CN117763941A - Gas storage well service life assessment method based on machine learning - Google Patents
Gas storage well service life assessment method based on machine learning Download PDFInfo
- Publication number
- CN117763941A CN117763941A CN202311513602.2A CN202311513602A CN117763941A CN 117763941 A CN117763941 A CN 117763941A CN 202311513602 A CN202311513602 A CN 202311513602A CN 117763941 A CN117763941 A CN 117763941A
- Authority
- CN
- China
- Prior art keywords
- failure
- gas storage
- storage well
- attribute index
- value
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 91
- 238000010801 machine learning Methods 0.000 title claims abstract description 26
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 238000012937 correction Methods 0.000 claims description 4
- 230000007547 defect Effects 0.000 abstract description 6
- 239000007789 gas Substances 0.000 description 212
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 8
- 230000005540 biological transmission Effects 0.000 description 5
- 239000003345 natural gas Substances 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000004590 computer program Methods 0.000 description 3
- 239000000523 sample Substances 0.000 description 3
- 239000002689 soil Substances 0.000 description 3
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012806 monitoring device Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000001902 propagating effect Effects 0.000 description 2
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 239000003570 air Substances 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009429 electrical wiring Methods 0.000 description 1
- 239000011261 inert gas Substances 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
A method for evaluating the service life of a gas storage well based on machine learning comprises the steps of collecting attribute indexes of the gas storage well, and setting critical quantities of all the attribute indexes; calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity; correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value; according to the method for predicting the residual time period of the gas storage well failure according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, the failure prediction can be performed on the gas storage wells with various specifications, hidden dangers caused by failure accidents of the gas storage wells are effectively reduced, workers can be prompted in advance to replace the gas storage wells which are about to fail, the defect of strong subjective randomness in the prior art is prevented by using a reasonable prediction method of machine learning, and the life precision of the predicted gas storage wells is high.
Description
Technical Field
The invention belongs to the technical field of gas storage well life assessment, and particularly relates to a gas storage well life assessment method based on machine learning.
Background
The gas storage well is a well-type tubular device which is vertically arranged underground and used for storing compressed air, and the gas storage well mainly stores compressed natural gas, so that the natural gas in the well is rarely expanded with heat and contracted with cold due to the change of the external temperature because the underground temperature is relatively constant; and natural gas is safer to store underground. The gas storage well is used for storing natural gas, nitrogen or inert gas and air as working mediums. Gas storage wells are being incorporated into the category of stationary pressure vessels in specialty equipment regulations.
In order to ensure safe operation of the gas storage well, it is necessary to predict the life of the gas storage well, so as to prevent the failure problem caused by aging of the gas storage well, as in the prior art scheme of the patent application No. CN202021512322.1 and the patent name of the gas storage well life monitoring device, the gas storage well life monitoring device obtains corresponding attribute values through a probe for collecting the performance values of the gas storage tank, and then the prediction of the life of the gas storage well is performed on the corresponding performance values by means of manpower.
Disclosure of Invention
In order to solve the defects in the prior art, the invention provides a gas storage well service life assessment method based on machine learning, which is implemented by collecting attribute indexes of a gas storage well and setting critical quantities of all the attribute indexes; calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity; correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value; according to the method for predicting the residual time period of the gas storage well failure according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, the failure prediction can be performed on the gas storage wells with various specifications, hidden dangers caused by failure accidents of the gas storage wells are effectively reduced, workers can be prompted in advance to replace the gas storage wells which are about to fail, the defect of strong subjective randomness in the prior art is prevented by using a reasonable prediction method of machine learning, and the life precision of the predicted gas storage wells is high.
The invention adopts the following technical scheme.
A method for estimating life of a gas storage well based on machine learning, comprising:
step 1: collecting attribute indexes of the gas storage well, and setting critical quantity of each attribute index;
Step 2: calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity;
step 3: correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
step 4: and constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode.
Preferably, the method for calculating the importance of each attribute index to cause the failure of the gas storage well comprises the following steps:
calculating the presumption probability of the failure of the gas storage well caused by the attribute index;
calculating the state probability of gas storage well failure when the value of the attribute index is y;
calculating the probability of failure of the gas storage well caused by each attribute index according to the speculated probability and the state probability;
and calculating the importance of the failure of the gas storage well caused by each attribute index according to the failure probability of the gas storage well caused by each attribute index.
Preferably, the method for calculating the estimated probability of the attribute index to cause the failure of the gas storage well comprises the following steps:
n gas storage wells are used as observation objects to calculate importance;
the equation is applied: q (d) = |n d The presumption probability that the gas storage well fails is caused by the I/N computing attribute index, Q (d) in the equation is the presumption probability, and N d The i is the number of observation objects whose attribute index d causes failure, and the i N is the total number of observation objects.
Preferably, the method for calculating the state probability of the gas storage well failure when the value of the attribute index is y comprises the following steps:
the equation is applied:calculating the state probability of gas storage well failure when the value of the attribute index is y, and calculating Q (y|d j ) Is attribute index d j When the value of (2) is y, the probability of a state that leads to failure of the gas storage well is +.>Is attribute index d j The number of observations that lead to failure of the gas storage well when the value of (2) is y, -, is given by>Is attribute index d j The number of observed objects that caused the failure of the gas storage well.
Preferably, the method for calculating the probability of failure of the gas storage well caused by each attribute index according to the estimated probability and the state probability comprises the following steps:
according to the equation:calculating the probability of failure of the gas storage well caused by each attribute index, wherein j is a preset sequence code of each attribute index in the equation, O is the number of the attribute indexes, and y is d j Z (j) is the probability of each attribute index causing the failure of the gas storage well, Q (d) is the probability of each attribute index causing the failure of the gas storage well, Q (y|d) j ) Is attribute index d j Where y is the probability of a condition that would lead to failure of the gas storage well.
