aerosol generating devices such as electronic cigarettes, which a user can operate for the purpose of inhaling an aerosol. The aerosol typically contains flavouring and optionally nicotine as a component. Background of the invention Heating devices for generating an aerosol or vapour for inhalation are known in the art. (In this specification, the term “aerosol” should hereafter be taken to include “vapour” as an alternative.) Such devices typically include a heater arranged to heat an aerosolizable product (i.e., an aerosol precursor), such as a “stick” that serves as an aerosol source of flavouring in a volatile form. In operation, the device heats the aerosol source with the heater to convert constituents of the product into an aerosol for the user to inhale. In some examples, the product may comprise tobacco and may be similar to a traditional cigarette; in other examples the product may be in liquid form. Such devices are powered using the energy stored in a rechargeable battery. Their users are often concerned that the stored energy might run out at a time or in a place where it is not convenient to recharge the battery so that they will be unable to make further use of the device. Accordingly, it is desirable to be able to provide the user with an accurate estimate of the number of heating sessions that they will be able to carry out based on the remaining capacity of the battery. This allows the user to plan their usage and / or recharging of the device in an optimal way. In some devices, the state of charge of the battery is indicated by a display, e.g. a row of LEDs on the housing of the device, which are all illuminated when the battery is full and are extinguished in turn as the remaining charge in the battery becomes depleted. This gives the user a general indication of when recharging might be required but cannot be converted directly into a remaining number of heating sessions. Published international patent application WO 2021 / 180815 A1 discloses an aerosol generating device that warns the user when the battery capacity falls below a level sufficient for a predetermined number of remaining heating sessions. It does this by using a look-up table, which correlates the measured battery voltage with the remaining number of heating sessions, based on average values over a large number of experiments. A more accurate prediction of the remaining number that is not based on only average values would be desirable. P51767WO / 6554 In this specification, the terms “heating session”, “smoking session” and “vaping session” are used interchangeably. Summary of the invention The invention provides an aerosol generating device comprising a heater for generating aerosol from an aerosol source during a heating session; a battery for supplying power to the heater; a monitoring circuit configured to output a parameter of the battery; a memory for storing information about the usage of the device in relation to one or more external conditions; and a controller configured to calculate a remaining number of heating sessions based on the parameter of the battery, on the stored usage information and on the values of the one or more external conditions at the time of the calculation, and to notify the calculated remaining number to a user via a user interface. By taking account not only of the parameter of the battery but also of the one or more external conditions, the controller can more accurately predict the number of vaping sessions that can be performed using the remaining energy stored in the battery. This is because the external conditions can directly affect the performance of the device and / or they can affect the behaviour of a user of the device. The calculated remaining number may differ when the one or more external condition is different even if both the parameter of the battery and the stored usage information are same. It means if the user changes the external condition (e.g., a location), the calculated remaining number may increase or decrease. The one or more external conditions may include the time of day. Preferably, the memory stores information about usage of the device in different time frames during the day. If the user has a reasonably regular daily routine, then their usage of the device typically varies over the course of a day. For example, they might be more likely to use the device for multiple consecutive heating sessions in the mornings rather than the afternoons, which affects how much each session depletes the charge stored in the battery. The ambient conditions in which the device is used, such as temperature or humidity, can also vary in a predictable manner over the course of a day and can affect how much each session depletes the energy stored in the battery. The time frames during the day may be useful to store the information about usage of the device in an efficient way. This approach is a sort of compression of data, and a reduction of a necessary capacity of the memory will be possible. Daily behaviours of the user may tend to be same or similar in the same time frames, so this compression may not lose an accuracy of the calculation. Additionally or alternatively, the one or more external conditions may include the day of the week. If the user has a reasonably regular weekly routine, then their usage of the device typically varies over the course of a week. For example, they might be more likely to use the device for multiple consecutive heating sessions on working days rather than at weekends, which affects how much each session depletes the energy stored in the battery. P51767WO / 6554 The one or more external conditions may include one or more ambient conditions. One such condition is ambient temperature. Unless the device has been used recently, at the start of a session the temperature of the heater must be raised from ambient temperature to the desired operating temperature; therefore, a lower ambient temperature may result in the session consuming more energy. Another of the external conditions on which the controller may base its calculation is ambient humidity. In humid conditions, the aerosol source can absorb more water from the atmosphere and its heat capacity will be increased so more energy from the battery is required to raise it to the desired operating temperature. Atmospheric humidity may also affect the amount of energy which is required to keep a temperature of the aerosol source around the desired operating temperature. The atmospheric temperature may also have an effect on these values. The parameter of the battery that is output by the monitoring circuit is preferably the remaining capacity of the battery, i.e. the charge stored in the battery that it is still capable of using to supply power to the heater. This is the most important factor required for calculating the remaining number of heating sessions. Alternatively, the parameter could be output in units of the energy stored in the battery on the assumption that it is operating at its nominal voltage. The remaining capacity of the battery may precisely indicate an available energy compared to its voltage. A capacity of the battery may be generally described by a watt-hour [Wh] or an ampere-hour [Ah] value. A calculation of the remaining capacity of the battery may be generally complex, so using a dedicated monitoring circuit, which is preferably commercially available, may be advantageous. The stored information about the usage of the device preferably includes information about the energy consumed per heating session in relation to different values of the one or more external conditions. For example, it may include information about the average amount of energy consumed if the heating session takes place between 9.00am and 10.00am or if the ambient temperature is between 20°and 22°C. By storing the information as values representing the energy consumed, the information can be used in a simple manner in calculating the number of heating sessions available from the remaining capacity of the battery. The calculation of the remaining number of heating sessions NSESSIONmay be based on the formula: where: is the remaining capacity of the battery in units of charge; is the energy consumed per heating session according to the stored information in relation to the value of the one or more external conditions at the time when the calculation is performed; and V is either the nominal voltage of the battery or a measured voltage of the battery at the time of the calculation. P51767WO / 6554 Multiplying the capacity in units of charge by the voltage of the battery gives the energy stored in the battery, which can be divided by the average energy per heating session under current conditions to calculate the remaining number of sessions that can be powered from the battery. Alternatively, the stored information about the usage of the device may include information about the power or current drawn from the battery during each heating session in relation to different values of the one or more external conditions. For example, if the operation of the device is controlled so as to raise the temperature of the heater to a desired value within a predetermined time, then under certain external conditions the battery might need to supply a higher (or lower) level of power in order to achieve this. The