CN-122014155-A - Sand control valve continuous sand removal and anti-sticking control system for oil well oil extraction

CN122014155ACN 122014155 ACN122014155 ACN 122014155ACN-122014155-A

Abstract

The invention relates to the technical field of sand prevention valves, in particular to a continuous sand removal and anti-sticking control system of a sand prevention valve for oil well oil extraction. The sand control system comprises a data acquisition unit for acquiring and normalizing sand setting thickness, sand concentration, torque and suspended weight parameters, a risk prediction unit for calculating the probability of stuck drilling risk by adopting a gradient lifting tree model, an anti-stuck intelligent control unit for generating a target opening instruction of a sand control valve through a fuzzy logic control algorithm, a sand removal optimization control unit for generating a target rotating speed instruction of a spiral separator through a PID control algorithm, and an execution unit for driving the sand control valve and the spiral separator to execute corresponding actions. The risk prediction is carried out by fusing multidimensional features through the gradient lifting tree model and the fuzzy logic and PID cooperative control are combined, so that the problem that continuous self-adaptive sand removal and anti-sticking control are difficult to realize in the prior art due to the fact that single threshold value or manual experience is relied on is solved.

Inventors

MENG XIANWEI
WANG JUNHUI
ZHANG XIN
SUN MINGJUN
SUN BINSHENG
YU BAISHENG
XU TONG

Assignees

吉林市天宇科技有限责任公司

Dates

Publication Date: 20260512
Application Date: 20260409

Claims (10)

1. A continuous sand-removing and anti-sticking control system of a sand-preventing valve for oil extraction of an oil well is characterized by comprising The data acquisition unit (1) is used for acquiring sand setting thickness, sand concentration, torque and suspended weight parameters through a sensor group, and carrying out normalization processing on the data to obtain valve parameter data; the risk prediction unit (2), the risk prediction unit (2) adopts a gradient lifting tree model to calculate the probability of stuck drilling risk based on valve parameter data; the intelligent anti-sticking control unit (3) is used for generating a target opening instruction of the sand control valve through a fuzzy logic control algorithm according to the probability of the drill sticking risk and the sand concentration; The sand removal optimization control unit (4) is used for generating a target rotating speed instruction of the spiral separator through a PID control algorithm by taking the sand concentration acquired in real time as a process variable; and the execution unit (5) is used for receiving the target opening instruction and the target rotating speed instruction and driving the sand control valve and the spiral separator to execute corresponding actions.
2. The continuous sand removal and anti-sticking control system of the sand prevention valve for oil well production according to claim 1, wherein the data acquisition unit (1) comprises a data acquisition module (11) and a normalization processing module (12); The data acquisition module (11) acquires sand setting thickness, sand concentration, torque and suspended weight parameters in real time through a sensor group consisting of a sand setting thickness sensor, a sand concentration sensor, a torque sensor and a suspended weight sensor; The normalization processing module (12) adopts a min-max normalization algorithm to map the acquired data to the [0,1] interval and outputs normalized valve parameter data.
3. The continuous sand removal and jamming prevention control system for the sand prevention valve for oil well production according to claim 1, wherein the risk prediction unit (2) comprises a model construction module (21) and a risk prediction module (22); The model construction module (21) constructs a historical feature vector based on the characteristics of the historical sand setting thickness, the sand concentration, the torque, the suspended weight and the change rate thereof to serve as input features, and takes whether drilling sticking events occur at the same historical moment as labels to form a training data set for constructing a gradient lifting tree model; the risk prediction module (22) constructs a real-time feature vector based on the valve parameter data as an input feature and outputs a stuck risk probability using a gradient lifting tree model.
4. The continuous sand removal and anti-sticking control system of the sand prevention valve for oil well production according to claim 3, wherein the specific steps involved in the construction of the gradient lifting tree model in the model construction module (21) are as follows: Extracting time sequence data of core parameters from a historical database, calculating change rate characteristics of each core parameter, splicing the core parameters and the change rate characteristics thereof in columns to form a historical characteristic vector At the same time, the historical feature vectors Each piece of data of (1) marks whether a drill sticking event occurs or not to obtain an event label Wherein the core parameters include sand setting thickness Sand concentration Torque and torque of the motor Suspended weight ; Setting the initial shallow decision tree number as The complexity of a single shallow decision tree is limited not to exceed the maximum depth of the tree While limiting each decision conclusion to be based on the number of samples exceeding the minimum number of samples for the leaf node ; At the position of In the round of iteration, each round of iteration calculates a logarithmic loss function based on a historical feature vector and an event label, calculates the negative gradient of a current gradient lifting tree model on each sample based on the logarithmic loss function, trains a new shallow decision tree according to the negative gradient to learn the correction direction; Calculating the loss on the validation set after each iteration if continuous The training is stopped in advance when the wheel loss is no longer reduced; training the multiple shallow decision trees to obtain a learning rate Combining into final gradient lifting tree model 。
5. The continuous sand removal and jamming prevention control system for a sand prevention valve for oil well production according to claim 3, wherein the specific steps involved in the risk prediction module (22) for constructing a real-time feature vector based on valve parameter data as an input feature and outputting a jamming risk probability by using a gradient lifting tree model are as follows: Calculating normalized parameter value change rate based on valve parameter data at the current moment and the previous moment, wherein the parameter value change rate comprises sand setting thickness change rate, sand concentration change rate, torque change rate and suspended weight change rate; Sequentially splicing valve parameter data and the change rate thereof to construct a real-time feature vector ; Real-time feature vector Inputting a gradient-lifted tree model In which the original predicted value is output ; Original predicted value through sigmoid function Conversion to stuck risk probability 。
6. The continuous sand-removing and anti-sticking control system of the sand-preventing valve for oil well production according to claim 1, wherein the specific steps involved in generating the target opening command of the sand-preventing valve by a fuzzy logic control algorithm in the anti-sticking intelligent control unit (3) are as follows: Probability of risk of sticking drill And real-time sand concentration Mapping the fuzzy language values into fuzzy language values through a preset triangle membership function respectively, wherein the fuzzy language values of the stuck risk probability are low, medium and high, and the fuzzy language values of the sand concentration are low, normal and high; based on a predefined fuzzy rule base, adopting a Mamdani reasoning method, and using fuzzy language values of the stuck risk probability and fuzzy language values of the sand concentration to change the opening degree; The fuzzy language value of the opening degree change is deblurred by adopting a gravity center method, and the accurate opening degree change quantity is obtained through calculation ; The current opening degree And the opening degree variation Adding to obtain a target opening command And instruct the target opening degree And performing amplitude limiting treatment to ensure that the effective opening degree of the sand control valve is not exceeded.
7. The continuous sand removal and anti-sticking control system of the sand prevention valve for oil well production according to claim 1, wherein the sand removal optimization control unit (4) comprises a concentration deviation calculation module (41) and a PID parameter adjustment module (42); The concentration deviation calculation module (41) compares the real-time sand concentration with the set target concentration, and calculates a concentration deviation value and a deviation change rate; and the PID parameter adjusting module (42) dynamically adjusts a target rotating speed instruction of the spiral separator based on a PID control algorithm according to the concentration deviation value and the deviation change rate.
8. The continuous sand-removing and anti-sticking control system of a sand-preventing valve for oil well production according to claim 7, wherein the concentration deviation calculating module (41) comprises the following specific steps of: Subtracting the real-time sand concentration from a preset target concentration value to obtain a concentration deviation value at the current moment ; The concentration deviation change rate is obtained by carrying out differential calculation on concentration deviation values in two continuous sampling periods.
9. The continuous sand removal and anti-sticking control system of a sand prevention valve for oil well production according to claim 1, wherein the specific steps involved in the PID parameter adjusting module (42) for dynamically adjusting the target rotating speed command of the spiral separator based on a PID control algorithm are as follows: receiving real-time concentration deviation value and deviation change rate, and using preset proportionality coefficient Multiplying the real-time concentration deviation to obtain a proportional term ; By cumulatively summing the historical concentration bias values and multiplying by a preset integral coefficient Obtaining integral terms ; From preset differential coefficients Multiplying the concentration deviation change rate to obtain a differential term ; Adding the proportional term, the integral term and the derivative term to obtain the output control quantity of the PID controller at the current moment ; Will output the control quantity With the current rotational speed of the spiral separator Adding to obtain a target rotating speed instruction Then, the target rotating speed is instructed And carrying out amplitude limiting treatment to ensure that the upper limit and the lower limit of the safe rotation speed allowed by the spiral separator are not exceeded.
10. The continuous sand-removing and anti-sticking control system of the sand-preventing valve for oil well production according to claim 1, wherein the execution unit (5) comprises the following specific steps of: The received target opening instruction Control voltage signal converted into sand control valve actuator ; The received target rotating speed command Control current signal converted into spiral separator driving motor ; Will control the voltage signal The electric actuator is output to the sand control valve to drive the valve to adjust to the target opening degree, and the control current signal is sent to the sand control valve Outputting to a variable frequency driver of the spiral separator, and driving a motor to reach a target rotating speed; Reading actual opening feedback value of sand control valve in real time And actual rotational speed feedback value of the spiral separator Comparing the command with the target opening command and the target rotating speed command, and triggering a command retransmission mechanism if any deviation exceeds the maximum allowable deviation; when the actual opening is consistent with the target opening and the actual rotating speed is consistent with the target rotating speed, judging that the action is completed, and enabling the system to enter a steady-state operation mode to continuously monitor and wait for a new control instruction.

Description

Sand control valve continuous sand removal and anti-sticking control system for oil well oil extraction Technical Field The invention relates to the technical field of sand prevention valves, in particular to a continuous sand removal and anti-sticking control system of a sand prevention valve for oil well oil extraction. Background In the oil extraction process of an oil well, stratum sand grains enter a shaft along with crude oil, are easy to deposit on a valve, a pipeline and other parts, so that the thickness of settled sand is increased, the sand concentration is increased, and serious faults such as valve blockage, drilling tool blockage and the like are caused. The existing sand control system is dependent on manual experience or single parameter threshold control, lacks intelligent prediction and cooperative regulation and control capability for multi-parameter coupling risk, and is difficult to realize continuous and self-adaptive sand removal and anti-sticking control, so that the continuous sand removal and anti-sticking control system of the sand control valve for oil well oil extraction is designed. Disclosure of Invention The invention aims to provide a continuous sand removal and anti-sticking control system of a sand control valve for oil well oil extraction, which aims to solve the problems that the existing sand control system provided in the background art is dependent on manual experience or single parameter threshold control, lacks intelligent prediction and cooperative regulation and control capability of multi-parameter coupling risks and is difficult to realize continuous and self-adaptive sand removal and anti-sticking control. The invention aims to provide a continuous sand-removing and anti-sticking control system of a sand-preventing valve for oil production of an oil well, which comprises The data acquisition unit acquires sand setting thickness, sand concentration, torque and suspended weight parameters through a sensor group, and performs normalization processing on the data to obtain valve parameter data; The risk prediction unit is used for calculating the drilling sticking risk probability by adopting a gradient lifting tree model based on valve parameter data; The intelligent anti-sticking control unit generates a target opening instruction of the sand control valve through a fuzzy logic control algorithm according to the drilling risk probability and the sand concentration; The sand removal optimization control unit takes the sand concentration acquired in real time as a process variable and generates a target rotating speed instruction of the spiral separator through a PID control algorithm; And the execution unit is used for receiving the target opening instruction and the target rotating speed instruction and driving the sand control valve and the spiral separator to execute corresponding actions. As a further improvement of the technical scheme, the data acquisition unit comprises a data acquisition module and a normalization processing module; the data acquisition module acquires the sand setting thickness, the sand concentration, the torque and the suspended weight parameters in real time through a sensor group consisting of a sand setting thickness sensor, a sand concentration sensor, a torque sensor and a suspended weight sensor; The normalization processing module adopts a min-max normalization algorithm to map the acquired data to the [0,1] interval and outputs normalized valve parameter data. As a further improvement of the technical scheme, the risk prediction unit comprises a model construction module and a risk prediction module; The model construction module constructs a historical feature vector based on the characteristics of the historical sand setting thickness, the sand concentration, the torque, the suspended weight and the change rate thereof to serve as input features, and whether drilling sticking events occur at the same historical moment serves as labels to form a training data set for constructing a gradient lifting tree model; The risk prediction module constructs a real-time feature vector based on the valve parameter data to serve as an input feature, and outputs the stuck risk probability by using the gradient lifting tree model. As a further improvement of the technical scheme, in the model construction module, the concrete steps involved in constructing the gradient lifting tree model are as follows: Extracting time sequence data of core parameters from a historical database, calculating change rate characteristics of each core parameter, splicing the core parameters and the change rate characteristics thereof in columns to form a historical characteristic vector At the same time, the historical feature vectorsEach piece of data of (1) marks whether a drill sticking event occurs or not to obtain an event labelWherein the core parameters include sand setting thicknessSand concentrationTorque and torque of the motorSuspended weight; Setting