Full-automatic smoke detection method and tester

Abstract

The invention relates to the technical field of physicochemical property detection, in particular to a full-automatic smoke dust detection method and a tester, and the method comprises the following steps: acquiring the concentration, temperature and relative humidity data of the inner wall of a flue, and preprocessing; dividing the temperature sequence into a plurality of temperature subsequences, and obtaining a concentration subsequence and a relative humidity subsequence of a corresponding bit sequence; obtaining a noise content index of a single temperature subsequence; obtaining the size of a filtering window of the filter in a single temperature sub-sequence; acquiring peak values and valley values of all the concentration subsequences and the relative humidity subsequences, and acquiring in-peak elements of all the peak values; acquiring a violent index of each peak value; acquiring a humidity irrelevant coefficient of a single concentration subsequence; acquiring a concentration correction factor of a single concentration subsequence; and correcting each concentration sub-sequence. According to the method, the concentration correction factor can be adaptively adjusted according to different process parameters, so that the detection precision of the smoke concentration is improved.

Description

Technical Field

[0001] This application relates to the technical field of physical and chemical property detection, and particularly relates to a full-automatic smoke and dust detection method and tester. Background Art

[0002] With the continuous advancement of the industrialization process, the smoke and dust emissions from industrial facilities such as coal-fired boilers, steel production, and blast furnaces have become important factors affecting the atmospheric environmental quality. In order to effectively control and reduce industrial smoke and dust emissions and ensure air quality and public health, real-time monitoring of smoke and dust emissions is particularly important. Among the existing non-contact smoke and dust detection technologies, the light scattering method is widely used in the field of smoke and dust detection due to its advantages such as continuous real-time detection, small size, low cost, and high sensitivity. However, when the light scattering method is used for smoke and dust detection, it is mainly affected by the relative humidity in the smoke and dust, resulting in errors. Therefore, it is necessary to correct the smoke and dust concentration results detected by the light scattering method.

[0003] Using a correction coefficient to correct the light scattering result is a common method to eliminate the system deviation of light scattering. However, in the existing technology, when using the correction coefficient for correction, although the influence of relative humidity on light scattering is considered, only fixed parameters are used to calculate the concentration correction factor, and the influence of different relative humidities on the smoke and dust concentration is not considered to be inconsistent, resulting in a poor correction effect on light scattering and unable to accurately obtain the smoke and dust concentration data. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a full-automatic smoke and dust detection method and tester, and the specific technical solutions adopted are as follows: In the first aspect, an embodiment of this application provides a full-automatic smoke and dust detection method, and the method includes the following steps: Obtain the concentration sequence, temperature sequence, and relative humidity sequence of the inner wall of the flue within the same time, and perform normalization processing on them respectively; Divide the temperature sequence into multiple temperature subsequences, and obtain the corresponding concentration subsequence and relative humidity subsequence of the corresponding sequence number according to the sequence number of a single temperature subsequence in the sequence; obtain the noise content index of a single temperature subsequence according to the autocorrelation degree and data dispersion degree of the single temperature subsequence; obtain the filtering window size of the filter for the single temperature subsequence according to the noise content index of the single temperature subsequence, and filter the single temperature subsequence and its corresponding concentration subsequence and relative humidity subsequence according to the filtering window size; Obtain the peaks and valleys of all concentration subsequences and relative humidity subsequences, and obtain the elements within the peaks of each peak; according to the number of elements within the peak of each peak, the coefficient of variation of the elements within the peak, and the difference in the coefficient of variation from other peaks in its corresponding subsequence, obtain the intensity index of each peak, and compare it with a preset threshold to obtain abnormal peaks; According to the difference in abnormal peaks between a single concentration subsequence and the relative humidity subsequence of the same ordinal position, and the ordinal distance between each concentration abnormal peak and its nearest neighbor humidity abnormal peak, obtain the humidity independence coefficient of a single concentration subsequence; according to the humidity independence coefficient of a single concentration subsequence and the element mean in the relative humidity subsequence of the same ordinal position, obtain the concentration correction factor of the corresponding concentration subsequence, and correct the corresponding concentration subsequence.

