SENSOR SYSTEM AND SENSING METHOD

Abstract

A sensor system is for sensing a substance in a fluid (1), comprising a first sensor (10) and a second sensor (12) of the same type. The first sensor (10) is used until a calibration event, during which the second sensor (12) is used. The second sensor (12) is used only during the calibration event. The first sensor (10) is calibrated using the sensor information from the second sensor (12).

Claims

1. A sensor system for determining the status of a filter, comprising: a first sensor for sensing the substance in the fluid; a second sensor for sensing the substance in the fluid, the second sensor being of the same type as the first sensor and for sensing the same substance; and a controller adapted to: control the first sensor to sense the substance in the fluid, until a calibration event; during the calibration event, control the second sensor to sense the substance in the fluid, the second sensor being operated only during the calibration event; calibrate the first sensor using the sensor information from the second sensor, a first filter coupled to the first sensor and a second filter coupled to the second sensor; the sensor system is configured to route fluid filtered by the first and second filters to one of the first and second sensors in turn, and to determine a status of the first filter based on a difference between signals from the first and the second sensor.

2. A sensor system as claimed in claim 1, wherein the sensor system is configured for sensing the presence and/or amount of a target substance in a fluid.

3. A sensor system as claimed in claim 2, wherein the sensor system is configured for sensing a pollutant amount or concentration in air.

4. A sensor system as claimed in claim 1, wherein the first and second sensors each comprise a mechanical sensor.

5. A sensor system as claimed in claim 4, wherein the mechanical sensors each comprise a sensing element, and a transducer adapted to drive the sensing element into resonance and to detect a resonance frequency of the sensing element, wherein the resonance frequency is dependent on a mass of particles deposited on the sensing element.

6. A sensor system as claimed in claim 1, wherein the first and second sensors each comprise a light scattering optical sensor.

7. A sensor system as claimed in claim 1, wherein the first and second sensors each comprise a gas sensor, or such as an electrochemical sensor or a MOX-based gas sensor.

8. A sensor system as claimed in claim 1, further comprising a sample intake device for driving the fluid monitored towards the sensor being used.

9. A sensor system as claimed in claim 1, wherein the first and second sensors are particle sensors, the first filter is for selecting a range of particle sizes from air for supply to the first sensor and the second filter is for selecting a range of particle sizes from air for supply to the second sensor.

10. A sensor system as claimed in claim 1, further comprising a valve arrangement for selectively routing fluid filtered by the first and second filters to a selected one of the first and second sensors.

11. An air treatment device, comprising a sensor system as claimed in claim 1.

12. A sensing method for determining the status of a filter, comprising: sensing a substance in a fluid using a first sensor until a calibration event; during the calibration event, sensing the substance using a second sensor, the second sensor being of the same type as the first sensor and for sensing the same substance, the second sensor being operated only during the calibration event; and calibrating the first sensor using the sensor information from the second sensor; and filtering the fluid before supply to the first sensor using a first filter and filtering the fluid before supply to the second sensor using a second filter, wherein the method further comprises: performing a test of the first filter by routing fluid filtered by the first and second filters to one of the first and second sensors in turn, and determining a first filter status based on a difference between signals from the first and the second sensors.

13. A method as claimed in claim 12, wherein the first and second sensors each comprise: a physical sensor, such as a mechanical, optical, thermal, acoustic, electrostatic, or electromagnetic sensor; or a chemical sensor, such as an electrochemical sensor or a MOX sensor; or a hybrid sensor, such as a flame ionization detector.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Examples of the invention will now be described in detail with reference to the accompanying drawings, in which:

[0063] FIG. 1 shows a first example of sensor system and associated timing diagram;

[0064] FIG. 2 shows a method of operating the sensor system of FIG. 1;

[0065] FIG. 3 shows a second example of sensor system;

[0066] FIG. 4 shows a method of operating the sensor system of FIG. 3; and

[0067] FIG. 5 shows how sensor response characteristics can change over time;

[0068] FIG. 6 shows measurements which can be taken for a baseline calibration;

[0069] FIG. 7 is used to explain a first approach for sensitivity calibration;

[0070] FIG. 8 is used to explain a second approach for sensitivity calibration; and

[0071] FIG. 9 shows one example of possible sensor structure.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The invention provides a sensor system for sensing a substance in a fluid, comprising first and second sensors of the same type. The first sensor is used until a calibration event, during which the second sensor is used. The second sensor is used only during the calibration event such that a duty cycle of operation of the second sensor is lower than the duty cycle of operation of the first sensor. The first sensor is recalibrated using the sensor information from the second sensor.

[0073] In this way, two sensors of the same kind are used in the sensing system. One working sensor operates in conventional manner for continuous sensing. The other, calibration, sensor works intermittently to perform real time internal calibration. Periodic calibration overcomes the reliability problem of the working sensor in long term operation. The calibration sensor (and any associated pre-filter) is only used for a short time, guaranteeing the availability of a long-term service as well. By using sensors of the same kind, responses from interfering substances can be balanced out.

[0074] The sensing thus has at least two modes: a sensing mode and a calibration mode. Signal subtraction between the two sensors can be used to eliminate interference from other substances during calibration. The time of calibration is shorter, and preferably much shorter, than the overall sensing time. The ratio between sensing time and calibration time may for example be tunable.

[0075] A diagnostic mode may also be implemented, as discussed below.

[0076] FIG. 1 shows a first example of a sensor system for sensing a substance in a fluid 1. The fluid may be an air stream which contains particulate matter to be sensed, or the sensing may be for detecting and/or measuring the quantity of a target pollutant. The “substance” to be sensed may thus comprise a known chemical or category of chemical or any particulate matter (which may be of known or unknown chemical composition).

[0077] The system comprises a first sensor 10 for sensing the substance in the fluid 1 and a second sensor 12 for sensing the substance in the fluid 1, the second sensor 12 being of the same type as the first sensor 10 and for sensing the same substance.

[0078] A controller 14 is adapted to:

[0079] sense the substance in the fluid using the first sensor 10, until a calibration event;

[0080] during the calibration event, sense the substance in the fluid using the second sensor 12, the second sensor being operated only during the calibration event such that a duty cycle of operation of the second sensor is lower than the duty cycle of operation of the first sensor; and

[0081] recalibrate the first sensor using the sensor information from the second sensor.

[0082] The first sensor 10 may provide continuous monitoring and/or measurement. Drift and/or sensitivity changes in the output signal from the first sensor 10 can be compensated by the second sensor during a discrete calibration measurement. The first sensor 10 may stop operating during the calibration events, or it may continue to monitor.

[0083] In the simplest embodiment, of FIG. 1, two cheap sensors can be used without pre-filters. The first sensor 10 functions as a working sensor and the second sensor 12 functions as a calibration sensor.

[0084] The air stream 1 can pass through both sensors, and it can be original air from the targeted environment, or else pre-filters can be used upstream. The controller 14 retrieves signals from both sensors and performs the calibration of the first sensor 10.

[0085] The timing diagram in FIG. 1 represents the timing for the operation of the sensors 10 and 12 as plots 100 and 120 respectively. Most of time, the second sensor 12 is separated from the air stream and sensor 10 is exposed to the air stream for sensing. Periodically, the second sensor 12 is turned on and exposed to the air stream 1 for calibration. After a short while, sensor 12 is turned off and separated from air stream again until the next cycle. The first sensor 10 is shown as deactivated during the calibration, but it may continue to operate, since the calibration is a signal processing modification rather than a modification to a physical configuration of the first sensor 12.

[0086] The difference between the signals as received from the two sensors 10, 12 can be used for baseline adjustment of the first sensor. The baseline adjustment comprises adjusting the zero point. However, the sensitivity may also be recalibrated and can be implemented by changing a calibration factor in the software used in operating the sensor.

[0087] FIG. 2 shows the method for using the sensor system of FIG. 1. In step 16 the sensor 10 is used to sense the substance in the fluid 1 (e.g. air stream) until a calibration event. During the calibration event, the sensing takes place with the second sensor 12 in step 18. The second sensor is operated only during the calibration event, i.e. in step 18, such that a duty cycle of operation of the second sensor is lower than a duty cycle of operation of the first sensor. Step 19 is the calibration of the first sensor.

[0088] The two sensors use the same sensing methodology. There are various possible types of sensor, and a non-exhaustive list is:

[0089] a mechanical sensor;

[0090] an optical sensor;

[0091] a chemical sensor, for example an electrochemical sensor or a MOX sensor;

[0092] an electrostatic sensor;

[0093] a sensor based on detecting differences in thermal behavior;

[0094] a flame ionization detector.

[0095] The two sensors are also configured for detection of the same substance, (e.g. the same chemical species or the same particle size range). The first and second sensors can be identical, but they may be scaled equivalents so that their output signals still respond in the same way to the target and to pollutants.

[0096] The calibration events may simply be periodically carried out. A long delay between calibration events will give a longer lifetime of the calibrating sensor but will result in more signal drift between the calibration events. The frequency of the calibration may for example by set by the user.

[0097] The timing of calibration events may instead be controlled based on information which is indicative of when a calibration may be needed. The sensing history of the first sensor may for example be used to determine when a calibration should be carried out.

[0098] FIG. 3 shows a second embodiment in which the first sensor 10 has a pre-filter 20 and the second sensor 12 has a pre-filter 22. The pre-filters are used to remove known expected pollutants, or to limit the particle size range reaching the sensors within a defined.

[0099] The output from the first pre-filter 20 is supplied to a first valve 24 which can supply the pre-filtered air to either one of the sensors 10,12. The output from the second pre-filter 22 is supplied to a second valve 26 which can allow or block the pre-filtered air passing to the second sensor 12.

[0100] The valve arrangement enables online diagnosis of the pre-filters as well as the deterioration of the performance of the working sensor 10. For the diagnosis functions, three series of data are used:

[0101] (i) the response signals of the working sensor 10 to working air,

[0102] (ii) the response signals of the calibration sensor 12 to working air (i.e. from the filter 20),

[0103] (iii) the response signals of the calibration sensor 12 to the calibration air (i.e. from the filter 22)

[0104] The working air from the pre-filter 20 may contain undesired pollutants if the pre-filter 20 fails. The calibration air from the pre-filter 22 is assumed to be free of the pollutant as a result of the low duty cycle with which the pre-filter 22 is used.

[0105] FIG. 4 shows the method.

[0106] In step 30 the working sensor 10 is used to sense air from the first pre-filter 20.

[0107] In step 32, the calibration sensor 12 is used to sense air from the first pre-filter 20. In step 34, the calibration sensor 12 is used to sense air from the second pre-filter 22. Steps 32 and 34 may be considered together to comprise part of a calibration event, and the duration of step 30 is again longer than the duration of steps 32 and 34.

[0108] If the signal in step 32 is very different from that in step 34 for the first time, it means the pre-filter 20 is running out, because the sensor response to the air flow filtered by the two pre-filters is very different.

[0109] The difference between the signals from steps 30 and 32 indicates the deterioration or baseline shift of the sensor 10, because the two sensors have different responses to the same air supply.

[0110] Before the pre-filter 20 is changed, the information gathered about the filter status can be used to modify the interpretation of the data from the first sensor 10. Thus, in the final processing step 36, the system can determine when the filter 20 needs to be changed, it can perform baseline correction based on the different sensor responses to the same stimulus, and it can perform linearity and stability adjustments to the signal processing based on the filter performance. The linearity of the working sensor can for example be recalibrated by correlating the trend lines of the two sensors.

[0111] The signal produced by each sensor comprises a value generated in the absence of a pollutant (the baseline) to which a value is added that depends on the pollutant concentration. The simplest correction is based on assuming that the baseline of the two sensors has not changed. Any change in the signal can then be corrected by changing a calibration factor (with which the sensor signal is multiplied). This is in general not sufficient as a full calibration operation, in which case the more complete calibration operations described below can be used.

[0112] There are many types of sensor which, when operated without fluid (i.e. air) flowing through, produce a signal that does not contain any pollutant information. This then gives a baseline signal. This applies to all the sensor principles mentioned above. In the case of such sensors, baseline differences are easily detected and corrected, for example in the software used to operate the sensor.

[0113] When operating the sensors with polluted air, differences in sensitivity, rather than baseline, can be detected and corrected in a similar manner, using the signals produced by the sensors. For optical sensors based on light scattering, the intensity of the light source can also be adjusted, within limits inherent to the operation principle of the light source. Alternatively, also, within limits inherent to the operation principle, the voltage over the photodetector can be adjusted. Similar measures can also be taken in case of sensors based on optical absorption.

[0114] One baseline calibration method has been described above. Some further possible calibration steps will now be discussed.

[0115] FIG. 5 shows how a linear sensor response (as concentration versus signal recorded) may vary over time for a new sensor (plot 0M), for a one month old sensor (plot 1M) and for a three month old sensor (plot 3M). The gradient changes, which is illustrative of the sensitivity, and the zero point at which there is no signal also changes.

[0116] Both of these properties of the sensor response should be recalibrated over time. As the sensor deteriorates, it may only be possible to obtain good signals at very high concentrations. In such a case, for some optical sensors, it is also possible to control the working conditions of the sensor (for example increase the power of the light source) to compensate for the loss of sensitivity.

[0117] FIG. 6 is used to illustrate another baseline calibration approach. Three signal measurements are made. Signal A is the response of the calibration sensor 12 to an input which has been pre-filtered by filter 22, which functions to remove the target substances of interest. Signal B is the response of the calibration sensor 12 to an input which has not been pre-filtered. Signal C is the response the working sensor 10 to an input which has not been pre-filtered and thus includes the target substances of interest.

[0118] The clean calibration sensor signal A can be used as a zero point baseline signal, and can thus be used to interpret the working sensor signal C. This approach assumes that the two sensors have similar baseline values, since the baseline is obtained using the calibration sensor, and it is then assumed to apply also to the working sensor. Signals B and C are the normal working and calibration sensor signals as used for the calibration process described above.

[0119] For a zero point and sensitivity calibration, a first approach is explained with reference to FIG. 7.

[0120] A number of calibration points are taken at different pollutant levels, for example three or five. Plot 100 is the working sensor signals and plot 120 is the calibration sensor signals. The requirement for different pollutant levels is generally met in daily life. This enables the sensor response curve to be formed as the line of best fit as shown as plot 37. This again shows the concentration versus signal recorded. The correct zero point and slope of the line can be calculated, which are used later to calculate other measurement results. When very poor linearity is found in the calibration, it means the working sensor is at the end of its life.

[0121] This assumes a linear sensor response.

[0122] If the kind of sensor used in principle has a non-linear response to different pollutant concentrations, an alternative approach is explained with reference to FIG. 8.

[0123] The correlation between the two sensors is first found, as shown in plot 38. This plots the working sensor signal W against the calibration sensor signal C. The working sensor function shown as plot 39 can be obtained by interpolation, with the working sensor response being calculated as a converted version of the calibration sensor response, which is assumed to be static. The plot 39 is again the concentration versus signal recorded.

[0124] If poor linearity is found between the two sensors, it means the working sensor is at the end of its life.

[0125] At the end of the life of the first sensor 10, for example detected as explained above, the second sensor 12 can be used in its place, and a new calibration sensor can be installed. In this way, the calibration sensor is always near the beginning of its lifetime.

[0126] Note that the simplest implementation of FIG. 1 does not use valves to allow air paths to be swapped. In that case, the diagnosis is based only on two measurements, one by each sensor of the same incoming air supply.

[0127] As mentioned above, the invention can be applied to different types of sensor.

[0128] One example is a mass sensor, for example for measuring the particle concentration of pollutants. Direct mass measurement using resonant devices is a known technique. It is based on the relationship between the resonance frequency (f.sub.0) and the mass of a resonator, as shown in FIG. 9.

[0129] In FIG. 9 a resonator mass 40 is represented schematically, with a mass m and spring constant k. The graph shows the amplitude of the resonant oscillations (on the y-axis) as a function of frequency (the x-axis). Plot 42 is for the basic resonator mass. If an additional mass 44 is added (Am), the oscillation curve shifts down in frequency to plot 46 with a frequency shift Δf.

[0130] The equations which govern the resonant vibrations are;

[00001]$\begin{array}{cc}{f}_0=\frac{1}{2.Math.\pi }.Math.\sqrt{\frac{k}{m}}& \left(1\right)\\ \Delta .Math..Math.f=-\frac{1}{2}.Math.\frac{\Delta .Math..Math.m}{m}.Math.f_0& \left(2\right)\\ \Delta .Math..Math.m_{\min }\propto \frac{m}{Q}& \left(3\right)\end{array}$

[0131] Equation (1) shows the relationship between the basic resonance frequency and the resonator characteristics. Equation (2) shows the change in frequency caused by a change in mass, and equation (3) shows the minimum mass (Δm.sub.min) that can be detected. The minimum depends on the mechanical quality factor Q of the resonator.

[0132] There are several examples of resonance based mass sensing for aerosol contamination monitoring in literature. For example, use of a micromachined silicon cantilever device with a picogram level of mass resolution for personal exposure monitoring has been proposed. Filters can be used for eliminating large particles and an electrostatic sampler can be provided for depositing nanoparticles on the cantilever.

[0133] As a rule of thumb, mechanical sensors which operate by monitoring changes in resonance frequency operate in a range where the added mass is small compared to the initial resonator mass. However, continuous mass accumulation during the lifetime of the sensor is inevitable. This problem is more pronounced for MEMS scale devices, in which mechanical and/or chemical cleaning of the accumulated mass is not possible at least for consumer applications. Therefore, the lifetime of a MEMS sensor can be roughly estimated by considering the initial mass and the approximate mass deposition per measurement cycle.

[0134] Optical sensing technologies (for PM2.5) have also been proposed for consumer level applications for air purifiers. The primary technology for consumer level applications is based on the optical scattering by suspended particles in air (e.g., nephelometry). The sensor accuracy depends on the quality of the optical pathway, for example depending on the presence and position of pollutant particles. Nephelometers use a light source and an optical detector. The method is essentially based on measuring the scattered light intensity by suspended particles in air (or other carrier gas).

[0135] The optical sensor measurement is based on the scattered light, in addition to other factors. The performance of this kind of sensor thus depends on the intensity of the scattered light. Over time, the brightness of the light source used by the optical sensor (for example an LED) decreases, and this gives rise to sensor drift. There is also an issue with contamination, which is more pronounced for indoor aerosols, especially oily aerosols which may form a layer on various optical components of the system (lenses, LED surface, etc.). This results in a decrease in the light intensity, hence false measurements.

[0136] The calibration enabled by the system addresses issues with long term sensor signal drift or non-linearity. The invention can be applied to the optical or mechanical sensors outlined above, but also to chemical sensors (for example electrochemical sensors), electrostatic sensors, sensors based on detecting differences in thermal behavior or flame ionization detectors.

[0137] To avoid the calibration sensor ageing at the same rate as the working sensor, it should not be exposed to the pollutant when not in use. The supply of the fluid being analyzed can be controlled by electrostatic or electrophoretic precipitation in the case of charged particles, on a grounded or oppositely biased resonator. Thermophoretic precipitation may instead be used which comprises creating a temperature difference between the sensor and a counter surface. A fan or pump for delivering a sampled fluid volume may also be used.

[0138] The sensor system may be for measuring a pollutant amount or concentration in air, or for sensing a target substance in a fluid such as air.

[0139] The duty cycle may be 0.1 or less, i.e. the calibration sensor is used for 10% or less of the time that the working sensor is used. This also means the calibration sensor is always at the beginning 10% of its lifetime, if it is then used to replace an expired working sensor and replaced by a new calibration sensor.

[0140] The calibration may be performed daily, weekly, or only after longer periods such as several weeks or months. The duty cycle may thus be extremely low so the calibration sensor is operated only for a very short period of time.

[0141] Applications of interest include air purifiers, stand-alone particle sensor units, personal exposure monitoring devices, vehicle cabin particle measurement sensors, particle sensors for outdoor use (as a standalone sensor unit or for example, sensors for lamp posts for city management), ventilation units, various parts of a building climate management system and in general various types of mechanical sensor which operate by detecting changes in resonance frequency. There are also medical applications in respiratory support and drug delivery applications.

[0142] The system makes use of a controller. Components that may be employed for the controller include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).

[0143] In various implementations, a processor or controller may be associated with one or more storage media such as volatile and non-volatile computer memory such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when executed on one or more processors and/or controllers, perform at the required functions. Various storage media may be fixed within a processor or controller or may be transportable, such that the one or more programs stored thereon can be loaded into a processor or controller.

[0144] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.