Dust concentration signal processing device and signal processing method therefor

Abstract

A dust concentration signal processing method is used to analyze a detection signal from a dust concentration meter in order to calculate a dust concentration. The method comprises the following steps: sampling the detection signal (Sd) to calculate several eigenvalues ​​(Es), determining a signal type of the detection signal (Sd) based on the eigenvalues ​​(Es), selecting a specific algorithm based on the eigenvalues ​​(Es), and calculating the eigenvalues ​​(Es) using the specific algorithm to obtain an estimated dust concentration value (Cc), performing a calibration compensation on the estimated dust concentration value (Cc) to obtain an accurate dust concentration value (Ce), and performing a numerical conversion on the accurate dust concentration value (Ce) to provide the dust concentration (Cd) in a specific data format.

Description

BACKGROUND Technical area

[0001] The present disclosure relates to a dust concentration signal processing device and to a method for processing the signal therefor, and in particular to a dust concentration signal processing device with reduction of the influence of noise and a method for processing the signal therefor. Description of the state of the art

[0002] The statements in this section merely provide background information regarding the present disclosure and do not necessarily represent the state of the art.

[0003] The so-called total suspended particulate matter (SPF) concentration refers to the dust content per unit volume of air. In factory buildings or other locations with specific requirements for SPF emission specifications, environmental regulations and other requirements apply, particularly when particles are emitted into the external environment. To measure whether the data conforms to the standard, specific calibration procedures and associated measuring equipment must be used to determine the measurement uncertainty and accuracy of the equipment installed in the exhaust duct.

[0004] The conventional optical measurement system, which uses the principle of light absorption or scattering to capture the signal from the light-emitting source at one end and the receiving source at the other for calibration, is installed on the exhaust duct. However, due to the ingress of foreign matter (e.g., leaves, paper scraps, steam, fly ash, etc.) into the exhaust duct, the noise signal along the optical path is higher than the actual dust concentration signal, and the frequency of recalibration by the maintenance equipment must be several hours to obtain a reliable reference measurement.

[0005] Furthermore, differences in the geometry of the exhaust duct (circular cross-section, rectangular cross-section), as well as various upstream and downstream braking elements (valve, throttle valve, orifice plate, etc.) and differences in flow field distribution and pressure, often cause uneven flow in the exhaust duct, which is prone to vortex shedding. Therefore, the signal received by the interpreting device is mixed with too much real-time noise, making it difficult to accurately derive the precise dust concentration.

[0006] Therefore, the present disclosure proposes a dust concentration signal processing device with noise reduction and a signal processing method for it to improve the influence of the noise generated by the eddy currents in the exhaust duct and to provide a more accurately calibrated dust concentration. SUMMARY

[0007] To solve the problems mentioned above, the present disclosure provides a method for processing a dust concentration signal. The method is used to analyze a detection signal from a dust concentration meter in order to calculate a dust concentration. The method comprises the following steps: sampling the detection signal to calculate several eigenvalues ​​and determining a signal type of the detection signal based on the eigenvalues; selecting a specific algorithm based on the eigenvalues ​​and calculating the eigenvalues ​​using the specific algorithm to obtain an estimated dust concentration value; performing a calibration compensation on the estimated dust concentration value to obtain an accurate dust concentration value; and performing a numerical conversion on the accurate dust concentration value to provide the dust concentration in a specific data format.

[0008] To solve the aforementioned problems, the present disclosure provides a dust concentration signal processing device. The device is used to analyze a detection signal from a dust concentration meter in order to calculate the dust concentration. The device comprises a feature processing module, a calculation module, a calibration compensation module, and a numerical implementation module. The feature processing module samples the detection signal to calculate several eigenvalues ​​and determine the signal type of the detection signal based on these eigenvalues. The calculation module selects a specific algorithm based on the eigenvalues ​​and uses this algorithm to calculate the eigenvalues ​​and obtain an estimated value of the dust concentration.The calibration compensation module performs a calibration compensation on the estimated dust concentration value to obtain an accurate dust concentration reading. The numerical conversion module performs a numerical conversion on the accurate dust concentration value to provide the dust concentration in a specific data format.

[0009] The main purpose and effect of the present disclosure is to provide a signal processing method for the concentration of total suspended particles (TSP) in the exhaust duct, which includes online sampling features and processing methods for asynchronous signals. Therefore, the changing response of the fluid vortex flow, the sampling frequency of the software signal, and the signal are subjected to wavelet filtering or sliding averaging to improve the noise effect caused by the exhaust vortex and to provide a more accurate dust concentration.

[0010] It is understood that both the preceding general description and the following detailed description are exemplary and intended to provide a further explanation of the present claimed disclosure. Further advantages and features of the present disclosure will become apparent from the following description, the drawings, and the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The present disclosure can be more fully understood by reading the following detailed description of the embodiment with reference to the accompanying drawings: Fig. Figure 1 is a schematic representation of dust monitoring equipment according to the present disclosure. Fig. Figure 2 is a schematic circuit block diagram of a detection circuit according to the present disclosure. Fig.Figure 3 is a schematic circuit block diagram of a dust concentration signal processing device according to the present disclosure. Fig. Figure 4A is a schematic representation of the continuous emission of dust according to the present disclosure. Fig. Figure 4B is a schematic representation of the temporary emission of dust according to the present disclosure. Fig. Figure 5 is a schematic waveform of filtering by a filter unit according to the present disclosure. Fig. Figure 6 is a schematic representation of a calibration test curve according to the present disclosure. Fig. Figure 7 is a block diagram of a unit for numerical implementation according to the present disclosure. Fig. Figure 8 is a schematic representation of a measurement range curve according to the present disclosure. Fig.Figure 9 is a flowchart of a method for processing a dust concentration signal according to the present disclosure. DETAILED DESCRIPTION

[0012] Reference will now be made to the figures in order to describe the present disclosure in detail. It is understood that the figures and exemplary embodiments of the present disclosure are not limited to its details.

[0013] With reference to Fig.Figure 1 shows a schematic representation of a dust monitoring device according to the present disclosure. The dust monitoring device 100 comprises a dust concentration meter 1 and a dust concentration signal processing device 2, and the dust concentration meter 1 is coupled to the dust concentration signal processing device 2. The exhaust duct 200 has a specific wind speed (e.g., without limitation, 3 meters, 40 meters, etc.) and blows from point A to point B to blow dust D into the air from point A to point B. The dust concentration meter 1 comprises a probe 12 and a detection circuit 14, and the probe 12 is inserted into the exhaust duct 200. The depth to which the probe 12 is inserted into the exhaust duct 200 is preferably 1 / 3 to 2 / 3 of the diameter of the section of the exhaust duct 200.When dust D (charged particles) comes into contact with, collides with, or rubs against probe 12, electrical charges are generated on probe 12. Therefore, a dust concentration Cd is detected by filtering, amplifying, analyzing, and processing the inducible charges by the detection circuit 14 and the dust concentration signal processing device 2. If the sampling rate is too high, all features generated by the turbulent flow in the exhaust duct 200 will be sampled, and the dust concentration signal processing device 2 will not be able to determine which signals are genuine and which are caused by the turbulent flow. Therefore, by using a suitable sampling rate, features of turbulent flow in the exhaust duct 200 can be filtered out, and sufficient sampled signals will be available for signal analysis.

[0014] When wind blows towards probe 12, the physical probe 12 also disturbs the fluid, likely causing an uneven distribution of suspended particles or eddy currents. Therefore, it is easy to distort the detected signal or generate unnecessary noise (for example, dust D repeatedly touches, hits, or rubs against probe 12, or the suspended particles may bypass probe 12). Accordingly, the main purpose and effect of the present disclosure is to provide a signal processing method for the concentration of all suspended particles (TSP) in the outlet channel, which includes online sampling features and asynchronous signal processing.Therefore, the changing response of the fluid vortex flow, the sampling frequency of the software signal, and the signal are subjected to wavelet filtering or sliding averaging (it can be adjusted for averaging from 1 second to 24 hours and then output) to improve the noise effect caused by the extraction vortex and to provide a more accurate dust concentration Cd.

[0015] With reference to Fig. 2 and additional reference to Fig.Figure 1 shows a schematic circuit block diagram of a detection circuit according to the present disclosure. The detection circuit 14 consists mainly of hardware circuits and includes an amplification circuit 142, a filter circuit 144, and an analog-to-digital converter 146. The amplification circuit 142 is coupled to the probe 12 and amplifies the electrical charges generated on the probe 12 (i.e., an alternating current response mainly due to the change in amplitude in response to the number of suspended particles) to an amplified signal Sa. The amplification circuit 142 includes a charge amplifier 142-1 and a programmable amplifier 142-2, and the charge amplifier 142-1 is coupled to the programmable amplifier 142-2. The charge amplifier 142-1 mainly performs signal amplification and elementary filtering of the electrical charges generated on the probe 12.The programmable amplifier 142-2 has a good signal-to-noise ratio (S / N) to specifically amplify the required signal and reduce the influence of noise. The filter circuit 144 is coupled to the programmable amplifier 142-2 and filters the amplified signal Sa to a filtered signal Sf. The filter circuit 144 can utilize the filter's attenuation rate above the third order, and the attenuation per octave is increased from the original attenuation of 6 dB to more than 18 dB to improve the signal-to-noise ratio. The filter circuit 144 also preferably increases the filtering capability from 50-60 Hz to reduce the influence of radiation interference caused by the general power supply. The analog-to-digital converter 146 is coupled to the filter circuit 144 and converts the filtered signal Sf into the detection signal Sd.In one embodiment, the internal structure of the aforementioned detection circuit 14 is only one of several detection circuits 14 and is not limited to them. Therefore, any detection circuit 14 that can be used to detect suspended particles and to generate the corresponding detection signal Sd should be included in this embodiment.

[0016] With reference to Fig. 3 and additional reference to Fig. 1 and Fig.Figure 2 shows a schematic circuit block diagram of a dust concentration signal processing device according to the present disclosure. The dust concentration signal processing device 2 can be a controller with data processing functions such as a microcontroller (MCU), but it is not excluded that part of it can consist of a hardware circuit to form a physical control circuit with the same function (such as, but not limited to, comparators, circuits consisting of amplifiers and logic gates). The dust concentration signal processing device 2 comprises a feature processing module 22, a computation module 24, a calibration compensation module 26, and a numerical implementation module 28. In particular, the modules mentioned above can consist of software programs or be partly software and partly hardware.The feature processing module 22 is coupled to the analog / digital converter 146 and receives the acquisition signal Sd, samples the acquisition signal Sd to provide several eigenvalues ​​Es, and determines a signal type of the acquisition signal Sd according to the eigenvalues ​​Es.

[0017] In particular, the feature processing module 22 samples the detection signal Sd according to a preset sampling rate, and a suitable sampling rate is preferably in the range of 10 Hz to 10 kHz, with 100 Hz being particularly optimal. Especially if the sampling rate is too low, the detected signals not only miss the situation of the instantaneous increase in dust in the actual outlet channel 200 due to the long time interval, but also reduce the immediacy and increase the error.

[0018] The feature processing module 22 comprises a sampling unit 222, a feature calculation unit 224, a type determination unit 226, and optionally a filter unit 228. The sampling unit 222 is coupled to the analog-to-digital converter 146 and samples the acquisition signal Sd to provide multiple samples Vs. The feature calculation unit 224 is coupled to the sampling unit 222 and calculates the eigenvalues ​​Es according to the samples Vs. In particular, the dust concentration signal processing device 2 sets a specific time unit (for example, without restriction, 1 second) in advance, and the feature calculation unit 224 calculates the samples Vs in each time unit (such as, for example, without restriction, 1 minute, 1 hour, etc.) to acquire an eigenvalue Es in order to form the eigenvalues ​​Es in a continuous time unit.In particular, the feature calculation unit 224 provides at least one mean calculation method, a square mean calculation method, a standard deviation calculation method and a peak-to-peak value calculation method to calculate the sampled values ​​Vs as eigenvalues ​​Es.

[0019] The method for calculating the mean is expressed as formula 1: Ave=1100×∑i=1100xi

[0020] The quadratic mean calculation method is expressed as formula 2: Rms=1100×∑i=1100xi2

[0021] The method for calculating the standard deviation is expressed as Formula 3 and Formula 4: σ=1100×∑i=1100(xi−x¯)2 x¯=1100×∑i=1100xi

[0022] The peak-to-peak value calculation method is expressed as formula 5: Vpp=[Array A(Maximum value)]−[Array A(Minimum value)]

[0023] The feature calculation unit 224 selects at least one of the mean, root mean square, standard deviation, and peak-to-peak calculation methods according to the aspect of the sample value Vs in order to calculate the sample values ​​Vs as eigenvalues ​​Es. For example, without restriction, the maximum and minimum values ​​are more obvious, and the peak-to-peak and / or standard deviation calculation methods can be selected to calculate the sample values ​​(Vs) as eigenvalues ​​(Es).