[0005] Preferably, the method for determining the noise content index is as follows: Take a single temperature subsequence as the input of the autocorrelation function, and output a set of autocorrelation coefficient values; Denote the mean value of the autocorrelation coefficient values as the autocorrelation index of the single temperature subsequence; Obtain the fitting line of the single temperature subsequence, obtain the shortest distance from each data point in the single temperature subsequence to the fitting line, and denote the degree of difference between all the shortest distances in the single temperature subsequence as the discrete index of the single temperature subsequence; Denote the sum of the autocorrelation index and the discrete index of the single temperature subsequence as the noise content index of the single temperature subsequence.

[0006] Preferably, the calculation method for the filter window size of the single temperature subsequence is as follows: , where represents the filter window size of the filter in the z-th temperature subsequence, is the rounding function, is the exponential function with the natural constant e as the base, is the first preset threshold, is the second preset value, represents the noise content index of the z-th temperature subsequence.

[0007] Preferably, the elements within the peak of each peak refer to all the elements between the previous nearest neighbor valley value and the next nearest neighbor valley value of each peak. If there is no valley value in front of or behind the peak, the elements within the peak of the peak refer to all the elements between the peak and the valley value at the head or tail of the sequence.

[0008] Preferably, the calculation method for the intensity index is as follows: , where is the intensity index of the i-th peak, is the coefficient of variation between all the elements within the peak of the i-th peak, is the mean of the absolute difference between the coefficient of variation of the i-th peak and the coefficients of variation of all peaks in the corresponding subsequence. is the number of elements within the i-th peak.

[0009] Preferably, the method for determining the humidity independence coefficient is as follows: Denote the abnormal peaks in the concentration subsequence and the relative humidity subsequence as concentration abnormal peaks and humidity abnormal peaks, respectively. Denote the sequence composed of all concentration abnormal peaks in a single concentration subsequence as the concentration abnormal peak sequence, and denote the sequence composed of all humidity abnormal peaks in a single relative humidity subsequence as the humidity abnormal peak sequence. Calculate the DTW distance between the concentration abnormal peak sequence of a single concentration subsequence and the humidity abnormal peak sequence of the relative humidity subsequence with the same ordinal position, and denote the DTW distance as the humidity asynchronous index of the single concentration subsequence. Obtain the humidity response interval of a single concentration subsequence according to the ordinal distance between each concentration abnormal peak in the single concentration subsequence and its nearest neighboring humidity abnormal peak. Denote the product of the humidity asynchronous index and the humidity response interval of a single concentration subsequence as the humidity independence coefficient of the single concentration subsequence.

[0010] Preferably, the method for determining the humidity response interval is as follows: In a single concentration subsequence and the relative humidity subsequence with the same ordinal position, denote the humidity abnormal peak that is the nearest neighbor of each concentration abnormal peak and has an ordinal position less than or equal to the ordinal position of the corresponding concentration abnormal peak as the pre-humidity peak of each concentration abnormal peak; denote the absolute difference between the ordinal position of each concentration abnormal peak and the ordinal position of its pre-humidity peak as the humidity response index of each concentration abnormal peak; denote the mean of the humidity response indices of all concentration abnormal peaks in a single concentration subsequence as the humidity response interval of the single concentration subsequence.

[0011] Preferably, the calculation method of the concentration correction factor is as follows: , where is the concentration correction factor of the z-th concentration subsequence; is the normalized humidity independence coefficient of the z-th concentration subsequence; is the mean of all elements in the z-th relative humidity subsequence.

[0012] Preferably, the process of correcting the corresponding concentration subsequence is as follows: Take the ratio of each element in each concentration subsequence to the concentration correction factor of the corresponding concentration subsequence as the corrected concentration subsequence.

[0013] In a second aspect, an embodiment of the present application further provides a fully automatic smoke and dust detection tester, which includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned fully automatic smoke and dust detection methods are implemented.

[0014] The present application has at least the following beneficial effects: Considering that when smoke and dust are emitted, electronic components will be affected by the outside world, resulting in errors in the collected data, the present application constructs a noise content index to filter the collected data, improve the data accuracy, and thus provide data support for subsequent calculations. Then, according to the influence of relative humidity on the smoke and dust concentration, a severity index is constructed to select significant data from the concentration subsequence; then, according to the difference in abnormal peaks between a single concentration subsequence and the relative humidity subsequence of the same ordinal position, and the ordinal distance between each concentration abnormal peak and its nearest neighbor humidity abnormal peak, a humidity independence coefficient is constructed to reflect the influence degree of relative humidity on concentration under the same process parameters; finally, the concentration correction factor is optimized according to the humidity independence coefficient. Compared with the prior art, the present application can adaptively adjust the concentration correction factor for different process parameters, thereby improving the detection accuracy of the smoke and dust concentration. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0016] Figure 1 It is a flowchart of the steps of a fully automatic smoke and dust detection method provided by an embodiment of the present application; Figure 2 It is a flowchart of the construction of a concentration correction factor provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] In order to further elaborate on the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following will, in conjunction with the accompanying drawings and preferred embodiments, describe in detail the specific implementation manners, structures, features, and effects of a fully automatic smoke and dust detection method and tester proposed according to the present application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this application belongs.

[0019] The following specifically describes the specific solutions of a full-automatic soot detection method and a detector provided by this application with reference to the accompanying drawings.

[0020] Please refer to Figure 1 , which shows a flowchart of the steps of a full-automatic soot detection method provided by an embodiment of this application. The method includes the following steps: Step 1: Obtain the concentration sequence, temperature sequence, and relative humidity sequence of the inner wall of the flue within the same time, and perform normalization processing on each of them.

[0021] The scenario targeted by this embodiment is to detect the discharged soot in the factory flue. An optoelectronic sensor, a temperature sensor, and a humidity sensor are respectively connected to the inner wall of the flue. When soot is discharged, the laser beam will be scattered by the soot particles in the detected flue. The backscattered light is collected by the optoelectronic sensor to obtain the concentration information of the soot. At the same time, the temperature and humidity data of the discharged soot are collected through the temperature and humidity sensors, and the relative humidity data at this moment is calculated based on the temperature and humidity data at the same moment. It should be noted that the relative humidity refers to the ratio of the actual content of water vapor in the current air to the maximum amount of water vapor that the air can hold at this temperature. The calculation of relative humidity is a well-known technology and will not be elaborated here. The acquisition interval of all data in this embodiment is 1 second, and the duration of each data acquisition is 30 minutes.

[0022] According to the time sequence of data acquisition, a concentration sequence, a temperature sequence, and a relative humidity sequence are respectively constructed. In order to eliminate the influence of the dimension between data, normalization processing is performed on all data. The normalization processing methods include but are not limited to Z-score standardization, maximum normalization, and maximum-minimum normalization.

[0023] Step 2: Divide the temperature sequence into multiple temperature subsequences, obtain the corresponding concentration subsequence and relative humidity subsequence of the corresponding sequence number according to the sequence number of a single temperature subsequence in the sequence; obtain the noise content index of a single temperature subsequence according to the autocorrelation degree and data dispersion degree of the single temperature subsequence; obtain the filtering window size of the filter for the single temperature subsequence according to the noise content index of the single temperature subsequence, and filter the single temperature subsequence and its corresponding concentration subsequence and relative humidity subsequence according to the filtering window size.

[0024] Under different process parameters, the concentration and temperature of industrial emissions of soot will also change. Since the soot emitted industrially usually has a high temperature, and both the high temperature and the emission rate of the soot will affect electronic components, the data collected by the sensors will have error noise. In addition, since the emission temperatures of the soot under different process parameters are inconsistent, and the high temperature will also cause changes in the humidity in the soot, thereby affecting the moisture content in the soot and causing fluctuations in the relative humidity. Therefore, first, the soot data under multiple different process parameters need to be divided, and then the data needs to be filtered to eliminate noise, so as to further accurately analyze the influence degree of relative humidity on the soot concentration.

[0025] Considering that when collecting data in the flue through an intelligent sensor, the flow rate and high temperature of the soot in the flue will cause random disturbances to the electronic components, which will result in large errors in the collected data. Therefore, first, the collected soot concentration data needs to be filtered, and then the influence degree of relative humidity on the concentration can be accurately calculated. However, when filtering, if the filtering window is too large, it may oversmooth the data and weaken the data features; if the filtering window is too small, effective noise reduction cannot be achieved. Therefore, first, a suitable filtering window needs to be set according to the degree of noise interference of the data.

[0026] Since the emission temperatures of the soot under different process parameters are inconsistent, and the temperature changes under the same process parameter are relatively stable, then under the state of switching between different process parameters, the soot temperature data will show periodic changes. Therefore, the emissions of the soot under different process parameters can be divided by the periodic changes of the soot temperature.

[0027] The temperature sequence is segmented into multiple temperature subsequences through the Bernaola Galvan segmentation algorithm (abbreviated as BG segmentation algorithm). The sequence segmentation algorithms include but are not limited to the BG segmentation algorithm, the PELT sequence segmentation algorithm, etc.

[0028] Taking the z-th temperature subsequence as an example, according to the position range of the z-th temperature subsequence in the temperature sequence, subsequences within the corresponding position range are respectively extracted from the concentration sequence and the relative humidity sequence, and are respectively denoted as the z-th concentration subsequence and the z-th relative humidity subsequence.

[0029] Under the same process parameter, in a short period of time, due to relatively consistent process control, the industrial production process is stable. Therefore, various collected data should have strong autocorrelation in a short period of time, that is, there is a short-term dependence relationship in the data. Considering that the soot concentration is affected by the relative humidity, and the relative humidity is affected by the soot temperature, in this embodiment, the temperature data less affected by environmental factors is used for analysis, and the autocorrelation of the soot temperature in a short period of time is detected to evaluate the noise impact of the emitted soot on the electronic components.

[0030] Take the z-th temperature subsequence as the input of the autocorrelation function, and output a series of autocorrelation coefficient values. Each coefficient corresponds to a different lag amount, and the difference between adjacent lag amounts is 1. In this embodiment, the value range of the lag amount takes the empirical value [1, 30]. It should be noted that if the data length of the z-th temperature subsequence is less than 30, then the value range of the lag amount is set to [1, N / 2], where N is the data length of the z-th temperature subsequence.

[0031] Denote the mean value of all autocorrelation coefficient values as the autocorrelation index of the z-th temperature subsequence; the autocorrelation index can reflect the time dependence relationship of the temperature of the emitted soot under the same process parameters in a short time; since the noise does not have obvious autocorrelation, therefore, the smaller the autocorrelation index, the worse the autocorrelation of the collected soot temperature in a short time, the less it conforms to the true physical characteristics of soot emission, indicating that the collected temperature data is more affected by noise.

[0032] Further, perform a linear fit on the z-th temperature subsequence, calculate the shortest distance from each data point in the z-th temperature subsequence to the fitted line, and then calculate the degree of difference between all the shortest distances. The degree of difference can be calculated by methods such as variance, standard deviation, and coefficient of variation.

[0033] Denote the degree of difference as the discrete index of the z-th temperature subsequence. The discrete index can reflect the degree of deviation of each temperature data from the overall temperature change trend. The larger the discrete index, the more obvious the difference between the data and the overall trend, indicating that the z-th temperature subsequence is more affected by noise interference.

[0034] Further, denote the filtering window size of the moving average filter as , , where represents the filtering window size of the filter in the z-th temperature subsequence, is the rounding function, is the exponential function with the natural constant e as the base, is the noise content index of the z-th temperature subsequence, representing the sum of the autocorrelation index and the discrete index, is the first preset value, is the second preset value. In this embodiment, 45 and 11 are taken respectively. It should be noted that in order to ensure that the core element of the filter is located at the center position and the weights on both sides are symmetrically distributed, the filtering window size is processed. If the calculation result is odd, then is used as the filtering window size. If the calculation result is even, then -1 is used as the filtering window size. According to the obtained filtering window size, the z-th temperature subsequence is filtered by using a moving average filter, and with the same filtering window size, the z-th concentration subsequence and the z-th relative humidity subsequence are filtered by using a moving average filter to eliminate the noise effects on electronic components caused by high temperature of soot and emission flow rate in the same period.

[0035] So far, the elimination process of noise in the collected data has been realized, thus providing accurate data support for the subsequent calculation of the influence of relative humidity on concentration.

[0036] Step 3: Obtain the peaks and valleys of all concentration subsequences and relative humidity subsequences, and obtain the elements within the peaks of each peak; according to the number of elements within the peaks of each peak, the coefficient of variation of the elements within the peaks, and the difference in the coefficient of variation from other peaks in its subsequence, obtain the intensity index of each peak, and compare it with a preset threshold to obtain abnormal peaks.

[0037] Furthermore, if the relative humidity in the soot is relatively high, it will cause water vapor in the air to condense on the surface of the particulate matter, resulting in a higher detected soot concentration. Moreover, the higher the relative humidity, the more serious the concentration deviation. As a result, abnormal peaks of increased concentration appear, and there are significant differences between these abnormal peaks and normal peaks. Therefore, it is necessary to first extract abnormal peaks from the concentration data to further analyze the influence degree of relative humidity.

[0038] Taking the filtered z-th concentration subsequence as an example. All peaks and valleys in the z-th concentration sequence are obtained through a peak-valley detection algorithm. The peak-valley detection algorithm is a well-known technology, and the specific process will not be elaborated here. Taking the i-th peak in the z-th concentration subsequence as an example for analysis. For the i-th peak, all elements between its previous nearest neighbor valley and its next nearest neighbor valley are called the elements within the peak of this peak. It should be noted that if the peak is located in the head or tail region of the sequence and there is no valley in front of or behind it, then all elements between this peak and the valley at the head or tail of the sequence are recorded as the elements within the peak of this peak.

[0039] Let . In the formula, is the intensity index of the i-th peak; is the coefficient of variation between all elements within the i-th peak; is the average value of the absolute difference between the coefficient of variation of the i-th peak and the coefficients of variation of all peaks in its subsequence; is the number of elements within the i-th peak, which reflects the continuous change time of the soot concentration within the range of the i-th peak. The calculation of the coefficient of variation is a well-known technology, and the specific process will not be elaborated here. The change index can reflect the severity of the change in the soot concentration within the peak region.

[0040] It can reflect the change rate of the soot concentration within the i-th peak range. If the change in the soot concentration within the i-th peak range is greater and the required change time is shorter, it indicates that the change rate of the soot concentration within the i-th peak range is greater. Average change difference It can measure the significance of the change of the i-th peak relative to other peaks. The greater it is, the more significant the change degree of the soot within the i-th peak range is among all peaks. Therefore, if the severity index is greater, it can reflect that the change degree of the soot within the i-th peak range is greater, and there are significant differences from the change degrees of the soot within other peak ranges. However, due to the strict regulations on soot emissions under the same process parameters, if is greater, it can reflect that the possibility of abnormal changes in the soot within the i-th peak range due to humidity influence is greater.

[0041] Furthermore, calculate the severity index of each peak in the z-th concentration subsequence, and use all the severity indices as the input of the Otsu threshold method. The output of the Otsu threshold method is the severity threshold. Mark the peaks with all severity indices greater than or equal to the severity threshold as the abnormal peaks of the z-th concentration subsequence.

[0042] Similarly, the abnormal peaks in all concentration subsequences can be obtained.

[0043] Similarly, the abnormal peaks in all relative humidity subsequences can be obtained.

[0044] So far, the abnormal peaks in all concentration subsequences and all relative humidity subsequences have been obtained.

[0045] Step 4: Obtain the humidity independence coefficient of a single concentration subsequence according to the difference between the abnormal peaks of a single concentration subsequence and the relative humidity subsequence with the same ordinal position, and the ordinal distance between each concentration abnormal peak and its nearest neighbor humidity abnormal peak; obtain the concentration correction factor of the single concentration subsequence according to the humidity independence coefficient of the single concentration subsequence and the element mean value in the relative humidity subsequence with the same ordinal position, and correct the corresponding concentration subsequence.

[0046] After extracting the abnormal peaks in each concentration subsequence and each relative humidity subsequence, the influence degree of relative humidity on the soot concentration can be calculated to further optimize the concentration correction factor.

[0047] Denote the abnormal peaks in the z-th concentration subsequence and the z-th relative humidity subsequence as the concentration abnormal peak and the humidity abnormal peak, respectively. Construct the concentration peak rank sequence and the humidity peak rank sequence of the z-th concentration subsequence and the z-th relative humidity subsequence respectively according to the ranks of each concentration abnormal peak and humidity abnormal peak in the z-th concentration subsequence and the z-th relative humidity subsequence.

[0048] Obtain the DTW distance between the concentration peak rank sequence and the humidity peak rank sequence, and denote it as the humidity asynchronous index of the z-th concentration subsequence. The calculation method of the DTW distance is a well-known technology and will not be elaborated here. The humidity asynchronous index can reflect whether the concentration data in the soot will change correspondingly after a drastic change in humidity; the smaller the humidity asynchronous index, the more consistent the change in soot concentration and relative humidity, and the greater the influence of relative humidity on soot concentration.

