CN107083951B - Oil and gas well monitoring method and device - Google Patents

Abstract

The invention discloses a method and a device for monitoring an oil-gas well, belonging to the technical field of development and production of oil-gas fields. The method comprises the following steps: acquiring data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set; obtaining at least one data change rate set according to at least one production parameter data set; determining a mathematical expectation of the rate of change of the data of the at least one production parameter; selecting production parameters corresponding to the problems to be diagnosed from at least one production parameter to form a first parameter set according to the diagnosis model; determining mathematical expected values of the data change rate of each production parameter in the first parameter set to form a mathematical expected value set; determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm; determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed; the monitoring efficiency is improved.

Description

Oil and gas well monitoring method and device

Technical Field

The invention relates to the technical field of oil and gas field development and production, in particular to a method and a device for monitoring an oil and gas well.

Background

In the production process of the oil and gas well, operators need to monitor the production performance of the oil and gas well, diagnose problems of the oil and gas well in the production process and pertinently adopt technological measures and production optimization means to solve the problems. For example, in the process of liquid drainage and gas production of an oil and gas well, produced gas often carries liquids such as water and/or condensate, and the liquids are easily condensed into a plug flow, so that the liquids are left at the bottom of the well to form bottom hole accumulated liquids, and the bottom hole accumulated liquids are too much to cause the abandonment of the oil and gas well, so that operators often need to monitor the condition of the oil and gas well in real time.

Whether there is shaft bottom hydrops in adopting the well testing curve method to monitor among the prior art in the oil gas well, specific process can be: the operating personnel firstly carry out open flow liquid discharge treatment on the oil-gas well to remove liquid in the gas pipe in the oil-gas well; then, placing a pressure gauge in the well, periodically measuring the pressure at the bottom of the well and the pressure of each gas producing layer in the well, drawing a curve by using a gas steady-state seepage equation according to the pressure at the bottom of the well and the pressure of each gas producing layer in the well, and comparing the drawn curve with a given normal theoretical curve; when the difference exists between the drawn curve and the normal theoretical curve, determining the pressure of the bottom of the well and the pressure change condition of each gas production layer in the oil-gas well by analyzing the difference between the drawn curve and the normal theoretical curve; judging whether bottom hole accumulated liquid exists in the oil-gas well or not according to the pressure at the bottom of the well and the change conditions of the pressure of each gas production layer; and when the pressure at the bottom of the well and the pressure of each gas producing layer are not in the stable attenuation state, determining that the bottom of the well is accumulated liquid in the oil-gas well.

In the process of implementing the invention, the inventor finds that the prior art has at least the following problems:

because the periodically measured bottom pressure and the pressure of each gas producing layer in the well are time sequence data, an operator needs to manually process the time sequence data and compare and analyze a drawn curve with a normal theoretical curve, and therefore the efficiency of monitoring the oil and gas well in the prior art is low and the timeliness is poor.

Disclosure of Invention

In order to solve the problems of the prior art, the invention provides a method and a device for monitoring an oil and gas well. The technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides a method for monitoring an oil and gas well, the method comprising:

acquiring data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set;

determining the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set;

determining a mathematical expectation of the data rate of change of the at least one production parameter based on the at least one set of data rates of change;

selecting production parameters corresponding to the problem to be diagnosed from the at least one production parameter to form a first parameter set according to a diagnosis model, wherein the diagnosis model is used for storing the corresponding relation between the problem to be diagnosed and the production parameters;

determining the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate;

determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm;

and determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

In one possible design, the determining a mathematical expectation of the data rate of change of the at least one production parameter from the at least one set of data rates of change includes:

acquiring the edge density of each data change rate in each data change rate set;

calculating a mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate, wherein the first data change rate set is any one of the at least one data change rate set:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the at least one set of data rates of change, x_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

In one possible design, before the selecting the first set of parameters corresponding to the problem to be diagnosed from the at least one production parameter, the method further includes:

determining a production parameter corresponding to each problem to be diagnosed in a problem set to be diagnosed, wherein the production parameter corresponding to each problem to be diagnosed is a production parameter defined based on domain knowledge and a preset rule;

and constructing the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

In one possible design, after determining whether the problem to be diagnosed exists in the oil and gas well according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed, the method further includes:

estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

according to the reliability index of the problem to be diagnosed, determining an early warning grade corresponding to the reliability index of the problem to be diagnosed from the corresponding relation between the reliability index and the early warning grade;

and carrying out early warning reminding on the problem to be diagnosed according to the early warning grade.

In one possible design, the predicting the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm includes:

selecting a first production parameter data set from the at least one production parameter data set, wherein the first production parameter data set is a data set of production parameters preset based on domain knowledge and preset rules;

according to the first production parameter data set, determining a clustering object meeting a preset condition through the following formula II, wherein the preset condition is that the value of a norm expression of the first production parameter data set is minimum, and the formula II is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 1, 2, … …, k) is the number one of k clustersj clusters comprise production parameter data, p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jClustered data corresponding to the included production parameter data;

determining trend distribution data of the first production parameter data set according to the clustering object;

determining the reliability index of the problem to be diagnosed according to the trend distribution data and through a third formula, wherein the third formula is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is_i(t) is the reliability index.

In a second aspect, embodiments of the present invention provide an apparatus for hydrocarbon well monitoring, the apparatus comprising:

the acquisition module is used for acquiring data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set;

the first determining module is used for determining the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set;

a second determining module for determining a mathematical expectation of the data rate of change of the at least one production parameter based on the at least one set of data rates of change;

the selection module is used for selecting the production parameters corresponding to the problems to be diagnosed from the at least one production parameter to form a first parameter set according to a diagnosis model, and the diagnosis model is used for storing the corresponding relation between the problems to be diagnosed and the production parameters;

the third determination module is used for determining the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate;

the fourth determining module is used for determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm;

and the fifth determining module is used for determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

In one possible design, the second determining module is further configured to obtain an edge density of each data change rate in each data change rate set; calculating a mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate, wherein the first data change rate set is any one of the at least one data change rate set:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the at least one set of data rates of change, x_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

In one possible design, the apparatus further includes:

a sixth determining module, configured to determine a production parameter corresponding to each to-be-diagnosed problem in the to-be-diagnosed problem set, where the production parameter corresponding to each to-be-diagnosed problem is a production parameter defined based on domain knowledge and a preset rule;

and the building module is used for building the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

In one possible design, the apparatus further includes:

the estimation module is used for estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

a seventh determining module, configured to determine, according to the reliability index of the problem to be diagnosed, an early warning level corresponding to the reliability index of the problem to be diagnosed from a correspondence between the reliability index and the early warning level;

and the early warning module is used for carrying out early warning reminding on the problem to be diagnosed according to the early warning grade.

In one possible design, the estimation module includes:

a selection unit, configured to select a first production parameter data set from the at least one production parameter data set, where the first production parameter data set is a data set of production parameters preset based on domain knowledge and preset rules;

a first determining unit, configured to determine, according to the first production parameter data set, a clustering object that meets a preset condition by using a following formula two, where the preset condition is that a norm expression of the first production parameter data set is minimum, and the formula two is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 6, 2, … …, k) is the production parameter data included in the jth cluster of the k clusters, and p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jClustered data corresponding to the included production parameter data;

the second determining unit is used for determining trend distribution data of the first production parameter data set according to the clustering object;

a third determining unit, configured to determine, according to the trend distribution data, a reliability index of the problem to be diagnosed through a third formula, where the third formula is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is_i(t) is the reliability index.

In the embodiment of the present invention, the server may calculate, after determining at least one data change rate set according to the at least one production parameter data set, a mathematical expected value of a production parameter data change rate corresponding to each data change rate set in the at least one data change rate set; then the server determines that the production parameters corresponding to the problems to be diagnosed form a first parameter set according to the diagnosis model stored in the server, and the mathematical expected values of the data change rate of each production parameter in the first parameter set form a mathematical expected value set; and the server calculates a target mathematical expected value of the first parameter set, and determines whether the oil and gas well has the problem to be diagnosed according to the target mathematical expected value of the first parameter set and a preset mathematical expected value of the problem to be diagnosed. The server converts a plurality of continuously observed production parameter data sets into a plurality of change rate sets with probability statistical characteristics, so that the server can directly calculate the plurality of data change rate sets to determine states such as mathematical expected values of the plurality of production parameter data change rates, the server can synthesize the states of the plurality of production parameters to diagnose the problem to be diagnosed directly according to a diagnosis model without manual processing, and the efficiency of monitoring the oil-gas well is improved; moreover, the server can timely judge whether the oil-gas well has the problem to be diagnosed, so that the manual processing time is saved, and the timeliness of diagnosing the oil-gas well in the process of monitoring the oil-gas well is improved.

Drawings

FIG. 1 is a flow chart of a method of oil and gas well monitoring provided by an embodiment of the present invention;

FIG. 2 is a flow chart of a method of oil and gas well monitoring provided by an embodiment of the present invention;

fig. 3 is a schematic diagram of a clustering object according to an embodiment of the present invention;

fig. 4 is a schematic structural diagram of an oil and gas well monitoring device provided by the embodiment of the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The embodiment of the invention provides a method for monitoring an oil and gas well, wherein an execution main body of the method can be a server, and as shown in figure 1, the method comprises the following steps:

step 101: and acquiring data of at least one production parameter of the oil-gas well every preset time to obtain at least one production parameter data set.

Step 102: and determining the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set.

Step 103: a mathematical expectation of the data rate of change of the at least one production parameter is determined based on the at least one set of data rates of change.

Step 104: and selecting the production parameters corresponding to the problem to be diagnosed from the at least one production parameter to form a first parameter set according to a diagnosis model, wherein the diagnosis model is used for storing the corresponding relation between the problem to be diagnosed and the production parameters.

Step 105: and determining the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate.

Step 106: and determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm.

Step 107: and determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

In one possible design, determining a mathematical expectation of the data rate of change of the at least one production parameter based on the at least one set of data rates of change comprises:

acquiring the edge density of each data change rate in each data change rate set;

calculating a mathematical expectation value of a first set of data change rates according to each data change rate in each set of data change rates and the edge density of each data change rate, wherein the first set of data change rates is any one of the at least one set of data change rates by the following formula one:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the data rate of change sets x in the at least one data rate of change set_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

In one possible design, before selecting the first set of parameters corresponding to the problem to be diagnosed from the at least one production parameter, the method further comprises:

determining a production parameter corresponding to each problem to be diagnosed in the problem set to be diagnosed, wherein the production parameter corresponding to each problem to be diagnosed is a production parameter defined based on domain knowledge and a preset rule;

and constructing the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

In one possible design, after determining whether the problem to be diagnosed exists in the oil and gas well according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed, the method further includes:

estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

according to the reliability index of the problem to be diagnosed, determining the early warning grade corresponding to the reliability index of the problem to be diagnosed from the corresponding relation between the reliability index and the early warning grade;

and carrying out early warning reminding on the problem to be diagnosed according to the early warning grade.

In one possible design, the predicting the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second predetermined algorithm includes:

selecting a first production parameter data set from the at least one production parameter data set, wherein the first production parameter data set is a data set of production parameters preset based on domain knowledge and preset rules;

according to the first production parameter data set, determining a clustering object meeting a preset condition through the following formula II, wherein the preset condition is that the value of a norm expression of the first production parameter data set is minimum, and the formula II is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 1, 2, … …, k) is production parameter data included in the jth cluster of the k clusters, and p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jAfter clustering corresponding to included production parameter dataThe data of (a);

determining trend distribution data of the first production parameter data set according to the clustering object;

according to the trend distribution data, determining the reliability index of the problem to be diagnosed by the following formula III, wherein the formula III is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is the acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is the maximum acquisition time corresponding to the trend distribution data when the value is zero_i(t) is the reliability index.

In the embodiment of the present invention, the server may calculate, after determining at least one data change rate set according to the at least one production parameter data set, a mathematical expected value of a production parameter data change rate corresponding to each data change rate set in the at least one data change rate set; then the server determines that the production parameters corresponding to the problems to be diagnosed form a first parameter set according to the diagnosis model stored in the server, and the mathematical expected values of the data change rate of each production parameter in the first parameter set form a mathematical expected value set; and the server calculates a target mathematical expected value of the first parameter set, and determines whether the oil and gas well has the problem to be diagnosed according to the target mathematical expected value of the first parameter set and a preset mathematical expected value of the problem to be diagnosed. The server converts a plurality of continuously observed production parameter data sets into a plurality of change rate sets with probability statistical characteristics, so that the server can directly calculate the plurality of data change rate sets to determine states such as mathematical expected values of the plurality of production parameter data change rates, the server can synthesize the states of the plurality of production parameters to diagnose the problem to be diagnosed directly according to a diagnosis model without manual processing, and the efficiency of monitoring the oil-gas well is improved; moreover, the server can timely judge whether the oil-gas well has the problem to be diagnosed, so that the manual processing time is saved, and the timeliness of diagnosing the oil-gas well in the process of monitoring the oil-gas well is improved.

The embodiment of the invention provides a method for monitoring an oil and gas well, wherein an execution main body of the method can be a server, and as shown in figure 2, the method comprises the following steps:

step 201: the server acquires data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set.

In the embodiment of the invention, the server diagnoses the possible problems of the oil-gas well in the production process by acquiring at least one production parameter of the oil-gas well in real time, so as to further pre-warn the problems, and workers can take technical measures and production optimization measures in a targeted manner to solve the problems.

For example, taking the problem of diagnosing whether the produced gas well has bottom hole liquid accumulation as an example, the problem of the bottom hole liquid accumulation is that in the production process of the gas well, liquid carried in the produced gas is condensed into a broken plug flow along with the reduction of the pressure and the speed of an oil pipe, and the liquid is left at the bottom hole to form the bottom hole liquid accumulation, so that the yield of the gas well is reduced and even the gas well is scrapped; after the sensor collects data of production parameters such as gas production rate, oil pipe pressure and the like of the production gas well in real time, the server can obtain the data in real time/quasi-real time and analyze the data of the production parameters such as the gas production rate, the oil pipe pressure and the like of the oil gas well, so that whether bottom hole liquid accumulation exists in the current gas well or not is diagnosed.

In this step, the server may obtain data of at least one production parameter of the oil and gas well through the acquisition device connected thereto, periodically obtain data of at least one production parameter of the oil and gas well every preset time period, and form a production parameter data set from the data of the at least one production parameter periodically obtained within the preset time period, thereby obtaining at least one production parameter data set.

Wherein, the production parameters can be the gas production rate, the liquid production amount, the oil pipe pressure, the casing pipe pressure and the like of the oil-gas well. The preset duration can be set and changed according to the user requirement, and the embodiment of the invention is not particularly limited to this. For example, the preset time period may be 1 day, 1 month, or the like.

Step 202: and the server determines the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set.

In this step, since the at least one production parameter data set is time series data periodically acquired by the server, for the convenience of subsequent analysis, the server transforms the at least one production parameter data set into a data set satisfying a probability statistical rule through a data transformation or calculation, such as a difference or differential algorithm, specifically, the server calculates the data change rate of two adjacent production parameters in each production parameter data set according to the at least one production parameter data set by the following formula four, thereby obtaining at least one data change rate set,

the formula four is as follows:

wherein y (t) is any one of the at least one production parameter data set, t is the acquisition time, dy (t) is the derivative of y (t), dt is the derivative of t, y (t)' is the derivative of y (t), Δ y (t)_iIs the ith production parameter data y in y (t)_iIncrement of, Δ t_iFor the ith acquisition time t in t_iIncrement of (a), y_iFor the ith production parameter data, t, in the production parameter data set_iIs the ith acquisition time in t; dy (t) ═ y (t)' dt is applied to the production parameter data set in which the data distribution is a continuous function among the at least one production parameter data set;

the method is suitable for the production parameter data set of which the data distribution in the at least one production parameter data set is a discrete function.

The data change rate set is determined according to the data change rates of two adjacent production parameters in each production parameter data set, and thus the data change rate set is also the data change rate set corresponding to the production parameters.

It should be noted that each data change rate set in the at least one data change rate set is data satisfying a normal distribution property. In this step, since the production parameter data set is a kind of time series data, it is difficult for the server to directly perform subsequent calculation analysis based on statistical significance and diagnose possible problems in the oil and gas well according to the time series data. Therefore, the server performs a change rate transformation on the at least one production parameter data set to obtain change rate data, and the change rate data has the property of probabilistic statistical analysis, that is, the continuous change rate data satisfies the property of normal distribution.

Thus, the server obtains at least one set of data rates of change over time, and obviously, based on the at least one set of data rates of change over time, the sensitivity of the data set to the event response can be improved.

In the embodiment of the present invention, the server may further convert the production parameter data set into other data sets, and the other data sets only need to have a property of probability statistical analysis, that is, a property of normal distribution is satisfied, which is not specifically limited in the embodiment of the present invention.

Step 203: the server determines a mathematical expectation of the data rate of change of the at least one production parameter based on the at least one set of data rates of change.

In the embodiment of the invention, under the condition that the data change rate sets meet the normal distribution property, the state of each data change rate set can be represented by the mathematical expectation of each data change rate set, so that the quantitative description of the data state of the production parameters is realized.

Specifically, the steps may be: the server obtains the edge density of each data change rate in each data change rate set, and calculates the mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate through the following formula one, wherein the first data change rate set is any one data change rate set in at least one data change rate set:

the formula I is as follows:

wherein m is_xMathematical expectation, x, for any of the at least one set of data rates of change, x_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

It should be noted that the mathematical expected value of the at least one data change rate of the production parameter is a mathematical expected value calculated according to a first data change rate set corresponding to a data set of the production parameter within a preset time period, and therefore, the mathematical expected value of the production parameter is a mathematical expected value of the data change rate set corresponding to the production parameter.

It should be noted that the mathematical expectation value of the data change rate set, i.e. the first moment of the data change rate set, herein represents the state of the data change rate set in the statistical sense, i.e. when m is_xIf the data change rate set is less than 0, the data change rate set is in a descending trend, namely the data in the data change rate set gradually decreases along with the time; when m is_xIf the data change rate set is more than 0, the data change rate set is in an ascending trend, namely the data in the data change rate set is gradually increased along with the time; when m is_x0 means that the trend of the data change rate set is stable, i.e. the data in the data change rate set remains unchanged.

In the embodiment of the present invention, the mathematical expected value of the data change rate set may be set and changed according to a user requirement, which is not specifically limited in the embodiment of the present invention. For example, the mathematical expectation value is represented by a second moment or the like.

Step 204: and the server selects the production parameters corresponding to the problems to be diagnosed from the at least one production parameter to form a first parameter set according to the diagnosis model, wherein the diagnosis model is used for storing the corresponding relation between the problems to be diagnosed and the production parameters.

In the embodiment of the invention, the server constructs the diagnosis model in advance, and the diagnosis model stores the corresponding relation between the problems to be diagnosed and the production parameters, so that the server can directly utilize the diagnosis model to diagnose the problems existing in the oil-gas well.

Thus, the server may construct the diagnostic model by: the server determines a production parameter corresponding to each problem to be diagnosed in the problem set to be diagnosed, wherein the production parameter corresponding to each problem to be diagnosed is a production parameter defined based on domain knowledge and a preset rule; and the server constructs the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

Specifically, in step 204, the server obtains the current problem to be diagnosed, and determines the production parameter corresponding to the problem to be diagnosed from the corresponding relationship between the problem to be diagnosed and the production parameter stored in the diagnosis model according to the problem to be diagnosed, and in order to improve the accuracy of diagnosis, the server can diagnose through a plurality of production parameters, so that the problem to be diagnosed can correspond to a plurality of production parameters, and the server forms the plurality of production parameters corresponding to the problem to be diagnosed into the first parameter set.

In a possible design provided by the embodiment of the present invention, the server may set a state layer, a symptom layer, and a diagnosis layer in the diagnosis model, and the server may store a mathematical expected value of at least one data change rate of the production parameter in the state layer, and store a mathematical expected value of a first parameter set corresponding to the problem to be diagnosed in the symptom layer, and in addition, the diagnosis layer mainly stores a corresponding relationship between the problem to be diagnosed and the production parameter, a corresponding relationship between the problem to be diagnosed and a preset algorithm, and a corresponding relationship between the problem to be diagnosed and a preset mathematical expected value, so that the server in the subsequent step 205 and 206 determines whether the problem to be diagnosed exists in the oil and gas well according to the mathematical expected value of the data change rate of the production parameter.

Correspondingly, the steps can be as follows: the server obtains the corresponding relation between the problem to be diagnosed and the production parameters from the diagnosis layer, determines the production parameters corresponding to the problem to be diagnosed from the corresponding relation between the problem to be diagnosed and the production parameters, and combines the production parameters of the problem to be diagnosed into a first parameter set.

Specifically, the diagnostic Model MOD includes a state layer S, a symptom layer a, and a diagnostic layer D, i.e., MOD ═ m<S，A，D>Wherein the server stores the mathematical expected value of the data change rate of the at least one production parameter determined in step 201-₁，S₂，…，S_i，…S_n}，S_iThe mathematical expectation value of the data change rate of the ith production parameter in the n production parameters is obtained. The diagnostic layer D stores the correspondence between the problem to be diagnosed and the production parameter, i.e., D ═ D₁，D₂，…D_q，…D_j"field knowledge and preset rules for j fields stored by the server, D_qThe q domain knowledge and the preset rule in the j domain knowledge and the preset rule are set; for the genus D_qThe server may define a binary formula Q ═ for the diagnosis of the problem Q to be diagnosed, Q being any one of the j problems to be diagnosed<MOD，R_d>Wherein R is_dA first set of parameters, R, corresponding to the problem Q to be diagnosed_dM production parameters, R_d＝{R₁，R₂，…，R_i，…，R_m}，R_iThe ith production parameter in the first parameter set corresponding to the problem Q to be diagnosed.

Step 205: the server determines the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate, and determines the target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm.

Since the server determines the mathematical expected values of the change rates of the at least one production parameter data in step 203, the step of determining, by the server, the mathematical expected values of the change rates of each production parameter data in the first parameter set to form the mathematical expected value set according to the mathematical expected values of the change rates of the at least one production parameter data may be: the server obtains the mathematical expectation value of the data change rate of the production parameter from the mathematical expectation value of the data change rate of at least one production parameter according to the production parameter included in the first parameter set, and the mathematical expectation value of each data change rate of the production parameter in the first parameter set is combined into a mathematical expectation value set.

In this step, the server defines a preset algorithm for calculating a target mathematical expected value of each parameter set in advance, and stores a corresponding relationship between the problem to be diagnosed and the preset algorithm, so that the step of determining the target mathematical expected value of the first parameter set according to the mathematical expected value set and the first preset algorithm may be: the server acquires a preset algorithm corresponding to the current problem to be diagnosed from the corresponding relation between the problem to be diagnosed and the preset algorithm according to the problem to be diagnosed, the preset algorithm is used as a first preset algorithm, a target mathematical expected value of the mathematical expected value set is calculated through the first preset algorithm, and the target mathematical expected value is used as a target mathematical expected value of a first parameter set corresponding to the mathematical expected value set.

In one possible design provided by the embodiments of the present invention, the diagnostic model MOD is<S，A，D>The server may store the mathematical expected value set corresponding to the first parameter set in the symptom layer, that is, symptom layer a ═ { a ═ a-₁，A₂，…A_i，…A_m}，A_iThe mathematical expectation value of the data rate of change of the ith production parameter is included for the m production parameters included in the first set of parameters.

Specifically, for the diagnosis of the problem Q to be diagnosed, Q is any one of j problems to be diagnosed, and the mathematical expected value set a corresponding to the first parameter set of the problem Q to be diagnosed is ═ a₁，A₂，…A_i，…A_mAnd if the preset algorithm corresponding to the problem Q to be diagnosed stored in the diagnosis layer is an averaging operation, the server calculates A as { A ═ A }₁，A₂，…A_i，…A_mMean number of

I.e. the first one corresponding to the problem Q to be diagnosedThe target mathematical expected value of the parameter set is

The first preset algorithm may be set and changed according to the user requirement, and the first preset algorithm is not specifically limited in the embodiment of the present invention. For example, the first predetermined algorithm may also be a summation operation, a weighted average operation, a logic operation, and the like.

In a possible design provided in the embodiment of the present invention, the first parameter set includes a plurality of parameters, and in order to improve efficiency on the premise of ensuring accuracy, the server may further continuously select, from the first parameter set, a parameter with a reference value to form a second parameter set, specifically, the step may be: the server selects parameters meeting preset rules from the first parameter set to form a second parameter set, determines the mathematical expectation value of each production parameter change rate in the second parameter set to form a reference mathematical expectation value set according to the mathematical expectation value of at least one production parameter change rate, and determines the target mathematical expectation value of the second parameter set according to the reference mathematical expectation value set and a third preset algorithm.

The preset rule may be set and changed according to a user requirement, which is not specifically limited in the embodiment of the present invention. For example, the preset rule may be a production parameter related to pressure, if the first parameter set is: { tubing pressure, casing pressure, gas production, production time }, then the second set of parameters may be: { oil pipe pressure, casing pressure }.

Step 206: and the server determines whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

In this step, the server stores the corresponding relationship between the problem to be diagnosed and the preset mathematical expected value, so that the server obtains the preset mathematical expected value of the problem to be diagnosed from the corresponding relationship between the problem to be diagnosed and the preset mathematical expected value, and determines whether the problem to be diagnosed exists in the oil and gas well by comparing the target mathematical expected value of the first parameter set with the preset mathematical expected value.

Specifically, when the server diagnoses different problems to be diagnosed, the parameters corresponding to the different problems to be diagnosed are different, and the way of comparing the target mathematical expected value of the first parameter set with the preset mathematical expected value is also different, which is not specifically limited in the embodiment of the present invention. For example, when the server diagnoses the bottom-hole liquid accumulation problem, the way for the server to compare the target mathematical expected value of the first parameter set corresponding to the bottom-hole liquid accumulation problem with the preset mathematical expected value of the bottom-hole liquid accumulation problem may be: the server judges whether a target mathematical expected value of a first parameter set corresponding to the bottom hole liquid accumulation problem is larger than or equal to a preset mathematical expected value of the bottom hole liquid accumulation problem, if not, the bottom hole liquid accumulation problem exists in the oil-gas well, and otherwise, the server determines that the bottom hole liquid accumulation problem does not exist in the oil-gas well.

Specifically, when the server diagnoses whether bottom hole liquid exists in the oil-gas well, the production parameters of the problem to be diagnosed can be temperature and oil pipe pressure, if the target mathematical expected value of the change rate data set corresponding to the temperature and oil pipe pressure parameters calculated by the server is 100, the preset mathematical expected value corresponding to the bottom hole liquid problem is 200, that is, the target mathematical expected value is smaller than the preset mathematical expected value, and therefore, the server determines that the bottom hole liquid does not exist in the oil-gas well.

It should be noted that the preset mathematical expected value is a mathematical expected value of the problem to be diagnosed, which is defined by the server based on the domain knowledge and the preset rule. The preset mathematical expected value may be set and changed according to the user's needs, which is not particularly limited in the embodiment of the present invention, for example, the preset mathematical expected value may be 100, 2000, and the like.

In the embodiment of the invention, the server determines the target mathematical expected value of the first parameter set according to the mathematical expected values of the data change rates of the multiple production parameters corresponding to the problem to be diagnosed, so that the server can comprehensively diagnose the problem to be diagnosed possibly existing in the oil and gas well according to the states corresponding to the multiple production parameters of the oil and gas well, thereby improving the accuracy of diagnosing the oil and gas well.

In one possible design provided in the embodiment of the present invention, it is determined that a target mathematical expected value of the second parameter set can also be determined, and in this case, the step may further be: and the server determines whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the second parameter set and the preset mathematical expected value of the problem to be diagnosed. The implementation mode of the method is consistent with the implementation mode of determining whether the oil-gas well has the problem to be diagnosed according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed, and the implementation mode is not repeated one by one.

Step 207: and the server predicts the reliability index of the problem according to the at least one production parameter data set and a second preset algorithm.

In the embodiment of the invention, in the process of comprehensively diagnosing a specific problem to be diagnosed by using the states of a plurality of production parameters of an oil and gas well, the production data is influenced by a plurality of factors such as an underground reservoir, a well condition, a production condition and the like, and the production parameter data used in the diagnosis process is an acquired production parameter data set within a preset time, so that different oil and gas wells have different response strengths for the same problem. Therefore, the embodiment of the invention also provides a method for quantitatively evaluating the problem to be diagnosed by estimating the reliability index through the trend analysis of a certain production parameter data set.

Specifically, this step can be realized by the following steps 2071 and 2072.

Step 2071: the server selects a first production parameter data set from the at least one production parameter data set, and determines a clustering object meeting a preset condition according to the first production parameter data set by the following formula II, wherein the preset condition is that the value of a norm expression of the first production parameter data set is minimum, and the formula II is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is the first lifeProducing a norm expression of the parameter data set under the k clustering meaning, k is the clustering number, c_j(j ═ 1, 2, … …, k) is production parameter data included in the jth cluster of the k clusters, and p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jThe clustered data corresponding to the included production parameter data.

It should be noted that the norm expression may be a first order norm expression of the first production parameter data set.

In the second step, the second formula is a fitting formula selected by the server according to the domain knowledge and the preset rule, and the server clusters and fits the change trend of the data in the first production parameter data set within a period of time through the fitting formula, so that the norm expression represents the difference between the first production parameter data set and the clustering object after being fitted through the clustering fitting formula; if the difference is smaller, it indicates that the overall degree of fit between the first production parameter data set and the cluster object is higher, here, we select the cluster object that minimizes the difference, i.e., the value of the norm expression.

If the first set of production parameter data is Y ═ Y (t)₁)，y(t₂)，……y(t_n) And calculating k clustering objects corresponding to min { E } by using loop iteration according to a second formula, wherein the corresponding clustering number of the clustering objects is k.

As shown in fig. 3, an image (a) in fig. 3 shows a first production parameter data set clustered and fitted by using a formula two provided by the embodiment of the present invention and a clustering object corresponding to the first production parameter data set; image (b) shows the fitting results for the first set of production parameters using a conventional linear fit; wherein, the solid line is a first production parameter data set to be fitted, and the dotted line is a fitting result. As can be seen from fig. 3, the fitting effect of the image (a) is significantly higher than that of the image (b) compared to the image (a), and the error generated by the fitting process of the image (b) is significantly larger than that of the image (a). Therefore, the clustering and fitting method using the formula two provided by the embodiment of the invention has strong self-adaptability, thereby improving the fitting degree and reducing the error generated in the fitting process.

The first production parameter is a production parameter selected by the server from at least one production parameter based on domain knowledge and a preset rule, and the first production parameter data set is time sequence data corresponding to the first production parameter periodically acquired by the server.

Step 2072: the server determines trend distribution data of the first production parameter data set according to the clustering object, and determines the reliability index of the problem to be diagnosed according to the trend distribution data through the following formula III, wherein the formula III is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is the acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is the maximum acquisition time corresponding to the trend distribution data when the value is zero_i(t) is the reliability index.

In the embodiment of the invention, the server connects the clustering objects included in each of the k clusters, extends and extrapolates the clustering objects to two ends until the clustering objects are extended to be intersected with the coordinate axes, determines the connected and extended clustering objects as the trend distribution data of each first production parameter data set, and determines the reliability index corresponding to the first production parameter data set corresponding to each problem to be diagnosed through the formula III.

As shown in image (a) in fig. 3, extending the clustering object to both ends, and intersecting with the coordinate axis; YF (t) reaches a maximum value max { YF } when t is 0, where Ri is 0; when YF is 0, t reaches a maximum value Tmax, and when Ri is 1. Therefore, the server can determine the reliability index of the problem to be diagnosed according to the current corresponding time t of the problem to be diagnosed.

Step 208: and the server determines the early warning grade corresponding to the reliability index of the problem to be diagnosed from the corresponding relation between the reliability index and the early warning grade according to the reliability index of the problem to be diagnosed, and performs early warning reminding on the problem according to the early warning grade.

In this step, the server stores the correspondence between the reliability index of the problem to be diagnosed and the early warning level, and the early warning level may be set and changed according to the user requirement, which is not specifically limited in the embodiment of the present invention. For example, the warning level may be classified into four levels of an emergency level, a high level, a middle level, and a low level.

It should be noted that when the state of each production parameter in the oil and gas well is monitored, the embodiment of the invention can not only diagnose whether the problem to be diagnosed exists in the oil and gas well in real time, but also early warn the problem to be diagnosed. According to the embodiment of the invention, the current state of the problem to be diagnosed in the oil-gas well is described and determined by determining the reliability index of the problem to be diagnosed, and the early warning grade of the oil-gas well is further determined according to the reliability index of the problem to be diagnosed, so that the problem to be diagnosed in the oil-gas well is early warned in time, the condition of the problem to be diagnosed in the oil-gas well is known in time by a worker, and the worker can take effective measures in time to deal with the problem to be diagnosed in the oil-gas well.

In the embodiment of the present invention, the server may calculate, after determining at least one data change rate set according to the at least one production parameter data set, a mathematical expected value of a production parameter data change rate corresponding to each data change rate set in the at least one data change rate set; then the server determines that the production parameters corresponding to the problems to be diagnosed form a first parameter set according to the diagnosis model stored in the server, and the mathematical expected values of the data change rate of each production parameter in the first parameter set form a mathematical expected value set; and the server calculates a target mathematical expected value of the first parameter set, and determines whether the oil and gas well has the problem to be diagnosed according to the target mathematical expected value of the first parameter set and a preset mathematical expected value of the problem to be diagnosed. The server converts a plurality of continuously observed production parameter data sets into a plurality of change rate sets with probability statistical characteristics, so that the server can directly calculate the plurality of data change rate sets to determine states such as mathematical expected values of the plurality of production parameter data change rates, the server can synthesize the states of the plurality of production parameters to diagnose the problem to be diagnosed directly according to a diagnosis model without manual processing, and the efficiency of monitoring the oil-gas well is improved; moreover, the server can timely judge whether the oil-gas well has the problem to be diagnosed, so that the manual processing time is saved, and the timeliness of diagnosing the oil-gas well in the process of monitoring the oil-gas well is improved.