information can be stored in a look-up table or alternative format, the stored values being derived from tests carried out in advance on similar devices operated under varying external conditions. The relationship between power and current is well known, depending on the voltage at which the current is supplied, so the information can alternatively be stored as measurements of current rather than power, in order to be used more directly to the battery capacity measured in units of charge. Each heating session may be divided into a plurality of phases and the stored information includes information about the power or current drawn from the battery during each phase. The general aerosol generating devices employ a heating session being able to be divided into the plurality of phases rather than a simple single phase, for supplying a sufficient amount of aerosol across an overall duration of the heating session. Thus, dividing each heating session into the plurality of phases may be advantageous in view of a precise calculation of the remaining number of heating sessions. For example, the session typically comprises an initial phase in which the battery supplies a high level of power to raise the temperature of the heater quickly, followed by one or more phases operated at lower power levels while the user is drawing vapour from the device. The stored information about the usage of the device may include information about the average interval between heating sessions in relation to different values of the one or more external conditions. If under certain conditions, such as at certain times of day, the user of the device typically carries out consecutive heating sessions – i.e. with a small interval between them – then the heater does not cool down fully between the sessions. Less energy is required in the subsequent session for the heater to reach its operating temperature, therefore having access to this information can make the calculation of the remaining number of heating sessions more accurate. Specifically, if the values of the one or more external conditions at the time the calculation is performed are associated with a shorter average interval between heating sessions, the controller may increase the calculated remaining number of heating sessions. This may lead to further improve the accuracy of the calculation of the remaining number of heating sessions and optimize the aerosol generating device to each user. P51767WO / 6554 The stored information about the usage of the device may also include information about whether cleaning of the device is typically performed in relation to different values of the one or more external conditions. Normally, these conditions will be the time of the day or the time of the week, depending on whether the user of the device has a regular daily or weekly routine for cleaning the device. If the heating chamber of the device, which receives the aerosol source, is dirty, more energy is required to raise the temperature of the heater to a desired operating temperature. It can be explained from a reduction of the contact area between the heating chamber and the aerosol source. Therefore, knowing whether the user is likely to clean the device during the time period under consideration can make the calculation of the remaining number of heating sessions more accurate. Specifically, if the values of the one or more external conditions at the time the calculation is performed are associated with cleaning of the device typically being performed, the controller increases the calculated remaining number of heating sessions. The drawings Figure 1 is a schematic diagram of an aerosol generating device in accordance with the present invention. Figure 2 is a block diagram showing the flow of information used in calculating the remaining number of heating sessions in accordance with the present invention. Figure 3 is a plot showing a temperature profile of a heating session. Figure 4 is a plot showing a first power profile of the heating session of Figure 3. Figure 5 is a plot showing a second power profile of the heating session of Figure 3. Figure 6 shows examples of usage profiles over the course of a week, for two different users of an aerosol generating device in accordance with the present invention. Figure 7 is a block diagram showing the overall flow of information in a device in accordance with a second embodiment of the present invention. Figure 8 is a flowchart showing the operation of the LSS algorithm module of Figure 7. Figure 9 is a circuit diagram to explain a battery model used in embodiments of the present invention. Figure 10 is a diagram to explain a model of a heater external to the aerosol source, which may be used in embodiments of the present invention. Figure 11 is a diagram to explain a model of a heater internal to the aerosol source, which may be used in alternative embodiments of the present invention. Figure 12 is a diagram to explain a battery heating model, which may be used in embodiments of the present invention. Detailed description of the drawings Figure 1 schematically shows an aerosol generating device enclosed by a housing 2. An airflow path extends from an inlet 4, via a heating chamber 6 and a conduit 8 to a mouthpiece 10. An aerosol source 12, which may be in the form of a stick, is received in the heating chamber 6 so as to be capable of replacement when it has been exhausted. A heater 14 surrounds the chamber 6 so that, when the heater 14 is connected to a supply of electrical current from a battery 16, it generates heat and increases P51767WO / 6554 the temperature of the aerosol source 12 to evolve an aerosol into the heating chamber 6. A user can put their lips against the mouthpiece 10 and breathe in to draw air along the airflow path, which they can inhale together with the aerosol from the source 12. The aerosol may comprise components such as nicotine and flavour compounds. It should be understood that Figure 1 is purely schematic and the device could be configured in many different ways. For instance, although the heater 14 is illustrated as a resistance coil heating the aerosol source 12 radiatively, it could alternatively heat the source 12 by conduction or induction and at least a portion of the heater 14 could directly contact or penetrate the aerosol source 12. The airflow path does not need to be straight; its inlet could be provided independently of the mouth 4 of the heating chamber 6; and the mouthpiece 10 does not need to project from the housing 2 but could comprise a simple hole in the housing 2. The device operates under the control of a controller 20, which is typically in the form of a microprocessor unit or a microcontroller unit. Data and control lines are shown connecting the controller 20 to a memory 22, a user interface 24 located on the exterior of the housing 2, an ambient temperature sensor 26 also located on the exterior of the housing 2, a puff sensor 28 located in the conduit 8, and a switch 30 that can selectively connect or disconnect the supply of current from the battery 16 to the heater 14. For simplicity, Figure 1 does not show the further circuits by which the battery 16 also supplies power to the controller 20, memory 22 and user interface 24. Again, it should be understood that Figure 1 is purely schematic and the electronic components of the device could be configured in many different ways. For instance, the memory 22 could be included on the same microchip as the controller 20; the puff sensor 28 could be located anywhere along the airflow path, in particular close to the inlet 4; and the ambient temperature sensor 26 could also be located in the airflow path close to the inlet 4. The switch 30 may not be a simple on / off switch but a device such as a transistor that allows the controller 20 to control the level of current that flows from the battery 16 to the heater 14, either in an analogue fashion or by using pulse width modulation (PWM) or pulse frequency modulation (PFM). The user interface 24 is schematically shown as a screen on the exterior of the housing 2, by which information can be displayed to the user. Alternatively, a limited range of information could be conveyed by one or more LEDs. For the user to input commands to the device, the interface 24 could be a touch screen or one or more control buttons (not illustrated) could be provided. Additionally or alternatively to the built in interface 24, the device could comprise an antenna (not illustrated), by which the controller 20 transmits information to and receives commands from a smart phone and / or a smart watch. The user interface could accordingly be provided via an app on the smart phone, which could also be used as a source of data such as ambient temperature and location. P51767WO / 6554 Figure 2 is a block diagram that shows one possible configuration of the flow of information in software executed by the controller 20 to carry out the present invention. The controller 20 is embodied as a microcontroller unit (MCU). From a fuel gauge 40, the controller 20 receives data that represents the remaining capacity of the battery 16, and from a temperature sensor 42, the controller 20 receives data that represents the temperature of the heater 14. The temperature sensor 42 may be a negative temperature coefficient (NTC) thermistor. The fuel gauge 40 may be a dedicated integrated chip (IC) to periodically calculate the parameter of the battery 16 (e.g., the remaining capacity) based on inputs from various sensors, and to subsequently store these into an internal register by updating previous values. By periodically communicating with the fuel gauge 40, the controller 20 may obtain the latest values of the parameters of battery 16. In Figure 2, the parameter of the battery 16 is labelled as “batteryMeasurements”. A communication between the fuel gauge 40 and the controller 20 may be realized via any type of serial communication (e.g., I2C, SPI or UART). An actual calculation of the remaining capacity is rather complicated, so using the fuel gauge 40 may be advantageous in view of development and production of the aerosol generating device. The fuel gauge 40 may be also named as a gas gauge. The temperature sensor 42 is not limited to the NTC thermistor. Either one of a positive temperature coefficient (PTC) thermistor, a thermocouple or a thermopile may be alternatively or additionally employed. In Figure 2, the data representing the temperature of the heater 14 is labelled as “heaterTemperature”. The controller 20 optionally passes the battery data to a battery observer module 44, which outputs battery parameters to a battery model 46. The battery observer module 44 may calculate a future value of the parameter of the battery 16 passed from the controller 20, and / or convert the parameter of the battery 16 into a secondary parameter being more suitable for the battery model 46. Since the fuel gauge 40 is the IC being commercially available, original outputs of the fuel gauge 40 are not always suitable for the battery model 46. In Figure 2, the outputs of the battery observer mo\dule 44 are labelled as “batteryModelParams”. Of course, the outputs of the battery observer module 44 may also contain the parameter of the battery 16 obtained from the fuel gauge 40 as original. Details of the battery observer module 44 are described later. The battery model 46 maintains a model of the condition of the battery 16 over time – both during a heating session and over the charging cycle of the battery – for example to estimate whether its core temperature, surface temperature or closed circuit voltage (CCV) exceeds certain thresholds, which would indicate to the controller 20 that the device should not be used. The simulation carried out by the battery model 46 may make use of ambient temperature data received from the temperature sensor 26. P51767WO / 6554 The battery model 46 may contain a thermal model to estimate a temperature at a specific point of the battery 16 at a specific timing, and an equivalent circuit as an electrical model to estimate the CCV at a specific timing. In Figure 2, one directional communication from the battery observer module 44 to the battery model 46 is depicted, but actual communication may be bi-directional. Details of the battery model 46 are described later. The controller 20 also passes the heater temperature data to a heater observer module 48, which output heater parameters to a heater model 50. In Figure 2, the output from the heater observer module 48 is labelled as “heaterModelParams”. The heater model 50 also receives data from a heating protocol output control logic block 52 (HPOCL), which provides the temperature setpoints that the heater 14 should reach as its temperature varies over the course of a heating session. In Figure 2, the data from the HPOCL block 52 is labelled as “temperatureSetpoints”. As shown in Figure 3, each heating session preferably comprises a series of successive phases: for example, a ramp-up phase during which the heater is brought up to its maximum temperature, followed by a phase in which the current is cut off to allow the temperature to fall to a desired operating temperature. In a third phase, during which the user commences vaping, sufficient current is supplied to maintain the operating temperature for a predetermined time (135s in this example). Further phases may follow, in which the temperature is ramped up again, then maintained at a new, higher level for a further predetermined time. The profile of temperature setpoints can be specified by a user or can be adapted to different circumstances, e.g. to maximize battery life. The heater observer module 48 may calculate a future value of the heater temperature passed from the controller 20, and / or convert the heater temperature into a secondary parameter being more suitable for the heater model 50. Of course, the outputs of the heater observer module 48 may also contain the heater temperature obtained from the temperature sensor 42 as original. Details of the heater observer module 48 are described later. The heater model 50 uses the received data about the actual heater temperature and the desired temperature profile to control the current supply to the heater 14 (e.g. by pulse width modulation) in order to match as closely as possible the desired temperature profile. Strictly, the specified temperature profile should refer to the temperature of the aerosol source 12, not the heater itself, and the heater model 50 can also simulate the flow of heat from the heater 14 to the aerosol source 12. Alternatively, a temperature of the heater 14 may be used instead of the temperature of the aerosol source 12 to simplify and accelerating a calculation. The level of current that is required to achieve and maintain the desired temperatures – and hence the amount of power demanded from the battery 16 – may vary depending on conditions external to the device: in particular, the initial temperature of the heater. The initial temperature will generally be close to the ambient temperature of the environment in which the device is being operated, unless only a short interval has elapsed since the previous heating session, in which case the initial temperature of the heater 14 may be higher and less energy will be required during the P51767WO / 6554 first phase of the next session. The level of current and power may also vary depending on the behaviour of the user: for example, the timing, strength, duration and frequency of their puffs. These conditions external to the device, comprising both ambient conditions and user behaviour, tend to vary in a predictable way at different times of day and, in some cases, on different days of the week. In Figure 2, one-directional communication from the heater observer module 48 to the heater model 50 is depicted, but actual communication may be bi-directional. Details of the heater model 50 are described later. Figures 4 and 5 show examples of how the power supply to the heater 14 might need to be varied with time in order to achieve the temperature profile of Figure 3. (The horizontal timescales of Figures 3, 4 and 5 are the same.) The plot in Figure 4 is an average of a number of sessions recorded during the afternoon, while the plot in Figure 5 is an average of a number of sessions recorded during night-time. With repeated use by the same user, the device can learn their behaviour and can store in its memory 22 details of a typical profile at different times of day: for example, in the early morning, at midday, in the afternoon, in the evening and at night. For each phase of each profile, the device might store the average and maximum values of either current or power. The stored profiles can be continuously updated, for example by the use of moving average values for the most recent n sessions. As described above, the controller 16 may control a power supplied from the battery 16 to the heater 14 by manipulating the switch 30 with the PWM manner. Thus, the stored profiles may be also referred as a PWM load profiles. Such a series of sequences may be realized by a user observer module 56. The heater model 50 can thus adapt its calculation of the estimated power profile of the heater 14 to take account of varying external conditions, which include aspects of the user’s behaviour. The heater model 50 calculates the estimated power profile of the heater 14 under the prevailing conditions at the time the calculation is made and outputs the result to an RSS (Remaining Smoking Sessions) estimation module 54. In other words, the heater model 50 may take account of the stored profiles passed from the user observer module 56 as a basis and subsequently calculates the estimated power profile of the heater 14 while referring to the outputs from the heater observer module 48 as arguments. The outputs from the heater model 50 and the user observer module 56 are respectively labelled as “PWMloadProfile (Estimated)” and “consumerBehaviour” in Figure 2. In addition to the stored profiles, the outputs from the user observer block may contain various parameter relating user behaviors (e.g., an average interval between sessions, or the device cleaning). The RSS estimation module 54 also receives data from the battery model 46. The RSS estimation module 54 uses the received data to perform a calculation of the remaining number of heating sessions NSESSION, which may be based on the formula: P51767WO / 6554 In this formula, R is the remaining capacity of the battery in units of charge, which is part of the data received by the RSS estimation module 54 from the battery model 46. V is the voltage of