Chemical production online fault detection method and system based on big data

Abstract

The invention relates to the technical field of fault detection, in particular to a chemical production online fault detection method and system based on big data, and the method comprises the steps: collecting operation data of production equipment, constructing real-time data matrixes with different sizes, intercepting historical data with the same size as the real-time data matrixes, and constructing a historical data matrix; and calculating the similarity between any real-time data matrix and each historical data matrix with the same size by using a sliding window. According to the method, the real-time data matrixes of different sizes are constructed to analyze the possible long and short term trends of the data, the similarity between the real-time data matrixes and the historical data matrixes is calculated by moving the sliding window, and the estimated value of the comprehensive control limit value of the real-time data is estimated through the control limit values of the similar historical data matrixes. And the fault condition of the equipment can be accurately judged according to the difference between the estimated value and the comprehensive control limit value of the real-time data.

Description

Technical Field

[0001] The present invention relates to the field of fault detection technology, and more specifically, to a chemical production online fault detection method and system based on big data. Background Art

[0002] There are multiple links and equipment designed in the chemical production process, such as reactors, pumps, valves, etc. The operating status of each link needs to be monitored in real time to detect potential faults in time. Traditional chemical production fault detection methods rely on the experience and subjective judgment of operators. When performing fault detection, they cannot detect the occurrence of faults in time, leading to danger. With the continuous development of big data technology, technicians use statistical models, machine learning models, etc. with physical models to build hybrid models for online fault detection in chemical production.

[0003] At present, the dynamic characteristics of chemical production processes are usually nonlinear and time-varying. There are complex nonlinear relationships between parameters such as temperature, pressure, and flow. The use of traditional PCA statistics, linear regression and other methods cannot capture the complex relationship of multiple variables in time, resulting in low accuracy in equipment fault detection. Therefore, the use of fault detection technology for online fault detection in chemical production is the problem solved by this patent. Summary of the invention

[0004] The present invention provides a chemical production online fault detection method and system based on big data, aiming to solve the problem that traditional PCA statistics, linear regression and other methods used in related technologies cannot capture the complex temporal relationship of multiple variables, resulting in low accuracy in equipment fault detection.

[0005] In a first aspect, the present invention provides a chemical production online fault detection method based on big data, comprising: collecting operating data of production equipment, constructing real-time data matrices of different sizes, and intercepting historical data of the same size as the real-time data matrix to construct a historical data matrix; using a sliding window to calculate the similarity between any real-time data matrix and each historical data matrix of the same size, the similarity is used to characterize the degree of difference between each data in the sliding window when the real-time data matrix and the historical data matrix are traversed using the sliding window, and screening out the historical data matrix with the greatest similarity to the real-time data matrix as the most similar matrix, and calculating the comprehensive control limit of the most similar matrix, the calculation formula is: ; In the formula, is the most similar matrix The comprehensive control limit of is the most similar matrix The Hotelling statistic control limits are is the most similar matrix The squared prediction error control limits, is the most similar matrix With the standard F distribution data matrix Statistics, Represents the size of the matrix; the average value of the product of the comprehensive control limit of the most similar matrices of different sizes and the similarity with the real-time data matrix is ​​used as the estimated value of the comprehensive control limit of the real-time data; based on the comparison result between the comprehensive control limit of the real-time data and the estimated value, fault detection is performed on the production equipment.

[0006] Furthermore, fault detection is performed on the production equipment, including: if the difference between the comprehensive control limit of the real-time data and the estimated value is greater than the alarm threshold, it is determined that a chemical production failure has occurred and an alarm is issued; if the difference between the comprehensive control limit of the real-time data and the estimated value is less than or equal to the alarm threshold, it is determined that a production equipment failure has occurred and an alarm is issued.

[0007] Furthermore, the estimated value of the comprehensive control limit of real-time data is calculated using the following formula: ; In the formula, is an estimate of the comprehensive control limit of real-time data, Represents real-time data matrix The most similar matrix The similarity of Represents the most similar matrix The comprehensive control limit of Indicates the number of real-time data matrices of different sizes.

[0008] Furthermore, the similarity between any real-time data matrix and its historical data matrices of the same size is calculated using the following formula: ; In the formula, Represents real-time data matrix With historical data matrix The similarity of represents the normalization function, Indicates the number of sliding windows. Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Representation Matrix In the The first sliding window Line The characteristics of the columns, Representation Matrix In the The first sliding window Line The features of the column, the size of the sliding window is .

[0009] Furthermore, real-time data matrices of different sizes are constructed, including: firstly centralizing the collected operation data of the production equipment, wherein the operation data includes temperature, pressure, flow, liquid level, vibration and current.

[0010] Furthermore, any type of data in the operating data is centrally processed, including: for any type of data, any data in the type minus the mean of the data in the type is used as the value after centralization.

[0011] Furthermore, the empirical value of the alarm threshold is 1.

[0012] In the second aspect, a chemical production online fault detection system based on big data is also provided, including a processor and a memory, the memory stores a computer program, and the processor executes the computer program to implement any of the above-mentioned chemical production online fault detection methods based on big data.