Preferably, the method for calculating the importance of each attribute index to cause the failure of the gas storage well according to the probability of each attribute index to cause the failure of the gas storage well comprises the following steps:
the equation is applied:calculating importance degree of failure of gas storage well caused by each attribute index, and Z is in equation j Is the probability of each attribute index to cause the failure of the gas storage well, j is the preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause failure of the gas storage well.
Preferably, the method for constructing the failure value equation according to the importance and the critical quantity comprises the following steps:
the failure value equation is constructed as:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is the nonce of the attribute index.
Preferably, the method for calculating the probability of failure of the gas storage well caused by each attribute index according to the estimated probability and the state probability comprises the following steps:
acquiring the value of the attribute index of the gas storage well with failure as a detection value;
Respectively calculating failure values according to the detection values and the collected time sequence;
gradually configuring the critical quantity of each attribute index until the failure value is zero when the failure value is gradually close to zero after being lower than zero;
and taking the configured critical quantity of each attribute index as the critical quantity of the modified attribute index.
Preferably, the method further comprises:
calculating the average of the fluctuation amounts of the attribute indexes after the collection time interval of C times
By usingRepresenting the variation of the attribute index after the collection time interval of C times.
The value of each collection attribute index has a collection time interval (namely, the time interval between the values of the adjacent 2 collection attribute indexes), the time interval is set to be U, the fluctuation amount of an attribute index d in an attribute index group in the time interval U is defined as delta y, the value of the attribute index is not increased according to the proportion of the first power, the value of the attribute index after the time interval U can be accurately predicted, the average of the fluctuation amount of the last C times can be obtained, and then the value of the attribute index after the time interval U can be accurately predicted, and the equation is as follows:within the equation is a variation of the attribute index after the time interval U, if the value of the current attribute index d is y 1 The value of d after the time interval U is y 2 The value of d after time distance U is then: / >The values of the attribute index d after the prediction minutes by the above method are: />
Preferably, the method for constructing the failure prediction mode according to the fluctuation amount, the failure value equation and the corrected critical amount of each attribute index comprises the following steps:
loss of gripThe effective prediction mode is configured to:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is the present value of the attribute index,/->Is the fluctuation of the attribute index after the collection time interval of C times.
Preferably, the method for predicting the residual time period of the service life of the gas storage well, which is the service life of the gas storage well, according to the failure prediction mode comprises the following steps:
calculating corresponding failure values of all the collecting time points;
constructing a regression line according to the calculated failure value and the corresponding collection time point;
calculating the mean value of failure values in a preset period;
and calculating the residual time period of the failure of the gas storage well according to the failure value average and the regression line.
Preferably, the method for calculating the corresponding failure value of each collection time point comprises the following steps:
collecting the value of the attribute index of the gas storage well again, and registering the collecting time point;
And calculating corresponding failure values of all the collection time points according to the failure prediction mode.
Preferably, the method for predicting the residual time period of the gas storage well failure as the life of the gas storage well according to the failure prediction mode comprises the following steps:
and constructing a regression line according to the calculated failure value and the corresponding collection time point.
Preferably, the method for calculating the failure value average in the preset time period comprises the following steps:
setting a variable time interval M according to the current time point;
summing up the average of all attribute indexes in the variable time interval L;
and calculating the mean of failure values according to the mean of the attribute indexes and the failure prediction mode.
Preferably, the method for calculating the mean of failure values according to the mean of the attribute indexes and the failure prediction mode comprises the following steps:
calculating a collection time point corresponding to the failure value average according to the failure value average and the regression line;
and calculating the residual time period of the gas storage well with failure according to the collection time point corresponding to the failure value average value and the regression line.
Preferably, the method for calculating the residual time period of the failure of the gas storage well according to the collection time point corresponding to the failure value average and the regression line comprises the following steps:
and calculating the time point when the failure value is zero according to the regression line.
Preferably, the method for setting the variable time interval M according to the current time point comprises the following steps:
the variable time interval M is set to { time instant-M/2 } to { time instant +M/2}.
Preferably, the method for setting the critical amount of each attribute index includes:
acquiring a preset number of failed gas storage wells;
the values of all attribute indexes when all the gas storage wells fail are summed;
the number of samples of each attribute index is summed;
the value with the highest number of one sample value of each attribute index is taken as the critical quantity of the corresponding attribute index.
A machine learning based gas storage well life assessment device comprising:
the setting module is used for collecting attribute indexes of the gas storage well and setting critical quantities of the attribute indexes;
the operation module is used for calculating the importance degree of the gas storage well failure caused by each attribute index and constructing a failure value equation according to the importance degree and the critical quantity;
the correction module is used for correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
and the prediction module is used for constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode.
Compared with the prior art, the method has the advantages that the attribute indexes of the gas storage well are collected, and the critical quantity of each attribute index is set; calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity; correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value; according to the method for predicting the residual time period of the gas storage well failure according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, the failure prediction can be performed on the gas storage wells with various specifications, hidden dangers caused by failure accidents of the gas storage wells are effectively reduced, workers can be prompted in advance to replace the gas storage wells which are about to fail, the defect of strong subjective randomness in the prior art is prevented by using a reasonable prediction method of machine learning, and the life precision of the predicted gas storage wells is high.
Drawings
FIG. 1 is a flow chart of a method for estimating life of a gas storage well based on machine learning according to the present invention;
fig. 2 is a block diagram of a gas storage well life assessment device based on machine learning according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely expressed with reference to the drawings in the embodiments of the present invention. The embodiments expressed in this application are merely examples of some, but not all, of the present invention. All other embodiments, which can be made by those skilled in the art without inventive faculty, are within the scope of the invention.
As shown in fig. 1, the method for evaluating the service life of a gas storage well based on machine learning according to the present invention is operated on a prediction terminal such as a computer, and includes:
step 1: collecting attribute indexes of the gas storage well, and setting critical quantity of each attribute index;
step 2: calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity;
step 3: correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
step 4: and constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode. That is, the attribute indexes are selected initially, the critical quantity of each attribute index is set, the importance of each attribute index is calculated, the critical quantity is corrected, the fluctuation quantity of each attribute index is calculated, the failure prediction mode is constructed according to the fluctuation quantity and the failure equation, and finally the residual time period of the failure of the gas storage well is predicted through the time interval mode.