the battery, for which may be used either a nominal voltage value of the battery 16, a measured voltage of the battery 16 at the time of the calculation, or a calculated voltage received from the battery model 46. Both the remaining capacity of the battery 16 (R) and the measured voltage of the battery 16 (V) may be able to be obtained directly from the fuel gauge 40. E is the energy consumed during each heating session, which is derived from the data received from the heater model 50, for example by integrating with respect to time the power profile that is the output of the heater model 50. In simple terms, the numerator of the formula is the total energy available from the battery 16 and the denominator is the energy required for each heating session. The skilled person will understand that the remaining number of heating sessions NSESSIONwill contain a decimal in almost cases, and it is preferably rounded down to an integer value when shown on the user interface 24. In performing its calculation, the RSS estimation module 54 may adapt its determination of the energy required for each heating session to take account of user behaviour data received from the user observer module 56. Figure 6 shows examples of usage profiles for two different users of the device, identified as User A and User B. Each row represents a day of the week and each column represents an hour of the day from 0 to 23. The main number in each cell represents the number of smoking sessions that the user performed during that hour and the cells are shaded accordingly. In some cells, a second, smaller number in a circle represents how many of those sessions were consecutive. A number “1” in the lower part of a cell indicates that the user cleaned their device during that hour. It will be recalled that, in the event of consecutive sessions, the device consumes less energy from the battery 16 because the heater 14 does not have time to cool down between the sessions. The consecutive session may be defined as a session which is performed without the aerosol generation device having transited to a sleep mode or the user having turned off the device after a previous session. It will also be recalled that, if the device has been cleaned, it will use less energy because the heater 14 will heat the aerosol source 12 more efficiently. The controller 20 may be able to determine when the device has been cleaned by detecting a reduction in its energy consumption after cleaning. Alternatively, cleaning can be detected by a dedicated tool and / or interface. The two users in Figure 6 show different patterns of behaviour across the day and week. For example, User A performed most smoking sessions between 6:00 and 7:00 on Friday and Saturday, including several consecutive sessions, but did not use the device before 15:00 on Sunday or Monday. They cleaned the device at quite irregular times. User B performed smoking sessions at a more uniform rate across each day and across the week, never consecutively, and they typically cleaned the device in the afternoons. If hours of the day are allocated to the several time frames, as a concrete example, 6:00 to 11:00 may consist of an early morning, 11:00 to 14:00 may consist of a midday time frame, 14:00 to 18:00 may consist of an afternoon time frame, 18:00 to 22:00 may consist of an evening time frame, and 22:00 to P51767WO / 6554 6:00 may consist of a night time frame. A number of time frames is not limited to five: more or fewer time frames may be employed. Borders of respective time frames and / or a number of time frames may vary depending on the day of week. The daily behaviors are likely to be different among users, so how to split the time frames can be optimized user by user. For example, a use of the aerosol generating device by user A can be observed from 5:00, while user B does not use the device until 7:00. A simple way to customize the time frames may be by subtracting the sleeping time. The remaining hours may be identified as respective activation hours, and each time frame may be set by evenly distributing the activation hours. For example, the sleeping times of users A and B will be respectively 23:00 to 5:00 and 23:00 to 7:00. A skilled person will be understood that this approach will provide a different pattern of time frames to the users A and B. Another approach will be finding a vacant hour when the user does not use or hardly uses the aerosol generating device. In the usage profile of the user A, 7:00 to 8:00 and / or 12:00 to 14:00 will be identified as the vacant hour. On the other hand, 10:00 to 11:00 and / or 13:00 to 14:00 will be identified as the vacant hour in the usage profile of the user B. It will be expected that user behavior changes before and after such the vacant hour, so different time frames can be allocated before and after the vacant hour. A cooperation with the user’s smart devices (e.g., smartphone or smartwatch) may be advantageous to design the further optimized time frames. Especially, an activity log will be useful. In accordance with the invention, these patterns of behaviour can be distilled by the user observer module 56 into consumer behaviour data, that can then be used by the RSS estimation module 54 to make its calculations of the remaining number of heating sessions more accurate. To the extent that the patterns affect the average power consumption of the device at different times of day, averaged across the week, they may be reflected in different stored power profiles, examples of which are shown in Figures 4 and 5. If the RSS estimation module 54 is capable of taking into account both the day of the week and the time of day among the external conditions that it uses in its calculation, then the result can be made more accurate still. For example, the device belonging to user A might predict a higher number of remaining sessions at the end of the day on Thursday, compared with its prediction at the end of the day on Sunday, in anticipation of many consecutive sessions being performed on Friday morning but few or none on Monday. The device belonging to user B might predict a higher number of remaining sessions in the evening than in the morning, because at those times the heating chamber 6 of the device has often been recently cleaned. Another way in which the patterns of usage might usefully be stored and summarized would be to calculate the average interval between sessions at different times of the day or week. The user observer module 56 can also analyse the use of the device at even shorter timescales by using data received from the puff sensor 28 to record the timing of puffs during a single session, which can in turn affect the rate at which energy needs to be supplied by the battery 16 to the heater. P51767WO / 6554 The user observer module 56 preferably continually updates the data that it stores in the memory 22. For example, after each session, it might combine the energy consumed during the latest session and the energies consumed during a predetermined number of preceding sessions to calculate a moving or weighted average of the typical energy that the device consumes. The calculated number of remaining sessions may be output by the RSS estimation module 54 to the controller 20 for notification to the user via the user interface 24 or via broadcasting the information to a remote device such as a smart phone. The result of the calculation is preferably displayed numerically. The calculation may be performed and the number re-notified at the end of each session; when the device is switched on (because the external conditions might then have changed); or at any time on demand by the user. Hereinafter another embodiment to explain how the usage information affects the calculation of the remaining number of heating sessions. Figure 7 is another block diagram that shows one possible configuration of the flow of information in software executed by the controller 20 to carry out the present invention. Most parts of this block diagram are same as the block diagram shown in Figure 2, thus we will mainly focus on explaining the differences. The outputs of the battery model 46 and the heater model 50 are received by an LSS (Last Smoking Session) assurance module 55. The LSS assurance module 55 also receives data from the user observer module 56, which data may include information about the power profile of the last heating session. Based on the data it receives from the various sources, the LSS assurance module 55 carries out a basic check of whether the remaining energy in the battery 16 is sufficient to carry out a further smoking session and, if so, it performs at least one simulation of the further smoking session to determine whether, according to predetermined criteria, it should be permitted to proceed. On the basis of its determination, the LSS assurance module 55 outputs information 60 about the outcome of the simulation, which is used to control the further behaviour of the device. If the simulation indicates that the further session should be permitted, it may commence straight away. If the determination is that the further session should not be permitted, then preferably it is blocked, with a suitable notification to the user via the interface 24. The notification may contain a further recommendation, for example that the battery 