Abnormality detection method and device for blast furnace thermocouple temperature measurement data and storage medium

Abstract

The application provides a blast furnace thermocouple temperature measurement data anomaly detection method and device and a storage medium. The method comprises the following steps: S1, periodically collecting temperature measurement time series data of a predetermined length for any selected thermocouple buried in a blast furnace body; S2, processing the collected temperature measurement time series data using a sliding window to obtain a stacked temperature measurement data two-dimensional array X; and S3, training the temperature measurement data in the two-dimensional array X using an isolation forest algorithm, wherein each row in the two-dimensional array X is used as a different sample for training the isolation forest, each column is used as a different feature for training the isolation forest, the anomaly scores of the different samples are calculated, the anomaly scores are compared with a predetermined anomaly score threshold, and the abnormal samples are determined according to the comparison result. The above technical solution can quickly find abnormal points in the temperature measurement data online.

Description

Technical Field

[0001] This invention relates to the field of blast furnace ironmaking detection technology, and in particular to a method, device and storage medium for detecting anomalies in blast furnace thermocouple temperature measurement data. Background Technology

[0002] Blast furnaces, as crucial equipment in steel production, incur enormous construction and maintenance costs, making their longevity a top priority for ironmaking professionals. Generally, achieving a long blast furnace lifespan involves two main challenges: the hearth and bottom are prone to burn-through, and the belly, waist, and lower part of the furnace have limited lifespan. To understand the working condition and erosion status of these areas, thermocouples are typically installed in the hearth and bottom lining and furnace cooling equipment to monitor temperature. Especially in the construction of large blast furnaces, hundreds of thermocouple temperature measurement points are buried in the hearth and bottom area; this online temperature measurement method has become the primary means of monitoring and calculating the erosion status of the hearth and bottom.

[0003] However, in production practice, as the blast furnace operates for an extended period, temperature fluctuations become increasingly frequent. Coupled with numerous external environmental factors, abnormal fluctuations in thermocouple readings can occur due to abnormal furnace conditions, instrument malfunctions, or unstable circuitry. Blast furnace operators need to be aware of these anomalies and determine whether they are due to changes in furnace conditions or other factors in order to formulate appropriate operational plans. Because there are numerous temperature measurement points, manual verification by blast furnace operators and maintenance personnel is insufficient for timely and comprehensive observation of these anomalies. Therefore, a detection method is needed to quickly and automatically observe abnormal fluctuations in thermocouple temperature data. Summary of the Invention

[0004] The embodiments of the present invention provide a method, device and storage medium for detecting anomalies in blast furnace thermocouple temperature measurement data, so as to realize the online and rapid detection of anomalies in thermocouple temperature measurement data.

[0005] To achieve the above objectives, a method for detecting anomalies in blast furnace thermocouple temperature measurement data is provided, comprising the following steps:

[0006] S1, For any selected thermocouple embedded in the blast furnace body, temperature measurement time series data are periodically collected for a predetermined duration, wherein the predetermined duration is divided into T time periods according to the collection period, wherein each time period corresponds to one temperature measurement data, and T is a positive integer;

[0007] S2, the collected temperature measurement time series data is processed using a sliding window method to obtain a stacked two-dimensional array X of temperature measurement data, wherein the length of the sliding window is w, w>1, the two-dimensional array X has T-w+1 rows and w columns, wherein each row of the two-dimensional array X contains temperature measurement data of one window length, and different rows contain temperature measurement data of different windows obtained after the window is slid; the columns of the two-dimensional array X contain temperature measurement data at the same position of different windows;

[0008] S3, the isolated forest algorithm is used to train the temperature measurement data in the two-dimensional array X. Each row in the two-dimensional array X is used as a different sample for training the isolated forest, and each column is used as a different feature for training the isolated forest. The abnormal scores of different samples are calculated, and the abnormal scores are compared with a predetermined abnormal score threshold. Abnormal samples are determined based on the comparison results.

[0009] Preferably, in the anomaly detection method, the thermocouples embedded in the blast furnace body are not limited to those embedded in the blast furnace throat, furnace body, furnace waist, furnace belly, furnace hearth, and furnace bottom areas.

[0010] Preferably, in the anomaly detection method, the arbitrarily selected thermocouple is any thermocouple embedded in the blast furnace body.

[0011] Preferably, in the anomaly detection method, determining the abnormal sample based on the comparison result includes:

[0012] Samples with anomaly scores lower than a predetermined anomaly score threshold are identified as anomalous samples.