The embodiment of the invention provides a device for monitoring an oil-gas well, which can be applied to a server and comprises the following components as shown in figure 4:

the acquisition module 401 is configured to acquire data of at least one production parameter of an oil and gas well every preset time period to obtain at least one production parameter data set;

a first determining module 402, configured to determine, according to the at least one production parameter data set, data change rates of two adjacent production parameter data in each production parameter data set, so as to obtain at least one data change rate set;

a second determining module 403, configured to determine a mathematical expectation value of the data change rate of the at least one production parameter according to the at least one data change rate set;

a selecting module 404, configured to select, according to a diagnostic model, a production parameter corresponding to a problem to be diagnosed from the at least one production parameter to form a first parameter set, where the diagnostic model is configured to store a corresponding relationship between the problem to be diagnosed and the production parameter;

a third determining module 405, configured to determine, according to the mathematical expectation value of the data change rate of the at least one production parameter, that the mathematical expectation value of each data change rate of the production parameter in the first parameter set constitutes a mathematical expectation value set;

a fourth determining module 406, configured to determine a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm;

and a fifth determining module 407, configured to determine whether the problem to be diagnosed exists in the oil and gas well according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

In one possible design, the second determining module 403 is further configured to obtain an edge density of each data change rate in each data change rate set; calculating a mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate, wherein the first data change rate set is any one of the at least one data change rate set, by the following formula one:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the data rate of change sets x in the at least one data rate of change set_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

In one possible design, the apparatus further includes:

the sixth determining module is used for determining the production parameter corresponding to each problem to be diagnosed in the problem set to be diagnosed, wherein the production parameter corresponding to each problem to be diagnosed is a production parameter defined based on domain knowledge and a preset rule;

and the building module is used for building the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

In one possible design, the apparatus further includes:

the estimation module is used for estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

a seventh determining module, configured to determine, according to the reliability index of the problem to be diagnosed, an early warning level corresponding to the reliability index of the problem to be diagnosed from a correspondence between the reliability index and the early warning level;

and the early warning module is used for carrying out early warning reminding on the problem to be diagnosed according to the early warning grade.

In one possible design, the estimation module includes:

a selection unit for selecting a first production parameter data set from the at least one production parameter data set, the first production parameter data set being a data set of production parameters preset based on domain knowledge and preset rules;

a first determining unit, configured to determine, according to the first production parameter data set, a clustering object that meets a preset condition by using a following formula two, where the preset condition is that a norm expression of the first production parameter data set is minimum, and the formula two is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 6, 2, … …, k) is the production parameter data included in the jth cluster of the k clusters, and p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jClustered data corresponding to the included production parameter data;

the second determining unit is used for determining trend distribution data of the first production parameter data set according to the clustering object;

a third determining unit, configured to determine, according to the trend distribution data, a reliability index of the problem to be diagnosed through a third formula, where the third formula is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is the acquisition time, YF (t) is the trend distribution data, and max { YF } is the trend distribution dataTmax is the maximum acquisition time corresponding to when the value of the trend distribution data is zero, R_i(t) is the reliability index.

In the embodiment of the present invention, the server may calculate, after determining at least one data change rate set according to the at least one production parameter data set, a mathematical expected value of a production parameter data change rate corresponding to each data change rate set in the at least one data change rate set; then the server determines that the production parameters corresponding to the problems to be diagnosed form a first parameter set according to the diagnosis model stored in the server, and the mathematical expected values of the data change rate of each production parameter in the first parameter set form a mathematical expected value set; and the server calculates a target mathematical expected value of the first parameter set, and determines whether the oil and gas well has the problem to be diagnosed according to the target mathematical expected value of the first parameter set and a preset mathematical expected value of the problem to be diagnosed. The server converts a plurality of continuously observed production parameter data sets into a plurality of change rate sets with probability statistical characteristics, so that the server can directly calculate the plurality of data change rate sets to determine states such as mathematical expected values of the plurality of production parameter data change rates, the server can synthesize the states of the plurality of production parameters to diagnose the problem to be diagnosed directly according to a diagnosis model without manual processing, and the efficiency of monitoring the oil-gas well is improved; moreover, the server can timely judge whether the oil-gas well has the problem to be diagnosed, so that the manual processing time is saved, and the timeliness of diagnosing the oil-gas well in the process of monitoring the oil-gas well is improved.

It should be noted that: in the oil and gas well monitoring device provided by the embodiment, when the oil and gas well is monitored, only the division of the functional modules is used for illustration, and in practical application, the function distribution can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules so as to complete all or part of the functions described above. In addition, the oil and gas well monitoring device provided by the embodiment and the oil and gas well monitoring method embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment and is not described again.

It will be understood by those skilled in the art that all or part of the steps for implementing the above embodiments may be implemented by hardware, or may be implemented by a program instructing relevant hardware, where the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a magnetic disk or an optical disk, etc.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (10)

1. A method of hydrocarbon well monitoring, the method comprising:

acquiring data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set;

determining the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set;

determining a mathematical expectation value of any one of the at least one set of data change rates according to the at least one set of data change rates;

selecting production parameters corresponding to the problem to be diagnosed from the at least one production parameter to form a first parameter set according to a diagnosis model, wherein the diagnosis model is used for storing the corresponding relation between the problem to be diagnosed and the production parameters;

determining the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate;

determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm;

and determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

2. The method of claim 1, wherein determining a mathematical expectation of any of the at least one set of data rates of change from the at least one set of data rates of change comprises:

acquiring the edge density of each data change rate in each data change rate set;

calculating a mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate, wherein the first data change rate set is any one of the at least one data change rate set:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the at least one set of data rates of change, x_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

3. The method of claim 1, wherein prior to selecting the production parameter corresponding to the problem to be diagnosed from the at least one production parameter into the first set of parameters, the method further comprises:

determining a production parameter corresponding to each problem to be diagnosed in a problem set to be diagnosed, wherein the production parameter corresponding to each problem to be diagnosed is a production parameter defined based on domain knowledge and a preset rule;

and constructing the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

4. The method of claim 1, wherein after determining whether the problem to be diagnosed exists in the hydrocarbon well according to the target mathematical expected value of the first set of parameters and the preset mathematical expected value of the problem to be diagnosed, the method further comprises:

estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

according to the reliability index of the problem to be diagnosed, determining an early warning grade corresponding to the reliability index of the problem to be diagnosed from the corresponding relation between the reliability index and the early warning grade;

and carrying out early warning reminding on the problem to be diagnosed according to the early warning grade corresponding to the reliability index of the problem to be diagnosed.

5. The method according to claim 4, wherein said predicting a reliability index of said problem to be diagnosed according to said at least one production parameter data set and a second predetermined algorithm comprises:

selecting a first production parameter data set from the at least one production parameter data set, wherein the first production parameter data set is a data set of production parameters preset based on domain knowledge and preset rules;

according to the first production parameter data set, determining a clustering object meeting a preset condition through the following formula II, wherein the preset condition is that the value of a norm expression of the first production parameter data set is minimum, and the formula II is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 1, 2, … …, k) isThe jth cluster of the k clusters contains production parameter data, p being cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jClustered data corresponding to the included production parameter data;

determining trend distribution data of the first production parameter data set according to the clustering object;

determining the reliability index of the problem to be diagnosed according to the trend distribution data and through a third formula, wherein the third formula is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is_i(t) is the reliability index.

6. An apparatus for hydrocarbon well monitoring, the apparatus comprising:

the acquisition module is used for acquiring data of at least one production parameter of the oil-gas well every preset time length to obtain at least one production parameter data set;

the first determining module is used for determining the data change rate of two adjacent production parameter data in each production parameter data set according to the at least one production parameter data set to obtain at least one data change rate set;

a second determining module, configured to determine, according to the at least one data change rate set, a mathematical expectation value of any one of the at least one data change rate set;

the selection module is used for selecting the production parameters corresponding to the problems to be diagnosed from the at least one production parameter to form a first parameter set according to a diagnosis model, and the diagnosis model is used for storing the corresponding relation between the problems to be diagnosed and the production parameters;

the third determination module is used for determining the mathematical expectation value of each production parameter data change rate in the first parameter set to form a mathematical expectation value set according to the mathematical expectation value of the at least one production parameter data change rate;

the fourth determining module is used for determining a target mathematical expectation value of the first parameter set according to the mathematical expectation value set and a first preset algorithm;

and the fifth determining module is used for determining whether the oil and gas well has the problem to be diagnosed or not according to the target mathematical expected value of the first parameter set and the preset mathematical expected value of the problem to be diagnosed.

7. The apparatus of claim 6,

the second determining module is further configured to obtain an edge density of each data change rate in each data change rate set; calculating a mathematical expectation value of a first data change rate set according to each data change rate in each data change rate set and the edge density of each data change rate, wherein the first data change rate set is any one of the at least one data change rate set:

the formula I is as follows:

wherein m is_xA mathematical expectation value, x, for any of the at least one set of data rates of change, x_iFor the ith data rate of change, p, in any data rate set x_x(x_i) Edge density, dx, for the ith data rate of change in any of the data rate sets x_iIs x_iDifferentiation of (2).

8. The apparatus of claim 6, further comprising:

a sixth determining module, configured to determine a production parameter corresponding to each to-be-diagnosed problem in the to-be-diagnosed problem set, where the production parameter corresponding to each to-be-diagnosed problem is a production parameter defined based on domain knowledge and a preset rule;

and the building module is used for building the diagnosis model according to each problem to be diagnosed and the production parameter corresponding to each problem to be diagnosed.

9. The apparatus of claim 6, further comprising:

the estimation module is used for estimating the reliability index of the problem to be diagnosed according to the at least one production parameter data set and a second preset algorithm;

a seventh determining module, configured to determine, according to the reliability index of the problem to be diagnosed, an early warning level corresponding to the reliability index of the problem to be diagnosed from a correspondence between the reliability index and the early warning level;

and the early warning module is used for carrying out early warning reminding on the problem to be diagnosed according to the early warning grade corresponding to the reliability index of the problem to be diagnosed.

10. The apparatus of claim 9, wherein the estimation module comprises:

a selection unit, configured to select a first production parameter data set from the at least one production parameter data set, where the first production parameter data set is a data set of production parameters preset based on domain knowledge and preset rules;

a first determining unit, configured to determine, according to the first production parameter data set, a clustering object that meets a preset condition by using a following formula two, where the preset condition is that a norm expression of the first production parameter data set is minimum, and the formula two is a formula preset in a second preset algorithm:

the formula II is as follows:

wherein E is a norm expression of the first production parameter data set in the k clustering meaning, k is a clustering number, c_j(j ═ 6, 2, … …, k) is the production parameter data included in the jth cluster of the k clusters, and p is cluster c_jProduction parameter data of (1), o_jIs to characterize the jth cluster c_jI.e. the jth cluster c_jClustered data corresponding to the included production parameter data;

the second determining unit is used for determining trend distribution data of the first production parameter data set according to the clustering object;

a third determining unit, configured to determine, according to the trend distribution data, a reliability index of the problem to be diagnosed through a third formula, where the third formula is a formula preset in a second preset algorithm:

the formula III is as follows:

wherein t is acquisition time, YF (t) is the trend distribution data, max { YF } is the maximum value of the trend distribution data, Tmax is the maximum acquisition time corresponding to the trend distribution data when the value is zero, and R is_i(t) is the reliability index.

Images