A method for identifying transient operating conditions during heating process of energy production equipment

Abstract

The present invention proposes a method for identifying transient conditions in the heating process of energy production equipment, comprising the following steps: obtaining heating condition data of the production equipment; performing outlier processing on the heating condition data and filling in the outliers; filtering out data from the heating condition data that meets the initial conditions of the production equipment's heating transient process, as transient initial condition data; filtering out data from the heating condition data that meets the termination conditions of the production equipment's heating transient process, as transient termination condition data; merging and filtering the transient initial condition data and the transient termination condition data to generate merged data, and identifying the transient process end time; using the merged data to identify the start time of the production equipment's heating transient process; recording the start time and end time of the production equipment's heating transient process; and performing transient event confirmation. The present invention can improve the timeliness and accuracy of transient identification, thereby ensuring the safe and stable operation of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment heating monitoring, and more particularly to a method for identifying transient operating conditions of a heating process of energy production equipment. Background Art

[0002] With the domestic substitution environment, there are more and more industrial independent heating equipment, and the demand for massive data analysis based on the equipment to ensure the stable operation of the equipment has emerged. To ensure the stable operation of the equipment, effective analysis methods are indispensable.

[0003] The process transient analysis of traditional energy production heating equipment has the following disadvantages:

[0004] 1) Manual identification of transient heating conditions;

[0005] 2) The transient recognition time period is long;

[0006] 3) The transient confirmation cycle is long. Summary of the Invention

[0007] In view of the problems existing in the prior art, the purpose of the present invention is to provide a method for identifying transient operating conditions of a heating process of energy production equipment.

[0008] To achieve the above-mentioned purpose, the present invention is implemented through the following technical solutions:

[0009] A method for identifying transient operating conditions during a heating process of energy production equipment comprises the following steps:

[0010] S1: Obtain heating condition data of production equipment;

[0011] S2: Process the outliers of the heating condition data and fill in the outliers;

[0012] S3: Filtering the data that meets the initial conditions of the heating transient process of the production equipment from the heating condition data as the transient initial condition data;

[0013] S4: Filtering data that meets the termination condition of the heating transient process of the production equipment from the heating condition data as the transient termination condition data;

[0014] S5: Merge and filter the transient initial condition data and the transient termination condition data to generate merged data, and identify the end time of the transient process;

[0015] S6: using the merged data to identify the start time of the heating transient process of the production equipment;

[0016] S7: Record the transient process start time and transient process end time of the production equipment heating;

[0017] S8: Perform transient event confirmation.

[0018] Further, step S1 includes:

[0019] The coolant temperature and hot section temperature data of the power plant including the complete heating condition of the production equipment are obtained, and the obtained data sequence is marked as sequence D.

[0020] Further, step S2 includes:

[0021] By calculating the median of the sequence D every ten minutes, the sequence D1 is obtained;

[0022] Subtract the median in the window from the current data to obtain the difference data, and get the sequence D2;

[0023] The outliers of sequence D2 are marked with the box plot method, the values ​​corresponding to the outlier data indexes are set to null values, and the null values ​​are filled with linear interpolation to obtain sequence D3.

[0024] Further, step S2 is specifically as follows:

[0025] For sequence D:

[0026]

[0027] First take the window data and sort it from small to large: , , m is the number of moving windows, and the formula for taking the median of the moving window is:

[0028] When m is an odd number:

[0029] When m is an even number:

[0030] N is the sequence element value after the moving window takes the median, and the new sequence is obtained after the moving window takes the median:

[0031]

[0032] Subtract the median value of the previous 10-minute moving window from the current value to form a new sequence D2. The calculation method is as follows:

[0033]

[0034] Use the box plot method to mark outliers for series D2. The algorithm is as follows:

[0035] Q3 = D2 (75% quantile)

[0036] Q1 = D2 (25% quantile)

[0037] IQR = Q3 – Q1

[0038] minobs = Q1 - 3.5 IQR

[0039] maxobs = Q3 - 3.5 IQR

[0040] Among them, Q3 is the upper quartile of sequence D2, Q1 is the lower quartile of sequence D2, IQR is the interquartile range of sequence D2, minobs is the minimum observed value of sequence D2, and maxobs is the maximum observed value of sequence D2;

[0041] In sequence D2, data greater than maxobs or less than minobs are filtered and recorded as abnormal data. In sequence D, the value corresponding to the abnormal data index is set to a null value, and the null value of the abnormal data is filled with linear interpolation. The resulting sequence is recorded as D3.

[0042] Further, step S3 includes:

[0043] For sequence D3, data with a temperature lower than a specific heating temperature are screened out to obtain sequence segment index S1. Further, step S4 includes:

[0044] Calculate the slope of the 10-minute moving window of sequence D3 as follows:

[0045] Assume that the variable relationship of the window is y=kx + b, and there are m groups of data with errors in the sequence D3 ( , ) ,in is the actual value of sequence D3, is the window sequence number value, =1,2,3…m;

[0046] Sequence D3 calculates the slope of the moving window every 10 minutes to obtain the sequence segment index S2. The slope calculation formula is as follows:

[0047]

[0048]

[0049] The data with sequence S2 slope less than 0 and sequence D3 measurement point value greater than 291.5°C were screened to obtain sequence fragment index S3.

[0050] Further, step S5 includes:

[0051] In sequence D3, create a feature named label and assign it an empty value.

[0052] For the sequence segment index S1 that satisfies the label value in sequence D3, it is reassigned to 0;

[0053] The label value of the sequence segment index S3 in sequence D3 is reassigned to 1;

[0054] Perform a forward difference on the label column and record the data where the forward difference result is 1;

[0055] Perform reverse difference on the label column and record the data whose reverse difference result is -1;

[0056] Pair the data whose forward difference result is 1 and the data whose reverse difference result is -1 in the adjacent time in the order of closest proximity. The first pairing label value is -1 and the second pairing label value is 1. The data that does not meet this relationship are deleted. Record each pairing result as , where the data is ( , ), and record them as the initial and end identification points of the heating transient of the production equipment, That is the transient end time of the heating process of the production equipment.

[0057] Further, step S6 includes:

[0058] by For the data window, calculate the slope of the moving window every 10 minutes, using the slope calculation method of step S4;

[0059] The last value that is less than the target device heating temperature and has a slope value less than 0 is recorded as , This is the start time of the heating process of the production equipment.

[0060] Compared with the existing technology, the beneficial effect of the present invention is that: the present invention provides a method for identifying transient operating conditions in the heating process of energy production equipment, which can provide engineers with equipment transient event data, improve the timeliness and accuracy of transient identification, and thus ensure the safe and stable operation of the equipment.

[0061] It can be seen that compared with the prior art, the present invention has outstanding substantial features and significant progress, and the beneficial effects of its implementation are also obvious. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0063] Attachment Figure 1 It is a method flow chart of a specific embodiment of the present invention.

[0064] Attachment Figure 2 This is a trend transient identification diagram of measuring point 1 according to a specific embodiment of the present invention.

[0065] Attachment Figure 3 This is a trend transient identification diagram of measuring point 2 according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0066] The specific embodiments of the present invention are described below with reference to the accompanying drawings.

[0067] like Figure 1 A method for identifying transient operating conditions during a heating process of an energy production device is shown, comprising the following steps:

[0068] S1: Obtain heating condition data of production equipment.

[0069] Two measurement points were selected for the coolant temperature, hot-segment temperature, and cold-segment temperature of a power plant, covering the complete heating operation of production equipment. Measurement point 1 measured the coolant temperature minus the first hot-segment temperature, and measurement point 2 measured the coolant temperature minus the second hot-segment temperature. The data spanned one year, and the two sequences of these measurement points were labeled Sequences C and D. Both sequences were processed in the same way; in the following examples, Sequence D will be used.

[0070] S2: Process the heating condition data for outliers and fill in the outliers.

[0071] The method for handling outliers is to calculate the median of a 10-minute moving window of sequence D to generate sequence D1. The difference data is obtained by subtracting the median within the window from the current data to generate sequence D2. Because this new series of difference data is more likely to detect outliers in time series data, outliers in sequence D2 are then marked using a boxplot. The values ​​corresponding to the outlier indexes are set to null and filled with linear interpolation to generate sequence D3.

[0072] The specific calculation method is as follows:

[0073] For the original sequence D:

[0074]

[0075] First take the window data and sort it from small to large: ,

[0076] , m is the number of moving windows, and the formula for taking the median of the moving window is:

[0077] When m is an odd number:

[0078]

[0079] When m is an even number:

[0080]

[0081] N is the sequence element value after the moving window takes the median, and the new sequence is obtained after the moving window takes the median.

[0082]

[0083] The current value minus the median value of the previous 10-minute moving window forms a new sequence D2, which is calculated as follows:

[0084]

[0085] Use the box plot method to mark outliers for series D2. The algorithm is as follows:

[0086] Q3 = D2 (75% quantile)

[0087] Q1 = D2 (25% quantile)

[0088] IQR = Q3 – Q1

[0089] minobs = Q1 - 3.5 IQR

[0090] maxobs = Q3 - 3.5 IQR

[0091] Q3 is the upper quartile of sequence D2, Q1 is the lower quartile of sequence D2, IQR is the interquartile range of sequence D2, minobs is the minimum observed value (lower edge) of sequence D2, and maxobs is the maximum observed value (upper edge) of sequence D2. Data greater than maxobs or less than minobs are filtered out of sequence D2 and recorded as outliers. In the original sequence D, the values ​​corresponding to the outlier indexes are set to null values ​​and linear interpolation is used to fill in the null values. The resulting sequence is recorded as D3.

[0092] S3: Filter out data that meets the initial conditions of the heating transient process of the production equipment from the heating condition data as transient initial condition data.

[0093] This step filters the data that meets the initial conditions of the transient heating process of the production equipment. That is, for the sequence D3 obtained in step S2, the data with a temperature less than 50°C (according to the target equipment heating temperature requirement) is filtered to obtain the sequence segment index S1.

[0094] S4: Filter out data that meets the termination condition of the transient heating process of the production equipment from the heating condition data as the transient termination condition data.

[0095] First, calculate the slope of the moving window every 10 minutes for the sequence D3 obtained in step S2. The calculation method is as follows:

[0096] Assume that the variable relationship of the window is , the present invention obtains m groups of data containing errors ( , ) ,in is the actual value of sequence D3 obtained in step S2, is the window sequence number value, =1,2,3…m, the sequence D3 obtained in step S2 calculates the slope of the moving window every 10 minutes to obtain the sequence segment index S2: the slope calculation formula is as follows:

[0097]

[0098]

[0099] The data with the original sequence S2 slope less than 0 and the sequence D3 measurement point value greater than 291.5°C were screened to obtain the sequence fragment index S3.

[0100] S5: Merge and filter the transient initial condition data and the transient termination condition data to generate merged data, and identify the end time of the transient process.

[0101] This step combines and filters the transient initial condition and termination condition time, and identifies the end time at the same time. Specifically:

[0102] 1. In the sequence D3 obtained in step S2, create a feature, name it label, and assign it a null value;

[0103] 2. Reassign the label value of the index S1 that satisfies the condition in the sequence D3 obtained in step S2 to 0;

[0104] 3. Reassign the label value that satisfies the S3 index in the sequence D3 obtained in step S2 to 1;

[0105] 4. Perform forward difference on the label column and record the data whose forward difference result is 1;

[0106] 5. Perform reverse difference on the label column and record the data whose reverse difference result is -1;

[0107] 6. Pair the data whose forward difference result is 1 and the data whose reverse difference result is -1 in the adjacent time in the order of closest proximity. The first pairing label value is -1 and the second pairing label value is 1. Delete the data that does not meet this relationship. Record each pairing result as , where the data is ( , ), and record them as the initial and end identification points of the heating transient of the production equipment, That is the transient end time of the heating process of the production equipment.

[0108] S6: Identify the start time of the heating transient process of the production equipment using the merged data.

[0109] Iterate each pairing result in the set H obtained in step S5 to For the data window, calculate the slope of the moving window every 10 minutes. The calculation method is the same as the slope calculation method in step S4. For the last value that is less than (according to the target device heating temperature regulation) and the slope value is less than 0, record it as , This is the start time of the heating process of the production equipment.

[0110] S7: Record the transient process start time and transient process end time of the heating of the production equipment.

[0111] Table 1 below shows the transient start and end times of heating of production equipment at two measuring points. Figure 2-3 The following is a trend transient identification diagram for the corresponding measuring points. The transient start time for the production equipment heating, as written for each measuring point, was between 8:24 and 9:16 on November 8, 2019. The transient end time for the production equipment heating was between 15:55 and 15:56 on November 10, 2019. Since the actual heating time for the production equipment was 8:24 on November 8, 2019, and ended at 15:55 on November 10, 2019, the coolant temperature-second hot section temperature can be used as the primary factor for identifying transient conditions during the production equipment heating process.

[0112] Table 1 Statistics of transient start and end times of heating of production equipment

[0113]

[0114] S8: Perform transient event confirmation.

[0115] Transient event confirmation is achieved through three methods: automatic judgment, manual entry, and judgment confirmation.

[0116] A. Automatic determination:

[0117] Automatically determine some transients and form transient records. The system implements this function based on the algorithm rules in the transient parameter configuration.

[0118] B. Manual entry:

[0119] The transient parameter configuration provides a manual entry method for some complex transients. The system uses algorithmic rules to identify transients and form preliminary transient record information. The transient manager can modify or supplement existing transient information on the manual entry function page, or directly enter new transient information.

[0120] C. Determination and Confirmation:

[0121] To ensure the accuracy of transients, the transient record information generated by the automatic judgment and manual entry function pages is in a pending confirmation state and is transferred to the judgment confirmation function page. The transient manager confirms and verifies that the information is correct and then includes it in the database management.

[0122] The present invention will be further described with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are intended only to illustrate the present invention and are not intended to limit the scope of the present invention. In addition, it should be understood that after reading the contents of the present invention, those skilled in the art may make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the present application.

Claims

1. A method for identifying transient operating conditions during a heating process of an energy production device, characterized in that: The steps include: S1: Obtain heating condition data of production equipment; S2: Process the outliers of the heating condition data and fill in the outliers; S3: Filtering the data that meets the initial conditions of the heating transient process of the production equipment from the heating condition data as the transient initial condition data; S4: Filtering data that meets the termination condition of the heating transient process of the production equipment from the heating condition data as the transient termination condition data; S5: Merge and filter the transient initial condition data and the transient termination condition data to generate merged data, and identify the end time of the transient process; S6: using the merged data to identify the start time of the heating transient process of the production equipment; S7: Record the transient process start time and transient process end time of the production equipment heating; S8: Perform transient event confirmation; The step S1 comprises: Obtain coolant temperature hot section temperature data for the power plant, including the complete heating condition of production equipment, and mark the obtained data sequence as sequence D; The step S2 comprises: By calculating the median of the sequence D every ten minutes, the sequence D1 is obtained; Subtract the median in the window from the current data to obtain the difference data, and get the sequence D2; Use the box plot method to mark the outliers in sequence D2, set the values ​​corresponding to the outlier indexes to null values, and fill the null values ​​with linear interpolation to obtain sequence D3; The step S2 is specifically as follows: For sequence D: First take the window data and sort it from small to large: , , m is the number of moving windows, and the formula for taking the median of the moving window is: When m is an odd number: When m is an even number: N is the sequence element value after the moving window takes the median, and the new sequence is obtained after the moving window takes the median: Subtract the median value of the previous 10-minute moving window from the current value to form a new sequence D2. The calculation method is as follows: Use the box plot method to mark outliers for series D2. The algorithm is as follows: Q3 = D2 (75% quantile) Q1 = D2 (25% quantile) IQR = Q3 – Q1 minobs = Q1 - 3.5 IQR maxobs = Q3 - 3.5 IQR Among them, Q3 is the upper quartile of sequence D2, Q1 is the lower quartile of sequence D2, IQR is the interquartile range of sequence D2, minobs is the minimum observed value of sequence D2, and maxobs is the maximum observed value of sequence D2; In sequence D2, filter out data that is greater than maxobs or less than minobs and record them as abnormal data. In sequence D, set the value corresponding to the abnormal data index to a null value and fill the null value of the abnormal data using linear interpolation. The resulting sequence is recorded as D3. The step S3 comprises: For sequence D3, data with a temperature lower than a specific heating temperature are screened to obtain sequence segment index S1; The step S4 comprises: Calculate the slope of the 10-minute moving window of sequence D3 as follows: Assume that the variable relationship of the window is y=kx + b, and there are m groups of data with errors in the sequence D3 ( , ) ,in is the actual value of sequence D3, is the window sequence number value, =1,2,3…m; Sequence D3 calculates the slope of the moving window every 10 minutes to obtain the sequence segment index S2. The slope calculation formula is as follows: The data with sequence S2 slope less than 0 and sequence D3 measurement point value greater than 291.5°C were screened to obtain sequence fragment index S3; The step S5 comprises: In sequence D3, create a feature named label and assign it an empty value. For the sequence segment index S1 that satisfies the label value in sequence D3, it is reassigned to 0; The label value of the sequence segment index S3 in sequence D3 is reassigned to 1; Perform a forward difference on the label column and record the data where the forward difference result is 1; Perform reverse difference on the label column and record the data whose reverse difference result is -1; Pair the data whose forward difference result is 1 and the data whose reverse difference result is -1 in the adjacent time in the order of closest proximity. The first pairing label value is -1 and the second pairing label value is 1. The data that does not meet this relationship are deleted. Record each pairing result as , where the data is ( , ), and record them as the initial and end identification points of the heating transient of the production equipment, That is, the transient end time of the heating process of the production equipment; The step S6 comprises: by For the data window, calculate the slope of the moving window every 10 minutes, using the slope calculation method of step S4; The last value that is less than the target device heating temperature and has a slope value less than 0 is recorded as , This is the start time of the heating process of the production equipment.

Images