[0049] Take the u-th concentration abnormal peak as an example. Among all the humidity abnormal peaks, find the humidity abnormal peak that is the nearest neighbor to the u-th concentration abnormal peak and whose rank is less than or equal to the rank of the u-th concentration abnormal peak, and denote it as the pre-humidity peak of the u-th concentration abnormal peak. Denote the absolute difference between the rank of the u-th concentration abnormal peak and the rank of its pre-humidity peak as the humidity response index of the u-th concentration abnormal peak.

[0050] In the same way, obtain the humidity response indexes of all concentration abnormal peaks and calculate the mean value, and denote the mean value as the humidity response interval of the z-th concentration subsequence. The smaller the humidity response interval, the greater the influence of relative humidity on soot concentration, and the faster the response speed of soot concentration to relative humidity. It should be noted that: if a certain concentration abnormal peak does not have a corresponding pre-humidity peak, then this concentration abnormal peak does not participate in the calculation.

[0051] Denote the product of the humidity asynchronous index and the humidity response interval of the z-th concentration subsequence as the humidity non-relationship coefficient of the z-th concentration subsequence. The humidity non-relationship coefficient can reflect whether the soot concentration will change rapidly and synchronously when the relative humidity changes drastically. The smaller the humidity non-relationship coefficient, the greater the influence degree of relative humidity on concentration.

[0052] Calculate the concentration correction factor of the z-th concentration subsequence according to the influence of the z-th relative humidity subsequence on the z-th concentration subsequence. The construction flow chart of the concentration correction factor is as Figure 2 shown. Specifically, let . In the formula, is the concentration correction factor of the z-th concentration subsequence; is the normalized humidity non-relationship coefficient of the z-th concentration subsequence; is the mean value of all elements in the z-th relative humidity subsequence.

[0053] When the relative humidity in the soot is higher and the influence of relative humidity on the concentration is greater, a larger concentration correction factor is required to correct the collected soot concentration; while if the relative humidity in the soot is lower and the humidity influence is smaller, the concentration correction factor is close to 1 and there is no need to over-correct the soot concentration data.

[0054] Divide each element in the z-th concentration subsequence by the concentration correction factor to obtain the corrected z-th concentration subsequence; in the same way, correct all the concentration subsequences to obtain accurate emission soot concentration data.

[0055] So far, the adaptive optimization of the concentration correction factor under various process parameters has been realized, thereby accurately correcting the soot concentration and improving the data detection accuracy of the emission soot concentration.

[0056] Based on the same inventive concept as the above method, an embodiment of the present application also provides a full-automatic soot detection tester, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above full-automatic soot detection methods are implemented.

[0057] It should be noted that: the above sequence of embodiments of the present application is only for description and does not represent the advantages or disadvantages of the embodiments. And the above specific embodiments of this specification have been described. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0058] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0059] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present application shall be included within the protection scope of the present application.

Claims

1. An automatic soot detection method, characterized in that, The method includes the following steps: Obtain the concentration sequence, temperature sequence, and relative humidity sequence of the inner wall of the flue gas duct within the same time, and perform normalization processing on each of them respectively; Divide the temperature sequence into multiple temperature subsequences, obtain the corresponding concentration subsequence and relative humidity subsequence of the corresponding ordinal position according to the ordinal position of a single temperature subsequence in the sequence; obtain the noise content index of a single temperature subsequence according to the autocorrelation degree and data dispersion degree of the single temperature subsequence; obtain the filtering window size of the filter for the single temperature subsequence according to the noise content index of the single temperature subsequence, and perform filtering on the single temperature subsequence and its corresponding concentration subsequence and relative humidity subsequence according to the filtering window size respectively; Obtain the peaks and valleys of all concentration subsequences and relative humidity subsequences, and obtain the in-peak elements of each peak; obtain the sharpness index of each peak according to the number of in-peak elements of each peak, the coefficient of variation of the in-peak elements, and the difference in the coefficient of variation with other peaks in its subsequence, and compare it with a preset threshold to obtain abnormal peaks; Obtain the humidity independence coefficient of a single concentration subsequence according to the difference in abnormal peaks between the single concentration subsequence and the relative humidity subsequence of the same ordinal position, and the ordinal distance between each concentration abnormal peak and its nearest neighbor humidity abnormal peak; obtain the concentration correction factor of the single concentration subsequence according to the humidity independence coefficient of the single concentration subsequence and the element mean value in the relative humidity subsequence of the same ordinal position, and correct the corresponding concentration subsequence.

