CN-122020494-A - Petrochemical equipment operation abnormality early warning system

CN122020494ACN 122020494 ACN122020494 ACN 122020494ACN-122020494-A

Abstract

The invention discloses an abnormal operation early warning system of petrochemical equipment, which relates to the technical field of abnormal equipment early warning, and comprises a matching module and an early warning module, wherein the matching module calls a deformation abnormal rule to carry out first matching on deformation parameters, if a first matching result is abnormal, then calls an image abnormal rule to carry out second matching on inner wall image data, if the first matching result is abnormal, only the first matching result is output, the early warning module builds a dynamic mapping relation based on the first matching result and the second matching result, a health degree evaluation value is obtained through coupling analysis, a hierarchical abnormal early warning signal is generated according to the health degree evaluation value, early warning operation of corresponding grades is carried out, the single parameter detection one-sidedness is solved through multi-dimensional collaborative matching of deformation and scaling images, the comprehensive identification of equipment abnormality is realized, a coupling degree function and a dynamic health degree evaluation system are built based on an entropy weight method, and the equipment state is accurately quantized by combining correction and attenuation coefficients, and the dynamic property and accuracy are improved.

Inventors

CHEN NAN
ZHANG XI
ZHAO YUNLONG
REN YUFENG
SU WENJING
HUANG YAN

Assignees

博识峰云(湖南)信息技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (10)

1. The petrochemical equipment operation abnormality early warning system is characterized by comprising a matching module and an early warning module; The matching module is used for calling a preset deformation abnormal rule to perform first matching on deformation parameters, calling a preset image abnormal rule to perform second matching on the inner wall image data if the first matching result is abnormal, and outputting only the first matching result if the first matching result is non-abnormal; the early warning module constructs a dynamic mapping relation based on the first matching result and the second matching result, obtains a health degree evaluation value through coupling analysis, generates a grading abnormal early warning signal according to the health degree evaluation value, and executes early warning operation of corresponding grades.
2. The petrochemical equipment operation abnormality early warning system according to claim 1, wherein the deformation abnormality rule is constructed based on historical deformation parameters of the reaction kettle, scaling parameters are extracted based on historical inner wall image data of the reaction kettle, and the image abnormality rule is constructed according to the scaling parameters; The configuration logic of the deformation abnormal rule comprises the steps of substituting the historical deformation parameters and the preset reference deformation parameters into a cosine similarity formula for calculation to obtain a similarity value; screening out abnormal deformation parameters according to the similarity value, wherein the screened logic is that the historical deformation parameters with the similarity value lower than a preset threshold value are judged to be the abnormal deformation parameters; Setting a first rule threshold according to the abnormal deformation parameters, and constructing a deformation abnormal rule based on the first rule threshold; The configuration logic of the image anomaly rule is that scaling parameters are extracted based on historical inner wall image data, the image anomaly rule is constructed based on the scaling parameters, the historical inner wall image data is segmented by adopting a self-adaptive threshold segmentation algorithm, a scaling region and an inner wall background region are obtained, and contour features of the scaling region are extracted by a Canny edge detection algorithm; calculating the scale thickness, scale area occupation ratio and scale distribution density by combining the profile features and the image pixel scale; the scaling parameters are configured as single-dimensional or multi-dimensional combinations; Obtaining a first abnormal characteristic quantity according to a numerical interval of the scaling thickness and the scaling area occupation ratio, wherein the first abnormal characteristic quantity comprises scaling thickness exceeding coefficient and scaling area occupation ratio exceeding value; Obtaining a second abnormal characteristic quantity according to the scale distribution density in the scale parameter, wherein the second abnormal characteristic quantity is the duty ratio of a scale dense area and scale non-uniformity; And combining the first abnormal characteristic quantity and the second abnormal characteristic quantity to obtain a judging characteristic quantity, converting the judging characteristic quantity into a second rule threshold value, and constructing an image abnormal rule after setting according to the scaling parameter safety standard.
3. The petrochemical equipment operation abnormity early warning system of claim 2, wherein the deformation abnormity rule is called to perform first matching on deformation parameters, if the first matching result is abnormal, the image abnormity rule is called to perform second matching on the inner wall image data to obtain a second matching result, the method specifically comprises the steps of obtaining deformation parameters of a reaction kettle, calling a preset deformation abnormity rule, comparing the deformation parameters with a first rule threshold in the deformation abnormity rule, and if the deformation parameters exceed the first rule threshold, judging that the first matching result is abnormal, and outputting a first matching result comprising the deformation parameters, a first threshold comparison difference value and an abnormity identifier; The first threshold comparison difference is a difference value of deformation parameters exceeding a first rule threshold; Synchronously acquiring inner wall image data of a reaction kettle, calling a preset image anomaly rule, extracting scaling parameters from the inner wall image data based on the image anomaly rule, and comparing the scaling parameters with a second rule threshold value in the image anomaly rule; If the scaling parameter exceeds the second rule threshold, judging that the second matching result is abnormal, and outputting a second matching result comprising the scaling parameter, a second threshold comparison difference value and an abnormal identifier; The second threshold comparison difference is a difference value of the scaling parameter exceeding a second rule threshold; If the scaling parameter does not exceed the second rule threshold, judging that the second matching result is abnormal-free, and outputting the second matching result comprising the scaling parameter and the abnormal-free mark.
4. The petrochemical equipment operation abnormality early-warning system according to claim 3, wherein if the first matching result is abnormal-free, the call of the abnormality rule is terminated, and only the first matching result is output, specifically comprising, if the deformation parameters do not exceed the first rule threshold, judging that the first matching result is abnormal-free, directly terminating the call flow of the image abnormal rule, and only outputting the first matching result comprising the deformation parameters and the abnormal-free mark.
5. The petrochemical equipment operation anomaly early warning system according to claim 4, wherein the dynamic mapping relationship comprises a first dynamic mapping relationship and a second dynamic mapping relationship; The first dynamic mapping relation construction logic is used for constructing a first dynamic mapping relation according to a first mapping dimension and a second mapping dimension by taking deformation parameters, a first threshold comparison difference value and an abnormal mark in the first matching result as a first mapping dimension and taking scaling parameters, a second threshold comparison difference value and an abnormal mark in the second matching result as a second mapping dimension when the matching result comprises the first matching result and the second matching result; The deformation abnormal grades are first divided based on the first threshold comparison difference value to obtain different deformation normal grades, the first division logic is that the first threshold comparison difference value corresponds to the deformation mild abnormal grade in the left-closed right-open interval from 0 to A, the deformation moderate abnormal grades are corresponding to the left-closing and right-opening sections of A to B, the deformation severe abnormal grades are corresponding to the left-closing and right-opening sections of B to + -infinity, A, B are preset positive numbers, and A is smaller than B; Performing second division on the scale abnormality grades based on the second threshold comparison difference value to obtain different scale abnormality grades, wherein the second division logic is that the second threshold comparison difference value is between 0 and C, the scale abnormality grades correspond to the scale light abnormality grades, the scale formation moderate abnormal grade is corresponding to the left-closing and right-opening section from C to D, the scale formation severe abnormal grade is corresponding to the left-closing and right-opening section from D to + -infinity, C, D is a preset positive number, and C is smaller than D; The first dynamic mapping relation comprises a dynamic mapping relation of deformation abnormal grades and scaling abnormal grades, and specifically comprises a deformation mild abnormal grade corresponding to scaling mild abnormal grades, a deformation moderate abnormal grade corresponding to scaling moderate abnormal grades, and a deformation severe abnormal grade corresponding to scaling severe abnormal grades.
6. The petrochemical equipment operation abnormality early warning system according to claim 5, wherein the second dynamic mapping relation construction logic is configured to set a second matching result corresponding to the non-invoked image abnormality rule as a default result when the matching result includes only the first matching result; The default result comprises scaling parameters of 0, abnormal marks of no abnormality and a second threshold comparison difference value of 0, deformation parameters, the first threshold comparison difference value and abnormal marks in the first matching result are used as main dimensions, and the default result is used as auxiliary dimensions to construct a second dynamic mapping relation; The first dynamic mapping relation and the second dynamic mapping relation take the abnormal dimension as a classification reference, and parameter values of the corresponding dimension are associated to form structural association, so that standardized data input is provided for subsequent coupling analysis; the abnormal dimension comprises a deformation dimension and a scaling dimension; The parameter values comprise deformation parameters, scaling parameters, a first threshold comparison difference value and a second threshold comparison difference value, and meanwhile, the abnormal identifications of the corresponding dimensions are bound.
7. The petrochemical equipment abnormal operation early warning system according to claim 6, wherein the coupling analysis comprises constructing a efficacy function and a coupling degree function; the efficacy function construction logic is that when the deformation parameter has positive efficacy, the expression of the deformation efficacy function is: ; Wherein, the Representing the deformation efficacy function value corresponding to the j-th deformation parameter, Representing the measured value of the j-th dimension of the deformation parameter, 、 Respectively representing the lower limit value and the upper limit value of the risk of the j-th dimension of the deformation parameter, The scaling parameter is a preset constant and is used for uniformly quantifying an efficacy function reference value when the deformation parameter and the scaling parameter reach health targets; When the deformation parameter has negative efficacy, the expression of the deformation efficacy function is: ; When the fouling parameters have positive efficacy, the expression of the fouling efficacy function is: ; Wherein, the Represents the scaling efficacy function value corresponding to the kth scaling parameter, An actual measurement representing the kth dimension of the fouling parameter, 、 Respectively representing a lower risk limit value and an upper risk limit value of a K dimension of the scaling parameter; when the fouling parameters have negative efficacy, the expression of the fouling efficacy function is: 。
8. The petrochemical equipment operation anomaly early warning system according to claim 7, wherein the coupling degree function construction logic is that the coupling degree function expression of the deformation parameter is: ; Wherein, for the first dynamic mapping relation, m represents m deformation parameter dimensions, Representing the contribution of the deformation parameters to the overall health by linear weighting, A first reference weight representing a j-th dimension of the corresponding deformation parameter, the first reference weight being determined from the historical fault data by an entropy weight method and satisfying: ; The coupling degree function expression of the scaling parameters is as follows: ; Wherein, for the first dynamic mapping relationship, n represents n scaling parameter dimensions, Indicating that the fouling parameters are linearly weighted to give a contribution to overall health, A second reference weight for the kth dimension of the corresponding scaling parameter, the second reference weight being determined from the historical fault data by an entropy weighting method and satisfying: ; When (when) And When any value of the deformation parameter is not 0, the coupling degree function expression of the deformation parameter and the scaling parameter is as follows: ; wherein the coupling degree C is a closed interval of 0 to 1; When (when) And When the values of the deformation parameter and the scaling parameter are all 0, the coupling degree function of the deformation parameter and the scaling parameter is 0.
9. The petrochemical equipment operation anomaly early warning system according to claim 8, wherein the health degree evaluation value calculation logic is that the expression of the health degree evaluation value is: ; based on the deformation abnormal grade and the scaling abnormal grade in the first dynamic mapping relation, the method is used for 、 Respectively giving dynamic correction coefficients 、 ; Every time the deformation abnormal level is increased by one level, An increase of 0.1, said A closed interval of 1 to 1.3; every time the scale abnormality level rises by one level, An increase of 0.1, said A closed interval of 1 to 1.3; Wherein T is the run time attenuation coefficient of the reaction kettle, and the expression of T is: ; wherein T is a double closed interval of 0.8 to 1; For the second dynamic mapping relation, the actual measurement value of the scaling parameter is set as a reference value of 0, and the weight of each dimension of the scaling parameter is set Setting a corresponding second reference weight, wherein the weight value is 1 when the scaling parameter is in a single dimension, and calculating according to the health evaluation value calculation logic to obtain H; the lower the value of the health degree evaluation value H is, the worse the operation state of the reaction kettle is.
10. The petrochemical equipment operation abnormality early warning system according to claim 9, wherein the health degree evaluation value H is divided into a first section, a second section, a third section, and a fourth section; the first interval is a first-level abnormal early warning level, wherein the H value belongs to a first-value-to-second-value double-closed interval; the second interval is a second-level abnormal early warning level corresponding to the H value belonging to the second-value-to-third-value double-closed interval; the third interval is a third-value-to-fourth-value double-closed interval in which the H value belongs to, and corresponds to three-level abnormal early warning grades; The fourth interval is a double-closed interval from the fourth value to the fifth value, and corresponds to a four-level abnormal early warning level; If the first-level early warning level is adopted, a red audible and visual alarm signal is generated, an emergency shutdown preparation program of the reaction kettle is triggered in advance, and meanwhile, an emergency maintenance short message instruction with fault positioning and trend prejudging is sent to operation and maintenance personnel; if the system is in the secondary early warning level, generating a yellow audible and visual alarm signal, triggering a start flow of a load reduction plan of the reaction kettle in advance, and sending an overhaul reminding short message instruction with abnormal parameter tracing and development trend analysis to operation and maintenance personnel; If the early warning level is three-level, generating a blue acousto-optic prompt signal, pushing an abnormal early warning mail and a planned inspection suggestion to operation and maintenance personnel in advance, prompting to pay attention to the operation state of the reaction kettle and starting preventive data monitoring; If the abnormal grade is not generated, an abnormal early warning signal is not generated, and only the health degree data of the reaction kettle are recorded for historical analysis and advanced optimization iteration of an early warning model.

Description

Petrochemical equipment operation abnormality early warning system Technical Field The invention relates to the technical field of equipment abnormality early warning, in particular to a petrochemical equipment operation abnormality early warning system. Background In recent years, petrochemical equipment, such as a reaction kettle, is used as a core device of a process industry, the stability of the running state of the petrochemical equipment is directly related to production safety, process efficiency and equipment service life, along with the acceleration of the industrial intelligent progress, the abnormal early warning technology of the reaction kettle equipment becomes a key support for guaranteeing continuous and reliable running of petrochemical production, and unplanned shutdown, equipment failure and even safety accidents can be effectively avoided by identifying the potential abnormality of the equipment in advance, so that the petrochemical equipment has irreplaceable significance for improving the reliability of the whole life cycle of the equipment and reducing the operation and maintenance cost. However, the conventional equipment abnormality early warning technology has a plurality of limitations, on one hand, multi-focus single-dimension detection lacks collaborative matching and dynamic association of various abnormal parameters of deformation parameters and scaling parameters, so that abnormality identification unilateral performance is easy to cause, on the other hand, even though multi-parameter fusion is involved, a simple threshold superposition or linear association mode is often adopted, a dynamic mapping relation and coupling analysis are not constructed by combining historical fault data, the whole health state of equipment is difficult to accurately quantify, in addition, early warning classification is dependent on empirical threshold division, the factors of parameter coupling degree and the running time attenuation of the reaction kettle equipment are not comprehensively considered, the early warning accuracy and timeliness are insufficient, and urgent requirements of petrochemical equipment on abnormality early warning comprehensiveness, precision and dynamics cannot be met. Disclosure of Invention The technical problems to be solved by the invention are that the prior equipment abnormality early warning technology has a plurality of limitations, on one hand, the multi-focus single-dimension detection lacks collaborative matching and dynamic association of a plurality of types of abnormal parameters of deformation parameters and scaling parameters, which is easy to cause abnormality identification unilateral performance, on the other hand, even if multi-parameter fusion is involved, a simple threshold superposition or linear association mode is adopted, dynamic mapping relation and coupling analysis are not built by combining historical fault data, the whole health state of equipment is difficult to accurately quantify, in addition, the early warning classification depends on empirical threshold division, the factors of parameter coupling degree and reaction kettle equipment operation time attenuation are not comprehensively considered, the early warning accuracy and timeliness are insufficient, and urgent requirements of petrochemical equipment on abnormality early warning comprehensiveness, precision and dynamics cannot be met. In order to solve the technical problems, the invention provides the technical scheme that the abnormal operation early warning system of petrochemical equipment comprises a matching module and an early warning module; The matching module is used for calling a preset deformation abnormal rule to perform first matching on deformation parameters, calling a preset image abnormal rule to perform second matching on the inner wall image data if the first matching result is abnormal, and outputting only the first matching result if the first matching result is non-abnormal; the early warning module constructs a dynamic mapping relation based on the first matching result and the second matching result, obtains a health degree evaluation value through coupling analysis, generates a grading abnormal early warning signal according to the health degree evaluation value, and executes early warning operation of corresponding grades. As a preferable scheme of the petrochemical equipment operation abnormality early warning system, the deformation abnormality rule is constructed based on historical deformation parameters of the reaction kettle, scaling parameters are extracted based on historical inner wall image data of the reaction kettle, and the image abnormality rule is constructed according to the scaling parameters; The configuration logic of the deformation abnormal rule comprises the steps of substituting the historical deformation parameters and the preset reference deformation parameters into a cosine similarity formula for calculation to obtain a similar