16 should be recharged or the heating chamber 6 should be cleaned. Devices according to some embodiments of the invention might only notify a warning to the user, without blocking the further smoking session, thereby allowing the user to make the decision about whether to proceed. This option P51767WO / 6554 could be enabled selectively, depending on which of the battery parameters has reached its respective threshold during the simulated session. Figure 8 is a flowchart that illustrates the functioning of the LSS assurance module 55. In a first step 102, data received from the battery observer module 44 is used to compute the state of energy (SoE) of the battery 16. The state of energy may be the remaining capacity, which is expressed in units of Watt-hour or Ampere-hour, of the battery 16 or the state of charge (SOC) of the battery 16, which is expressed in units of percentage. This value is supplied to a comparator 104, which also receives from the user observer module 56 a value representing the energy used during the last heating session performed by the device. This representing value may be expressed in the same units as the value supplied in the step 102, and may be recorded by the battery observer module 44 into the memory 22 in advance. If the comparator 104 determines that the state of energy of the battery 16 is not greater than the energy used during the last session, this suggests that a further session will not be possible and the LSS assurance module 55 outputs a value of “False” (step 106). This in turn causes the controller 20 to generate a warning (step 108) that a further session cannot be started and preferably, as discussed above, actually to prevent the user from initiating the session. If the comparator 104 determines that the state of energy of the battery 16 is greater than the energy used during the last session, the process continues with step 110, in which the further smoking session is simulated. The comparator 104 (and optionally the step 102) may be omitted in an alternative embodiment. In the alternative embodiment, the simulation in the step 110 runs regardless of the values of battery parameters. The simulation in the step 110 also uses data about the battery load profile, i.e. the profile of current or power that the battery must supply in order to heat the heater 14 to the desired temperature setpoints over the duration of the session. The simulation 110 further receives the current values of one or more battery parameters such as its open circuit voltage from the battery observer module 44. For each phase during the profile, the simulator 110 calculates the values of the prescribed battery parameters. It also stores the calculated values at predetermined times, which typically correspond to the end of each phase, when the battery parameters are likely to take their maximum or minimum values. At the end of the simulation or during the simulation, the stored values are output to a further comparator 112, which determines whether any of the battery parameters have reached or passed their respective thresholds. This output of the stored values may be triggered by a flag which is labelled as “ExitFlag” in Figure 8. The flag may be a completion of the simulation 110 or reaching a predetermined time in the simulation 110. In the illustrated example, the battery parameter of interest is the closed circuit voltage (CCV) of the battery 16. The comparator 112 compares each of the stored values of CCV with a predetermined minimum threshold value. If the CCV has not fallen below the threshold during the simulation, then the LSS P51767WO / 6554 assurance module 55 outputs a value of “True” (step 114). This in turn causes the controller 20 to operate the device to start the further smoking session (step 116). If the comparator 112 determines that the simulated value of the CCV has fallen below the minimum threshold at one of the predetermined times, then the further session cannot proceed on the basis of the load profile that was used during the simulation. In some embodiments of the invention, that may be the end of the process: the LSS assurance module will generate an output of “False” (step 106) and the further session will not start (step 108). In the illustrated embodiment, however, the device attempts to improve the user experience by trying to find an alternative load profile that will enable the further heating session to be performed without any of the battery parameters breaching their respective thresholds. Accordingly, the process passes to step 118, in which information received from the user observer module 56 is used to update the load profile, for example by reducing one or more of the temperature setpoints or by increasing the time that is taken for the heater 14 to heat up, which will reduce the maximum level of power that the battery 16 needs to supply. The updated heating profile may be of short duration and / or low power density. The simulation process 110 is then repeated using the updated profile and the test by the comparator 112 is carried out again. If one of the battery parameters has again reached or passed its threshold, the process returns again to step 118, where a further iteration of updating the load profile may be carried out. On each pass through this cycle, a comparator 120 determines whether the number of iterations, which is labelled as “iter” in Figure 8, has exceeded a predetermined maximum number which is labelled as “maxIter” in Figure 8. If not, then the further simulation using the updated profile is allowed to continue in step 110. If the maximum number of iterations has been exceeded then the comparator 120 causes the LSS assurance module to generate a “False” output (step 106) and control passes to step 108 to notify the user that no further session can be performed. In the illustrated embodiment of the invention, the principal battery parameter that is calculated during the simulation is the closed circuit voltage (CCV) of the battery 16. When the battery is delivering current to a load such as the heater 14, its open circuit voltage (OCV) is reduced as a result of the internal impedance of the battery. This reduction of voltage is also referred as an IR drop. This may be modelled using an equivalent circuit such as the one shown in Figure 9, in which the impedance comprises a series connection of an ohmic resistor and one or more parallel RC circuits; in this case there are two such RC circuits. This equivalent circuit may be stored in the battery model 46, and parameters of the equivalent circuit may be provided from the battery observer module 44. The voltage across the series resistor Rohmic is proportional to the current supplied to the heater but the voltages V1, V2 across the RC circuits are transient, decaying at a rate that depends on the respective time constants R1C1 and R2C2. On the assumption that the supplied current remains constant during each phase of the heating profile, an analytical approach may be used and the change in CCV over the duration of that phase can be calculated as: P51767WO / 6554 where: ^^̅^^is the current supplied to the heater Vi(0) is the voltage across circuit RiCiat the beginning of the phase t is the time since the beginning of the phase. This can clearly be generalized to any number of RC circuits. In Figure 9, Rohmicrepresents an ohmic resistance coming from a solution resistance, a first parallel RC circuit consisted by R1and C1represents an electric double layer on a surface of electrodes, and a second parallel RC circuit consisted by R2 and C2 represents a diffusion reaction inside electrodes. Hereinafter, a derivation of Equation (1) is described in detail. First, the CCV is a voltage which subtracts losses while discharging from the OCV. The losses of the equivalent circuit in Figure 9 are sum of voltage drops across the ohmic resistance, the first parallel RC circuit and second parallel RC circuit. It may be expressed by Equation (2) below.^^^^^^ = ^^^^^^ − ^^^^^ − ^^^^^ − ^^^^^^ According to Kirchhoff's first law, a current flowing through the equivalent circuit at time t can be described as follows: A time-varying current following through a capacitor can be described as follow: Following Equation (5) can be derived by substituting Equation (4) into Equation (3). P51767WO / 6554 Respective voltage V1, V2across the first and second parallel RC circuits can be replaced into a current as follows: For solving the derivative equation, Equation (6) can be transformed as follows: Equation (6) can be generalized as follows: Now, parameters a, b are introduced for simplifying. Equation (8) can be transformed by substituting Equation (9) as follows: P51767WO / 6554 If the value of the heater current between 0 to t is constant or substantially constant, ^^=^ can be replaced By multiplying a resistance Rionto Equation (11), a voltage Vican be obtained.^^^^^ = ^^ ∙ ^^%^^^P51767WO / 6554 By substituting Equation (12) into Equation (2), Equation (1) can be derived as follows: For running the simulation, it’s preferable that Equation (1) is transformed into a discrete format. In the discrete format, the CCV may be calculated at every ∆^ elapsed, and time t is handled as the k th step.Thus, F ∙ ∆^ is the elapsed time when the k th step is run. Equation (1) with the discrete format may beas follows: It’s known that the OCV depends on the stage of charge (SOC), so the OCV can be also described as follows:^^^GFH → ^^^^L^^GFH^ …(14)A change of the SOC can be described by below recurrence equation.L^^GFH = L^^GF − 1H − 3600 ∙ O^^…(15) The FCC stands for the fully charged capacity of the battery 16 and is expressed by a unit of Ampere- hour [Ah]. A recurrence equation suitable for the simulation can be derived by substituting Equation (15) into Equation (13). P51767WO / 6554 An initial value of the SOC may be obtained from the fuel gauge 40. The FCC may be recorded into either one of the fuel gauge 40, the controller 20 or the memory 22 in advance. The user observer module 56 may store and update the constant current ^^̅^^in the memory 22 by averaging or smoothing a battery current measured by the fuel gauge 40, and may provide it to the LSS assurance module 55. Adopting several values into the constant current ^^̅^^may improve an accuracy of the simulation. Such a time-varying constant current ^^̅^^may be set based the temperature profile shown on Figure 3. The user observer module 56 may divide the temperature profile into several sections, and subsequently store the respective constant current ^^̅^^for each section. Skilled person will be understood that such time-varying constant current ^^̅^^may show a similar or same trend as the power profile shown on Figures 4 or 5. Here, the constant current ^^̅^^provided from the user observer module 56 may be set based on the battery parameters provided from the fuel gauge 40 or the battery observer module 44 in previous sessions. Additionally, the constant current ^^̅^^may be linked to the one or more external conditions as well as the profiles shown in Figures 4 and 5. Thus, the user observer module 56 may provide an appropriate constant current ^^̅^^based on the external condition. It means that an output of the LSS assurance module 55 may be varied based on the user behavior and the external condition. The battery observer module 44 may provide the latest OCV value by inputting the SOC into a relationship, with the SOC and the OCV (SOC-OCV curve) being recorded in advance. The SOC-OCV curve may be provided as a function or a look-up table. For the sake of precise calculation, a plurality of the functions or the look-up tables may be prepared for different temperatures of the battery 16 and / or states of health (SOH) of the battery 16. Otherwise, the function or the look-up table may become a multi-variable type including the temperature and / or the SOH of the battery 16. The temperature and / or the SOH of the battery 16 may be obtained from the fuel gauge 40. The respective values of parameters in the equivalent circuit Ri, Ci, and Rohmicmay be provided from the battery observer module 44 to the battery model 46. If these parameters are handled as constant across whole of life of the battery 16, these parameters may be alternatively set in the battery model 46 in advance. In this alternative embodiment, the providing these parameters from the battery observer module 44 to the battery model 46 is unnecessary. The battery observer module 44 may vary the values of these parameters based on any parameter (e.g., temperature, SOH, or combination thereof) of the P51767WO / 6554 battery 16. If the simulation is successively performed across multiple heating sessions, these parameters may be kept to constant across the multiple heating sessions, otherwise these parameters may be updated once simulation for single heating session is completed. Such update of these parameters may be based on the expected value of temperature or SOH of the battery 16 at end timing of the simulation for previous heating session. The SOH may be updated based on an integration amount of discharged charges so far. The SOH may be defined as a ratio between a remaining useful capacity of the aged and new batteries. Here, the word “remaining useful capacity” may mean that available capacity when battery is fully charged. Details of a simulation algorithm for the battery temperature are described in later. An initial value of a voltage across the parallel RC circuit ^^^0^ may be deemed as zero. Without the assumption of constant current, it may be more practical to calculate the CCV by dividing the heating profile into small time intervals and using discrete variables to model how the current through each of the resistors develops from one interval to the next. Hereinafter, detail of this alternative approach is described in detail. First, Equation (10) is expressed using discrete variables as follows: In Equation (17), an integral term can be separated into a small fragment, which is from ^F − 1^ ∙ ∆^ toF ∙ ∆^, and a remaining. Equation (17) at F − 1 is as below:P51767WO / 6554 Equation (19) can be substituted into Equation (18) as follows: In Equation (20), if ∆^ is sufficiently small so that ^^=^ can be deemed as constant ^R^^^GF − 1H from^F − 1^ ∙ ∆^ to F ∙ ∆^, Equation (20) can be transformed as follows: The constants a, b can be replaced in accordance with definition of Equation (9). Here, Equation (2) can be made discrete and generalized as follows:^^^GFH = ^^^GFH − ^^GFH − ^^GFH − ^^^^^^ ∙ ^GFHP51767WO / 6554 The respective voltage can be described by a product of a resistance and a current, and Equation (22) can be substituted into Equation (23) as follows: By substituting Equation (15) into Equation (24), a recurrence equation suitable for the simulation can be derived. ^R^^^GF − 1H ∙ ∆^^^^GFH = ^^^ @L^^GF − 1H −3600 ∙ O^^ B …(25) Since most parameters may be obtained in the same way as with Equation (16), only differences are explained. A current initially flows into only a capacitor in parallel RC circuit, thus an initial value of a current flowing through a resistor in parallel RC circuit ^^(G0Hcan be deemed as zero. A total current supplied from the battery 16 between a small fragment ^R^^^may be provided from the heater model 50. Hereinafter, the heater model 50 in described in detail. The heater model 50 may be also explained as a heat transfer model among three bodies as shown on Figure 10. In this model, it’s treated that Joule heat generated in the heater 14 is transported to a surface of the aerosol source 12 by mean of a gap between the heater 14 and the aerosol source 12; and heat received at the surface of the aerosol source 12 is P51767WO / 6554 transferred into an inside of the aerosol source 12. Such a heat transfer model among three bodies (i.e., the heater 14, the surface of the aerosol source 12, and the inside of the aerosol source 12) may be described as follows: where: ^Y^^^ , ^\̂, and ^Y`-a^I are respectively heat capacities of the heater 14, the surface of the aerosolsource 12, and the inside of the aerosol source 12 Z^^^, Z\, and Z^_are respectively temperatures of the heater 14, the surface of the aerosol source 12, and the inside of the aerosol source 12 [^^^is the power consumed at the heater 14 ^^]̂ and ^Y`-a^I are respectively thermal resistance of the gap and the aerosol source 12.Equation (26) can be made discrete as follows: Equation (27) can be respectively solved for Z^^^, Z\, and Z^_as follows: P51767WO / 6554The values of thermal parameters ^Y^^^ , ^\̂ , ^Y`-a^I , ^^]̂ and ^Y`-a^I may be provided from the heaterobserver module 48 to the heater model 50, otherwise these are embedded into the heater model 50 in advance if they are treated as constant. Alternatively, they may be handled as variant changed by any user action (e.g., a puffing action or a cleaning of the heating chamber 6). In this alternative embodiment, the heater observer module 48 may update these values in response to a detection of a user action. The cleaning of the heating chamber 6 may reduce the thermal resistance of the gap, and the user observer module 56 may provide to the heater observer module 48 that the cleaning is expected based on the external condition. Such prediction may be based on the user profiles on Figure 6. It means that the output of the LSS assurance module 55 may be varied based on the user behavior and the external condition. The initial values of temperatures Z^^^, Z\, and Z^_may be obtained from sensors including the temperature sensor 42. The initial values of temperatures of the surface and the inside of the aerosol source 12 may be deemed as to be same or be close to an atmosphere temperature. The atmosphere temperature is labelled as “externalTemperature” in Figure 7. If the simulation is successively performed across multiple heating sessions, the initial temperature of the heater 14 at each successive session may be set based on the last temperature of the heater 14 of the previous session. More concretely, the initial temperature of the heater 14 at successive session may be the same as the last temperature of the heater 14 of the previous session, otherwise it may be bit lower than the last temperature of the heater 14 of the previous session considering a cooling effect coming from the interval between sessions. A frequency and / or a duration of the intervals may be set based on the user profiles on Figure 6. It means that the output of the LSS assurance module 55 may be varied based on the user behavior and the external condition. A power [^^^G^_H supplied to the heater 14 may be calculated based on the power profile shown