2. The fully automatic smoke and dust detection method according to claim 1, characterized in that The method for determining the noise content index is: use a single temperature subsequence as the input of the autocorrelation function, and output a set of autocorrelation coefficient values; record the mean value of the autocorrelation coefficient values as the autocorrelation index of the single temperature subsequence; obtain the fitting straight line of the single temperature subsequence, obtain the shortest distance from each data point in the single temperature subsequence to the fitting straight line, and record the difference degree between all the shortest distances in the single temperature subsequence as the dispersion index of the single temperature subsequence; record the sum of the autocorrelation index and the dispersion index of the single temperature subsequence as the noise content index of the single temperature subsequence.

3. The fully automatic soot detection method according to claim 1, characterized in that, The calculation method for the filtering window size of the single temperature subsequence is as follows: , where represents the filtering window size of the filter for the z-th temperature subsequence, is the rounding function, is the exponential function with the natural constant e as the base, is the first preset value, is the second preset value, represents the noise content index of the z-th temperature subsequence.

4. The fully automatic soot detection method according to claim 1, characterized in that The in-peak elements of each peak refer to all elements between the previous nearest neighbor valley and the next nearest neighbor valley of each peak. If there is no valley in front of or behind the peak, the in-peak elements of the peak refer to all elements between the peak and the valley at the head or tail of the sequence.

5. The fully automatic smoke and dust detection method according to claim 1, characterized in that, The calculation method of the severity index is as follows: , where is the severity index of the i-th peak, is the coefficient of variation between all elements within the i-th peak, is the mean of the absolute differences between the coefficient of variation of the i-th peak and the coefficients of variation of all peaks in the subsequence where it is located, is the number of elements within the i-th peak.

6. The fully automatic soot detection method according to claim 1, characterized in that The method for determining the humidity independence coefficient is: Record the abnormal peaks in the concentration subsequence and the relative humidity subsequence as concentration abnormal peaks and humidity abnormal peaks respectively; Record the sequence composed of all concentration abnormal peaks in a single concentration subsequence as the concentration abnormal peak sequence, and record the sequence composed of all humidity abnormal peaks in a single relative humidity subsequence as the humidity abnormal peak sequence; Calculate the DTW distance between the concentration abnormal peak sequence of a single concentration subsequence and the humidity abnormal peak sequence of the relative humidity subsequence of the same ordinal position, and record the DTW distance as the humidity asynchrony index of the single concentration subsequence; Obtain the humidity response interval of a single concentration subsequence according to the ordinal distance between each concentration anomaly peak and its nearest neighbor humidity anomaly peak in the single concentration subsequence; Denote the product of the humidity asynchronous index of a single concentration subsequence and the humidity response interval as the humidity independence coefficient of the single concentration subsequence.

7. The fully automatic soot detection method according to claim 6, wherein, The method for determining the humidity response interval is as follows: in a single concentration subsequence and the relative humidity subsequence of the same ordinal number, the humidity anomaly peak that is the nearest neighbor to each concentration anomaly peak and whose ordinal number is less than or equal to the ordinal number of the corresponding concentration anomaly peak is denoted as the pre-humidity peak of each concentration anomaly peak; the absolute difference between the ordinal number of each concentration anomaly peak and the ordinal number of its pre-humidity peak is denoted as the humidity response index of each concentration anomaly peak; the mean value of the humidity response indexes of all concentration anomaly peaks in the single concentration subsequence is denoted as the humidity response interval of the single concentration subsequence.

8. The full-automatic soot detection method according to claim 1, characterized in that, The calculation method of the concentration correction factor is as follows: , where is the concentration correction factor of the z-th concentration subsequence; is the humidity-independent coefficient after normalization of the z-th concentration subsequence; is the mean value of all elements in the z-th relative humidity subsequence.

9. The full-automatic soot detection method according to claim 1, characterized in that, The process of correcting the corresponding concentration subsequence is as follows: take the ratio of each element in each concentration subsequence to the concentration correction factor of the corresponding concentration subsequence as the corrected concentration subsequences.

10. A fully automatic smoke and dust detection tester, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of a fully automatic soot detection method according to any one of claims 1-9.

Images