[0013] The anomaly detection method according to claim 1 is characterized in that the temperature measurement data collected at T time points of a predetermined duration are y1, y2, ... y T Step S2 includes:

[0014] S201, the default temporary array Z is an array of length w;

[0015] S202, using any time t as the left boundary point of the sliding window, where the time range of t is 1,…,T-w+1, performs a cyclic assignment operation according to the following steps:

[0016] Let Z = [y t ,y t+1 ,…,y t+w-1 ], Z is stacked and filled into the two-dimensional array X as the t-th row, and the window slides to the right by one time point, i.e., t = t + 1. The time range of the sliding window is r = (t), ..., (t + w - 1);

[0017] S203, obtain a two-dimensional array X filled with all temperature measurement data, where any row of data represents a segment of temperature measurement data corresponding to any time t, where the time range of t is 1,…,T-w+1.

[0018] On the other hand, an apparatus for detecting anomalies in blast furnace thermocouple temperature measurement data is provided, including a memory and a processor, wherein the memory stores at least one program, which is executed by the processor to implement any of the methods described above.

[0019] In another aspect, a computer-readable storage medium is provided, wherein at least one program is stored therein, the at least one program being executed by a processor to implement any of the methods described above.

[0020] The above technical solution has the following technical effects:

[0021] The technical solution of this invention collects temperature measurement data from each thermocouple for a certain period of time, obtains a stacked data stream of a two-dimensional array using a sliding window, and trains the obtained stacked data stream using the isolated forest algorithm. Each row of the two-dimensional array, i.e., a segment of temperature measurement data corresponding to any time t, is used as a different sample to train the isolated forest algorithm, and each column is used as a different feature to train the isolated forest. This allows for anomaly analysis of the data stream, thereby enabling online and rapid detection of anomalies in the temperature measurement data. The anomaly detection method of this invention can detect anomalies in the time-series data stream of a thermocouple, such as a temperature measurement point. It is applicable to anomaly detection scenarios for thermocouple temperature measurement data, which helps in judging the furnace condition during blast furnace operation. For example, when abnormal furnace conditions lead to abnormalities related to the blast furnace structure itself, such as flow interruption, material wiping, or pipeline issues, it also helps in the maintenance of on-site equipment. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating an anomaly detection method for blast furnace thermocouple temperature measurement data according to an embodiment of the present invention.

[0023] Figure 2 This is a schematic diagram of the structure of an anomaly detection device for blast furnace thermocouple temperature measurement data according to an embodiment of the present invention. Detailed Implementation

[0024] To further illustrate the various embodiments, the present invention provides accompanying drawings. These drawings are part of the disclosure of the present invention, primarily used to illustrate the embodiments and to explain the operating principles of the embodiments in conjunction with the relevant descriptions in the specification. With reference to these drawings, those skilled in the art should be able to understand other possible implementations and the advantages of the present invention. Components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0025] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0026] Example 1:

[0027] Figure 1 This is a flowchart illustrating a method for detecting anomalies in blast furnace thermocouple temperature measurement data according to an embodiment of the present invention. Figure 1 The method for detecting anomalies in blast furnace thermocouple temperature measurement data in this embodiment includes the following steps:

[0028] S1. For any selected thermocouple embedded in the blast furnace body, temperature measurement time series data are periodically collected for a predetermined duration. The predetermined duration is divided into T time periods according to the collection period, i.e. from the 1st to the Tth time period. Each time period corresponds to one temperature measurement data, and T is a pre-set positive integer.

[0029] The main function of this step is to collect data. The thermocouples embedded in the blast furnace body are not limited to those embedded in the throat, body, waist, belly, hearth, or bottom. Any selected thermocouple can be used; that is, temperature data from all thermocouples can be collected. The technical solution of this invention is used to detect whether all thermocouples are abnormal.

[0030] For example, a sampling period of 20 seconds is used, meaning the time interval between every two data points is 20 seconds, and temperature measurement time-series data is collected over a predetermined duration of 30 minutes. In this example, the number of sampling points T = 90 within 30 minutes. The values ​​for the sampling period and predetermined duration here are merely examples and can be adjusted according to actual needs. The temperature measurement data collected from time 1 to T is represented as y. 1…T That is, y1, y2, ... y T Each time t corresponds to a temperature measurement value y. t The time series data length is T.

[0031] S2. The collected temperature measurement time series data is processed using a sliding window method to obtain a stacked two-dimensional array X of temperature measurement data. Let the length of the sliding window be w, where w>1. The two-dimensional array X has T-w+1 rows and w columns. Each row of the two-dimensional array X contains the temperature measurement data at w time points of a window. Different rows contain the temperature measurement data of different windows obtained after the window is slid. The columns of the two-dimensional array X contain the temperature measurement data at the same position in different windows.

[0032] Specifically, the temperature measurement data collected from time 1 to time T is y 1…T For example, the window length w can be set to 10, and step S2 specifically includes:

[0033] S201, the default temporary array Z is an array of length w;

[0034] S202, using any time t as the left boundary point of the sliding window, where the time range of t is 1,…,T–w+1, performs a cyclic assignment operation according to the following steps:

[0035] Let Z = [y t ,y t+1 ,…,y t+w-1 ], Z is stacked and filled into the two-dimensional array X as the t-th row, and the window slides to the right by one time point, i.e., t = t + 1. The time range of the sliding window is r = (t), ..., (t + w - 1);

[0036] S203 yields a two-dimensional array X filled with all temperature measurement data. Each row of data represents a segment of temperature measurement data corresponding to any time t, where the time range of t is 1,…,T–w+1.

[0037] S3. The isolated forest algorithm is used to train the temperature measurement data in the two-dimensional array X. Each row in the two-dimensional array X is used as a different sample for training the isolated forest, and each column is used as a different feature for training the isolated forest. The anomaly score corresponding to different samples is calculated, and the anomaly score is compared with a predetermined anomaly score threshold. The abnormal samples are determined based on the comparison result.

[0038] The isolated forest algorithm used is as follows:

[0039] a) Given an input two-dimensional array X, the number of trees n t Subsampling size r s Initialize an isolated forest

[0040] b) Set the tree height limit to l = ceiling(log2 r) s Here, ceiling() refers to the integer part that is rounded up.

[0041] c) For i = 1…t, perform the following operation in a loop:

[0042] Randomly sample X according to r to obtain subsampled data X';

[0043] Construct an isolated tree iTree(X', 0, 1) and execute iForest → iForest ∪ iTree(X′, 0, 1).

[0044] d) After the loop is complete, the isolated forest iForest is obtained, and the anomaly score of each sample is calculated.

[0045] The steps involved in constructing the isolated iTree in this isolated forest algorithm are as follows:

[0046] a) Given an input two-dimensional array X, the current tree height e, and a tree height limit of 1;

[0047] b) Determine if the condition is true: e≥l or|X|≤1, where |X| represents the sample size of X; if the condition is true, proceed to step c), otherwise proceed to step d);

[0048] c) Let X be the tail node, output the node, and then end;

[0049] d) Let Q be a list representing all features of X. Randomly select a feature q in Q, and take the maximum and minimum values ​​of all sample features q in X as boundaries, and randomly select a value p as the dividing point.

[0050] e) Collect samples in X whose feature q values ​​are less than p into a left set array X. l Collect samples in X whose feature q value is greater than p into a right set array Xr;

[0051] f) Create a node inNode whose left subtree is iTree(X l The right subtree is iTree(X, e+1, l). r The node is defined as q, with a boundary value of p. The node is then output, and the process ends. The two subtrees are established by recursively calling this algorithm.

[0052] The anomaly score for each sample is calculated using the following formula:

[0053]

[0054] In the formula, x represents any sample; n represents the number of samples; Here, H() is the harmonic series, H(n-1) = ln(n-1) + ξ, where ξ is Euler's constant, c(n) represents the average path length of a binary tree constructed from n samples; h(x) = e(x) + c(T.size), where e(x) represents the number of boundaries that sample x experiences on its journey from the root node to a leaf node, T.size represents the number of samples that share a leaf node with sample x; E(h(x)) represents the mean of h(x) for the sample across all trees in the isolated forest.

[0055] The abnormality score obtained by the above formula has a range of -0.5 to 0.5.

[0056] It is important to note that each row in the two-dimensional array, corresponding to a segment of temperature measurement data at any time t, is considered as a different sample in the training isolated forest, while each column is considered as a different feature in the training isolated forest.

[0057] Specifically, for each thermocouple temperature measurement point, a certain anomaly scoring threshold is pre-set based on its location, temperature change characteristics, and other factors. Then, based on this threshold, abnormal samples are identified to obtain anomalies in the temperature measurement data. For example, different thermocouples can have different anomaly scoring thresholds set.

[0058] Preferably, samples in the Isolation Forest algorithm with abnormal scores lower than a pre-set score threshold are identified as abnormal samples. In this example, the abnormal score threshold is set to -0.3, meaning that when the abnormal score of an input sample is less than -0.3, the sample is identified as an abnormal sample, i.e., abnormal temperature measurement data.

[0059] By following the steps above, you can complete the anomaly detection of any blast furnace thermocouple temperature measurement data.

[0060] Example 2:

[0061] The present invention also provides a device for detecting anomalies in blast furnace thermocouple temperature measurement data, such as... Figure 2 As shown, the device includes a processor 201, a memory 202, a bus 203, and a computer program stored in the memory 202 and executable on the processor 201. The processor 201 includes one or more processing cores. The memory 202 is connected to the processor 201 via the bus 203. The memory 202 is used to store program instructions. When the processor executes the computer program, it implements the steps in the above-described method embodiment of Embodiment 1 of the present invention.

[0062] Furthermore, as an executable solution, the device for detecting anomalies in the blast furnace thermocouple temperature measurement data can be a computer unit, which can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer unit may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the above-described computer unit structure is merely an example and does not constitute a limitation on the computer unit. It may include more or fewer components, or combine certain components, or use different components. For example, the computer unit may also include input / output devices, network access devices, buses, etc., and this embodiment of the invention does not limit this.

[0063] Furthermore, as an executable solution, the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc. The processor is the control center of the computer unit, connecting various parts of the entire computer unit via various interfaces and lines.

[0064] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the computer unit by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0065] Example 3:

[0066] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method described in the embodiments of the present invention.

[0067] If the modules / units integrated in the computer unit are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), and software distribution media, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0068] Although the invention has been specifically shown and described in conjunction with preferred embodiments, those skilled in the art should understand that various changes in form and detail may be made to the invention without departing from the spirit and scope of the invention as defined in the appended claims, all of which shall be within the scope of protection of the invention.

Claims

1. A blast furnace thermocouple temperature measurement data anomaly detection method, characterized in that, The method comprises the following steps: S1, periodically collecting temperature time series data of a predetermined length for any selected thermocouple buried in the blast furnace body, wherein the predetermined length is divided into T time points according to the collection period, wherein each time point corresponds to a temperature data, and T is a positive integer; S2, processing the collected temperature time series data using a sliding window method to obtain a stacked temperature data two-dimensional array X, wherein the length of the sliding window is w, w>1, the two-dimensional array X has T-w+1 rows and w columns, wherein each row of the two-dimensional array X contains temperature data of a window length, and different rows contain temperature data of different windows obtained after window sliding; the columns of the two-dimensional array X contain temperature data at the same position of different windows; S3, training the temperature data in the two-dimensional array X using an isolation forest algorithm, wherein each row in the two-dimensional array X is used as a different sample for training the isolation forest, and each column is used as a different feature for training the isolation forest, the abnormal score of different samples is calculated, and the abnormal score is compared with a predetermined abnormal score threshold to determine abnormal samples according to the comparison result.

2. The abnormality detection method according to claim 1, characterized by, The thermocouple buried in the blast furnace body is not limited to the thermocouple buried in the blast furnace throat, body, waist, belly, cylinder and bottom region.

3. The abnormality detection method according to claim 1, characterized by, The selected thermocouple is any thermocouple buried in the blast furnace body.

4. The abnormality detection method according to claim 1, characterized by, According to the comparison result, the abnormal sample is determined. The sample with an abnormal score less than the predetermined abnormal score threshold is determined as an abnormal sample.

5. The anomaly detection method according to claim 1, characterized by, The temperature measurement data collected at T time points of a predetermined time length are y1, y2, … yT T The step S2 comprises: S201, the temporary storage array Z is pre-set as an array with a length of w; S202, taking any time point t as the left boundary point of the sliding window, wherein the time range of t is 1,…,T-w+1, and the following steps are performed in a loop assignment operation: Let Z = [y t ,y t+1 ,…,y t+w-1 ], Z is filled into the two-dimensional array X as the t-th row stack, the window slides right one time, that is, t = t + 1, the time range of the sliding window is r = (t),…,(t+w-1) S203, obtaining a two-dimensional array X filled with all temperature data, wherein any row of data represents a segment of temperature data corresponding to any time point t, and the time range of t is 1,…,T-w+1.

6. A device for detecting abnormality of temperature measurement data of a blast furnace thermocouple, characterized by comprising: The memory stores at least one program, and the at least one program is executed by the processor to realize the method of any one of claims 1 to 5.

7. A computer readable storage medium characterized in that, The storage medium stores at least one program, and the at least one program is executed by the processor to realize the method of any one of claims 1 to 5.

Images