[0013] Beneficial effects: By constructing real-time data matrices of different sizes to analyze the possible long-term and short-term trends of the data, the similarity between the real-time data matrix and the historical data matrix is ​​calculated through a moving sliding window. The estimated value of the comprehensive control limit of the real-time data is estimated through the control limit of the similar historical data matrix. Based on the difference between the estimated value and the comprehensive control limit of the real-time data, the fault condition of the equipment can be accurately judged. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The following detailed description is read with reference to the accompanying drawings, which illustrate several embodiments of the present invention in an exemplary and non-limiting manner, and in which like or corresponding reference numerals represent like or corresponding parts, wherein: Figure 1 Detailed description is a flow chart schematically illustrating a method of calculating an estimated value of a comprehensive control limit of real-time data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0016] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0017] like Figure 1As shown, S101: collecting operation data of production equipment.

[0018] In one embodiment, sensors are installed at key nodes of chemical production equipment, such as reactors, pumps, valves, pipelines, etc., and corresponding sensors are used to collect temperature, pressure, flow, liquid level, vibration and current of different equipment respectively. The collected data is then transmitted to a central data platform through a real-time data acquisition system for storage and processing.

[0019] In one embodiment, the collected operating data of the production equipment is also centrally processed. Specifically, for any type of data, any data in the type minus the mean of the data in the type is used as the value after central processing. , , , ,right The value after centralization is At this point, the processed values ​​in all running data sets can be obtained.

[0020] In one embodiment, the statistics in the prior art The control limits are calculated under the premise that normal data obeys the multivariate normal distribution and are fixed values. However, in the actual data distribution, due to the influence of environmental noise, the data does not conform to the multivariate normal distribution and there is a periodic trend, which leads to control limit errors, resulting in untimely identification of production equipment failures or false warnings. Therefore, this embodiment constructs real-time data matrices of different sizes, analyzes the possible long-term and short-term trends of the data, calculates the similarity between the real-time data matrix and the historical data matrix through a moving sliding window, predicts the estimated value of the comprehensive control limit of the real-time data through the control limit values ​​of the similar historical data matrices, and accurately judges whether the production equipment has a failure based on the difference between the comprehensive value of the real-time data and the estimated value.

[0021] S102: Calculate the similarity between any real-time data matrix and each historical data matrix of the same size.

[0022] In one embodiment, after obtaining the processed values ​​in all running data sets, real-time data matrices of different sizes are constructed. For example, matrices of different sizes are constructed as follows: ,common Real-time data matrices of different sizes, The maximum value of is the length of time from the start of data collection to the current time. An exemplary matrix size is Size ,matrix Including the past From time to the present time Then, the historical data of the same size as the real-time data matrix is ​​intercepted to construct the historical data matrix. The historical data refers to the operating data of the production equipment before this time.

[0023] In one embodiment, a sliding window is used to calculate the similarity between any real-time data matrix and each historical data matrix of the same size. The similarity is used to characterize the degree of difference between each data in the sliding window when the sliding window is used to traverse the real-time data matrix and the historical data matrix.

[0024] In one embodiment, the calculation formula for the similarity between any real-time data matrix and each historical data matrix of the same size is: ; In the formula, Represents real-time data matrix With historical data matrix The similarity of represents the normalization function, Indicates the number of sliding windows. Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Representation Matrix In the The first sliding window Line The characteristics of the columns, Representation Matrix In the The first sliding window Line The features of the column, the size of the sliding window is .

[0025] S103: Filter out the historical data matrix with the greatest similarity to the real-time data matrix as the most similar matrix.

[0026] In one embodiment, for a real-time data matrix of any size, the historical data matrix with the greatest similarity is selected as the most similar matrix to the real-time data matrix. Most similar matrix .

[0027] S104: Calculate the comprehensive control limit of the most similar matrix.

[0028] In one embodiment, according to the most similar matrix , and Statistics, to calculate the comprehensive control limits The calculation formula is: ; In the formula, is the most similar matrix The comprehensive control limit of is the most similar matrix The Hotelling statistic control limits are is the most similar matrix The squared prediction error control limits, is the most similar matrix With the standard F distribution data matrix Statistics, Represents the matrix size. Calculate the most similar matrix , and Statistics belong to the existing technology and will not be described in detail again.

[0029] In one embodiment, due to the traditional fault identification, based on the assumption that the data satisfies the F distribution, only or Whether the corresponding control limit is exceeded is used to determine whether a fault has occurred. Data affected by noise will affect the distribution of data, resulting in a single statistical indicator being unable to determine whether a fault has occurred. Test the difference between the calculated most similar matrix and the standard F-distributed data matrix Statistics, through , and Statistics Calculation Comprehensive Control Limits , estimate the real-time comprehensive control limit through the historical comprehensive control limit, that is, obtain the estimated value of the comprehensive control limit of the real-time data.

[0030] S105: Calculate an estimated value of the comprehensive control limit of the real-time data.

[0031] In one embodiment, the average value of the product of the comprehensive control limit of the most similar matrices of different sizes and the similarity with the real-time data matrix is ​​used as the estimated value of the comprehensive control limit of the real-time data. The calculation formula of the estimated value is: ; In the formula, is an estimate of the comprehensive control limit of real-time data, Represents real-time data matrix The most similar matrix The similarity of Represents the most similar matrix The comprehensive control limit of Indicates the number of real-time data matrices of different sizes.

[0032] S106: Perform fault detection on production equipment based on the comparison result of the comprehensive control limit of the real-time data and the estimated value.

[0033] In one embodiment, if the difference between the comprehensive control limit of the real-time data and the estimated value is greater than the alarm threshold, it is determined that the chemical production has failed and an alarm is issued; if the difference between the comprehensive control limit of the real-time data and the estimated value is less than or equal to the alarm threshold, it is determined that the production equipment has failed and an alarm is issued. In this embodiment, the empirical value of the alarm threshold is 1, and in other embodiments, the empirical value of the alarm threshold is 1.1 or 1.2, etc., which can be adjusted according to specific implementation conditions.

[0034] By constructing real-time data matrices of different sizes to analyze the possible long-term and short-term trends of the data, the similarity between the real-time data matrix and the historical data matrix is ​​calculated through a moving sliding window. The estimated value of the comprehensive control limit of the real-time data is estimated through the control limit of the similar historical data matrix. Based on the difference between the estimated value and the comprehensive control limit of the real-time data, the fault condition of the equipment can be accurately judged.

[0035] The present invention also provides a chemical production online fault detection system based on big data. The system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a chemical production online fault detection method based on big data according to the first aspect of the present invention is implemented.

[0036] The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface, whose configuration and functions are known in the art and thus will not be described in detail here.

[0037] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions stored or otherwise maintained in such a computer-readable medium.

[0038] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0039] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0040] The above-mentioned embodiments only express several implementation methods of the present invention, and the description is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent application. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A chemical production online fault detection method based on big data, characterized in that: include: Collect the operating data of production equipment, build real-time data matrices of different sizes, and intercept historical data of the same size as the real-time data matrix to build a historical data matrix; The sliding window is used to calculate the similarity between any real-time data matrix and each historical data matrix of the same size. The similarity is used to characterize the degree of difference between the data in the sliding window when the real-time data matrix and the historical data matrix are traversed by the sliding window, and the historical data matrix with the greatest similarity to the real-time data matrix is ​​selected as the most similar matrix, and the comprehensive control limit of the most similar matrix is ​​calculated. The calculation formula is: ; In the formula, is the most similar matrix The comprehensive control limit of is the most similar matrix The Hotelling statistic control limits are is the most similar matrix The squared prediction error control limits, is the most similar matrix With the standard F distribution data matrix Statistics, Indicates the matrix size; The average value of the product of the comprehensive control limit of the most similar matrices of different sizes and the similarity with the real-time data matrix is ​​used as the estimated value of the comprehensive control limit of the real-time data. Based on the comparison result between the comprehensive control limit of the real-time data and the estimated value, fault detection is performed on the production equipment.

2. The method for online fault detection in chemical production based on big data according to claim 1 is characterized in that: Perform fault detection on production equipment, including: If the difference between the comprehensive control limit value of the real-time data and the estimated value is greater than the alarm threshold, it is determined that a chemical production failure has occurred and an alarm is issued; If the difference between the comprehensive control limit of the real-time data and the estimated value is less than or equal to the alarm threshold, it is determined that the production equipment has failed and an alarm is issued.

3. The method for online fault detection in chemical production based on big data according to claim 1 is characterized in that: Calculate the estimated value of the comprehensive control limit of real-time data using the following formula: ； In the formula, is an estimate of the comprehensive control limit of real-time data, Represents real-time data matrix The most similar matrix The similarity of Represents the most similar matrix The comprehensive control limit of Indicates the number of real-time data matrices of different sizes.

4. The method for online fault detection in chemical production based on big data according to claim 1, characterized in that: The similarity between any real-time data matrix and its historical data matrices of the same size is calculated using the following formula: ； In the formula, Represents real-time data matrix With historical data matrix The similarity of represents the normalization function, Indicates the number of sliding windows. Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Represents the sliding window in the matrix Previous The characteristic standard deviation of the sliding Representation Matrix In the The first sliding window Line The characteristics of the columns, Representation Matrix In the The first sliding window Line The features of the column, the size of the sliding window is .

5. The method for online fault detection in chemical production based on big data according to claim 1, characterized in that: Build real-time data matrices of varying sizes, including: The collected operating data of the production equipment is first centrally processed, wherein the operating data includes temperature, pressure, flow, liquid level, vibration and current.

6. The method for online fault detection in chemical production based on big data according to claim 5 is characterized in that: Centralized processing of any type of the operating data includes: For any type of data, the value after centralization is obtained by subtracting the mean of any data in that type of data from that type of data.

7. The method for online fault detection in chemical production based on big data according to claim 2 is characterized in that: The empirical value of the alarm threshold is 1.

8. A chemical production online fault detection system based on big data, comprising a processor and a memory, characterized in that: The memory stores a computer program, and the processor executes the computer program to implement the chemical production online fault detection method based on big data as described in any one of claims 1 to 7.

Images