CN-121994299-A - Fault detection method for oil gas recovery treatment equipment by compression condensation membrane separation method

CN121994299ACN 121994299 ACN121994299 ACN 121994299ACN-121994299-A

Abstract

The application relates to a fault detection method of oil gas recovery processing equipment by a compression condensation membrane separation method, which relates to the technical field of oil gas recovery processing and comprises the following steps that firstly, multi-parameter data including at least power supply voltage, temperature, pressure, flow, liquid level and vibration acceleration are continuously collected by a plurality of sensors when the oil gas recovery processing equipment operates; and step two, inputting the collected multi-parameter data into a fault detection system. The application realizes early and accurate detection of equipment operation abnormality by carrying out multi-parameter state estimation and calculating the Mahalanobis distance of residual errors based on an extended Kalman filtering algorithm, can rapidly distinguish various fault types by combining a support vector machine classifier, and automatically executes corresponding optimization strategies or safety interlocking protection measures such as nonlinear PID regulation, model predictive control and the like according to different fault types based on a finite state machine model, thereby remarkably improving the accuracy, the instantaneity and the automation level of a system of fault detection.

Inventors

SUN DAQIAN
FAN XIAOJIE
ZHANG LUJUN
HAN JUNSHENG

Assignees

北京恒合信业技术股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260108

Claims (8)

1. The fault detection method of the oil gas recovery treatment equipment by the compression condensation membrane separation method is characterized by comprising the following steps of: Continuously acquiring multi-parameter data of the oil gas recovery processing equipment during operation through a plurality of sensors, wherein the multi-parameter data at least comprises power supply voltage, temperature, pressure, flow, liquid level and vibration acceleration; Inputting the acquired multi-parameter data to a fault detection system, performing state estimation based on an extended Kalman filtering algorithm, and calculating an observation residual error and a Mahalanobis distance thereof; Step three, if the Mahalanobis distance exceeds a preset threshold value, triggering a fault classifier, and distinguishing fault types by adopting a support vector machine; And step four, executing corresponding safety protection measures according to the fault type, wherein the safety protection measures comprise alarming and interlocking parking.
2. The fault detection method of oil gas recovery processing equipment by a compression condensation membrane separation method according to claim 1, wherein a state vector of an extended kalman filter algorithm is defined as: Wherein In the case of a pressure force, the pressure, In order to be able to determine the temperature, In order to be a flow rate, Is vibration acceleration; The system state equation and the observation equation are respectively: Wherein In order to control the input of the device, And Is Gaussian noise; the extended kalman filter algorithm includes a prediction step and an update step: And a prediction step: ; updating: , wherein, For a state transition jacobian matrix, In order to process noise covariance(s), In order to observe the jacobian matrix, To observe noise covariance.
3. The fault detection method for oil and gas recovery processing equipment by compression condensation membrane separation method according to claim 1, wherein the fault detection is performed by residual error Is realized by the Mahalanobis distance: If (if) Triggering a fault classifier; The fault classifier adopts a support vector machine, and the decision function is as follows: Wherein As a function of the radial basis kernel, In order to be a lagrange multiplier, In the case of a category label, Is a bias term.
4. The fault detection method of a compression condensation membrane separation process oil gas recovery processing device according to claim 1, wherein the fault type includes any one or more of the following: Emergency stop faults, including power failure or emergency button triggering; The temperature alarm is stopped, and the temperature alarm comprises that the exhaust temperature of the induced draft fan/compressor is too high or the water inlet temperature of cooling water is too high; the pressure alarm is stopped, and the pressure alarm comprises overhigh discharge pressure of an induced draft fan/a compressor or overhigh outlet pressure of a vacuum pump; The liquid level gives an alarm to stop the vehicle, comprises an oil storage temporary storage tank with ultrahigh liquid level; the motor is stopped when the fault happens, including the faults of an induced draft fan, a compressor, a refrigeration compressor, a vacuum pump or a secondary refrigerant pump; And (5) manually stopping.
5. The method for detecting a fault in a condensed-film-separation-process oil and gas recovery processing apparatus according to claim 4, wherein the detection logic of the emergency stop fault is based on a voltage energy accumulation model: , wherein, For the rated voltage to be the rated voltage, As a result of the energy threshold value, Is the time window length; The detection of the temperature alarm parking is based on the thermal equilibrium equation: The triggering condition is that if And is also provided with The vehicle is stopped in an alarm way, wherein, In order to be able to achieve a heat capacity, In order to be a thermal conductivity coefficient, For the heat dissipation area to be a thermal dissipation area, The upper limit of the temperature is set to be the upper limit of the temperature, Is a rate of change threshold; the detection of the pressure alarm parking is based on an ideal gas dynamic model: The triggering conditions are as follows: And is also provided with Then the vehicle is stopped, wherein, In order to set the pressure to be applied, R is a gas constant, and V is a system volume; Detection of liquid level alarm parking is based on mass balance equation: , wherein, Is the height of the liquid level, and the liquid level is the liquid level, Is the cross-sectional area of the storage tank; detection of motor fault parking, monitoring based on vibration acceleration: the triggering condition is that And stopping the vehicle.
6. The fault detection method of a compression condensation membrane separation process oil and gas recovery processing device according to claim 1, wherein the execution of the safety protection measures is based on a finite state machine model, comprising the following states: state one: data acquisition, sampling and filtering at 100Hz frequency; The second state is fault diagnosis, calculating fault indexes and judging whether the fault indexes exceed a threshold; A third state is a control decision, wherein an adjustment strategy is selected according to the fault type, and the adjustment strategy comprises nonlinear PID adjustment of the refrigerant flow or model prediction control adjustment of the compressor frequency; and fourthly, executing control, and sending a control signal through the OP CUA protocol, wherein the response time is less than 100ms.
7. The fault detection method of oil and gas recovery processing equipment by a compression condensation membrane separation method according to claim 6, wherein the control decision comprises adjustment of abnormal condensation efficiency, specifically: and adjusting the opening degree of the condensing agent flow valve by adopting a nonlinear PID: Wherein , Is a nonlinear gain.
8. The method for detecting the failure of the oil and gas recovery processing equipment by the compression condensation membrane separation method according to claim 6, wherein the control decision comprises the adjustment of the over-pressure, in particular: The compressor frequency is adjusted by model predictive control: The constraint is that 。