Therefore, failure prediction can be performed on gas storage wells with various specifications, hidden danger caused by failure accidents of the gas storage wells is effectively reduced, workers can be prompted in advance to replace the gas storage wells which are about to fail, and the defect of strong subjective randomness in the prior art is prevented by using a reasonable prediction method of machine learning, so that the life precision of the predicted gas storage wells is high.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the importance of each attribute index to cause the failure of the gas storage well comprises:
calculating the presumption probability of the failure of the gas storage well caused by the attribute index;
calculating the state probability of gas storage well failure when the value of the attribute index is y;
calculating the probability of failure of the gas storage well caused by each attribute index according to the speculated probability and the state probability;
and calculating the importance of the failure of the gas storage well caused by each attribute index according to the failure probability of the gas storage well caused by each attribute index.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the probability of the attribute index causing the failure of the gas storage well comprises the following steps:
n gas storage wells are used as observation objects to calculate importance; the observation object can comprise a gas storage well that has failed over time.
The equation is applied: q (d) = |n d The presumption probability that the gas storage well fails is caused by the I/N computing attribute index, Q (d) in the equation is the presumption probability, and N d The i is the number of observation objects whose attribute index d causes failure, and the i N is the total number of observation objects. The probability of N number of observed objects to infer attribute indexes to cause the failure of the gas storage well is defined as Q (d) j ) By observing object N d The representative attribute index is an object group formed by an observation object which causes failure, and the value of the probability before calculation under the condition that the observation object is enough to be observed singly and in the same specification is: q (d) = |n d I/N, Q (d) is the probability of speculation in the equation, N d The i is the number of observation objects whose attribute index d causes failure, and the i N is the total number of observation objects. The attribute indexes include: the method comprises the steps of measuring well wall thickness indexes of different heights of a gas storage well by a plurality of thickness measuring instruments which are equidistantly arranged in the gas storage well from top to bottom, measuring air pressure indexes of different heights of the gas storage well by a plurality of pressure sensors which are equidistantly arranged in the gas storage well from top to bottom, measuring intensity indexes of different heights of the gas storage well by a plurality of intensity measuring instruments which are equidistantly arranged on the inner wall of the gas storage well from top to bottom, measuring soil humidity of different heights of the soil of the gas storage well by a plurality of humidity sensors which are equidistantly arranged in the soil of the gas storage well along the outer wall direction of the gas storage well from top to bottom, wherein the thickness measuring instruments, the pressure sensors, the humidity sensors and the intensity measuring instruments are all connected with a PLC or a single chip microcomputer, the PLC or the single chip microcomputer is also connected with a 4G module, and the PLC or the single chip microcomputer can transmit values measured and transmitted by the thickness measuring instruments, the pressure sensors and the humidity sensors and the intensity measuring instruments to a prediction terminal in the 4G network to perform prediction.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the probability of a state of a gas storage well failure when the value of the attribute index is y comprises:
the equation is applied:calculating the state probability of gas storage well failure when the value of the attribute index is y, and calculating Q (y|d j ) Is attribute index d j When the value of (2) is y, the probability of a state that leads to failure of the gas storage well is +.>Is attribute index d j The number of observations that lead to failure of the gas storage well when the value of (2) is y, -, is given by>Is attribute index d j The number of observed objects that caused the failure of the gas storage well. The probability of the attribute index causing the failure of the gas storage well is defined as Q (d j ) Then the probability of the speculative state of each attribute indicator can be characterized as Q (y|d j ) This Q (y|d) j ) The attribute index for causing the failure of the gas storage well is d j At this time d j The probability that the value of (a) is y is Q (y|d j ) Use->Represents N d The internal attribute index is d j And d j If y is an observation object that causes the gas storage well to fail, then the probability of state can be estimated as: />
In a preferred but non-limiting embodiment of the present invention, the method for calculating the probability of failure of the gas storage well caused by each attribute index according to the estimated probability and the state probability comprises:
according to the equation:calculating the probability of failure of the gas storage well caused by each attribute index, wherein j in the equation is a preset sequence code of each attribute index, and O is the attribute index And y is d j Z (j) is the probability of each attribute index causing the failure of the gas storage well, Q (d) is the probability of each attribute index causing the failure of the gas storage well, Q (y|d) j ) Is attribute index d j Where y is the probability of a condition that would lead to failure of the gas storage well. The situation that the gas storage well is out of order is usually caused by more than one attribute index in the attribute index group, under normal conditions, if the gas storage well is out of order caused by one attribute index, whether the gas storage well is out of order can be easily determined, if the gas storage well is out of order caused by a plurality of attribute indexes, the moment that the gas storage well is out of order is difficult to determine, because the value of each attribute index does not reach the critical quantity epsilon of the index, the gas storage well is often provided with the situations of insufficient functions, gas storage well leakage and the like, so that the situation of failure is presented; to overcome the problem that a plurality of attribute indexes together cause the failure of the gas storage well, the state setting that the states among the attribute indexes are independent is applied, namely, each attribute index in the attribute index group is independent, and the probability that one index in the attribute index group is mapped to the gas storage well to fail is as follows: The value of Z (j) can be represented as the probability that an attribute index in the attribute index group is mapped to the failure occurrence probability of the gas storage well, y in the equation represents the value of the attribute index d, y is higher than zero and lower than a critical quantity, and Q (d) and Q (y|d) are calculated j ) The feed equation can obtain a single attribute index to cause the failure probability of the gas storage well.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the importance of each attribute index to the failure of the gas storage well according to the probability of each attribute index to the failure of the gas storage well comprises:
the equation is applied:calculating importance degree of failure of gas storage well caused by each attribute index, and Z is in equation j Is the probability of each attribute index to cause the failure of the gas storage well, and j is each attribute indexIs the number of attribute indexes, X j Is the importance of each attribute index to cause failure of the gas storage well. Through the above operation, the probability of the failure of the gas storage well caused by each attribute index can be obtained, and the probability is defined as follows: z is Z 1 、Z 2 …, can calculate when the gas storage well became invalid, each attribute index can map the importance that becomes invalid and be: />
In a preferred but non-limiting embodiment of the present invention, the method for constructing a failure value equation according to importance and critical quantity comprises:
The failure value equation is constructed as:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is the present value of the attribute index (the present value can be the value of the corresponding attribute index transmitted by the controller at present). The probability of failure of the gas storage well caused by each attribute index is obtained through the operation, so the equation for detecting the failure value of the gas storage well is as follows: />The failure value of the current gas storage well can be obtained by the operation of the present value and the critical value of each attribute index according to the failure value equation, when the condition of the gas storage well is ordinary (i.e. not failed), the values of the attribute indexes are all lower than the critical value epsilon j So a < 0, then the condition of the reservoir becomes a progression of failure via the ordinary, as the value of a approaches zero progressively.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the probability of failure of the gas storage well caused by each attribute index according to the estimated probability and the state probability comprises:
acquiring the value of the attribute index of the gas storage well with failure as a detection value;
Respectively calculating failure values according to the detection values and the collected time sequence;
gradually configuring the critical quantity of each attribute index until the failure value is zero when the failure value is gradually close to zero after being lower than zero;
and taking the configured critical quantity of each attribute index as the critical quantity of the modified attribute index. The method for setting the critical quantity of the attribute index is very brief, so that the accuracy of the detected value is not high, and the critical quantity is corrected to improve the prediction accuracy of the gas storage well; selecting Y ordinary gas storage wells with function conditions, and collecting the values of the attribute indexes of the single gas storage well from each interval U; the values of the attribute indexes collected for each time are used for calculating the failure value by the equation, the value of the failure value is changed to be close to zero or higher than zero through A < 0 after a period of time, and the following 3 types of conditions often occur in the process:
when A is less than 0, the gas storage well fails; when a=0, the gas storage well fails; when A is more than 0, the gas storage well fails.
In the case of A < 0 and A > 0, if the failure prediction of the gas storage well has errors, the configuration critical quantity is executed according to the collected value of the attribute index when the failure of the gas storage well occurs, so that the failure value is maximized and is close to zero, and only the failure prediction party of the gas storage well can be more accurate; the critical quantity correction method is that the value of the attribute index of the gas storage well, which is collected in the above, is extracted to be defined as a reference F, the invalid value in the reference F is calculated one by one according to the time sequence of the value collection of the attribute index by using the equation, in the process of calculation according to the time sequence of the reference data, the invalid value A is gradually close to zero after the initial value A < 0, and the (critical quantity) of each attribute index is gradually configured to be consistent with that when the gas storage well is invalid, the value is extremely close to zero.
In a preferred but non-limiting embodiment of the present invention, further comprising:
calculating the average of the fluctuation amounts of the attribute indexes after the collection time interval of C times
By usingRepresenting the variation of the attribute index after the collection time interval of C times.
The value of each collection attribute index has a collection time interval (namely, the time interval between the values of the adjacent 2 collection attribute indexes), the time interval is set to be U, the fluctuation amount of an attribute index d in an attribute index group in the time interval U is defined as delta y, the value of the attribute index is not increased according to the proportion of the first power, the value of the attribute index after the time interval U can be accurately predicted, the average of the fluctuation amount of the last C times can be obtained, and then the value of the attribute index after the time interval U can be accurately predicted, and the equation is as follows:within the equation is a variation of the attribute index after the time interval U, if the value of the current attribute index d is y 1 The value of d after the time interval U is y 2 The value of d after time distance U is then: />The values of the attribute index d after the prediction minutes by the above method are: />
In a preferred but non-limiting embodiment of the present invention, a method for constructing a failure prediction mode according to a variation amount of each attribute index, a failure value equation and a corrected critical amount includes:
The failure prediction mode is configured to:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is the present value of the attribute index,/->Is the fluctuation of the attribute index after the collection time interval of C times. The fluctuation of the attribute indexes is obtained by the method, the value of random attribute indexes in the attribute index group can be predicted by the same method, namely the value of A after N multiplied by U can be predicted, so that the failure prediction mode of the gas storage well can be characterized as follows: />When n=0, the above equation obtains an actual failure value at the current time point, and when N > 0, the above equation obtains a predicted failure value after n×u.
In a preferred but non-limiting embodiment of the present invention, the method for predicting the size of a residual period of time as the life of a gas storage well in which a failure occurs according to a failure prediction mode comprises:
calculating corresponding failure values of all the collecting time points; the collection time point is the time point when the PLC or the singlechip transmits the value of the attribute index to the prediction terminal.
Constructing a regression line according to the calculated failure value and the corresponding collection time point;
Calculating the mean value of failure values in a preset period;
and calculating the residual time period of the failure of the gas storage well according to the failure value average and the regression line.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the corresponding failure value of each collection time point includes:
collecting the value of the attribute index of the gas storage well again, and registering the collecting time point;
and calculating corresponding failure values of all the collection time points according to the failure prediction mode.
Collecting the value of the attribute index once every time interval, continuously collecting the value of the attribute index of a cluster of gas storage wells, and registering the value of the attribute index to obtain a time point u; the collected values of the attribute indexes can be used for obtaining corresponding failure values by the operation of the failure prediction mode, and the time point u obtained by the collection and operation of the values of the attribute indexes at a certain period is connected with the failure value A of the gas storage well by the operation ofThe upper failure prediction mode can calculate the failure value A after a period of time N 。
In a preferred but non-limiting embodiment of the present invention, a method for predicting the magnitude of a residual period of time as the life of a gas storage well in which a gas storage well fails according to a failure prediction mode comprises:
and constructing a regression line according to the calculated failure value and the corresponding collection time point.
Through the connection between the failure value and the collection time point, a regression line can be obtained by using a least square method, wherein the value of the X axis of the rectangular coordinate system is u, and the value of the Y axis of the rectangular coordinate system is A.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the mean of failure values in a preset time period includes:
setting a variable time interval M according to the current time point;
summing up the average of all attribute indexes in the variable time interval L;
and calculating the mean of failure values according to the mean of the attribute indexes and the failure prediction mode.
The value of the attribute index of the gas storage well is increased when the gas storage well is accompanied, the increase of the index value is not proportional to the increase at one time, and the gas storage well is only increased by one point (u j ，A j ) To predict that the magnitude of the time period remaining from the time of failure of the gas storage well will have a small deviation.
The method uses the value of the attribute index in a period to execute the calculation of the size of the period in which the failure occurs and the period remains, and the size of the period remains can be used as a time interval, namely a variable time interval, wherein the size of the time interval is M, and the variable time interval is added in the regression line; if the values of the attribute indexes are collected in the variable time interval M for L times, then the average of one attribute index is: The above prediction modes can be equivalent to:
according to the equation, the mean value of failure value can be calculated>
In a preferred but non-limiting embodiment of the present invention, the method for calculating the mean of failure values according to the mean of the attribute indexes and the failure prediction mode includes:
calculating a collection time point corresponding to the failure value average according to the failure value average and the regression line;
and calculating the residual time period of the gas storage well with failure according to the collection time point corresponding to the failure value average value and the regression line.
In a preferred but non-limiting embodiment of the present invention, the method for calculating the residual time period of the gas storage well failing according to the collection time point corresponding to the failure value average and the regression line includes:
calculating the time point when the failure value is zero according to the regression line;
the difference obtained by subtracting the corresponding collection time point of the mean value of the failure value from the time point when the failure value is zero is taken as the residual time period of the failure of the gas storage well. According to the mean value of the failure values obtained by the above operation, the corresponding time points can be obtained through the regression line, so the time points in a variable time interval and the point of the failure values can be obtained as follows:at the time point->Towards point (u) k 0) (the point is the time point when the gas storage well fails), wherein the time span required during the movement is the time period left by the failure of the gas storage well distance; the time point that the failure value of the gas storage well is zero can be obtained through the regression line, and the time period from the time point of the average value of the failure value to the time point that the failure value is zero is the residual time period of the gas storage well to be failed; the value of the gas storage well attribute index collected at random time point is that the failure value of the gas storage well is A, if the time period of the gas storage well interval failure is predicted, the point (u, A) is fed into a variable time interval M, wherein the size of the variable time interval is u-M 2 to u+M/2, and calculating corresponding failure values of the attribute indexes collected in the variable time interval M by using an equation prediction mode as a basis +.>The corresponding time point of the regression line can be obtained>At->The time period required by the gas storage well is the residual time period for the gas storage well to fail.
In a preferred but non-limiting embodiment of the present invention, a method for setting a variable time interval M according to a current time point includes:
the variable time interval M is set to { time instant-M/2 } to { time instant +M/2}.
In a preferred but non-limiting embodiment of the present invention, a method for setting a critical amount of each attribute index comprises:
acquiring a preset number of failed gas storage wells;
the values of all attribute indexes when all the gas storage wells fail are summed;
the number of samples of each attribute index is summed;
the value with the highest number of one sample value of each attribute index is taken as the critical quantity of the corresponding attribute index.
The method comprises the steps that a set number of invalid gas storage well equipment is arranged in an observed object, if the number of the observed objects is high enough, only one index in an attribute index set is considered, the invalid of a gas storage well is determined only by an index d, and the value of the index d is y when the gas storage well is registered to be invalid; the number of gas storage wells registered as dead at this time is not high when the value of y is not high, or else the number of gas storage wells registered as dead is not high when the value of y is high, since most gas storage wells are already dead when the value is reached. Through the analysis, the index value and the failure number of the gas storage well are connected, and the connection is that: failure occurs when the value of y is closer to the average The higher the number of gas storage wells, the greater the number of gas storage wells that fail when y=v, the greater the number of gas storage wells that fail zg The highest failure number is that the mapping is closer to the critical quantity of the attribute indexes, and the critical quantity of each attribute index can be roughly determined by using the method.
By applying the scheme of the invention, failure prediction can be performed on the gas storage wells with various specifications, the hidden trouble that gas leakage of the gas storage well and downtime of a server occur due to the failure of the gas storage well is effectively reduced, and workers can be prompted in advance to replace the gas storage well which is about to fail.
As shown in fig. 2, the device for evaluating the service life of a gas storage well based on machine learning according to the present invention comprises:
the setting module is used for collecting attribute indexes of the gas storage well and setting critical quantities of the attribute indexes;
the operation module is used for calculating the importance degree of the gas storage well failure caused by each attribute index and constructing a failure value equation according to the importance degree and the critical quantity;
the correction module is used for correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
and the prediction module is used for constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode.
Compared with the prior art, the method has the advantages that the attribute indexes of the gas storage well are collected, and the critical quantity of each attribute index is set; calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity; correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value; according to the method for predicting the residual time period of the gas storage well failure according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, the failure prediction can be performed on the gas storage wells with various specifications, hidden dangers caused by failure accidents of the gas storage wells are effectively reduced, workers can be prompted in advance to replace the gas storage wells which are about to fail, the defect of strong subjective randomness in the prior art is prevented by using a reasonable prediction method of machine learning, and the life precision of the predicted gas storage wells is high.
The present disclosure can be a system, method, and/or computer program product. The computer program product can include a computer-readable backup medium having computer-readable program instructions embodied thereon for causing a processor to accomplish each aspect of the present disclosure.