[0024] The type determination unit 226 is coupled with the feature calculation unit 224 and determines, according to a feature of the eigenvalues ​​Es, whether the sampled values ​​Vs belong to a continuous dust emission or a temporary dust emission. In particular, they show Fig. 4A and Fig.Figure 4B shows schematic diagrams of continuous and intermittent dust emission according to the present disclosure. Continuous dust emission typically occurs in locations such as thermal power plants, waste incineration plants, paper mills, and other places where dust is continuously burned and released, but is not limited to these locations. The dust emission is usually continuous, and the dust concentration is only temporarily reduced when the equipment is not operating. In contrast, intermittent dust emission usually occurs in factories and other locations, and the dust emission is usually discontinuous. Only when, for example, bulk materials (plasticizing agents, grain, concrete, iron sand) are unloaded from vehicles or during construction work does the dust concentration increase significantly.

[0025] On the other hand, the filter unit 228 is coupled between the sampling unit 222 and the feature calculation unit 224 and selectively performs a filtering process on the sampled values ​​Vs according to a noise level of the sampled values ​​Vs. With reference to Fig. Figure 5 shows a schematic waveform of filtering by a filter unit according to the present disclosure. When the sampled values ​​V SThe resulting waveform, as shown in curve C1, means that even after filtering by the physical circuitry of the detection circuit 14, high-frequency harmonics are still present. Therefore, the filter unit 228 primarily uses a wavelet filter to perform first- to fifth-order filtering on curve C1, generating curves C2 to C6 respectively. The wavelet filter primarily uses the standard deviation method to filter out the high-frequency harmonics on curve C1, and the third-order wavelet filter (i.e., curve C4) is considered the upper limit. If too many orders of filtering are performed, the curve becomes distorted (as shown by curves C5 to C6).However, if fewer high-frequency harmonics are present after filtering by the physical circuitry of the detection circuit 14, or even if no high-frequency harmonics are present (i.e., the function of the detection circuit 14 is better), it may not be necessary to use the filter unit 228 to filter the waveform, such as the curve formed by the sampled values ​​Vs.

[0026] The computation module 24 is coupled to the type determination unit 226 and selects a specific algorithm according to the eigenvalues ​​Es belonging to the continuous dust emission or the intermittent dust emission (i.e., signal types) such that the eigenvalues ​​Es are calculated by the specific algorithm to obtain an estimated value of the dust concentration Cc. Specifically, the computation module 24 comprises a first computation unit 242 and a second computation unit 244. The first computation unit 242 and the second computation unit 244 are coupled to the type determination unit 226. If the eigenvalues ​​Es belong to the continuous dust emission, the first computation unit 242 performs a weighted moving average calculation to determine the eigenvalues ​​Es as the estimated value of the dust concentration Cc.In particular, the type determination unit 226, for example, without restriction, provides a corresponding code to the computation module 24, depending on whether the eigenvalues ​​Es belong to the continuous dust emission or the intermittent dust emission (i.e., signal types), so that the computation module 24 selects the first computation unit 242 or the second computation unit 244 to perform subsequent operations.

[0027] The weighted moving average calculation is expressed as formula 6 and formula 7: Ft=w1At−1+w2At−2+w3At−3+…+wnAt−n ∑i=1nwi=1

[0028] Here, Ft represents the current mean value, and Wi represents the weight of each data point. The sum of Wi is equal to 1, and the closer the value is to the current time, the greater the weight. At-n is set as the number of dust concentration data points in the time interval of the moving average (for example, without restriction, 100 data points per second, i.e., 100 Hz). At-1 represents the current data, At-2 represents the data from the last second, At-3 represents the data from the last two seconds, and so on.

[0029] For example, if 5 seconds are used to calculate the weighted moving average, the data is shown in the following table: Time Dust concentration Weighting Previous 4 seconds 8 0,05 Previous 3 seconds 11 0,05 Previous 2 seconds 15 0,15 Previous 1 second 12 0,25 Current 10 0,50

[0030] The calculated result is (8 × 0.05) + (11 × 0.05) + (15 × 0.15) + (12 × 0.25) + (10 × 0.5) = 11.2, i.e. 11.2 is the estimated value of the dust concentration Cc of the dust concentration recorded after the calculation.