onFigure 4 or 5, otherwise an actual output of PID controller when the previous heater temperatureis inputted may be used. By dividing [^^^G^_H by the voltage of the battery 16 (either theCCV or OCV), ^R^^^G^_H may be derived. Equation (28) can be transformed so that cooling effect coming from the puffing action is taken into account as follows: P51767WO / 6554 where: b^^^, b\, and b^_are respectively coefficients for the heater 14, the surface of the aerosol source 12, and the inside of the aerosol source 12. These coefficients reflect how strongly the puffing action affects each body, and may have a negative value. cdeffG^_H is a time-varying puff profile. The user observer module 56 may record the actual time-varying puff profiles in previous sessions with linking the external condition, and subsequently provide an appropriate time-varying puff profile to the heater observer module 48 based on the external condition. It means that the output of the LSS assurance module 55 may be varied based on the user behavior and the external condition. If the heater 14 alternatively heats the aerosol source 12 from its inside as shown in Figure 11, Equations (26) and (28) are respectively described as follows: where: ^^]̂is the heat capacity of the heating chamber 6 Z^]is the temperature of the heating chamber 6 ^^]̂is the thermal resistance of the gap between the heating chamber 6 and the aerosol source 12. Equation (31) may be transformed so that the puffing action is taken into account, in the same way as Equation (29). P51767WO / 6554 In the above formulae, the open circuit voltage (OCV) of the battery 16 is shown as being dependent on time. In fact, the OCV is known to vary in a predicable way with the state of charge of the battery, which is known in turn by measuring – or, in the case of a simulation, calculating – the current that flows out of the battery. The battery model may further include hysteresis effects to accurately compute the battery OCV and available power at low state-of-charge. The battery parameters that are included in the simulation and are used for comparison to determine whether the further heating session should be permitted may additionally or alternatively include values of the core and / or surface temperature of the battery 16. Heat is generated in the battery because of current flowing through its internal impedance. The voltage across that impedance is the difference between CCV and OCV in the formulae above and multiplying it by the simulated current gives the power developed as heat. Heat is simultaneously lost by conduction to the surroundings, which may be assumed to include a flow of cooling air at approximately constant temperature while the device is in use. Otherwise, the flow of cooling air may be further precisely taken account by using a temperature sensor which measures an atmosphere temperature. If the initial temperature, thermal capacity and thermal conductivity of the battery 16 are known, as well as the temperature and thermal conductivity of its surroundings, then the change in core and surface temperatures with time can be modelled as part of the simulation. In alternative embodiments, it may be sufficient to estimate the core and / or surface temperature of the battery for the simulated current supply by using look-up tables derived from tests that have been carried out previously. This will avoid the need to perform a full simulation of the heat flow in the battery. Hereinafter, another aspect of the battery model 46 is described in detail. The battery model 50 may comprise a thermal model, which may be also referred as “battery thermal model” hereinafter, as shown in Figure 12. The battery thermal model may also be explained as a heat transfer model among three bodies. In this model, it’s treated that Joule heat due to an internal resistance of the battery occurs at only the core of the battery 16, and it’s successively transported to a surface of battery 16 and to the atmosphere. It may be also deemed that the atmosphere temperature may be brought by air flow, and the temperature of the battery 16 may not affect the atmosphere temperature. In this embodiment, a cross- sectional shape of the battery 16 may be deemed as a circle. If the cross-sectional shape of the battery 16 has a different shape (e.g., rectangular), the following explanation will be adapted to a main surface of the battery 16. First, Equation (2) may imply that a difference between the OCV and the CCV represents a voltage drop inside the battery 16 as follows:^^^^^^ − ^^^^^^ = ^^^^^+^^^^^ + ^^^^^^ ∙ ^^^^ …(32)P51767WO / 6554 An energy lost inside the battery 16 can be described as follows:[G^_H = ^^^^G^_H − ^^^G^_H^ ∙ ^^̅^^ …(33)Instead of the constant current ^^̅^^, the simulated current ^G^_Hexplained in Equation (25) may be used. A respective derivative equation for a core and a surface temperature may be described as follows: where: ^^̂ and ^g̀ are respectively the thermal capacities of the core of the battery 16 and the surface ofthe battery 16 Z^, Z̀ , and Zf are respectively the temperatures of the core of the battery 16, the surface of thebattery 16, and the atmosphere ^^̂is the thermal resistance between the core and the surface of the battery 16 ^êis the thermal resistance between the surface of the battery 16 and the atmosphere. Equation (34) may be transformed into discrete format as follows: Equation (35) can be respectively solved for Z^G^_H and Z̀ G^_H as follows: The values of thermal parameters ^^̂, ^g̀ , ^^̂ and ^ê may be provided from the battery observer module44 to the battery model 46, otherwise these are embedded into the battery model 46 in advance if these P51767WO / 6554 are treated as constant. Alternatively, these may be treated as variants changed by any parameter of the battery 16 (e.g., SOC, SOH, or combination thereof). In this alternative embodiment, the battery observer module 44 may update these values based on the parameter of the battery 16. The initial temperature of the core and the surface of the battery 16 may be deemed as the same and be set based on the temperature of the battery 16 provided from the fuel gauge 40. If the simulation is successively performed across multiple heating sessions, these battery temperatures at each successive session may be set based on the last battery temperatures of the previous session. More concretely, the initial battery temperatures at each successive session may be same as the last battery temperatures of the previous session, otherwise they may be a bit lower than the last battery temperature of the previous session considering a cooling effect coming from the interval between sessions. A frequency and / or a duration of the intervals may be set based on the user profiles on Figure 6. It means that the output of the LSS assurance module 55 may be varied based on the user behavior and the external condition. Although not shown in the block diagram or flowchart for the illustrated embodiment of the invention, it will readily be understood that, instead of carrying out a single simulation of the next further heating session, the controller 20 could carry out simulations of multiple, successive heating sessions until a session is reached in which the charge of the battery is depleted or one of the other battery parameters passes its respective threshold. In this way, the simulation can provide a highly accurate estimate of the number of sessions that it will be possible to perform using the remaining capacity of the battery 16. Some assumptions may need to be made about the conditions under which the sessions will be carried out. In particular, if the sessions are performed consecutively, the heater 14 and aerosol source 12 will not have time to return fully to the ambient temperature in the interval between sessions, therefore less energy will be required to attain the desired operating temperature. In some embodiments of the invention, the device may record details of past sessions in order to develop a model of the user’s behaviour regarding parameters such as how frequently they perform consecutive smoking sessions. Although exemplary embodiments have been described in the preceding paragraphs, it should be understood that various modifications may be made to those embodiments without departing from the scope of the appended claims. Thus, the breadth and scope of the claims should not be limited to the above-described exemplary embodiments. In Figures 2 and 7, the battery observer module 44, the battery model 46, the heater observer module 48, the heater model 50, the heating protocol output control logic block 52, the RSS estimation module 54, the LSS assurance module 55, and the user observer module 56 are depicted separately from the controller 20. However, at least one of these may be realized as a function implemented in the controller 20. P51767WO / 6554 Any combination of the above-described features in all possible variations thereof is encompassed by the present disclosure unless otherwise indicated herein or otherwise clearly contradicted by context. Unless the context clearly requires otherwise, throughout the description and the claims, the words “comprise”, “comprising”, and the like, are to be construed in an inclusive as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to”. P51767WO / 6554

Claims

CLAIMS 1. An aerosol generating device comprising: a heater (14) for generating aerosol from an aerosol source (12) during a heating session; a battery (16) for supplying power to the heater (14); a monitoring circuit (46) configured to output a parameter of the battery (16); a memory (22) for storing information about the usage of the device in relation to one or more external conditions; and a controller (20) configured to calculate a remaining number of heating sessions based on the parameter of the battery (16), on the stored usage information and on the values of the one or more external conditions at the time of the calculation, and to notify the calculated remaining number to a user via a user interface (24).

2. An aerosol generating device according to claim 1, wherein the one or more external conditions include the time of day.

3. An aerosol generating device according to claim 2, wherein the memory (22) stores information about usage of the device in different time frames during the day.

4. An aerosol generating device according to any preceding claim, wherein the one or more external conditions include the day of the week.

5. An aerosol generating device according to any preceding claim, wherein the one or more external conditions include one or more ambient conditions.

6. An aerosol generating device according to claim 5, wherein the one or more ambient conditions include ambient temperature.

7. An aerosol generating device according to claim 5 or claim 6, wherein the one or more ambient conditions include ambient humidity.

8. An aerosol generating device according to any preceding claim, wherein the parameter of the battery (16) is the remaining capacity of the battery (16).

9. An aerosol generating device according to claim 8, wherein the stored information about the usage of the device includes information about the energy consumed per heating session in relation to different values of the one or more external conditions. P51767WO / 655410. An aerosol generating device according to claim 9, wherein the calculation of the remaining number of heating sessions NSESSIONis based on the formula:where: R is the remaining capacity of the battery in units of charge; E is the energy consumed per heating session according to the stored information in relation to the value of the one or more external conditions at the time when the calculation is performed; and V is either the nominal voltage of the battery or a measured voltage of the battery (16) at the time of the calculation.

11. An aerosol generating device according to any preceding claim, wherein the stored information about the usage of the device includes information about the power or current drawn from the battery (16) during each heating session in relation to different values of the one or more external conditions.

12. An aerosol generating device according to claim 11, wherein each heating session is divided into a plurality of phases and the stored information includes information about the power or current drawn from the battery (16) during each phase.

13. An aerosol generating device according to any preceding claim, wherein the stored information about the usage of the device includes information about the average interval between heating sessions in relation to different values of the one or more external conditions.

14. An aerosol generating device according to claim 13, wherein, if the values of the one or more external conditions at the time the calculation is performed are associated with a shorter average interval between heating sessions, the controller (20) increases the calculated remaining number of heating sessions.

15. An aerosol generating device according to any preceding claim, wherein: the stored information about the usage of the device includes information about whether cleaning of the device is typically performed in relation to different values of the one or more external conditions; and if the values of the one or more external conditions at the time the calculation is performed are associated with cleaning of the device typically being performed, the controller (20) increases the calculated remaining number of heating sessions. P51767WO / 6554

Images