Description

Fault detection method for oil gas recovery treatment equipment by compression condensation membrane separation method Technical Field The application relates to the technical field of oil gas recovery treatment, in particular to a fault detection method of oil gas recovery treatment equipment by a compression condensation membrane separation method. Background In the operation process of the traditional oil gas recovery treatment equipment by the compression condensation membrane separation method, due to the lack of an intelligent fault diagnosis means, the method generally only depends on simple threshold value alarming on single parameters (such as temperature and pressure), the early fault and complex fault types are difficult to accurately identify, false alarm and missing alarm are easy to occur, and effective regulation and control measures cannot be adopted before the fault occurs, so that the continuity and the reliability of the operation of the equipment are affected, and a certain safety risk exists. Content of the application In order to solve the problems, the application provides a fault detection method of oil gas recovery processing equipment by a compression condensation membrane separation method. The application provides a fault detection method of oil gas recovery processing equipment by a compression condensation membrane separation method, which adopts the following technical scheme: the fault detection method of the oil gas recovery treatment equipment by the compression condensation membrane separation method comprises the following steps of: Continuously acquiring multi-parameter data of the oil gas recovery processing equipment during operation through a plurality of sensors, wherein the multi-parameter data at least comprises power supply voltage, temperature, pressure, flow, liquid level and vibration acceleration; Inputting the acquired multi-parameter data to a fault detection system, performing state estimation based on an extended Kalman filtering algorithm, and calculating an observation residual error and a Mahalanobis distance thereof; Step three, if the Mahalanobis distance exceeds a preset threshold value, triggering a fault classifier, and distinguishing fault types by adopting a support vector machine; And step four, executing corresponding safety protection measures according to the fault type, wherein the safety protection measures comprise alarming and interlocking parking. As a preferred technical solution of the present application, the state vector of the extended kalman filter algorithm is defined as: Wherein In the case of a pressure force, the pressure,In order to be able to determine the temperature,In order to be a flow rate,Is vibration acceleration; The system state equation and the observation equation are respectively: Wherein In order to control the input of the device,AndIs Gaussian noise; the extended kalman filter algorithm includes a prediction step and an update step: And a prediction step: ; updating: , wherein, For a state transition jacobian matrix,In order to process noise covariance(s),In order to observe the jacobian matrix,To observe noise covariance. As a preferable technical scheme of the application, the fault detection passes through residual errorIs realized by the Mahalanobis distance: If (if) Triggering a fault classifier; The fault classifier adopts a support vector machine, and the decision function is as follows: Wherein As a function of the radial basis kernel,In order to be a lagrange multiplier,In the case of a category label,Is a bias term. As a preferred technical solution of the present application, the fault types include any one or more of the following: Emergency stop faults, including power failure or emergency button triggering; The temperature alarm is stopped, and the temperature alarm comprises that the exhaust temperature of the induced draft fan/compressor is too high or the water inlet temperature of cooling water is too high; the pressure alarm is stopped, and the pressure alarm comprises overhigh discharge pressure of an induced draft fan/a compressor or overhigh outlet pressure of a vacuum pump; The liquid level gives an alarm to stop the vehicle, comprises an oil storage temporary storage tank with ultrahigh liquid level; the motor is stopped when the fault happens, including the faults of an induced draft fan, a compressor, a refrigeration compressor, a vacuum pump or a secondary refrigerant pump; And (5) manually stopping. As a preferred technical solution of the present application, the detection logic of the emergency stop fault is based on a voltage energy accumulation model: , wherein, For the rated voltage to be the rated voltage,As a result of the energy threshold value,Is the time window length; The detection of the temperature alarm parking is based on the thermal equilibrium equation: The triggering condition is that if And is also provided withThe vehicle is stopped in an alarm way, wherein,In