[0031] If the eigenvalues ​​Es belong to the temporary dust emission, the second calculation unit 244 provides an integral calculation to calculate the eigenvalues ​​Es as the estimated value of the dust concentration Cc.

[0032] The integral calculation is expressed as formula 8: is=t1A1+t2A2+t3A3+⋯+tnAn

[0033] This presents i s represents the integral calculation range for the specified time interval, A n represents the current dust concentration at each unit of time and t n represents the unit of time. Preferably, if the concentration value (i.e., the eigenvalue Es) is more pronounced, the time before and after can be taken as a time interval.

[0034] The calibration compensation module 26 is coupled to the first computation unit 242 and the second computation unit 244 and performs a calibration compensation on the estimated value of the dust concentration Cc to obtain an accurate value of the dust concentration Ce. Specifically, the calibration compensation module 26 comprises a calibration unit 262 and a compensation unit 264. The calibration unit 262 is coupled to the first computation unit 242 and the second computation unit 244, and the compensation unit 264 is coupled to the calibration unit 262. The calibration unit 262 acquires a dust concentration calibration value Ccv by calculating the estimated value of the dust concentration Cc using a calibration constant value. With reference to Fig.Figure 6 shows a schematic representation of a calibration test curve according to the present disclosure. The calibration unit 262 calculates the dust concentration calibration value Ccv by performing multi-point calibration and interpolation. The calibration unit 262 performs calibration tests (by formula 9) in a test environment (such as, without limitation, a standard laboratory) to determine calibration constant values ​​before actual operation. ƒ(x)=((y1−y2)(x1−x2)×x+y1−(x1×(y1−y2)(x1−x2)))

[0035] After calibration, the calibration points (x1, y1) and (x2, y2) can be recorded, and the interval range is a linear characteristic curve. The calibration points are used in Formula 9 to obtain the calibration constant value x, which represents the estimated value of the dust concentration Cc. This is the actual value (unit is V) calculated by the calibration unit 262 during actual operation and can be converted into the calibrated value of the dust concentration (i.e., the dust concentration calibration value Ccv, the unit is mg / m³). 3 ) can be implemented by substituting it into the formula above.

[0036] The compensation unit 264 determines the precise dust concentration value Ce by calculating the dust concentration calibration value Ccv using an environmental constant value. Specifically, the compensation unit 264 sets the environmental constant value according to the test environment and an actual operating environment. For example, as long as the actual operating environment is the same as the test environment (i.e., the standard laboratory), this means that Kt (i.e., the environmental constant value) is 1. The composition of the environmental constant value includes the difference in the diameter of the dust powder, the difference in the shape of the outlet duct 200, the wind speed in the pipe, the temperature and humidity, the depth of the probe 12, the industrial safety factors at the actual site, etc.

[0037] Due to the influence of environmental differences, the composition of the environmental constant value can, without restriction, be a combination of any of the following relationships (K1, K2, K3 to Kn represent the factors above) in Formula 10 to Formula 14: Kt=K1+K2+K3⋯+Kn Kt=|K1−K2−K3⋯−Kn| Kt=K1×K2×K3×Kn Kt=1K1+K2+K3⋯+Kn Kt=K12+K22+K32+⋯+Kn2

[0038] By calculating the environmental constant value obtained using the formulas above and the dust concentration calibration value Ccv, the exact value of the dust concentration Ce can be obtained.

[0039] With reference to Fig.Figure 7 shows a block diagram of a numerical conversion unit according to the present disclosure. The numerical conversion module 28 is coupled to the compensation unit 264 and performs a numerical conversion on the exact value of the dust concentration Ce to provide the dust concentration Cd in a specific data format. In particular, the numerical conversion module 28 comprises an output processing unit 282 and a format conversion unit 284. The output processing unit 282 is coupled to the compensation unit 264, and the format conversion unit 284 is coupled to the output processing unit 282. The output processing unit 282 includes a formula calculation for displaying the dust concentration percentage (%) and a formula calculation for the output current signal.The format conversion unit 284 provides the exact value of the dust concentration Ce in a display output format Fd, the exact value of the dust concentration Ce in an analog output format Fa, the exact value of the dust concentration Ce in the relay output format Fr, or the exact value of the dust concentration Ce in an RS485 output format Frs through format conversion. The numerical conversion module 28 converts the exact value of the dust concentration Ce into the dust concentration Cd in the specific format through the two calculations mentioned above and by data conversion of the specific data format.Since the calculation module 24 can in particular use the weighted moving average calculation or the integral calculation to calculate the dust concentration in a certain time interval (such as, for example, without restriction, the time interval between the current time and the preceding 12 hours), the dust concentration signal processing device 2 provides the asynchronous output dust concentration Cd.

