Particle sensing device performing calibration operation, and operation method thereof

Abstract

A particle sensing device, according to the present technique, comprises: a photodetector for detecting light scattered from particles; an analog-to-digital converter for generating output data corresponding to an analog signal outputted from the photodetector; and a control circuit performing a normal sensing operation for generating a first data set by applying a threshold value set to an output data set which is obtained by accumulating the output data for a predetermined time, and generating a second data set by applying a transform matrix to the first data set, wherein the control circuit further performs a calibration operation for updating the threshold value set and the transform matrix by using a first reference data set and a second reference data set.

Description

Particle detection device performing calibration operation and method of operation thereof

[0001] The present technology relates to a particle detection device and a method of operation thereof, and more specifically, to a particle detection device that performs a calibration operation and a method of operation thereof.

[0002] Particle detection devices are sensor technologies that measure the size and concentration of particles in the air, and are used in various fields such as environmental sensors, smart home systems, and industrial sensor networks.

[0003] Conventional optical-based sensors have limited variability correction and limitations in guaranteeing the reliability of particle size and concentration data.

[0004] Instability in laser diode output and variability in fan speed are major causes of this data distortion.

[0005] In addition, conventional optical-based sensors have a problem in that it is difficult to maintain data consistency during production and distribution due to performance differences between sensor modules.

[0006] The present technology provides a particle detection device capable of performing a calibration operation and a method of operating the same.

[0007] A particle detection device according to one embodiment of the present invention includes: a photodetector that detects light scattered from a particle; an analog-to-digital converter that generates output data corresponding to an analog signal output from the photodetector; and a control circuit that performs a normal detection operation by applying a threshold set to an output data set accumulated for a certain period of time to generate a first data set and applying a transformation matrix to the first data set to generate a second data set, wherein the control circuit further performs a calibration operation that updates the threshold set and the transformation matrix using the first reference data set and the second reference data set.

[0008] A method of operation of a particle detection device according to an embodiment of the present invention is a method of operation of a particle detection device that generates digital output data corresponding to the size of a particle, comprising: a step of performing a normal detection operation by applying a threshold set to an output data set accumulated for a certain period of time to generate a first data set and applying a transformation matrix to the first data set to generate a second data set; and a step of performing a calibration operation by updating the threshold set and the transformation matrix using the first reference data set and the second reference data set.

[0009] In this technology, measurement errors between particle detection devices can be minimized by performing calibration operations using reference data provided by reference particle detection devices, and accordingly, the quality of mass-produced particle detection devices can be maintained consistently.

[0010] The particle detection device based on this technology is capable of self-calibrating operation without being connected to a separate external computing device.

[0011] FIG. 1 is a block diagram showing a particle detection device according to one embodiment of the present invention.

[0012] FIG. 2 is a flowchart showing the operation of a particle detection device according to an embodiment of the present invention.

[0013] FIG. 3 is a graph showing an output data set output from a particle detection device according to one embodiment of the present invention.

[0014] FIG. 4 is a flowchart illustrating a method for generating a PC data set according to an embodiment of the present invention.

[0015] FIG. 5 is an explanatory diagram showing a method for generating a PC data set according to an embodiment of the present invention.

[0016] Embodiments of the present invention are disclosed below with reference to the attached drawings.

[0017] FIG. 1 is a block diagram showing a particle detection device (1000) according to one embodiment of the present invention.

[0018] The particle detection device (1000) includes a case (1) in which air passages (2, 3) are formed and a fan (4) that generates air flow in the air passages (2, 3).

[0019] The airflow passing through the air passages (2, 3) contains particles (5) such as dust.

[0020] Light generated from a light source (11), such as a laser diode, collides with a particle (5) and is scattered, and a light detector (21), such as a photodiode, detects the scattered light. At this time, the light source (11) is controlled by a driving circuit (12).

[0021] In this embodiment, the photodetector (21) generates a current signal corresponding to the detected light, but the present invention is not limited thereto.

[0022] A particle detection device (1000) according to one embodiment of the present invention includes a transimpedance amplifier (TIA, 22) that amplifies a current signal generated from a photodetector (21) and outputs a voltage signal, an amplifier (23) that amplifies the output of the TIA (22), a filter (24) that filters the output of the amplifier (23), an analog-to-digital converter (ADC, 25) that generates a digital signal (D) from the output of the filter (24), and a control circuit (100) that performs signal processing operations and control operations according to the digital signal (D).

[0023] The TIA (22), amplifier (23), filter (24), ADC (25), and control circuit (100) can be implemented as a single integrated circuit and may be referred to as a read circuit or read-out circuit (200, ROIC).

[0024] The technology itself of using a read-out circuit to output an analog signal from a sensor as a digital signal is well known.

[0025] The circuit used to generate the digital signal (D) input to the control circuit (100) in the read-out circuit (200) of the present invention can be implemented by applying conventional technology.

[0026] The TIA (22), amplifier (23), filter (24), and ADC (25) connected sequentially in FIG. 1 are an example using such prior art, and their specific configuration and operation are omitted.

[0027] The control circuit (100) can control the operation of the driving circuit (12) or the fan (4) for the operation of the particle detection device (1000).

[0028] The control circuit (100) generates a PC (Particle Count) data set and a PM (Particulate Matter) data set through the operation of the particle detection device (1000).

[0029] The PC dataset includes information on particle size and the number of particles with a size greater than or equal to that size, and the PM dataset includes information on particle size and the mass concentration of particles with a size greater than or equal to that size.

[0030] In this embodiment, the control circuit (100) generates a PC data set and generates a PM data set by applying a transformation matrix to the PC data set, which is disclosed in detail below.

[0031] In the following, the PC data set may be referred to as the first data set and the PM data set as the second data set.

[0032] The control circuit (100) controls the calibration operation using a reference PC data set and a reference PM data set.

[0033] In the following, the reference PC data set may be referred to as the first reference data set, and the PM data set as the second reference data set.

[0034] A calibration operation can be performed when the particle detection device (1000) is used for the first time or reset during use. Accordingly, the calibration operation can be referred to as an initialization operation.

[0035] The reference PC data set and reference PM data set used in the calibration process can be provided from measurement results from an external reference particle detection device.

[0036] The type of reference particle detection device is not limited to a specific one, and any normally functioning particle detection device capable of generating PC data sets and PM data sets is sufficient.

[0037] Since the operating performance of the particle detection device (1000) is reset based on the reference particle detection device through calibration, it is desirable that the reference particle detection device have performance equivalent to or better than that of the particle detection device (1000) to be calibrated, but the present invention is not limited thereto.

[0038] FIG. 2 is a flowchart showing the calibration operation of a particle detection device (1000) according to one embodiment of the present invention.

[0039] The calibration operation can be performed by the control of the control circuit (100).

[0040] The calibration operation includes a first step (S100) of generating an output data set, a second step (S200) of generating PC data by applying an arbitrary threshold set to the output data set, a third step (S300) of determining an optimal threshold set and optimal PC data by comparing the PC data and reference PC data, and a fourth step (S400) of determining a transformation matrix using the optimal PC data and reference PM data.

[0041] Once the transformation matrix is ​​determined, the calibration operation is terminated and the normal detection operation is performed.

[0042] In normal detection operation, the particle detection device (1000) can generate a PC data set by applying an optimal threshold set to a measurement data set and generate a PM data set by applying the PC data set to a transformation matrix.

[0043] In the first step (S100), the particle detection device (1000) is operated for a certain period of time, and data (D) output from the ADC (25) is accumulated to generate an output data set.

[0044] The graph in Figure 3 illustrates a dataset containing nine data points measured over a certain period of time.

[0045] In the second step (S200), an arbitrary threshold set is applied to the output data set, and PC data is generated from it.

[0046] An arbitrary set of threshold values ​​includes multiple threshold values ​​determined by dust size.

[0047] Table 1 illustrates a predetermined set of threshold values.

[0048] Symbol Dust Size (㎛) Threshold TH10.3 10 TH20.5 20 TH31.0 50 TH42.5 100

[0049] A set of threshold values ​​includes multiple threshold values. For example, the threshold value corresponding to TH1 is 10, which serves as a criterion for identifying particles of 0.3 micrometers in size. The remaining TH2, TH3, and TH4 also serve as criteria for identifying particles of corresponding sizes, as shown in Table 1.

[0050] The graph in Figure 3 shows the four critical points TH1, TH2, TH3, and TH4 shown in Table 1 together with dotted lines.

[0051] The number of particles per category can be determined by applying a threshold set to the output data set.

[0052] Table 2 illustrates determining the category to which output data (D) belongs in correspondence with a threshold set.

[0053] Category Output Data (D) Range X (DON'T CARE) D < TH1 C1 TH1 ≤ D < TH2 C2 TH2 ≤ D < TH3 C3 TH3 ≤ D < TH4 C4 TH4 ≤ D

[0054] X is a category corresponding to a minimum measurement size less than the performance of the particle detection device (1000) and can be ignored as noise.

[0055] In this embodiment, output data containing 9 pieces of data is exemplified, but generally, it is common to use an output data set containing a larger number of pieces of data.

[0056] FIG. 4 is a flowchart specifically illustrating a method for generating a PC data set according to an embodiment of the present invention.

[0057] First, initialize the data pointer for the output data set (S210).

[0058] In this embodiment, the output data set is assumed to be sorted in the order in which the data was output, and in this case, the data pointer can correspond to the sequence number in which the data was output. By initializing the data pointer, it points to the first output data.

[0059] Afterwards, it is determined whether there is output data corresponding to the data pointer (S220).

[0060] If output data exists, determine whether the current state is rising (S230).

[0061] The current state is determined by comparing the previous category corresponding to the previous output data with the current category corresponding to the current output data.

[0062] If the current category corresponds to a larger particle than the previous category, it is determined to be in an "increasing" state; if the current category corresponds to a smaller particle than the previous category, it is determined to be in a "falling" state; and in all other cases, it is determined to be in a "maintaining" state.

[0063] If the current state is rising in step (S230), the waiting count is initialized (S231), the data pointer is moved to the next step (S260), and the process is returned to step (S220) to repeat the aforementioned operation.

[0064] In step (S230), if the current state is not rising, determine whether the current state is falling (S240).

[0065] In step (S240), if the current state is lower, determine whether the previous state was lower (S241).

[0066] The previous state refers to the current state corresponding to the previous output data.

[0067] If the previous state is lowered in step (S241), the wait count is initialized (S231), the data pointer is moved to the next (S260), and the process is moved to step (S220) to repeat the aforementioned operation.

[0068] In step (S241), if the previous state is not a decline, the category count corresponding to the previous category is increased (S242).

[0069] The previous category refers to the category corresponding to the previous output data, and in this embodiment, the category corresponding to the output data can be determined according to the rules shown in Table 2.

[0070] Afterwards, the waiting count is initialized (S231), the data pointer is moved to the next step (S260), and the process moves to step (S220) to repeat the aforementioned operation.

[0071] In step (S240), if the current state is not falling, the current state corresponds to the maintenance state.

[0072] At this time, the waiting count (S250) is increased, and it is determined whether the waiting count is greater than the waiting threshold (WIDTH) (S251).

[0073] If the waiting count is greater than the waiting threshold, the category count corresponding to the previous category is increased (S242).

[0074] Afterwards, the waiting count is initialized (S231), the data pointer is moved to the next step (S260), and the process moves to step (S220) to repeat the aforementioned operation.

[0075] In this embodiment, the category count is intermittently increased when the current state persists in a maintenance state. To this end, a wait count and a wait threshold are used.

[0076] The wait count corresponds to the number of times the current state remains in a hold state, and the wait threshold indicates the number of times to wait until the category count is incremented while maintaining the current state.

[0077] The standby threshold can be changed according to the embodiment, but in this embodiment, 2 is used.

[0078] If the waiting count in step (S251) is not greater than the waiting threshold, the data pointer is moved to the next step (S260), and the process moves to step (S220) to repeat the aforementioned operation.

[0079] If there is no output data corresponding to the data pointer in step (S220), output the PC data generated from the category count (S221) and terminate.

[0080] Figure 5 is an explanatory diagram showing the process of generating PC data according to the sequence of Figure 4 using the output data set of Figure 3 and the threshold set of Table 1.

[0081] In the flowchart of Fig. 4, the data pointer corresponds to the time (t) of Fig. 5.

[0082] At t=1, the output data (D) is 5, so it is classified as category X according to Table 2.

[0083] In the same way, the category corresponding to the entire output data, the corresponding current state, and the previous state can be determined.

[0084] Since the current state is set to rising in the initial state t=1, it is determined to be Yes in step (S230) of Fig. 4, and the waiting count is initialized to 0.

[0085] Since the current state is rising at t=2, it is determined to be Yes in step (S230) of Fig. 4, and the waiting count is initialized to 0.

[0086] At t=3, the current state is maintained and the previous state is rising, so it is determined to be No in step (S240) of Fig. 4, and the waiting count is increased to 1.

[0087] At t=4, since the current state is maintained and the previous state is maintained, it is determined as No in step (S240) of Fig. 4, and the waiting count is increased to 2.

[0088] At t=5, since the current state is maintained and the previous state is maintained, it is determined to be No in step (S240) of Fig. 4, and the waiting count is increased to 3.

[0089] If the wait count is 3, it is greater than the wait threshold of 2, so the previous category count is increased by 1. Since the previous category is C1, the category count corresponding to C1 is increased by 1. After that, the wait count is initialized to 0.

[0090] Since the current state is rising at t=6, it is determined to be Yes in step (S230) of Fig. 4, and the waiting count is initialized to 0.

[0091] Since the current state is rising at t=7, it is determined to be Yes in step (S230) of Fig. 4, and the waiting count is initialized to 0.

[0092] At t=8, since the current state is rising and the previous state is falling, it is determined to be No in step (S241) of Fig. 4, and the previous category count is increased by 1.

[0093] Since the previous category is C3, the category count corresponding to C3 increases to 1. Afterwards, the waiting count is initialized to 0.

[0094] At t=9, since the current state is falling and the previous state is falling, it is determined to be Yes in step (S241) of Fig. 4, and the waiting count is initialized to 0.

[0095] When the operation corresponding to t=9 is finished, since there is no data corresponding to t=10, the PC data generated from the category counter is output.

[0096] PC data is the accumulated number of dust particles larger than a certain size.

[0097] In Fig. 5, the final generated category count set is {1, 0, 1, 0}, the corresponding categories are C1, C2, C3, C4 in order, and the corresponding particle sizes are 0.3, 0.5, 1, 2 μm in order.

[0098] PC data can be generated by accumulating category counts and corresponds to {2, 1, 1, 0}.

[0099] The first element of the PC data corresponds to 2 as the number of particles with a size of 0.3 μm or more, the second element corresponds to 1 as the number of particles with a size of 0.5 μm or more, the third element corresponds to 1 as the number of particles with a size of 1 μm or more, and the fourth element corresponds to 0 as the number of particles with a size of 2 μm or more.

[0100] Figure 5 shows a category counter that is updated over time and corresponding PC data together, but this is only for the purpose of disclosing the invention.

[0101] It is sufficient to generate the PC dataset using the finally generated category count set.

[0102] By repeating the second step (S200), multiple threshold sets and multiple corresponding PC data sets can be generated.

[0103] In step 3 (S300), multiple PC data sets are compared with a reference PC data set to determine the optimal threshold set and the optimal PC data set.

[0104] The Least Mean Square (LMS) algorithm can be applied to select the optimal threshold set.

[0105] That is, among multiple PC data sets, the PC data set with the minimum distance from the reference PC data set can be determined as the optimal PC data set, and the corresponding threshold set can be determined as the optimal threshold set.

[0106] The method for determining the optimal threshold set can be modified in various ways.

[0107] For example, to determine the optimal threshold set, multiple threshold sets may be selected in advance and one of them selected as the optimal threshold set, but the optimal threshold set may also be determined by adjusting the threshold sets while considering the change in distance between the PC data set and the reference PC data set.

[0108] In this way, the third step (S300) determines an optimal threshold set using a reference PC data set and can be referred to as the first calibration operation.

[0109] In step 4 (S400), a transformation matrix is ​​determined using the optimal PC dataset and the reference PM dataset.

[0110] As described above, the present invention uses a transformation matrix to convert a PC data set into a PM data set.

[0111] Accordingly, if we denote the PM dataset as Y, the PC dataset as X, and the transformation matrix as H, the relationship as in Equation 1 holds.

[0112]

[0113] In mathematical formula 1, X and Y are row vectors each with m elements, and H is a matrix with dimensions of mxm (m is a natural number greater than or equal to 2).

[0114] To determine the transformation matrix, the optimal PC dataset determined in the previous step and the externally input reference PM dataset are used.

[0115] Optimal PC data set X O , reference PM dataset Y R In this embodiment, the transformation matrix (H) is derived by Equation 2.

[0116]

[0117] X in mathematical formula 2 O + is X O This represents the pseudo-inverse of . Since deriving the pseudo-inverse is well known in linear algebra, a detailed explanation is omitted.

[0118] In this way, the fourth step (S400) determines a transformation matrix using the reference PM data set and the reference PC data set, and can be referred to as the second calibration operation.

[0119] Once the transformation matrix is ​​determined as in Equation 2, the calibration operation is completed and the normal detection operation can then be performed.

[0120] In the normal detection operation, an output data set is generated, an optimal threshold set is applied to the output data set to generate a PC data set, and a transformation matrix is ​​applied to the PC data set to generate a PM data set.

[0121] As such, the present technology allows for easy calibration using the reference PC data set and reference PM data set provided by the reference particle detection device.

[0122] The scope of the present invention is not limited to the disclosure above. The scope of the present invention should be interpreted based on the scope literally described in the claims and their equivalents.

Claims

1. A photodetector that detects light scattered from particles; An analog-to-digital converter that generates output data corresponding to the analog signal output from the above photodetector; and A control circuit that performs a normal detection operation by applying a threshold set to an output data set accumulated over a certain period of time to generate a first data set, and applying a transformation matrix to the first data set to generate a second data set. Includes, A particle detection device in which the above control circuit further performs a calibration operation to update the threshold set and the transformation matrix using the first reference data set and the second reference data set.

2. A particle detection device according to claim 1, wherein the control circuit performs a first calibration operation in which, during the calibration operation, the control circuit updates the threshold set using an optimal threshold set selected using the first reference data set, and performs a second calibration operation in which the control circuit updates the transformation matrix using an optimal first data set corresponding to the optimal threshold set and the second reference data set.

3. In claim 2, the control circuit during the first calibration operation A particle detection device that generates a plurality of first data sets by applying a plurality of threshold set to the output data set, and determines the optimal threshold set among the plurality of threshold set according to the distance between the plurality of first data sets and the first reference data set.

4. A particle detection device according to claim 2, wherein the control circuit updates the transformation matrix by calculating a pseudo-inverse matrix corresponding to the optimal first data set and the reference second data set during the second calibration operation.

5. In claim 1, the control circuit A particle detection device that determines a plurality of categories to which a plurality of output data belongs by sequentially comparing a plurality of output data included in the output data set with a plurality of threshold points included in the threshold set, determines a category count set corresponding to the plurality of categories according to a change in the category corresponding to the plurality of output data, and determines a first data set corresponding to the threshold set from the category count set.

6. A particle detection device according to claim 1, wherein the first data set includes information on the number of particles greater than or equal to the corresponding particle size for each of a plurality of particle sizes, and the second data set includes information on the mass concentration of particles corresponding to each of a plurality of particle sizes.

7. A method of operation of a particle detection device that generates digital output data corresponding to the size of a particle, wherein A step of performing a normal detection operation to generate a first data set by applying a threshold set to an output data set accumulated over a certain period of time and to generate a second data set by applying a transformation matrix to the first data set; and A step of performing a calibration operation to update the threshold set and the transformation matrix using the first reference data set and the second reference data set. Includes, A method of operation of a particle detection device in which the first data set includes information on the number of particles greater than or equal to the corresponding particle size for each of a plurality of particle sizes, and the second data set includes information on the mass concentration of particles corresponding to each of a plurality of particle sizes.

8. In claim 7, the step of performing the calibration operation A first calibration step of updating the threshold set using an optimal threshold set selected using the first reference data set; and A second calibration step of updating the transformation matrix using the optimal first data set corresponding to the optimal threshold set and the second reference data set. A method of operation of a particle detection device including 9. In claim 7, the first calibration step A step of generating a plurality of first data sets by applying a plurality of threshold set to the above output data set; and A step of determining the optimal threshold set among the plurality of threshold sets according to the distance between the plurality of first data sets and the first reference data set. A method of operation of a particle detection device including 10. A method of operation of a particle detection device according to claim 7, wherein the second calibration step comprises the step of updating the transformation matrix by performing calculations on a pseudo-inverse matrix corresponding to the optimal first data set and the reference second data set.

11. In claim 7, the step of performing the normal detection operation A step of determining a plurality of categories to which a plurality of output data belongs by sequentially comparing a plurality of output data included in the above output data set with a plurality of threshold points included in the above threshold set; A step of determining a category count set corresponding to the plurality of categories according to a change in the category corresponding to the plurality of output data; and Step of determining the first data set corresponding to the threshold set from the category count set A method of operation of a particle detection device including 12. A method of operation of a particle detection device according to claim 7, wherein the first data set includes information on the number of particles greater than or equal to the corresponding particle size for each of a plurality of particle sizes, and the second data set includes information on the mass concentration of particles corresponding to each of a plurality of particle sizes.

Images