CN-115496142-B - Abnormality detection method and device for blast furnace thermocouple and storage medium

CN115496142BCN 115496142 BCN115496142 BCN 115496142BCN-115496142-B

Abstract

The invention provides an anomaly detection method, device and storage medium for a blast furnace thermocouple, which comprises the steps of S1, selecting a plurality of thermocouples in the blast furnace, which need anomaly detection, periodically collecting temperature measurement data of preset time to obtain a temperature measurement data sequence of each thermocouple temperature measurement point, S2, grouping the plurality of thermocouples according to installation position attributes, respectively packing the obtained temperature measurement data sequences of each thermocouple temperature measurement point according to the grouping to obtain two-dimensional arrays representing temperature measurement data of each group of thermocouple temperature measurement points, S3, training the temperature measurement data by using an isolated forest algorithm to obtain anomaly scores of each temperature measurement point, comparing the anomaly scores with preset anomaly score thresholds, and determining the anomaly temperature measurement points in each group of temperature measurement points according to comparison results. By utilizing the technical scheme, the abnormal temperature measuring point can be quickly found on line.

Inventors

FENG XIAOJUN
YAN HAN
WANG WEI
OU YAN
YE LIDE

Assignees

中冶南方工程技术有限公司

Dates

Publication Date: 20260508
Application Date: 20220921

Claims (9)

1. The abnormality detection method of the blast furnace thermocouple is characterized by comprising the following steps of: S1, selecting a plurality of thermocouples in a blast furnace, which need to be subjected to abnormality detection, and periodically collecting temperature measurement data of a preset time length for temperature measurement points of the thermocouples to obtain a temperature measurement data sequence of each thermocouple temperature measurement point; S2, grouping the selected thermocouples according to the installation position attribute of the thermocouples, and respectively packing the obtained temperature measurement data sequences of the thermocouple temperature measurement points according to the grouping to obtain one or more two-dimensional arrays representing the temperature measurement data of each group of thermocouple temperature measurement points, wherein one two-dimensional array corresponds to one group of thermocouple temperature measurement points, the row of each two-dimensional array contains the temperature measurement data of each thermocouple temperature measurement point of the corresponding group at the same moment, and the column of each two-dimensional array contains the temperature measurement data of the same thermocouple temperature measurement point at different moments; And S3, training the temperature measurement data in the two-dimensional arrays by using an isolated forest algorithm for each two-dimensional array, wherein the temperature measurement data of each column in the two-dimensional array is used as different samples in the training independent forest, the temperature measurement data of each row is used as different features in the training independent forest, the abnormal score of each temperature measurement point is obtained, the abnormal score is compared with a preset abnormal score threshold, and the abnormal temperature measurement points in each group of temperature measurement points are determined according to the comparison result.
2. The abnormality detection method according to claim 1, wherein the selected thermocouple is not limited to a thermocouple buried in a blast furnace throat, furnace body, furnace waist, furnace belly, furnace hearth, furnace bottom region.
3. The abnormality detection method according to claim 1, wherein the grouping according to the mounting position attribute of the thermocouples includes grouping all of the plurality of thermocouples according to a height, an angle, and/or an insertion depth at which the thermocouples are located.
4. The abnormality detection method according to claim 3, wherein the grouping according to the mounting position attribute of the thermocouples includes grouping thermocouples having the same height and the same insertion depth into groups, and the temperature measuring points in each group are different only in angle.
5. The abnormality detection method according to claim 1, wherein the number of thermocouple temperature measurement points in each group is 3 or more, and thermocouple temperature measurement points that cannot be grouped are discarded.
6. The abnormality detection method according to claim 1, wherein the step S2 of packing the obtained temperature measurement data sequences of the respective thermocouple temperature measurement points in groups includes: And sequentially stacking the thermocouple temperature measurement data sequences in the same group according to the time sequence to form a two-dimensional array, wherein the two-dimensional array comprises temperature measurement data sequences of all thermocouple temperature measurement points in the corresponding group.
7. The abnormality detection method according to claim 1, characterized in that the selected thermocouples are all thermocouples in a selected area, the number of the selected thermocouples being 3 or more, wherein the thermocouples in the selected area include: A plurality of thermocouples which are arranged in the cooling wall of the blast furnace body and have the same height, the same insertion depth and different angles; All furnace hearth temperature thermocouples; all thermocouples of 5-13 layers of side walls of the hearth; a plurality of thermocouples with the same height, the same insertion depth and different angles in all thermocouples of 5-13 layers of side walls of the hearth, or All furnace body temperature thermocouples.
8. An apparatus for anomaly detection of blast furnace thermocouple temperature measurement data, comprising a memory and a processor, the memory storing at least one program, the at least one program being executed by the processor to implement the method of any one of claims 1 to 7.
9. A computer readable storage medium, characterized in that at least one program is stored in the storage medium, the at least one program being executed by a processor to implement the method of any one of claims 1 to 7.

Description

Abnormality detection method and device for blast furnace thermocouple and storage medium Technical Field The invention relates to the technical field of blast furnace ironmaking detection, in particular to an abnormality detection method and device for a blast furnace thermocouple and a storage medium. Background The blast furnace has huge construction and maintenance costs as an important equipment facility in steel production, so the problem of long service life of the blast furnace is always highly valued by ironmaking workers. The long life of blast furnace mainly faces two problems, one is that the hearth and the bottom of the furnace are easy to burn through, and the other is that the life of the furnace belly and the lower part of the furnace body is limited. In order to know the working state and erosion condition of the parts, thermocouples are generally arranged in a hearth bottom brick lining of a hearth and a furnace body cooling device to detect the temperature. Particularly in the construction of large blast furnaces, hundreds of thermocouple temperature measuring points are buried in the hearth and bottom area of the hearth, and the online temperature measuring mode has become a main means for monitoring and calculating the erosion state of the hearth and the bottom of the hearth. However, because of the large number of thermocouples buried from the furnace body to the furnace bottom of the blast furnace, complicated installation of instruments, circuits and the like, the conditions of incorrect wiring, unstable wiring and the like of measuring points can often occur on each blast furnace site, and the blast furnace has long running time, large temperature change and various interference factors in the production of the blast furnace, so that the thermocouples can be broken, fused and the like in use. Some thermocouples fail to measure temperature for a number of reasons, which is detrimental to the actual operation of the blast furnace. Because of too many temperature measuring points, the condition of detecting or observing all thermocouples in the maintenance of the blast furnace is quite labor-consuming. Therefore, a detection means is needed to quickly and automatically find abnormal thermocouple temperature measurement points. Disclosure of Invention The embodiment of the invention provides an abnormality detection method and device for a blast furnace thermocouple and a storage medium, so as to realize quick detection of abnormal thermocouple temperature measuring points. In order to achieve the above object, in one aspect, there is provided an abnormality detection method of a blast furnace thermocouple, comprising the steps of: S1, selecting a plurality of thermocouples in a blast furnace, which need to be subjected to abnormality detection, and periodically collecting temperature measurement data of a preset time length for temperature measurement points of the thermocouples to obtain a temperature measurement data sequence of each thermocouple temperature measurement point; S2, grouping the selected thermocouples according to the installation position attribute of the thermocouples, and respectively packing the obtained temperature measurement data sequences of the thermocouple temperature measurement points according to the grouping to obtain one or more two-dimensional arrays representing the temperature measurement data of each group of thermocouple temperature measurement points, wherein one two-dimensional array corresponds to one group of thermocouple temperature measurement points, the row of each two-dimensional array contains the temperature measurement data of each thermocouple temperature measurement point of the corresponding group at the same moment, and the column of each two-dimensional array contains the temperature measurement data of the same thermocouple temperature measurement point at different moments; And S3, training the temperature measurement data in the two-dimensional arrays by using an isolated forest algorithm for each two-dimensional array, wherein the temperature measurement data of each column in the two-dimensional array is used as different samples in the training independent forest, the temperature measurement data of each row is used as different features in the training independent forest, the abnormal score of each temperature measurement point is obtained, the abnormal score is compared with a preset abnormal score threshold, and the abnormal temperature measurement points in each group of temperature measurement points are determined according to the comparison result. Preferably, in the abnormality detection method, the selected thermocouple is not limited to a thermocouple buried in a throat, a shaft, a waist, a belly, a hearth, or a hearth region of the blast furnace. Preferably, the abnormality detection method, wherein grouping according to the installation position attribute of the thermocouples comprises grouping all of the plurality