[0040] With reference to Fig.Figure 8 shows a schematic representation of a measurement range curve according to the present disclosure. The formula for calculating the dust concentration percentage (%) is calculated according to the set measurement range (e.g., 0.1 to A mg / m³), and the dust concentration percentage (%) = (instantaneous value / A) * 100%. The instantaneous value is the exact value of the dust concentration Ce. A represents the measurement range defined in the dust concentration signal processing device 2. For example, points C1 and C2 in Fig. 8 the upper and lower limits of the measuring range. For example, C1 to C2 is 0.1 mg / m³ 3 up to 10 mg / m² 3 , i.e., C1 is 0.1 and C2 is 10, or C1 to C2 is 0.1 mg / m³ 3 up to 500 mg / m³ 3, i.e., C1 is 0.1 and C2 is 500. The formula for calculating the output current signal corresponds to the industrial output format of 40 mA to 20 mA, which primarily converts the precise value of the dust concentration Ce into an analog output current signal. Specifically, the formula for calculating the output current signal is performed according to the defined measuring range (e.g., 0.1 to A mg / m³), and the output current signal (mA) = [(instantaneous value / A) * 16] + 4. The instantaneous value is the precise value of the dust concentration Ce, and A represents the measuring range (i.e., 4 to 20) defined in the dust concentration signal processing device 2.

[0041] The format conversion unit 284 comprises a display conversion unit 284-1, an analog conversion unit 284-2, a threshold comparison unit 284-3, and an RS485 conversion unit 284-4. These four distinct conversion units are coupled to the output processing unit 282 and equipment (such as, but not limited to, LCM displays, computers, relays, etc.) connected to the backend. The display conversion unit 284-1 converts the value processed by the output processing unit 282 into the dust concentration Cd in the display output format Fd, thus providing a more accurate dust concentration Cd for display by the liquid crystal module (LCM). The analog conversion unit 284-2 converts the value processed by the output processing unit 282 into the dust concentration Cd in the analog output format Fa, thus providing a more accurate dust concentration Cd for the backend equipment.In particular, the analog conversion unit 284-2 mainly converts the value into an analog output current signal of 4 mA to 20 mA.

[0042] The threshold comparator 284-3 converts the value processed by the output processing unit 282 into the dust concentration Cd in the relay output format Fr, thus providing a more accurate dust concentration Cd for switching the relay on and off. If the dust concentration Cd is greater than the threshold, the relay indicates that the dust concentration Cd is too high by triggering; otherwise, the relay indicates that the dust concentration Cd has not exceeded the standard by not triggering. The RS485 conversion unit 284-4 converts the value processed by the output processing unit 282 into the dust concentration Cd in the RS485 output format Frs, thus providing a more accurate dust concentration Cd for the backend equipment.In one embodiment, the output formats of the numerical conversion module28 are not limited to the four formats mentioned above, and one of the four output formats mentioned above can be selected according to the actual requirements, or other output formats can be added. Furthermore, the four output types mentioned above provide not only the asynchronous dust concentration Cd, but also the real-time dust concentration Cd, so that the user can intuitively distinguish between the current values ​​and the asynchronous values.

[0043] With reference to Fig. 9 and further reference to Fig. 1 to Fig.Figure 8 shows a flowchart of a method for processing a dust concentration signal according to the present disclosure. The method is used to analyze the detection signal Sd of a dust concentration meter 1 in order to calculate a dust concentration Cd, thereby reducing the influence of noise generated by the eddy current in the outlet channel. The method comprises the following steps: Sampling the detection signal Sd (S 100). In one embodiment, the feature processing module 22 samples the detection signal Sd according to a preset sampling rate, and a suitable sampling rate is preferably in the range of 10 Hz to 10 kHz, with 100 Hz being particularly optimal. Then, it is determined whether a filtering process is performed (S 120). If the determination result is "YES", a filtering process is performed on the sampled values ​​Vs (S 140).In one embodiment, the filter unit 228 mainly uses a wavelet filtering method to perform first- to fifth-order filtering on curve C1 to filter out the high-frequency harmonics on curve C1. If the determination result is "NO", the eigenvalues ​​Es are calculated according to the sampled values ​​Vs (S 160). In another embodiment, the dust concentration signal processing device 2 pre-sets a specific time unit (e.g., without restriction, 1 second), and the feature calculation unit 224 calculates the sampled values ​​Vs in each time unit (e.g., without restriction, 1 minute, 1 hour, etc.) to obtain an eigenvalue Es in order to form the eigenvalues ​​Es in a continuous time unit. In particular, if few high-frequency harmonics are present, or even if no high-frequency harmonics are present, steps (S120) to (S140) can be omitted.

[0044] Emission types are then determined (S180). In one embodiment, the type determination unit 226 determines, based on a feature of the eigenvalues ​​Es, whether the sampled values ​​Vs belong to a continuous or a temporary dust emission. A weighted moving average calculation is then performed (S200). If the sampled values ​​Vs belong to the continuous dust emission, the first calculation unit 242, in one embodiment, performs the weighted moving average calculation to calculate the eigenvalues ​​Es as the estimated value of the dust concentration Cc. An integral calculation is then performed (S220). If the sampled values ​​Vs belong to the temporary dust emission, the second calculation unit 244, in one embodiment, performs the integral calculation to calculate the eigenvalues ​​Es as the estimated value of the dust concentration Cc.

[0045] Subsequently, calibration is performed on the estimated dust concentration value Cc (S240). In one embodiment, the calibration unit 262 calculates the dust concentration calibration value Ccv by performing multi-point calibration and interpolation. The dust concentration signal processing device 2 performs calibration tests in a test environment (such as, without limitation, a standard laboratory) to obtain calibration constant values ​​prior to actual operation and then calculates the dust concentration calibration value Ccv according to the calibration constant value and the estimated dust concentration value Cc. Subsequently, compensation of the dust concentration calibration value Ccv is performed (S260). In one embodiment, the compensation unit 264 ascertains the exact dust concentration value Ce by calculating the dust concentration calibration value Ccv using an environmental constant value.In particular, the compensation unit 264 sets the ambient constant value according to the test environment and an actual operating environment. By calculating the ambient constant value, the exact value of the dust concentration Ce can be obtained.

[0046] Finally, a numerical conversion is performed to obtain the dust concentration (S280). In one embodiment, the numerical conversion module 28 performs a calculation and format conversion of the exact value of the dust concentration Ce to convert the exact value of the dust concentration Ce into the dust concentration Cd in a specific format. The numerical conversion module 28 can convert the estimated value of the dust concentration Ce into at least one of the display output format Fd, the analog output format Fa, the relay output format Fr, and the RS485 output format Frs. In one embodiment, the above-mentioned process can be performed not only using the formats specified in Fig.The 3 modules shown can be achieved, and all modules / circuits / programs that can achieve the same function should be included within the scope of this embodiment.

Claims

[1] Method for processing a dust concentration signal designed to analyze a detection signal (Sd) of a dust concentration meter (1) in order to calculate a dust concentration (Cd), the method comprising the following steps: Sampling the detection signal (Sd) to calculate several eigenvalues ​​(Es), and determining a signal type of the detection signal (Sd) according to the eigenvalues ​​(Es), Selecting a specific algorithm according to the eigenvalues ​​(Es) and calculating the eigenvalues ​​(Es) using the specific algorithm to obtain an estimated value of the dust concentration (Cc), Performing a calibration compensation on the estimated dust concentration value (Cc) to obtain an accurate dust concentration value (Ce), and Performing a numerical conversion on the exact value of the dust concentration (Ce) to provide the dust concentration (Cd) in a specific data format. [2] The method of claim 1, further comprising the following steps: Sampling the detection signal (Sd) at a sampling rate to provide multiple samples (Vs), Calculating the sample values ​​(Vs) in each time unit to capture an eigenvalue (Es), to form the eigenvalues ​​(Es) in a continuous time unit, and Determine, according to a characteristic of the eigenvalues ​​(Es), that the sampled values ​​(Vs) belong to a continuous dust emission or a temporary dust emission. [3] The method of claim 2, further comprising the following step: selectively performing a filtering process on the sampled values ​​(Vs) according to a noise of the sampled values ​​(Vs). [4] The method of claim 2, further comprising the following step: Selecting at least one of a mean calculation method, a root mean square calculation method, a standard deviation calculation method, and a peak-to-peak calculation method to calculate the sampled values ​​(Vs) as the eigenvalues ​​(Es). [5] The method of claim 2, further comprising the following step: Providing a weighted moving average calculation when the eigenvalues ​​(Es) belong to the continuous dust emission, in order to calculate the eigenvalues ​​(Es) as the estimated value of the dust concentration (Cc); or Providing an integral calculation when the eigenvalues ​​(Es) belong to the temporary dust emission, in order to calculate the eigenvalues ​​(Es) as the estimated value of the dust concentration (Cc). [6] The method of claim 1 further comprising the following steps: Calculating the estimated dust concentration (Cc) value using a calibration constant value to obtain a dust concentration calibration value (Ccv), Calculating the dust concentration calibration value (Ccv) using an ambient constant value to obtain the exact dust concentration value (Ce), Performing a calibration test in a test environment in advance to obtain the calibration constant value before actual operation, and Setting the environment constant value according to the test environment and an actual operating environment. [7] Dust concentration signal processing device (2) designed to analyze a detection signal (Sd) from a dust concentration meter (1) in order to calculate a dust concentration (Cd), the device (2) comprising: a feature processing module (22) designed to sample the detection signal (Sd) in order to calculate several eigenvalues ​​(Es) and to determine a signal type of the detection signal (Sd) according to the eigenvalues ​​(Es), a calculation module (24) designed to select a specific algorithm according to the eigenvalues ​​(Es) and to calculate the eigenvalues ​​(Es) using the specific algorithm to obtain an estimated value of the dust concentration (Cc), a calibration compensation module (26) designed to perform calibration compensation on the estimated value of the dust concentration (Cc) in order to obtain an accurate value of the dust concentration (Ce), and a module for numerical conversion (28) designed to perform a numerical conversion on the exact value of the dust concentration (Ce) in order to provide the dust concentration (Cd) in a specific data format. [8] Device (2) according to claim 7, wherein the feature processing module (22) comprises: a sampling unit (222) designed to sample the detection signal (Sd) at a sampling rate in order to provide multiple samples (Vs), a feature calculation unit (224) designed to calculate the eigenvalues ​​(Es) according to the sampled values ​​(Vs), and a type determination unit (226) designed to determine, according to a characteristic of the eigenvalues ​​(Es), that the detection signal (Sd) belongs to a continuous dust emission or a temporary dust emission, wherein the feature calculation unit (224) calculates the sample values ​​(Vs) in each time unit to capture an eigenvalue (Es) in order to form the eigenvalues ​​(Es) in a continuous time unit. [9] Device (2) according to claim 8, wherein the feature processing module (22) further comprises: a filter unit (228) designed to selectively perform a filtering process on the samples (Vs) according to a noise of the samples (Vs). [10] Device (2) according to claim 8, wherein the feature calculation unit (224) provides at least one of a mean value calculation method, a root mean square calculation method, a standard deviation calculation method and a peak-to-peak value calculation method to calculate the sampled values ​​(Vs) as the eigenvalues ​​(Es). [11] Device (2) according to claim 8, wherein the calculation module (24) comprises: a first computation unit (242) designed to provide a weighted moving average calculation when the eigenvalues ​​(Es) are part of the continuous dust emission, to calculate the eigenvalues ​​(Es) as the estimated value of the dust concentration (Cc), and a second computation unit (244) designed to provide an integral calculation when the eigenvalues ​​(Es) are part of the temporary dust emission, in order to calculate the eigenvalues ​​(Es) as the estimated value of the dust concentration (Cc). [12] Device (2) according to claim 7, wherein the calibration compensation module (26) comprises: a calibration unit (262) designed to obtain a dust concentration calibration value (Ccv) by calculating the estimated value of the dust concentration (Cc) using a calibration constant value, and a compensation unit (264) designed to obtain the exact value of the dust concentration (Ce) by calculating the dust concentration calibration value (Ccv) using an ambient constant value. [13] Device (2) according to claim 12, wherein the calibration unit (262) performs a calibration test in a test environment in advance to determine the calibration constant value prior to actual operation. [14] Device (2) according to claim 13, wherein the compensation unit (264) sets the environmental constant value according to the test environment and an actual operating environment.