The computer readable backup medium can be a tangible power grid line capable of holding and backing up instructions for execution of the power grid line exercise by the instructions. The computer readable backup medium can be, but is not limited to, an electrical backup power grid line, a magnetic backup power grid line, an optical backup power grid line, an electromagnetic backup power grid line, a semiconductor backup power grid line, or any suitable combination of the foregoing. Still further examples (non-enumerated list) of the computer-readable backup medium include: portable computer disk, hard disk, random access backup (RAM), read-only backup (ROM), erasable programmable read-only backup (EPROM or flash memory), static random access backup (SRAM), portable compact disk read-only backup (HD-ROM), digital versatile disk (DXD), memory stick, floppy disk, mechanical coded electrical wiring, punch card like with instructions backed up thereon, or bump structures in grooves, optionally in combination with the above. The computer-readable backup medium as used herein is not to be construed as a transitory signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (as an optical pulse through a transmission line cable), or an electrical signal transmitted through an electrical wire.
The computer readable program instructions expressed herein can be downloaded from a computer readable backup medium to each of the extrapolated/processed power grid lines, or downloaded to an external computer or external backup power grid line via a network, like the internet, a local area network, a wide area network, and/or a wireless network. The network can include copper transmission cables, transmission lines, wireless transmissions, routers, firewalls, switches, gateway computers and/or edge servers. The network interface card or network interface in each of the extrapolated/processed power grid lines receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable backup medium in each of the extrapolated/processed power grid lines.
The computer program instructions for performing the operations of the present disclosure can be assembler instructions, instruction set architecture (lSA) instructions, machine-related instructions, microcode, firmware instructions, conditional setting values, or source or destination code written in a random convergence of one or more programming languages, including an object oriented programming language such as Sdalltara, H++ or the like, as opposed to conventional procedural programming languages, such as the "H" programming language or similar programming languages. The computer readable program instructions can be executed entirely on the client computer, partly on the client computer, as a stand-alone software package, partly on the client computer and partly on a remote computer or entirely on the remote computer or server. In the case of remote computers, the remote computer can be connected to the client computer through a random network, including a local area network (LAb) or a wide area network (WAb), or can be connected to an external computer (as if an internet service provider were used to connect through the internet). In some embodiments, each aspect of the disclosure is achieved by personalizing an electronic circuit, like a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA), with a status value of computer readable program instructions, the electronic circuit being capable of executing the computer readable program instructions.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the above embodiments, it should be understood by those of ordinary skill in the art that: modifications and equivalents may be made to the specific embodiments of the invention without departing from the spirit and scope of the invention, and any modifications and equivalents are intended to be encompassed within the scope of the claims.
Claims (10)
1. A machine learning based gas storage well life assessment method, comprising:
step 1: collecting attribute indexes of the gas storage well, and setting critical quantity of each attribute index;
step 2: calculating the importance degree of the gas storage well failure caused by each attribute index, and constructing a failure value equation according to the importance degree and the critical quantity;
step 3: correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
step 4: and constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode.
2. The machine learning based method for estimating life of a gas storage well according to claim 1, wherein the method for calculating importance of each attribute index causing failure of the gas storage well comprises:
Calculating the presumption probability of the failure of the gas storage well caused by the attribute index;
calculating the state probability of gas storage well failure when the value of the attribute index is y;
calculating the probability of failure of the gas storage well caused by each attribute index according to the speculated probability and the state probability;
and calculating the importance of the failure of the gas storage well caused by each attribute index according to the failure probability of the gas storage well caused by each attribute index.
3. The machine learning based method for estimating life of a gas storage well according to claim 2, wherein the method for computing the probability of the gas storage well failing due to the attribute index comprises:
n gas storage wells are used as observation objects to calculate importance;
the equation is applied: q (d) = |n d The presumption probability that the gas storage well fails is caused by the I/N computing attribute index, Q (d) in the equation is the presumption probability, and N d I is the number of observations that the attribute index d causes to fail, |n| is the total number of objects observed;
a method of calculating a probability of a condition that results in a failure of a gas storage well when the value of the attribute index is y, comprising:
the equation is applied:calculating the state probability of gas storage well failure when the value of the attribute index is y, and calculating Q (y|d j ) Is attribute index d j When the value of (2) is y, the probability of a state that leads to failure of the gas storage well is +. >Is attribute index d j The number of observations that lead to failure of the gas storage well when the value of (2) is y, -, is given by>Is attribute index d j The number of observed objects that caused the failure of the gas storage well.
4. The method for estimating life of a gas storage well based on machine learning according to claim 2, wherein the method for calculating the probability of failure of the gas storage well by each attribute index according to the estimated probability and the state probability comprises:
according to the equation:calculating the probability of failure of the gas storage well caused by each attribute index, wherein j is a preset sequence code of each attribute index in the equation, O is the number of the attribute indexes, and y is d j Z (j) is the probability of each attribute index causing the failure of the gas storage well, Q (d) is the probability of each attribute index causing the failure of the gas storage well, Q (y|d) j ) Is attribute index d j When the value of (2) is y, the state probability of causing the failure of the gas storage well is provided;
the method for calculating the importance degree of each attribute index to the failure of the gas storage well according to the probability of each attribute index to the failure of the gas storage well comprises the following steps:
the equation is applied:calculating importance degree of failure of gas storage well caused by each attribute index, and Z is in equation j Is the probability of each attribute index to cause the failure of the gas storage well, j is the preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause failure of the gas storage well.
5. The machine learning based gas storage well life assessment method of claim 1, wherein the method for constructing the failure value equation according to the importance and the critical quantity comprises:
the failure value equation is constructed as:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is a present value of the attribute index;
the method for calculating the probability of failure of the gas storage well caused by each attribute index according to the speculated probability and the state probability comprises the following steps:
acquiring the value of the attribute index of the gas storage well with failure as a detection value;
respectively calculating failure values according to the detection values and the collected time sequence;
gradually configuring the critical quantity of each attribute index until the failure value is zero when the failure value is gradually close to zero after being lower than zero;
and taking the configured critical quantity of each attribute index as the critical quantity of the modified attribute index.
6. The machine learning based gas storage well life assessment method of claim 1, further comprising:
C times of calculationAverage of fluctuation amounts of attribute indexes after time interval
By usingRepresenting the variation of the attribute index after the collection time interval of C times.
7. The machine learning based gas storage well life assessment method according to claim 1, wherein the method for constructing a failure prediction mode according to the variation amount, the failure value equation and the corrected critical amount of each attribute index comprises:
the failure prediction mode is configured to:a in the equation is a failure value, j is a preset sequence code of each attribute index, O is the number of the attribute indexes and X j Is the importance of each attribute index to cause the failure of the gas storage well, epsilon j Is the critical quantity of each attribute index, y j Is the present value of the attribute index,/->The fluctuation of the attribute index after the collection time interval is C times;
a method of predicting a residual period of time for a gas storage well to fail as a life of the gas storage well according to a failure prediction mode, comprising:
calculating corresponding failure values of all the collecting time points;
constructing a regression line according to the calculated failure value and the corresponding collection time point;
calculating the mean value of failure values in a preset period;
calculating the residual time period of the gas storage well with failure according to the mean value of the failure values and the regression line;
The method for calculating the corresponding failure value of each collecting time point comprises the following steps:
collecting the value of the attribute index of the gas storage well again, and registering the collecting time point;
and calculating corresponding failure values of all the collection time points according to the failure prediction mode.
8. The machine learning based method for estimating life of a gas storage well according to claim 1, wherein the method for predicting the residual time period as the life of the gas storage well in which the gas storage well fails according to the failure prediction mode comprises:
and constructing a regression line according to the calculated failure value and the corresponding collection time point.
Preferably, the method for calculating the mean of failure values in a preset period comprises the following steps:
setting a variable time interval M according to the current time point;
aggregate the mean of the various attribute indicators in the variable time interval L;
calculating failure value average according to the average of all attribute indexes and the failure prediction mode;
the method for calculating the mean of failure values according to the mean of all attribute indexes and the failure prediction mode comprises the following steps:
calculating a collection time point corresponding to the failure value average according to the failure value average and the regression line;
and calculating the residual time period of the gas storage well with failure according to the collection time point corresponding to the average value of the failure values and the regression line.
9. The method for estimating a lifetime of a gas storage well based on machine learning according to claim 8, wherein the method for calculating the residual time period of the gas storage well failure according to the collection time point corresponding to the mean value of the failure values and the regression line comprises:
calculating the time point when the failure value is zero according to the regression line;
the method for setting the variable time interval M according to the current time point comprises the following steps:
setting a variable time interval M to { time point-M/2 } to { time point +M/2};
a method of setting a threshold amount for each attribute indicator, comprising:
acquiring a preset number of failed gas storage wells;
the values of all attribute indexes when all the gas storage wells fail are summed;
the number of samples of each attribute index is summed;
the value with the highest number of one sample value of each attribute index is taken as the critical quantity of the corresponding attribute index.
10. A machine learning based gas storage well life assessment device, comprising:
the setting module is used for collecting attribute indexes of the gas storage well and setting critical quantities of the attribute indexes;
the operation module is used for calculating the importance degree of the gas storage well failure caused by each attribute index and constructing a failure value equation according to the importance degree and the critical quantity;
The correction module is used for correcting the critical quantity of each attribute index according to the equation of the detection value and the failure value;
and the prediction module is used for constructing a failure prediction mode according to the fluctuation quantity, the failure value equation and the corrected critical quantity of each attribute index, and predicting the residual time period of the gas storage well, which is the service life of the gas storage well and is invalid, according to the failure prediction mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311513602.2A CN117763941B (en) | 2023-11-14 | 2023-11-14 | Gas storage well service life assessment method based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311513602.2A CN117763941B (en) | 2023-11-14 | 2023-11-14 | Gas storage well service life assessment method based on machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117763941A true CN117763941A (en) | 2024-03-26 |
| CN117763941B CN117763941B (en) | 2024-06-21 |
Family
ID=90324384
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311513602.2A Active CN117763941B (en) | 2023-11-14 | 2023-11-14 | Gas storage well service life assessment method based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117763941B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2896683A1 (en) * | 2014-07-11 | 2016-01-11 | Frantz Maerten | Regional stress inversion using frictional faults |
| US20160146709A1 (en) * | 2014-11-21 | 2016-05-26 | Satyadeep Dey | System for preparing time series data for failure prediction |
| US20160145994A1 (en) * | 2014-11-20 | 2016-05-26 | Petrochina Company Limited | Evaluation Method and Evaluation Device for Water Breakthrough Risk of Production Wells in Aquifer Drive Gas Reservoirs |
| US20170364795A1 (en) * | 2016-06-15 | 2017-12-21 | Akw Analytics Inc. | Petroleum analytics learning machine system with machine learning analytics applications for upstream and midstream oil and gas industry |
| CN108170974A (en) * | 2018-01-10 | 2018-06-15 | 石家庄爱科特科技开发有限公司 | A kind of system efficiency of pumping well durability analysis method |
| CN108170047A (en) * | 2017-12-18 | 2018-06-15 | 华润电力湖北有限公司 | Unit Boundary Failure switching method, device, equipment and computer storage media |
| CN112204225A (en) * | 2018-05-30 | 2021-01-08 | 沙特阿拉伯石油公司 | System and method for predicting shear failure of a formation |
| WO2021096569A1 (en) * | 2019-11-15 | 2021-05-20 | Halliburton Energy Services, Inc. | Value balancing for oil or gas drilling and recovery equipment using machine learning models |
| CN213398329U (en) * | 2020-07-27 | 2021-06-08 | 江苏省特种设备安全监督检验研究院 | In-use gas storage well life monitoring device |
| US20210304056A1 (en) * | 2020-03-25 | 2021-09-30 | International Business Machines Corporation | Learning Parameter Sampling Configuration for Automated Machine Learning |
| US20220066906A1 (en) * | 2020-09-01 | 2022-03-03 | Bmc Software, Inc. | Application state prediction using component state |
| CN114328513A (en) * | 2021-12-20 | 2022-04-12 | 国网天津市电力公司营销服务中心 | Big data attribute importance and identification degree early warning method based on clustering |
| WO2022112991A1 (en) * | 2020-11-30 | 2022-06-02 | Jio Platforms Limited | System and method of predicting failures |
| CN115481548A (en) * | 2021-05-31 | 2022-12-16 | 中国石油天然气股份有限公司 | Method for predicting residual life of oil well casing in variable corrosion environment |
| WO2023093874A1 (en) * | 2022-07-27 | 2023-06-01 | 深圳市城市公共安全技术研究院有限公司 | Building collapse risk assessment method and apparatus, computer device, and storage medium |
-
2023
- 2023-11-14 CN CN202311513602.2A patent/CN117763941B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2896683A1 (en) * | 2014-07-11 | 2016-01-11 | Frantz Maerten | Regional stress inversion using frictional faults |
| US20160145994A1 (en) * | 2014-11-20 | 2016-05-26 | Petrochina Company Limited | Evaluation Method and Evaluation Device for Water Breakthrough Risk of Production Wells in Aquifer Drive Gas Reservoirs |
| US20160146709A1 (en) * | 2014-11-21 | 2016-05-26 | Satyadeep Dey | System for preparing time series data for failure prediction |
| US20170364795A1 (en) * | 2016-06-15 | 2017-12-21 | Akw Analytics Inc. | Petroleum analytics learning machine system with machine learning analytics applications for upstream and midstream oil and gas industry |
| CN108170047A (en) * | 2017-12-18 | 2018-06-15 | 华润电力湖北有限公司 | Unit Boundary Failure switching method, device, equipment and computer storage media |
| CN108170974A (en) * | 2018-01-10 | 2018-06-15 | 石家庄爱科特科技开发有限公司 | A kind of system efficiency of pumping well durability analysis method |
| CN112204225A (en) * | 2018-05-30 | 2021-01-08 | 沙特阿拉伯石油公司 | System and method for predicting shear failure of a formation |
| WO2021096569A1 (en) * | 2019-11-15 | 2021-05-20 | Halliburton Energy Services, Inc. | Value balancing for oil or gas drilling and recovery equipment using machine learning models |
| US20210304056A1 (en) * | 2020-03-25 | 2021-09-30 | International Business Machines Corporation | Learning Parameter Sampling Configuration for Automated Machine Learning |
| CN213398329U (en) * | 2020-07-27 | 2021-06-08 | 江苏省特种设备安全监督检验研究院 | In-use gas storage well life monitoring device |
| US20220066906A1 (en) * | 2020-09-01 | 2022-03-03 | Bmc Software, Inc. | Application state prediction using component state |
| WO2022112991A1 (en) * | 2020-11-30 | 2022-06-02 | Jio Platforms Limited | System and method of predicting failures |
| CN115481548A (en) * | 2021-05-31 | 2022-12-16 | 中国石油天然气股份有限公司 | Method for predicting residual life of oil well casing in variable corrosion environment |
| CN114328513A (en) * | 2021-12-20 | 2022-04-12 | 国网天津市电力公司营销服务中心 | Big data attribute importance and identification degree early warning method based on clustering |
| WO2023093874A1 (en) * | 2022-07-27 | 2023-06-01 | 深圳市城市公共安全技术研究院有限公司 | Building collapse risk assessment method and apparatus, computer device, and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| YIFAN WU等: "Fluctuations at Metal Halide Perovskite Grain Boundaries Create Transient Trap States：Machine Learning Assisted Ab Initio Analysis", ACS PUBLICATIONS, 31 December 2022 (2022-12-31) * |
| 宋成立;淡勇;: "基于有限元分析的CNG储气井疲劳设计计算", 化工机械, no. 03, 15 June 2013 (2013-06-15) * |
| 陆百川;舒芹;马广露;何相;: "基于多属性TOPSIS决策的交通网络路段重要度计算", 浙江工业大学学报, no. 03, 12 June 2020 (2020-06-12) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117763941B (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3903112B1 (en) | System and method for evaluating models for predictive failure of renewable energy assets | |
| CN110868241B (en) | Underground cable fault early warning method and device based on multiple models | |
| US20090096406A1 (en) | Method and system for determining the reliability of a dc motor system | |
| CN110334816B (en) | Industrial equipment detection method, device, equipment and readable storage medium | |
| US10884404B2 (en) | Method of predicting plant data and apparatus using the same | |
| JP2014533484A (en) | Method and apparatus for monitoring high voltage current transmission lines | |
| EP3584657B1 (en) | Risk assessment device, risk assessment method, and risk assessment program | |
| CN110146128A (en) | Nuclear piping monitoring method, system, device, equipment and storage medium | |
| CN114994460A (en) | Cable insulation performance prediction device and method | |
| KR20190124307A (en) | Risk Assessment Apparatus, Risk Assessment Method and Risk Assessment Program | |
| CN114779747B (en) | Vehicle fault cause determining system and method | |
| CN110866634A (en) | Underground cable fault early warning method and device based on model selection | |
| CN114895163A (en) | Cable inspection positioning device and method based on cable insulation performance | |
| JP5582891B2 (en) | Dam measurement data evaluation method and dam measurement data evaluation system | |
| CN117763941B (en) | Gas storage well service life assessment method based on machine learning | |
| CN116165279B (en) | Composite insulator defect detection method, device, electronic equipment and storage medium | |
| CN117171366A (en) | Knowledge graph construction method and system for power grid dispatching operation situation | |
| CN117129815A (en) | Comprehensive detection method and system for multi-degradation insulator based on Internet of things | |
| CN116429161B (en) | Dynamic calibration method for online chemical instrument of thermal power plant | |
| KR102309979B1 (en) | Asset management method for electric power equipment | |
| JP4863918B2 (en) | OF cable abnormality detection device and oil leakage detection method of OF cable | |
| JP2013140080A (en) | Instrument soundness determination device and method | |
| CN116011171A (en) | Transformer service life estimation method and system | |
| CN113960965B (en) | Data processing method, device, electronic equipment and readable storage medium | |
| CN117109523B (en) | GIS equipment settlement monitoring and early warning method and system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |