CN118114063A - Fault diagnosis method and system for thermal power generating unit rotating equipment - Google Patents
Fault diagnosis method and system for thermal power generating unit rotating equipment Download PDFInfo
- Publication number
- CN118114063A CN118114063A CN202410209988.6A CN202410209988A CN118114063A CN 118114063 A CN118114063 A CN 118114063A CN 202410209988 A CN202410209988 A CN 202410209988A CN 118114063 A CN118114063 A CN 118114063A
- Authority
- CN
- China
- Prior art keywords
- parameter
- fault
- vector
- early warning
- memory matrix
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 79
- 238000003745 diagnosis Methods 0.000 title claims abstract description 41
- 239000013598 vector Substances 0.000 claims abstract description 106
- 239000011159 matrix material Substances 0.000 claims abstract description 62
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000006870 function Effects 0.000 claims description 27
- 238000004364 calculation method Methods 0.000 claims description 25
- 238000004590 computer program Methods 0.000 claims description 19
- 238000012795 verification Methods 0.000 claims description 18
- 238000012216 screening Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 3
- 238000012544 monitoring process Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 7
- 238000005461 lubrication Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- 238000010248 power generation Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a fault diagnosis method and a fault diagnosis system for rotating equipment of a thermal power generating unit, wherein the method comprises the steps of selecting historical normal operation data and preprocessing the data to obtain a historical normal data set; selecting characteristic parameters according to the running characteristics of the rotating equipment; selecting partial data from the historical normal data set according to the numerical range of each characteristic parameter to form a memory matrix D i; inputting an observation vector, constructing a dynamic memory matrix D and calculating an estimation vector by adopting an MSET; performing fault early warning and diagnosis by adopting two indexes of overall similarity and parameter similarity, and determining an early warning threshold value of a fault by introducing an early warning threshold value coefficient; and when the overall similarity is lower than the early warning threshold value, performing equipment fault warning, and if the similarity of a certain parameter is warning, indicating that the parameter is likely to be a fault parameter. The invention aims to reduce the calculated amount of the fault diagnosis method, improve the robustness and pollution resistance of the fault diagnosis method and realize the real-time and accurate fault identification of the fault of the rotating equipment.
Description
Technical Field
The invention relates to the technical field of power station rotating equipment fault early warning and diagnosis, in particular to a fault diagnosis method and system for thermal power generating unit rotating equipment.
Background
The thermal power generating unit has extremely complex structure, and comprises three main units of a boiler, a steam turbine, a generator and the like, and also comprises key auxiliary units of a coal mill, a fan, a water supply pump, a condensate pump, a circulating water pump and the like. A large number of equipment in main and auxiliary machines of the thermal power generating unit are rotary equipment, such as a turbo generator unit, various fans, water pumps and the like. The safe and stable operation of each rotary device is an important factor for guaranteeing the operation safety of the thermal power generating unit. Therefore, it becomes particularly important to develop state monitoring and preventive maintenance and overhaul in the aspect of efficient and safe operation of the rotating equipment of the thermal power generating unit.
The degradation process from early development to alarm tripping of the fault of the rotating equipment generally needs to be carried out for a period of time, and the existing equipment monitoring systems of the thermal power generating unit such as DCS (distributed control system ), TSI (steam turbine safety monitoring instrument system, turbine Supervisory Instrumentation) and the like usually only have single-point alarm and related protection tripping functions, focus on emergency treatment after accidents or analysis after the accidents, cannot evaluate the running state of the whole equipment in real time, and have very limited time for operators to intervene and rescue after finding alarm signals. In order to reduce unplanned unit outage and equipment damage accidents caused by equipment faults and improve the availability of equipment, a method for giving an early warning signal at the early stage of small degradation of the equipment is needed.
With the high-speed development of information technology, the method is widely applied to the large-scale construction of a DCS control system, an SIS system (plant-level monitoring information system, supervisory Information System) and various other information systems of power generation enterprises, and helps each power generation enterprise to accumulate mass production data. Lays a foundation for the analysis and research work of big data of the rotating equipment of the power generation enterprises. The thermal power generating unit equipment state monitoring and fault early warning system developed through the special and efficient real-time data mining technology can help a power plant to realize intelligent management of equipment states, improve equipment management efficiency, change fault post-treatment into pre-prevention, grasp overall dynamic change in equipment operation in real time, greatly improve operation safety level of the equipment, reduce non-abnormal stop caused by equipment reasons and reduce unnecessary planned maintenance cost of the equipment.
The state monitoring and the fault early warning are integrated technologies integrating a data prediction technology and a fault diagnosis technology. The effective state monitoring can be generally used for evaluating the health condition of equipment and discovering potential faults in the equipment in advance, the common state monitoring method is to establish a model to predict the reaching value of the relevant monitoring parameter under the normal state of the equipment, judge the running state of the unit by comparing the deviation between the reaching value of the relevant monitoring parameter and the actual running value, and consider that a certain equipment is possibly abnormal if the deviation between the reaching value and the actual value is larger. At present, the modeling method mainly comprises a mechanism modeling method and a data driving modeling method, and for complex dynamic objects such as a thermal power generating unit, the coupling relevance among parameters of each system and equipment is strong, and an accurate mechanism model is difficult to build. The existing monitoring system of the thermal power generating unit stores massive historical data of relevant monitoring parameters of each device of the unit, the modeling method based on data driving can fully mine information stored in massive historical operation data of the unit, excessive assumption conditions are not needed, and the modeling feasibility is high and is more suitable for practical application.
In the prior art, a plurality of methods are used for realizing the reconstruction of the normal running state of the thermal power generating unit, such as a neural network method, a multivariate state estimation algorithm, a regression model and the like, but the defects of large calculation amount or poor robustness and pollution resistance are remarkable in a plurality of methods, so that the model has certain limitation in terms of real-time performance and accurate fault identification.
Disclosure of Invention
The invention aims to solve the technical problems: aiming at the problems in the prior art, the invention provides a fault diagnosis method and a fault diagnosis system for rotating equipment of a thermal power generating unit, which aim to reduce the calculated amount of the fault diagnosis method, improve the robustness and pollution resistance of the fault diagnosis method and realize real-time and accurate fault identification of the faults of the rotating equipment.
In order to solve the technical problems, the invention adopts the following technical scheme:
a fault diagnosis method for thermal power generating unit rotating equipment comprises the following steps:
s1, selecting historical normal operation data covering all operation conditions of power station rotating equipment, and preprocessing to obtain a historical normal data set representing the normal operation state of the equipment;
S2, selecting parameters representing the output, the running state and the boundary conditions of the equipment as selected characteristic parameters according to the running characteristics of the rotating equipment;
S3, selecting partial data from the historical normal data set according to the numerical range of each characteristic parameter to form a memory matrix D i, and forming verification set data by the rest data;
s4, inputting an observation vector, and screening partial data in a memory matrix D i according to the numerical value of the characteristic parameter in the observation vector to form a dynamic memory matrix D;
S5, constructing a weighted MSET algorithm by adopting a weight coefficient based on the dynamic memory matrix D, calculating a prediction result of the observation vector, and recording the prediction result as an estimated vector;
S6, performing fault early warning and diagnosis by adopting two indexes of overall similarity and parameter similarity, calculating verification set data to obtain minimum values of the overall similarity and the parameter similarity indexes, and introducing an early warning threshold coefficient to determine an early warning threshold of a fault; and calculating the overall similarity and the similarity of each parameter according to the values of each parameter in the observation vector and the estimation vector, carrying out equipment fault alarm when the overall similarity is lower than an early warning threshold value, and indicating that a certain parameter is likely to be a fault parameter if the similarity alarm of the certain parameter.
Optionally, step S3 includes the steps of:
S31, selecting one parameter which can represent the operation condition of the equipment most from the selected characteristic parameters as a main characteristic parameter;
S32, dividing the historical normal data set into a plurality of groups by adopting an equal step sampling method according to the maximum value and the minimum value of the main characteristic parameters in the historical normal data set;
S33, dividing each group of data obtained by dividing into a plurality of subgroups by adopting an equal step sampling method according to the maximum value and the minimum value of each other characteristic parameter, screening a vector with the most central numerical value of each characteristic parameter in each subgroup, and adding the vector into a memory matrix;
S34, removing repeated vectors in the memory matrix, so as to obtain a final memory matrix D i, and forming the rest data in the historical normal data set into verification set data.
Optionally, step S4 includes the steps of:
S41, inputting an observation vector, obtaining the numerical value of each characteristic parameter in the observation vector, and screening partial vectors from the memory matrix D i within a positive and negative certain range according to the numerical value of each characteristic parameter to form a corresponding sub-data set;
S42, integrating all the sub-data sets and removing the repeated vectors to obtain a new data set, and calculating the occurrence times of each vector in the new data set in all the sub-data sets;
S43, selecting vectors with the same frequency as the number k of the characteristic parameters in each sub-data set, and adding the vectors into a dynamic memory matrix D of the observation vector;
S44, checking the number of vectors contained in the dynamic memory matrix D, if the number of vectors contained in the dynamic memory matrix D is smaller than a preset value, continuously selecting the vector with the occurrence number of k-1, adding the vector to the dynamic memory matrix D, and the like until the number of vectors contained in the dynamic memory matrix D is larger than the preset value.
Optionally, the functional expression of the dynamic memory matrix D obtained in step S4 is:
，
In the above-mentioned method, the step of, ～/>The n×m elements in the dynamic memory matrix D are respectively represented, where the number of rows n represents the number of parameters included in the state vector, and the number of columns m represents the number of stored historical state vectors.
Optionally, in step S5, the prediction result of the observation vector calculated by the weighting algorithm MSET based on the dynamic memory matrix D and using the weighting coefficient is recorded as a function expression of the estimated vector, where the function expression is as follows:
，
In the above formula, X est is an estimated vector, X obs is an observed vector, The method is characterized in that the method is a nonlinear operator, D is a dynamic memory matrix, and the function expression of the nonlinear operator is as follows:
，
In the above-mentioned method, the step of, For the calculation result of the nonlinear operator on the vectors X and Y, w i is the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model, and is/areFor the i-th element in vector X,/>For the i-th element in vector Y,/>Is the dimension of vectors X and Y.
Optionally, the calculation function expression of the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model is:
，
In the above formula, f i is the i-th element in the intermediate value f 1～fn, and there are:
，
In the above formula, x i,est is a predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is an actual measurement value of the ith parameter, and sigma i is a standard deviation of the ith parameter.
Optionally, the overall similarity calculation function in step S6 is expressed as:
，
In the above-mentioned method, the step of, Representing the overall similarity, x i,est is the predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is the measured value of the ith parameter,/>Is the dimension of vectors X and Y;
the calculation function expression of the parameter similarity is as follows:
，
In the above-mentioned method, the step of, Parameter similarity representing the i-th parameter;
In the step S6, the function expression of the early warning threshold value of the early warning threshold coefficient determination fault is as follows:
，
In the above-mentioned method, the step of, Is a fault early warning threshold value,/>K is an early warning threshold coefficient for the minimum value of the similarity S of the verification set, and k is greater than 1.
In addition, the invention also provides a fault diagnosis system for the thermal power generating unit rotating equipment, which comprises a microprocessor and a memory which are connected with each other, wherein the microprocessor is programmed or configured to execute the fault diagnosis method for the thermal power generating unit rotating equipment.
Furthermore, the invention also provides a computer readable storage medium, in which a computer program/instruction is stored, the computer program/instruction being programmed or configured to execute the fault diagnosis method for the thermal power generating unit rotating equipment by a processor.
Furthermore, the invention also provides a computer program product comprising a computer program/instruction programmed or configured to execute the fault diagnosis method for the thermal power generating unit rotating equipment by a processor.
Compared with the prior art, the invention has the following advantages:
1. The construction method of the historical memory matrix D i is improved, multiple characteristic parameters representing the output, the running state and the boundary conditions of equipment are introduced, the data are screened from a historical normal data set by adopting an equal step sampling method according to the characteristic parameters to form the memory matrix D i, and compared with the existing method, the method can ensure the full coverage of the memory matrix D i on the running working condition of the equipment, and effectively improve the accuracy of the MSET model prediction result. The data is further screened in the memory matrix D i according to the numerical value of each characteristic parameter in each observation vector to form a dynamic memory matrix D, the calculation amount of the model can be further reduced on the basis of guaranteeing the accuracy of calculation results, the calculation speed of the model is improved, and the model is guaranteed to meet the requirement of real-time performance in engineering operation.
2. The power station rotating equipment fault early warning and diagnosing method provided by the invention comprises the steps of constructing a weighted MSET algorithm based on a dynamic memory matrix D by adopting a weight coefficient, calculating a prediction result of an observation vector, recording the prediction result as an estimated vector, improving an MSET model by introducing the weight coefficient, improving the prediction precision of fault parameters under abnormal conditions, avoiding the influence of the fault parameters on the prediction results of other normal parameters, and effectively reducing the false alarm rate and the false alarm rate of the model. By monitoring the overall similarity of the equipment and the similarity of all parameters, early warning of equipment faults can be realized, fault parameters can be determined, and diagnosis of equipment fault points can be realized.
Drawings
FIG. 1 is a schematic diagram of a basic flow of a method according to an embodiment of the present invention.
FIG. 2 is a graph showing absolute values of relative errors of prediction results of validation set data in accordance with an embodiment of the present invention.
FIG. 3 shows overall similarity of the prediction results of the verification set data in an embodiment of the present invention.
FIG. 4 shows the similarity of parameters of the prediction result of the verification set data in the embodiment of the invention.
FIG. 5 is a comparison of the prediction results of the fault parameters in the embodiment of the present invention.
FIG. 6 is a comparison of the predicted results of normal parameters in an embodiment of the present invention.
Fig. 7 shows overall similarity in an embodiment of the present invention.
FIG. 8 is a graph showing the similarity of abnormal parameter bearing temperatures in an embodiment of the present invention.
Detailed Description
The invention will be described in further detail below with an example of a power station blower as a thermal power unit rotating device, but the thermal power unit rotating device of the invention can be applied to various rotating devices such as a turbo generator set, various fans, a water pump, and the like.
As shown in fig. 1, the fault diagnosis method for the thermal power generating unit rotating equipment according to the embodiment includes the following steps:
s1, selecting historical normal operation data covering all operation conditions of power station rotating equipment, and preprocessing to obtain a historical normal data set representing the normal operation state of the equipment;
S2, selecting parameters representing the output, the running state and the boundary conditions of the equipment as selected characteristic parameters according to the running characteristics of the rotating equipment;
S3, selecting partial data from the historical normal data set according to the numerical range of each characteristic parameter to form a memory matrix D i, and forming verification set data by the rest data;
s4, inputting an observation vector, and screening partial data in a memory matrix D i according to the numerical value of the characteristic parameter in the observation vector to form a dynamic memory matrix D;
S5, calculating a prediction result of the observation vector by adopting a weighted MSET (multivariate state estimation technique ) algorithm based on the dynamic memory matrix D and adopting a weight coefficient to be recorded as an estimation vector;
S6, performing fault early warning and diagnosis by adopting two indexes of overall similarity and parameter similarity, calculating verification set data to obtain minimum values of the overall similarity and the parameter similarity indexes, and introducing an early warning threshold coefficient to determine an early warning threshold of a fault; and calculating the overall similarity and the similarity of each parameter according to the values of each parameter in the observation vector and the estimation vector, carrying out equipment fault alarm when the overall similarity is lower than an early warning threshold value, and indicating that a certain parameter is likely to be a fault parameter if the similarity alarm of the certain parameter.
In this embodiment, the thermal power generating unit rotating device takes a power station blower as an example, and in step S1, historical normal operation data covering all operation conditions of the power station rotating device is selected, wherein the monitoring parameters include blower bearing temperatures 1-9, blower motor front/rear bearing temperatures, blower X/Y bearing vibration, blower lubrication oil pressure, blower lubrication oil temperature, blower inlet air temperature, blower movable vane opening, blower motor current and blower outlet air pressure.
The preprocessing in step S1 may include normalized processing of deleting null values, outliers, and data, and interpolation of the missing values may be performed as needed. As an alternative implementation, the normalization process in this embodiment uses a z-core normalization process, whose functional expression is:
，
in the above formula, x is a variable, μ is the mean value of x, and σ is the standard deviation of x.
In step S2 of this embodiment, when parameters characterizing the output, the running state and the boundary conditions of the rotating device are selected as selected feature parameters according to the running features of the rotating device, the selected feature parameters include: blower motor current, blower blade opening, blower outlet wind pressure, blower inlet wind temperature, blower lubrication oil pressure, blower lubrication oil temperature. The three characteristic parameters of blower motor current, blower movable vane opening and blower outlet wind pressure can be used for representing equipment output force and running state; the blower inlet air temperature characterizes the environmental boundary conditions, and the blower lubrication oil pressure and the blower lubrication oil temperature are used for characterizing the boundary conditions and are related to an external lubrication oil system.
Step S3 of this embodiment includes the steps of:
S31, selecting one parameter which can represent the operation condition of the equipment most from the selected characteristic parameters as a main characteristic parameter;
S32, dividing the historical normal data set into a plurality of groups by adopting an equal step sampling method according to the maximum value and the minimum value of the main characteristic parameters in the historical normal data set;
S33, dividing each group of data obtained by dividing into a plurality of subgroups by adopting an equal step sampling method according to the maximum value and the minimum value of each other characteristic parameter, screening a vector with the most central numerical value of each characteristic parameter in each subgroup, and adding the vector into a memory matrix;
S34, removing repeated vectors in the memory matrix, so as to obtain a final memory matrix D i, and forming the rest data in the historical normal data set into verification set data.
In this embodiment, step S4 includes the following steps:
S41, inputting an observation vector, obtaining the numerical value of each characteristic parameter in the observation vector, and screening partial vectors from the memory matrix D i within a positive and negative certain range according to the numerical value of each characteristic parameter to form a corresponding sub-data set;
S42, integrating all the sub-data sets and removing the repeated vectors to obtain a new data set, and calculating the occurrence times of each vector in the new data set in all the sub-data sets;
S43, selecting vectors with the same frequency as the number k of the characteristic parameters in each sub-data set, and adding the vectors into a dynamic memory matrix D of the observation vector;
S44, checking the number of vectors contained in the dynamic memory matrix D, if the number of vectors contained in the dynamic memory matrix D is smaller than a preset value, continuously selecting the vector with the occurrence number of k-1, adding the vector to the dynamic memory matrix D, and the like until the number of vectors contained in the dynamic memory matrix D is larger than the preset value.
In this embodiment, the function expression of the dynamic memory matrix D obtained in step S4 is:
，
In the above-mentioned method, the step of, ～/>The n×m elements in the dynamic memory matrix D are respectively represented, where the number of rows n represents the number of parameters included in the state vector, and the number of columns m represents the number of stored historical state vectors.
In this embodiment, in step S5, the function expression for calculating the prediction result of the observation vector by using the weighted MSET algorithm based on the dynamic memory matrix D and using the weight coefficient is shown as the estimated vector:
，
In the above formula, X est is an estimated vector, X obs is an observed vector, The method is characterized in that the method is a nonlinear operator, D is a dynamic memory matrix, and the function expression of the nonlinear operator is as follows:
，
In the above-mentioned method, the step of, For the calculation result of the nonlinear operator on the vectors X and Y, w i is the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model, and is/areFor the i-th element in vector X,/>For the i-th element in vector Y,/>Is the dimension of vectors X and Y.
In this embodiment, the calculation function expression of the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model is:
，
In the above formula, f i is the i-th element in the intermediate value f 1～fn, and there are:
，
In the above formula, x i,est is a predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is an actual measurement value of the ith parameter, and sigma i is a standard deviation of the ith parameter. In this embodiment, after the weight coefficient of each parameter is calculated, the weighted MSET model is used again to calculate the observation vector, and the prediction result is output as the final prediction result. In this embodiment, the verification set data is calculated by using a conventional MSET model, the relative error of each parameter prediction result is shown in fig. 2 (only one is listed in fig. 2 as an example), and the calculation results of the overall similarity of the device and each parameter similarity are shown in fig. 3 and fig. 4 (only one is taken as an example). It can be seen that for verification set data, the relative error of each parameter is within 2%, so that the prediction result of the normal data by the conventional MSET model is high in accuracy, and the requirements of engineering application can be met. In this embodiment, the sliding window method is adopted to perform smoothing processing on the data, so that negative effects caused by noise and interference signals can be reduced to a certain extent, and the accuracy of fault early warning is improved.
In this embodiment, the expression of the calculation function of the overall similarity in step S6 is:
，
In the above-mentioned method, the step of, Representing the overall similarity, x i,est is the predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is the measured value of the ith parameter,/>Is the dimension of vectors X and Y;
the calculation function expression of the parameter similarity is as follows:
，
In the above-mentioned method, the step of, Parameter similarity representing the i-th parameter;
In the step S6, the function expression of the early warning threshold value of the early warning threshold coefficient determination fault is as follows:
，
In the above-mentioned method, the step of, Is a fault early warning threshold value,/>K is an early warning threshold coefficient for the minimum value of the similarity S of the verification set, and k is greater than 1.
In this embodiment, 300 sets of data are selected from the historical normal operation database as a test set for calculation, and the accumulated offset with the step length of 0.1 ℃ is artificially added to the 1 st measure point of the blower bearing from the 101 st set of data, so that the early warning simulation test is performed on the abnormal fault of the blower bearing temperature 1 by using the fault early warning and diagnosis method of the rotating equipment provided by the application. Fig. 5 and 6 show the prediction results of the bearing temperature 1 and the blower outlet wind pressure by the conventional MSET early warning model and the weighted MSET early warning model provided by the application. It can be seen that for the data of the previous 100 groups without disturbance, the accuracy of the prediction results of the two models is higher; however, for 200 sets of data after disturbance, the difference between the value of the blower bearing temperature 1 predicted by the conventional MSET model and the actual value is larger, and when the disturbance is larger, the difference between the predicted result and the actual value of the blower outlet wind pressure is also gradually increased. However, the weighted MSET model provided by the application can accurately predict the normal value of the fault parameter bearing temperature 1, can reduce the influence of the abnormality of the bearing temperature 1 on the wind pressure prediction result of the outlet of the blower, and can effectively improve the robustness and pollution resistance of the MSET model.
Fig. 7 and 8 show the calculation results of the overall similarity of the device and the similarity of the bearing temperature 1 under the condition of the fault of the bearing temperature 1, and it can be seen that the similarity of the device is smaller than the early warning threshold value at the 130 th sample point, and an alarm is sent. In addition, the similarity of the bearing temperature 1 also gives an alarm, so that the fault point can be determined to be the bearing temperature 1. The bearing temperature 1 of the blower is only 48 ℃ when the overall similarity of the equipment is alarmed, and the alarm value of the measuring point DCS is 90 ℃. Therefore, the fault early warning model provided by the embodiment can give early warning signals in the early stage of the performance reduction of the equipment, accurately diagnose specific fault point parameters and provide sufficient time margin for field personnel to process faults.
In summary, the embodiment includes determining modeling variables according to conditions and operation characteristics of monitoring points of the power station rotating equipment; acquiring historical normal operation data covering all operation conditions of the rotating equipment, kicking off abnormal values and carrying out standardized processing to form a historical normal data set; determining characteristic parameters according to the operation characteristics of the rotating equipment, screening partial data based on the range of each characteristic parameter in the historical normal data set to form a memory matrix, and taking the rest data as verification set data; screening partial data in the memory matrix based on the numerical value of each characteristic parameter in the input observation vector to form a dynamic memory matrix; constructing a weighted MSET algorithm by introducing weight coefficients, and establishing a weighted MSET fault early warning model based on a dynamic memory matrix; and introducing the overall similarity of the equipment and the similarity function of each parameter as evaluation indexes of fault early warning and diagnosis, determining the minimum value of the average similarity in the verification set as a fault early warning threshold value, and alarming when the similarity value is lower than the fault early warning threshold value. The invention has the advantages of wide applicability, small calculated amount, high accuracy, low false alarm rate and fault point positioning capability.
In addition, the embodiment also provides a fault diagnosis system for the thermal power generating unit rotating equipment, which comprises a microprocessor and a memory which are connected with each other, wherein the microprocessor is programmed or configured to execute the fault diagnosis method for the thermal power generating unit rotating equipment. The present embodiment also provides a computer-readable storage medium having stored therein a computer program/instruction programmed or configured to execute the fault diagnosis method for thermal power generating unit rotating equipment by a processor. The present embodiment also provides a computer program product comprising a computer program/instructions programmed or configured to execute the fault diagnosis method for a thermal power generating unit rotating equipment by a processor.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein. The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above examples, and all technical solutions belonging to the concept of the present invention belong to the protection scope of the present invention. It should be noted that modifications and adaptations to the present invention may occur to one skilled in the art without departing from the principles of the present invention and are intended to be within the scope of the present invention.
Claims (10)
1. The fault diagnosis method for the thermal power generating unit rotating equipment is characterized by comprising the following steps of:
s1, selecting historical normal operation data covering all operation conditions of power station rotating equipment, and preprocessing to obtain a historical normal data set representing the normal operation state of the equipment;
S2, selecting parameters representing the output, the running state and the boundary conditions of the equipment as selected characteristic parameters according to the running characteristics of the rotating equipment;
S3, selecting partial data from the historical normal data set according to the numerical range of each characteristic parameter to form a memory matrix D i, and forming verification set data by the rest data;
s4, inputting an observation vector, and screening partial data in a memory matrix D i according to the numerical value of the characteristic parameter in the observation vector to form a dynamic memory matrix D;
S5, constructing a weighted MSET algorithm by adopting a weight coefficient based on the dynamic memory matrix D, calculating a prediction result of the observation vector, and recording the prediction result as an estimated vector;
S6, performing fault early warning and diagnosis by adopting two indexes of overall similarity and parameter similarity, calculating verification set data to obtain minimum values of the overall similarity and the parameter similarity indexes, and introducing an early warning threshold coefficient to determine an early warning threshold of a fault; and calculating the overall similarity and the similarity of each parameter according to the values of each parameter in the observation vector and the estimation vector, carrying out equipment fault alarm when the overall similarity is lower than an early warning threshold value, and indicating that a certain parameter is likely to be a fault parameter if the similarity alarm of the certain parameter.
2. The fault diagnosis method for a thermal power generating unit rotating apparatus according to claim 1, wherein step S3 comprises the steps of:
S31, selecting one parameter which can represent the operation condition of the equipment most from the selected characteristic parameters as a main characteristic parameter;
S32, dividing the historical normal data set into a plurality of groups by adopting an equal step sampling method according to the maximum value and the minimum value of the main characteristic parameters in the historical normal data set;
S33, dividing each group of data obtained by dividing into a plurality of subgroups by adopting an equal step sampling method according to the maximum value and the minimum value of each other characteristic parameter, screening a vector with the most central numerical value of each characteristic parameter in each subgroup, and adding the vector into a memory matrix;
S34, removing repeated vectors in the memory matrix, so as to obtain a final memory matrix D i, and forming the rest data in the historical normal data set into verification set data.
3. The fault diagnosis method for a thermal power generating unit rotating apparatus according to claim 1, wherein step S4 comprises the steps of:
S41, inputting an observation vector, obtaining the numerical value of each characteristic parameter in the observation vector, and screening partial vectors from the memory matrix D i within a positive and negative certain range according to the numerical value of each characteristic parameter to form a corresponding sub-data set;
S42, integrating all the sub-data sets and removing the repeated vectors to obtain a new data set, and calculating the occurrence times of each vector in the new data set in all the sub-data sets;
S43, selecting vectors with the same frequency as the number k of the characteristic parameters in each sub-data set, and adding the vectors into a dynamic memory matrix D of the observation vector;
s44, checking the number of vectors contained in the dynamic memory matrix D, if the number of vectors is smaller than a preset value, continuously selecting the vector with the occurrence number of k-1, adding the vector to the dynamic memory matrix D, and the like until the number of vectors contained in the dynamic memory matrix D is larger than the preset value.
4. The fault diagnosis method for thermal power generating unit rotating equipment according to claim 3, wherein the function expression of the dynamic memory matrix D obtained in step S4 is:
，
In the above-mentioned method, the step of, ～/>The n×m elements in the dynamic memory matrix D are respectively represented, where the number of rows n represents the number of parameters included in the state vector, and the number of columns m represents the number of stored historical state vectors.
5. The fault diagnosis method for thermal power generating unit rotating equipment according to claim 1, wherein in step S5, a function expression for calculating a prediction result of an observation vector by using a weighted MSET algorithm based on a dynamic memory matrix D and adopting a weight coefficient is expressed as an estimated vector:
，
In the above formula, X est is an estimated vector, X obs is an observed vector, The method is characterized in that the method is a nonlinear operator, D is a dynamic memory matrix, and the function expression of the nonlinear operator is as follows:
，
In the above-mentioned method, the step of, For the calculation result of the nonlinear operator on the vectors X and Y, w i is the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model, and is/areFor the i-th element in vector X,/>For the i-th element in the vector Y,Is the dimension of vectors X and Y.
6. The fault diagnosis method for thermal power generating unit rotating equipment according to claim 5, wherein the calculation function expression of the weight coefficient calculated by the ith parameter based on the calculation result of the MSET early warning model is:
，
In the above formula, f i is the i-th element in the intermediate value f 1～fn, and there are:
，
In the above formula, x i,est is a predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is an actual measurement value of the ith parameter, and sigma i is a standard deviation of the ith parameter.
7. The fault diagnosis method for thermal power generating unit rotating equipment according to claim 1, wherein the calculation function expression of the overall similarity in step S6 is:
，
In the above-mentioned method, the step of, Representing the overall similarity, x i,est is the predicted value of the ith parameter calculated by adopting the MSET early warning model, x i,obs is the measured value of the ith parameter,/>Is the dimension of vectors X and Y;
the calculation function expression of the parameter similarity is as follows:
，
In the above-mentioned method, the step of, Parameter similarity representing the i-th parameter;
In the step S6, the function expression of the early warning threshold value of the early warning threshold coefficient determination fault is as follows:
，
In the above-mentioned method, the step of, Is a fault early warning threshold value,/>K is an early warning threshold coefficient for the minimum value of the similarity S of the verification set, and k is greater than 1.
8. A fault diagnosis system for a thermal power plant rotating apparatus, comprising a microprocessor and a memory connected to each other, characterized in that the microprocessor is programmed or configured to perform the fault diagnosis method for a thermal power plant rotating apparatus according to any one of claims 1 to 7.
9. A computer readable storage medium having stored therein a computer program/instruction, characterized in that the computer program/instruction is programmed or configured to execute the fault diagnosis method for a thermal power generating unit rotating apparatus according to any one of claims 1 to 7 by a processor.
10. A computer program product comprising computer programs/instructions, characterized in that the computer programs/instructions are programmed or configured to execute the fault diagnosis method for a rotating equipment of a thermal power generating unit according to any one of claims 1 to 7 by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410209988.6A CN118114063A (en) | 2024-02-26 | 2024-02-26 | Fault diagnosis method and system for thermal power generating unit rotating equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410209988.6A CN118114063A (en) | 2024-02-26 | 2024-02-26 | Fault diagnosis method and system for thermal power generating unit rotating equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118114063A true CN118114063A (en) | 2024-05-31 |
Family
ID=91216645
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410209988.6A Pending CN118114063A (en) | 2024-02-26 | 2024-02-26 | Fault diagnosis method and system for thermal power generating unit rotating equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118114063A (en) |
-
2024
- 2024-02-26 CN CN202410209988.6A patent/CN118114063A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112784373B (en) | Fault early warning method for wind turbine generator gearbox | |
| CN110298455B (en) | Mechanical equipment fault intelligent early warning method based on multivariate estimation prediction | |
| CN110414155B (en) | Fan component temperature abnormity detection and alarm method with single measuring point | |
| CN109298697B (en) | Method for evaluating working state of each part of thermal power plant system based on dynamic baseline model | |
| CN107291991B (en) | Early defect early warning method for wind turbine generator based on dynamic network sign | |
| US6587737B2 (en) | Method for the monitoring of a plant | |
| CN106404403B (en) | Method and system for analysis of a turbomachine | |
| Wu et al. | A hybrid LSTM-KLD approach to condition monitoring of operational wind turbines | |
| CN112067335A (en) | Power plant blower fault early warning method based on multivariate state estimation | |
| CN110414154B (en) | Fan component temperature abnormity detection and alarm method with double measuring points | |
| CN113705924B (en) | Intelligent diagnosis method and system for thermal control equipment | |
| CN112487910A (en) | Fault early warning method and system for nuclear turbine system | |
| CN110469496B (en) | Intelligent early warning method and system for water pump | |
| CN112000654B (en) | High-speed railway strong wind monitoring data quality control method and device | |
| CN111855219A (en) | Diesel engine lubricating oil entering security parameter prediction method based on grey theory | |
| CN111563685B (en) | Power generation equipment state early warning method based on auto-associative kernel regression algorithm | |
| CN111192163B (en) | Generator reliability medium-short term prediction method based on wind turbine generator operating data | |
| CN113313365A (en) | Degradation early warning method and device for primary air fan | |
| CN118114063A (en) | Fault diagnosis method and system for thermal power generating unit rotating equipment | |
| US11339763B2 (en) | Method for windmill farm monitoring | |
| Zhang | Comparison of data-driven and model-based methodologies of wind turbine fault detection with SCADA data | |
| Li et al. | Combining canonical variate analysis, probability approach and support vector regression for failure time prediction | |
| Hu et al. | An Artificial Neural Network Model for Monitoring Real-Time Parameters and Detecting Early Warnings in Induced Draft Fan | |
| Dehghanabandaki et al. | Fault Detection of Wind Turbine System Based on Deep Learning and System Identification | |
| CN114151291B (en) | Early fault monitoring method for wind turbine generator |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |