Spray oil-gas separation self-adaptive degassing control method and system based on air pressure monitoring

Abstract

This invention relates to the field of oil-gas separation technology, and particularly to an adaptive degassing control method and system for spray oil-gas separation based on gas pressure monitoring. The method includes: synchronously collecting data from multiple dimensions to comprehensively reflect the state of the degassing process and provide a reliable basis for subsequent analysis and control; deeply mining the characteristics of gas pressure data to extract fluctuation feature vectors and grasp the dynamic change law of gas pressure; comprehensively analyzing multiple factors to output multi-dimensional stability coefficients and assess the stability of the degassing process; determining core parameters through quantitative relationships to locate key factors affecting degassing efficiency and provide direction for optimized control; constructing a dynamic degassing efficiency field to predict the degassing effect in advance, facilitating timely adjustment of strategies to improve efficiency; and planning and compensating for efficiency differences and determining termination conditions to specifically improve degassing unevenness, enhance degassing quality and uniformity, and accurately control the degassing endpoint.

Description

Technical Field

[0001] This invention relates to the field of oil-gas separation technology, and in particular to an adaptive degassing control method and system for spray oil-gas separation based on gas pressure monitoring. Background Technology

[0002] Existing spray degassing control methods have limitations. Their control logic is usually based on fixed timing or simple flow regulation, without deeply sensing and responding to the dynamic physical processes within the degassing chamber. Under real-world conditions, the differences in solubility of multiple gas components and the interaction between spraying and pumping actions jointly cause complex nonlinear fluctuations in the gas pressure within the degassing chamber. Existing methods lack differentiated sensing and coordinated control mechanisms for the degassing kinetics of each gas component, making it impossible to identify the degassing stability information contained in the pressure fluctuations. Furthermore, they cannot precisely enhance the degassing of difficult-to-degas components when multiple gases are simultaneously degassed. When dealing with changes in degassing conditions, existing methods generally face problems such as unstable degassing efficiency and asynchronous degassing of different gases, leading to distortions in the proportions of components in the degassing gas or incomplete degassing of certain key characteristic gases, ultimately severely affecting the accuracy of subsequent spectral analysis. Therefore, this invention proposes an adaptive degassing control method and system for spray oil-gas separation based on gas pressure monitoring. Summary of the Invention

[0003] The purpose of this invention is to solve the problems in the background art, and to propose an adaptive degassing control method and system for spray oil-gas separation based on air pressure monitoring.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] The first aspect of this invention provides an adaptive degassing control method for spray oil-gas separation based on gas pressure monitoring, comprising:

[0006] S1. Acquire the gas pressure time series data in the degassing chamber during the spray degassing process, and simultaneously record the spray flow time series data and oil temperature time series data at the corresponding time points.

[0007] S2. Construct a pressure-time curve by combining pressure time series data, and extract the fluctuation feature vector from the pressure-time curve;

[0008] S3. Perform multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and output the multidimensional stability coefficient of the current degassing process.

[0009] S4. Quantify the mapping relationship between the multidimensional stability coefficient and the gas component extraction to determine the core correlation parameters that affect the degassing efficiency;

[0010] S5. Based on the core correlation parameters, construct a coupled correlation map between the concentration of the extracted gas components and the rate of change of gas pressure to predict the dynamic degassing efficiency field of each gas component.

[0011] S6. Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, plan the virtual spray path and perform hierarchical regional compensation.

[0012] S7. Based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, determine whether the degassing termination condition is met.

[0013] Further, S1 includes:

[0014] Several high-precision pressure sensors are distributed and deployed inside the degassing chamber to synchronously collect time-series pressure data;

[0015] The total flow rate and branch flow ratio of the spray pipeline are monitored in real time by electromagnetic flowmeter to obtain spray flow time series data;

[0016] The real-time temperature of the oil is collected by a platinum resistance temperature sensor attached to the wall of the oil chamber, and oil temperature time series data is obtained.

[0017] Further, S2 includes:

[0018] Wavelet packet decomposition is performed on the barometric pressure time series data to extract wavelet coefficients within a preset frequency band, and the energy of the wavelet packet coefficients within each time window is calculated.

[0019] The energy ratio is calculated based on the wavelet packet coefficients of adjacent time windows, and the energy ratios are sorted by time to form an energy ratio sequence.

[0020] A linear fit is performed on the energy ratio sequence, and the slope obtained from the linear fit is used as the energy distribution gradient characterizing the trend of energy change in air pressure fluctuations during that period.

[0021] Furthermore, S2 also includes:

[0022] The derivative calculation and analysis of the pressure-time curve are performed to automatically identify and mark candidate abrupt change points;

[0023] Within the set analysis time window, the absolute values ​​of the first derivatives of all candidate mutation points are accumulated, and this accumulated value is used as the cumulative intensity of mutation points in that analysis time window.

[0024] The energy distribution gradient calculation process is repeatedly performed for the continuous analysis time window to generate an energy distribution gradient sequence; at the same time, the cumulative intensity calculation process of abrupt change points is repeatedly performed for the continuous analysis time window to generate a cumulative intensity sequence of abrupt change points.

[0025] The energy distribution gradient sequence and the cumulative intensity sequence of abrupt change points were normalized respectively.

[0026] The normalized energy distribution gradient sequence and the cumulative intensity sequence of abrupt change points are spliced ​​together according to the same time index to form a fluctuation feature vector that comprehensively characterizes the dynamic characteristics of the current degassing process.

[0027] Further, S3 includes:

[0028] Based on the fluctuation feature vector, the data components corresponding to the energy distribution gradient and the data components of the cumulative intensity of the mutation point are extracted and used as the pressure fluctuation stability component and the event sudden stability component of the current degassing process, respectively.

[0029] Acquire the current spray flow rate time series data and oil temperature time series data, and calculate their relative deviation from the corresponding preset reference values, which are used as the flow stability component and temperature stability component; wherein, the optimal flow rate reference value and the optimal temperature reference value constitute the preset reference value, the optimal flow rate reference value corresponds to the current spray flow rate time series data, and the optimal temperature reference value corresponds to the current oil temperature time series data;

[0030] Obtain the stability components of N consecutive backtracking time windows during the current degassing process, and construct an N×4 comprehensive stability component matrix;

[0031] Using the information entropy method, by analyzing the uncertainty of each column of the comprehensive stability component matrix, a stability contribution entropy weight representing its relative importance is assigned to each stability component.

[0032] The multidimensional stability coefficient of the current degassing process is obtained by multiplying each stability component value of the current backtracking time window by its corresponding stability contribution entropy weight and summing the results.

[0033] Further, S4 includes:

[0034] Real-time acquisition of the concentration time series data of various gas components dissolved in the oil during the degassing process, including various characteristic gases extracted from the oil, such as hydrogen, carbon monoxide, methane, ethylene, acetylene, ethane, etc.

[0035] Based on the concentration time series data of multiple gas components, the concentration change of each gas component per unit time is calculated to obtain the instantaneous extraction rate of the gas component; the calculation is repeated for continuous time points to form a sequence of instantaneous extraction rates of each gas component.

[0036] A multivariate nonlinear regression model was constructed using the multidimensional stability coefficient as the independent variable and the instantaneous extraction rate sequence of each gas component as the dependent variable.

[0037] By fitting a multivariate nonlinear regression model, the influence of different stability dimensions contained in the multidimensional stability coefficient on the desorption rate of each gas component is analyzed. The regression coefficients corresponding to the stability dimensions with the highest influence are extracted and normalized to obtain the concentration sensitivity factor of the gas component.

[0038] Obtain time-series air pressure data and calculate its first-order difference to obtain the first-order rate of change sequence of air pressure.

[0039] The mutual information value between the instantaneous extraction rate sequence of each gas component and the first-order rate of change of gas pressure is calculated separately. After normalizing the mutual information value, it is defined as the influence weight of the rate of change of gas pressure.

[0040] The gas component concentration sensitivity factor and the influence weight of the gas pressure change rate together constitute the core correlation parameter.

[0041] Further, S5 includes:

[0042] A coupled correlation map was constructed by using time as the horizontal axis, the rate of change of air pressure as the vertical axis, and the influence weights of the concentration sensitivity factor and the rate of change of air pressure of each gas component as coupling modulation parameters.

[0043] In the state space represented by the coupling correlation map, the concentration of each gas component is predicted and analyzed in the next K time steps, forming the concentration prediction trajectory of each gas component.

[0044] Based on the current oil temperature time series data, the concentration prediction trajectory of each gas component is converted into the corresponding oil-gas two-phase equilibrium partial pressure prediction value.

[0045] Based on the relationship between the predicted partial pressure of the oil and gas two-phase equilibrium and the current total pressure of the degassing chamber, the predicted degassing efficiency of each gas component at each future time point is calculated.

[0046] The predicted degassing efficiency values ​​of different gas components are mapped to a preset virtual degassing space, thereby generating a dynamic degassing efficiency field.

[0047] Further, S6 includes:

[0048] The dynamic degassing efficiency field and the preset target efficiency field are compared in each spatial unit of the virtual degassing space, and the efficiency difference between the two in each spatial unit is calculated.

[0049] Summarize the efficiency differences of all spatial units to form an efficiency difference matrix;

[0050] Traverse the efficiency difference matrix and identify all spatial units in which the efficiency difference exceeds the set compensation threshold;

[0051] Based on the inefficient gas components corresponding to all spatial units, spatial clustering analysis is performed to form several independent regions to be compensated.

[0052] For each area to be compensated, one or more virtual spray paths that effectively cover the area to be compensated are planned based on its spatial geometry and the distribution of internal efficiency differences.

[0053] The compensation level is set based on the average value of the efficiency difference in each region to be compensated;

[0054] Based on the virtual spray path planned for each area to be compensated, and in conjunction with the set compensation level, corresponding spray control commands and temperature control commands are generated, driving the actuator to perform graded area compensation operations sequentially along the virtual spray path.

[0055] Further, S7 includes:

[0056] After performing hierarchical regional compensation, the dynamic degassing efficiency field is updated; the root mean square error of all spatial units between the updated dynamic degassing efficiency field and the preset target efficiency field is calculated to determine whether the convergence of the dynamic degassing efficiency field meets the standard.

[0057] Extract the newly generated air pressure time series data after the compensation operation begins, and construct a new air pressure-time curve; calculate the fluctuation characteristic vector of the new air pressure-time curve to determine whether the morphological recovery of the air pressure-time curve meets the standard.

[0058] If the efficiency field convergence and morphological restoration both meet the standards, and the real-time degassing efficiency of all monitored gas components has entered their respective preset qualified ranges, then the degassing termination condition is met; otherwise, return to S4 for iterative optimization.

[0059] A second aspect of the present invention provides an adaptive degassing control system for spray oil-gas separation based on air pressure monitoring, comprising:

[0060] Data synchronization acquisition module: acquires the time-series data of air pressure in the degassing chamber during the spray degassing process, and synchronously records the time-series data of spray flow rate and oil temperature at the corresponding time points;

[0061] Pressure fluctuation extraction module: Constructs a pressure-time curve by combining pressure time series data, and extracts the fluctuation feature vector from the pressure-time curve;

[0062] Multivariate coupling analysis module: Performs multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and outputs the multidimensional stability coefficient of the current degassing process;

[0063] Core parameter determination module: Quantifies the mapping relationship between multidimensional stability coefficient and gas component degassing, thereby determining the core correlation parameters affecting degassing efficiency;

[0064] Degassing efficiency prediction module: Based on the core correlation parameters, construct a coupled correlation map between the concentration of degassing gas components and the rate of change of gas pressure to predict the dynamic degassing efficiency field of each gas component;

[0065] Spray path planning module: Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, it plans the virtual spray path and performs hierarchical regional compensation;

[0066] Degassing termination judgment module: Based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, it determines whether the degassing termination condition is met.

[0067] Compared with existing technologies, the beneficial effects of the adaptive degassing control method and system for spray oil-gas separation based on air pressure monitoring provided by this invention are as follows:

[0068] 1) By comprehensively acquiring multi-dimensional data, we can provide rich and accurate basic information for subsequent analysis, ensuring accurate grasp of the dynamic changes of various parameters during the degassing process; through calculation and analysis, we can construct a fluctuation feature vector that can comprehensively characterize the dynamic characteristics of gas pressure, effectively mine key information in the gas pressure data, and provide a basis for evaluating the stability of the degassing process.

[0069] 2) By integrating multiple stability components and assigning weights using the information entropy method, a multidimensional stability coefficient is obtained, which comprehensively evaluates the stability of the degassing process and lays the foundation for determining the core correlation parameters. By quantifying the dynamic relationship between the multidimensional stability coefficient and the gas component degassing, the core correlation parameters are determined, the key factors affecting the degassing efficiency are identified, and key support is provided for predicting the degassing efficiency.

[0070] 3) By constructing a coupling correlation map, the concentration of each gas component is estimated and a dynamic degassing efficiency field is generated, so as to grasp the trend of degassing efficiency changes in advance and facilitate timely adjustment of control strategies; by planning virtual spray paths based on efficiency field differences and performing hierarchical regional compensation, the degassing efficiency is improved in a targeted manner, and the uniformity and effectiveness of the degassing process are enhanced; by comprehensively judging the efficiency field convergence, morphological recovery degree and real-time degassing efficiency, it is possible to accurately determine whether the degassing termination condition is met, so as to avoid over-degassing or under-degassing. Attached Figure Description

[0071] Figure 1 This is a flowchart of the adaptive degassing control method for spray oil-gas separation based on air pressure monitoring proposed in this invention.

[0072] Figure 2This is a block diagram of the adaptive degassing control system for spray oil-gas separation based on air pressure monitoring proposed in this invention. Detailed Implementation

[0073] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example 1:

[0074] Please see Figure 1 This invention provides an adaptive degassing control method for spray oil-gas separation based on air pressure monitoring, comprising:

[0075] S1. Acquire the gas pressure time series data in the degassing chamber during the spray degassing process, and simultaneously record the spray flow time series data and oil temperature time series data at the corresponding time points.

[0076] S2. Construct a pressure-time curve by combining pressure time series data, and extract the fluctuation feature vector from the pressure-time curve;

[0077] S3. Perform multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and output the multidimensional stability coefficient of the current degassing process.

[0078] S4. Quantify the mapping relationship between the multidimensional stability coefficient and the gas component extraction to determine the core correlation parameters that affect the degassing efficiency;

[0079] S5. Based on the core correlation parameters, construct a coupled correlation map between the concentration of the extracted gas components and the rate of change of gas pressure to predict the dynamic degassing efficiency field of each gas component.

[0080] S6. Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, plan the virtual spray path and perform hierarchical regional compensation.

[0081] S7. Based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, determine whether the degassing termination condition is met.

[0082] It should be further explained that the acquisition of gas pressure time-series data in the degassing chamber during the spray degassing process, and the simultaneous recording of spray flow rate time-series data and oil temperature time-series data at the corresponding time points, are specifically implemented as follows:

[0083] Several high-precision pressure sensors are distributed and deployed inside the degassing chamber to synchronously collect time-series pressure data, including:

[0084] At least three high-precision pressure sensors are evenly arranged at key locations such as the top of the degassing chamber, the middle of the side wall, and the lower part near the oil spray inlet. The pressure sensors are triggered by a unified clock signal to ensure strict synchronization of the acquisition time points. The sampling frequency is set to 50Hz to capture the rapid pressure fluctuations that may occur during the spray degassing process. The raw voltage signal acquired by each pressure sensor is converted into a digital pressure value by an analog-to-digital converter and arranged in chronological order to form pressure time series data.

[0085] The total flow rate and branch flow ratio of the spray pipeline are monitored in real time using an electromagnetic flow meter to obtain time-series data of the spray flow rate, including:

[0086] An electromagnetic flowmeter is installed on the main circulation pipeline of the spray system to continuously measure the volumetric flow rate of transformer oil flowing through the pipeline. At the same time, proportional valves are installed on the branch pipelines leading to different spray areas. By reading the opening signal of the proportional valves and combining it with the main pipeline flow, the flow distribution ratio of each branch is calculated. The total flow data of the main pipeline and the calculated flow data of each branch are recorded with the same time base as the air pressure data to form the spray flow time series data.

[0087] Real-time oil temperature is collected by a platinum resistance temperature sensor attached to the oil chamber wall, yielding oil temperature time-series data, including:

[0088] A PT100 platinum resistance temperature sensor is used, which is tightly attached to a designated measurement point on the outer wall of the lower oil chamber of the degassing chamber with thermally conductive silicone grease to indirectly and stably measure the main body temperature of the oil. The measurement signal of the PT100 platinum resistance temperature sensor is processed by a temperature transmitter and output in digital form. It is also collected at a frequency of 50Hz and timestamped with the air pressure and flow rate to form oil temperature time series data.

[0089] It should be further explained that the construction of a pressure-time curve by combining pressure time series data and the extraction of fluctuation feature vectors from the pressure-time curve are specifically implemented as follows:

[0090] Wavelet packet decomposition is performed on the barometric pressure time series data to extract wavelet packet coefficients within a preset frequency band, and the energy of the wavelet packet coefficients within each time window is calculated, including:

[0091] The 'db4' wavelet basis function is selected to perform multi-level wavelet packet decomposition on continuous pressure time series data, with a decomposition level of 5. After multi-level wavelet packet decomposition, based on the physical characteristics of the degassing process, the frequency band from 0.1Hz to 2Hz is preset as the fluctuation frequency band that needs to be analyzed in detail. From all the wavelet packet nodes obtained by decomposition, all wavelet packet coefficients corresponding to the nodes whose frequencies cover the preset frequency band are extracted. The sum of squares of the extracted wavelet packet coefficients within the preset frequency band is calculated, and the sum of squares obtained is the wavelet packet coefficient energy.

[0092] The energy ratio is calculated based on the wavelet packet coefficients of adjacent time windows, resulting in energy ratios sorted by time, forming an energy ratio sequence, including:

[0093] Set a sliding time window with a length of 2 seconds, containing 100 sampling points at a sampling rate of 50Hz; extract wavelet packet coefficient energy within each sliding time window; calculate the ratio of the wavelet packet coefficient energy value of the current sliding time window to the wavelet packet coefficient energy value of the immediately preceding sliding time window; slide the calculation process across all atmospheric pressure time series data to obtain energy ratio values ​​arranged in chronological order, where the ordered set of obtained energy ratio values ​​constitutes the energy ratio sequence.

[0094] A linear fit is performed on the energy ratio sequence, and the slope obtained from the linear fit is used as the energy distribution gradient characterizing the trend of pressure fluctuation energy changes during that time period, including:

[0095] A relatively long analysis time window is selected, such as 30 seconds. Within the analysis time window, for the calculated energy ratio sequence data points, a linear fit is performed using the least squares method with time as the independent variable and energy ratio as the dependent variable. The linear fit yields a straight line, the slope of which reflects whether the pressure fluctuation energy shows an increasing trend, a decreasing trend, or remains stable within the analysis time window. The obtained slope value is the energy distribution gradient within the analysis time window, and its absolute value represents the severity of the trend change, while the positive or negative sign represents the direction of pressure fluctuation energy change.

[0096] The derivative calculation and analysis of the pressure-time curves are performed to automatically identify and mark candidate abrupt change points, including:

[0097] The central difference method is applied to calculate the first derivative of the time series pressure data point by point to obtain the pressure change rate sequence. The change rate threshold is determined according to the multiple of the standard deviation of the pressure change rate sequence in the current analysis period, for example, it is set to 1.5 times the standard deviation. By traversing the pressure change rate sequence, all points in the pressure change rate sequence whose absolute value exceeds the change rate threshold are identified. All identified points correspond to instantaneous events of a sharp rise or fall in pressure and are marked as candidate abrupt change points.

[0098] Within the set analysis time window, the absolute values ​​of the first derivatives of all candidate mutation points are accumulated, and this accumulated value is used as the cumulative intensity of mutation points within that time window, including:

[0099] Set an analysis time window of the same length as the energy distribution gradient being analyzed, such as 30 seconds; within the analysis time window, select all candidate mutation points that fall within the analysis time window; sum the absolute values ​​of the first derivatives corresponding to the selected candidate mutation points, where the summation value comprehensively reflects the frequency and amplitude of drastic pressure changes within the analysis time window. The larger the summation value, the more or more severe the sudden disturbance events within the analysis time window. This summation value is defined as the cumulative intensity of mutation points within the analysis time window.

[0100] The energy distribution gradient calculation process is repeatedly performed for a continuous analysis time window to generate an energy distribution gradient sequence; the cumulative intensity calculation process for abrupt change points is also repeatedly performed for a continuous analysis time window to generate a cumulative intensity sequence for abrupt change points, including:

[0101] A sliding window approach is used, sliding the 30-second analysis time window in a certain step (e.g., 1 second). For each newly covered 30-second data segment, the calculation process of energy distribution gradient and cumulative intensity of mutation points is repeated. Each time the window slides, an energy distribution gradient value and a cumulative intensity value of mutation points are calculated. Periodic monitoring is performed along the time axis to obtain a set of energy distribution gradients arranged in chronological order, forming a discrete-time energy distribution gradient sequence. At the same time, a set of cumulative intensities of mutation points arranged in chronological order is obtained, forming a cumulative intensity sequence of mutation points. The energy distribution gradient sequence and the cumulative intensity sequence of mutation points are then normalized.

[0102] The normalized energy distribution gradient sequence and the cumulative intensity sequence of abrupt change points are concatenated according to the same time index to form a fluctuation feature vector that comprehensively characterizes the dynamic characteristics of the current degassing process, including:

[0103] Since the two energy distribution gradient sequences and the cumulative intensity sequence of abrupt change points are calculated based on the same time base and sliding window, the two sequences have corresponding time indices. The normalized energy distribution gradient values ​​and the normalized cumulative intensity values ​​of abrupt change points under the same time index are combined into a two-dimensional vector. The two-dimensional vectors corresponding to all time points are arranged in chronological order to form a two-dimensional fluctuation feature vector. The fluctuation feature vector comprehensively characterizes the dynamic behavior of the gas pressure in the degassing chamber from two dimensions: energy change trend and event abrupt change intensity.

[0104] It should be further explained that the fluctuation feature vector is coupled with the acquired spray flow time series data and oil temperature time series data for multivariate analysis to output the multidimensional stability coefficient of the current degassing process. The specific implementation is as follows:

[0105] Based on the fluctuation feature vector, the data components corresponding to the energy distribution gradient and the data components of the cumulative intensity of the abrupt change point are extracted and used as the pressure fluctuation stability components and event sudden stability components of the current degassing process, respectively, including:

[0106] For the current preset time to be analyzed, a two-dimensional vector corresponding to the current preset time is extracted from the constructed fluctuation feature vector. The first element of the two-dimensional vector, namely the normalized energy distribution gradient value, is directly extracted as an indicator of whether the air pressure fluctuation is smooth or violent at the current preset time, and is defined as the air pressure fluctuation stability component. The second element of the two-dimensional vector, namely the normalized cumulative intensity value of the abrupt change point, is directly extracted as an indicator of the intensity of sudden disturbance events in the neighborhood of the current preset time, and is defined as the event sudden stability component. The smaller the values ​​of the two components, the more stable the air pressure state at the current preset time.

[0107] Acquire the current spray flow rate time series data and oil temperature time series data, and calculate their relative deviations from the corresponding preset reference values, which are used as flow rate stability components and temperature stability components, including:

[0108] Based on the spray flow time-series data, the instantaneous spray flow value corresponding to the current preset time is extracted. The instantaneous spray flow value is the flow rate measured and output by the electromagnetic flowmeter at the current sampling time. The preset optimal flow reference value is the theoretically optimal flow rate value pre-calibrated based on the current oil characteristics and degassing stage. The absolute difference between the current instantaneous spray flow value and the optimal flow reference value is calculated and then divided by the optimal flow reference value to obtain the relative deviation in percentage form. This relative deviation is the flow stability component. Similarly, the current instantaneous oil temperature value is extracted. The current instantaneous oil temperature value is the temperature value measured by the platinum resistance temperature sensor at the current sampling time and output by the transmitter. The relative deviation between the current instantaneous oil temperature value and the preset optimal temperature reference value is calculated as the temperature stability component. The pressure fluctuation stability component, the event sudden stability component, the flow stability component, and the temperature stability component constitute the stability component.

[0109] Obtain the stability components of N consecutive backtracking time windows during the current degassing process, and construct an N×4 comprehensive stability component matrix, including:

[0110] Set a backtracking analysis window with a time length that includes N consecutive sampling times, for example, N=20, which represents backtracking the most recent 20 sampling times. Extract the pressure fluctuation stability component, event sudden stability component, flow stability component, and temperature stability component from the N sampling times included in the backtracking analysis window from the stored time series data, for a total of N sets of data. Arrange the N sets of data into a matrix, where each row represents a historical time within the backtracking window, and each column represents a stability component, thus forming an N-row, 4-column comprehensive stability component matrix.

[0111] Using the information entropy method, by analyzing the uncertainty of each column of the comprehensive stability component matrix, a stability contribution entropy weight representing its relative importance is assigned to each stability component, including:

[0112] Normalize each column of the comprehensive stability component matrix composed of backtracking analysis window data so that the sum of all stability components in that column is 1, forming a probability distribution; calculate the information entropy value of each column of stability component according to the information entropy formula. The formula for information entropy is:

[0113] ;

[0114] In the formula, For the column index of the comprehensive stability component matrix, This is the index for the backtracking analysis window, where N is the number of backtracking time windows. For the first Liede The normalized probability values ​​of each backtracking analysis window; the larger the information entropy value, the higher the uncertainty and the more uniform the distribution of the stability component in that column; calculate the difference coefficient of each stability component in the column. : ;

[0115] Dividing the difference coefficient of each column by the sum of the difference coefficients of all columns yields the weight of the stability component represented by that column, i.e., the stability contribution entropy weight. The stability contribution entropy weight is calculated based on the distribution of historical data within the backtracking analysis window, reflecting the amount of information provided by each stability component in the recent period. Stability components with a larger amount of information receive higher weights in the comprehensive evaluation.

[0116] The multidimensional stability coefficients of the current degassing process are obtained by multiplying each stability component value in the current backtracking time window by its corresponding stability contribution entropy weight and then summing the results. These coefficients include:

[0117] The current assessment time is determined, i.e., the latest time at the end of the backtracking analysis window; the four stability component values ​​at the current assessment time are obtained; simultaneously, the stability contribution entropy weights corresponding to the four stability components calculated using the information entropy method based on the data from the same backtracking analysis window are obtained; each stability component value at the current assessment time is multiplied by its corresponding stability contribution entropy weight to obtain the weighted contribution value of that stability component; the four weighted contribution values ​​are added together, and the sum is the multidimensional stability coefficient characterizing the overall state of the degassing process at the current assessment time. The multidimensional stability coefficient is a comprehensive scalar value. The lower the multidimensional stability coefficient value, the more stable the current degassing process is in terms of pressure fluctuations, sudden events, flow rate, and temperature; the higher the multidimensional stability coefficient value, the worse the overall stability, which requires attention.

[0118] It needs further explanation that the mapping relationship between the quantified multidimensional stability coefficient and the gas component extraction is used to determine the core correlation parameters affecting the degassing efficiency. The specific implementation is as follows:

[0119] Real-time acquisition of concentration time-series data of various gaseous components dissolved in the oil during the degassing process, including:

[0120] The gas path system integrated at the top of the degassing chamber continuously guides the escaping mixed gas into an online photoacoustic spectroscopy detection device. This device has a built-in multi-component analysis algorithm that can measure and output the volume concentration percentage of various characteristic gases such as hydrogen, carbon monoxide, methane, ethylene, acetylene, and ethane in the mixed gas in real time. The collected concentration data is collected and recorded with the same time reference and frequency as the gas pressure and flow rate data, forming an independent time-series data sequence of concentration for each gas.

[0121] Based on the concentration time-series data of multiple gas components, the concentration change of each gas component per unit time is calculated to obtain the instantaneous extraction rate of that gas component; the calculation is repeated for consecutive time points to form a sequence of instantaneous extraction rates for each gas component, including:

[0122] For each gas component, such as hydrogen, its concentration time series data is taken; the difference between the gas component concentration value at the current sampling time and the gas component concentration value at the previous sampling time is calculated. The concentration difference approximately represents the amount of gas concentration change in that short period of time, i.e., the instantaneous degassing rate; the above calculation is performed sequentially for each sampling time to obtain the instantaneous degassing rate sequence of the gas component, which reflects the real-time fluctuation of the degassing rate.

[0123] Using the multidimensional stability coefficient as the independent variable and the instantaneous extraction rate sequence of each gas component as the dependent variable, a multivariate nonlinear regression model was constructed, including:

[0124] In modeling, the four dimensions of pressure fluctuation stability, event-induced stability, flow rate stability, and temperature stability are used as independent variables in a multivariate nonlinear regression model, denoted as X1, X2, X3, and X4. For each gas component, its instantaneous extraction rate is used as the dependent variable Y. A multinomial regression model is used to establish the mathematical relationship between the extraction rate of the gas component and the four stability dimensions, with the specific formula as follows:

[0125] ;

[0126] In the formula, β0 is the constant term, β1 to β8 are the regression coefficients, and ε is the error term.

[0127] By fitting a multivariate nonlinear regression model, the influence of different stability dimensions included in the multidimensional stability coefficient on the release rate of each gas component was analyzed. The regression coefficients corresponding to the stability dimensions with the highest influence were extracted and normalized to obtain the concentration-sensitive factors of the gas component, including:

[0128] The constructed multivariate nonlinear regression model was fitted using recent historical data, which included the values ​​of pressure fluctuation stability components, event-induced stability components, flow rate stability components, and temperature stability components at continuous time points, as well as the instantaneous gas component elution rates at the corresponding moments. After fitting, the regression coefficients β1 to β8 corresponding to the four independent variables in the multivariate nonlinear regression model were analyzed. The larger the absolute value of the regression coefficient, the greater the influence of the stability dimension represented by that independent variable on the gas elution rate. For each gas component, the regression coefficient with the largest absolute value was identified. This regression coefficient was extracted and normalized using the sum of the absolute values ​​of all regression coefficients in the multivariate nonlinear regression model, i.e., divided by this sum, so that the result is between zero and one. The normalized value is defined as the concentration sensitivity factor of that gas component, quantifying the relative influence intensity of the system stability most sensitive to the gas elution dynamics.

[0129] Acquire time-series air pressure data and calculate its first-order difference to obtain the first-order rate of change sequence of air pressure, including:

[0130] Based on the time series data of air pressure, the difference between adjacent sampling points is calculated, that is, the first-order difference operation is performed; the calculated difference sequence is the change of air pressure in each sampling interval, that is, the first-order rate of change of air pressure sequence, which directly reflects the instantaneous change rate of pressure in the degassing chamber.

[0131] Calculate the mutual information value between the instantaneous extraction rate sequence and the first-order rate of change of pressure sequence for each gas component. Normalize this mutual information value and define it as the pressure change rate influence weight, including:

[0132] Mutual information is used to measure the degree of interdependence between two random variables. For each gas component, the mutual information value of its instantaneous escape rate sequence and its first-order pressure change rate sequence is calculated. The calculation method includes estimating the joint probability distribution and the respective marginal probability distributions of the two sequences. Specifically, for each gas component, its instantaneous escape rate sequence and its first-order pressure change rate sequence are discretized, for example, by using equal-width or equal-frequency binning to map continuous values ​​into several discrete two-dimensional intervals. The frequency of the instantaneous escape rate sequence and the first-order pressure change rate sequence falling into each two-dimensional interval combination is counted and divided by the total number of data points to obtain the estimated value of the joint probability distribution Cr(a,b), where a represents the two-dimensional interval of the escape rate and b represents the two-dimensional interval of the pressure change rate. The frequency of each sequence falling into each two-dimensional interval is counted separately and divided by the total number of data points to obtain the estimated values ​​of their respective marginal probability distributions Cr(a) and Cr(b).

[0133] The larger the mutual information value, the stronger the correlation between the change in the gas component's extraction rate and the change in gas pressure. The mutual information values ​​of all gas components are normalized so that their sum is 1. The normalized mutual information value of each gas component is defined as the influence weight of the gas pressure change rate, that is, the influence weight of the gas pressure change rate on the gas extraction dynamics. The larger the influence weight of the gas pressure change rate, the more attention needs to be paid to the impact of gas pressure changes on the gas component in the control strategy. The gas component concentration sensitivity factor and the influence weight of the gas pressure change rate together constitute the core correlation parameters.

[0134] It should be further explained that, based on the core correlation parameters, a coupled correlation map of the concentration of the extracted gas components and the rate of change of gas pressure is constructed to predict the dynamic degassing efficiency field of each gas component. The specific implementation is as follows:

[0135] A coupling correlation map is constructed by plotting time on the horizontal axis and the rate of change of air pressure on the vertical axis, and using the influence weights of the concentration sensitivity factor and the rate of change of air pressure for each gas component as coupling modulation parameters. The map includes:

[0136] The coupling correlation map is a two-dimensional state plane with time as the horizontal axis and real-time pressure change rate as the vertical axis. The dynamic behavior of each gas component on this state plane is modulated by its two core correlation parameters: concentration sensitivity factor and pressure change rate influence weight. The concentration sensitivity factor determines the sensitivity of the gas component concentration evolution trajectory to stability perturbations when the stability state (reflected by the four stability components) changes. The pressure change rate influence weight determines the strength of the influence of pressure change rate fluctuations on the gas component concentration evolution trajectory along the vertical axis of the state plane. By integrating the concentration sensitivity factor and pressure change rate influence weight of each gas component as parameters into this state plane, a correlation map of coupling time, pressure change rate, and its own characteristics is constructed for each gas. The maps of all gas components together constitute a multi-gas coupling correlation map.

[0137] In the state space represented by the coupling correlation map, the concentrations of each gas component are predicted and analyzed for the next K time steps, forming the concentration prediction trajectory of each gas component, including:

[0138] In the state space defined by the coupling correlation map, a state equation and an observation equation are established for each gas component;

[0139] The equation of state characterizes the evolution of gas component concentrations over time, the rate of change of current pressure, the multidimensional stability coefficient at the current moment, and the core correlation parameters; the discrete-time form of the equation of state is:

[0140] Cu(t)=Au(t)×Cu(t-1)+Bu(t)×ΔP(t)×Wu+Ru×S(t)×Eu+mu(t);

[0141] In the formula, Cu(t) represents the concentration state of gas component u at the current time t; Cu(t-1) represents the concentration state of gas component u at the previous time t-1; Au represents the state transition matrix, used to characterize the natural decay or maintenance trend of concentration under no external disturbance; ΔP(t) is the rate of change of gas pressure at the current time t; Bu represents the influence coefficient of gas pressure change; Wu represents the influence weight of the rate of change of gas pressure of gas component u; S(t) represents the multidimensional stability coefficient at the current time t; Ru represents the stability influence coefficient matrix, used to adjust the contribution scale of the multidimensional stability coefficient to the concentration change; Eu represents the concentration sensitivity factor of gas component u; and mu(t) is the process noise.

[0142] The correlation between gas component concentrations and measured concentration values ​​is established through observation equations; the form of the observation equations is:

[0143] Du(t)=Hu×Cu(t)+Lu(t);

[0144] In the formula, Du(t) is the measured concentration of gas component u; Hu is the observation matrix, set as scalar 1; Cu(t) is the concentration state at the current time t; and Lu(t) is the observation noise.

[0145] The concentration prior is predicted using the equation of state. This prior estimate is then corrected by combining the actual observed concentrations of the gas components at the current moment, yielding the optimal, or posterior, estimate of the gas concentration at the current moment. Based on this optimal estimate, the concentration values ​​of the gas components at the next K time steps are recursively predicted using the equation of state and the predicted rate of change of pressure over those K time steps. This process is repeated as time progresses to the next moment, thus creating a rolling, updated concentration prediction trajectory for each gas component over time.

[0146] Based on the current oil temperature time-series data, the concentration prediction trajectories of each gas component are converted into corresponding predicted values ​​of the oil-gas two-phase equilibrium partial pressures, including:

[0147] According to Henry's Law, the dissolved concentration of a gaseous component in oil is directly proportional to its equilibrium partial pressure above the liquid surface. The proportionality constant is the dissolution constant, which is a function of temperature. Using real-time monitored oil temperature time-series data, and through a pre-stored table of dissolution constants for different gases at different temperatures, the dissolution constant of each gaseous component at the current oil temperature can be found. By multiplying the predicted concentration value at each future moment on the concentration prediction trajectory by the corresponding dissolution constant at that future moment, the partial pressure of the gaseous component in the gas phase when it reaches dissolution equilibrium at the future moment can be calculated, i.e., the predicted value of the equilibrium partial pressure of the oil and gas two phases.

[0148] Based on the relationship between the predicted partial pressures of the oil and gas two-phase equilibrium and the current total pressure of the degassing chamber, the predicted degassing efficiency of each gas component at various future time points is calculated, including:

[0149] Degassing efficiency is defined as the ratio of the actual amount of gas component removed from the oil to the theoretical maximum amount that can be removed; where the theoretical maximum amount that can be removed corresponds to the state when the gas reaches complete equilibrium; for a future predicted time, the predicted value of the oil-gas two-phase equilibrium partial pressure of the gas component is regarded as the theoretical partial pressure that the gas component may reach in the gas phase under the predicted conditions; assuming that the total pressure of the degassing chamber remains relatively stable or predictable, the predicted degassing efficiency of the gas component is approximately calculated as the predicted value of the oil-gas two-phase equilibrium partial pressure / the total pressure of the degassing chamber; for each gas component, the above calculation is performed at all future predicted time points to obtain a set of predicted degassing efficiency values.

[0150] The predicted degassing efficiency values ​​of different gas components are mapped to a preset virtual degassing space to generate a dynamic degassing efficiency field, including:

[0151] The virtual degassing space is a three-dimensional discrete grid space used to structure and organize predictive data. Its three dimensions are defined as follows: the time dimension corresponding to the future prediction time point, the gas type dimension corresponding to different gas components, and the spatial unit index dimension used to distinguish different logical regions or data channels. The virtual degassing space is composed of several basic units, namely spatial units, each of which has its own unique three-dimensional coordinate index. The predicted degassing efficiency values ​​of all gas components at each future time point are filled into the spatial units with corresponding coordinate indices in the virtual degassing space according to their corresponding time point and gas component type. This forms a dynamically updated structured data set containing multi-gas dimension information, namely the dynamic degassing efficiency field.

[0152] It should be further explained that, based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, a virtual spray path is planned and graded regional compensation is performed. The specific implementation is as follows:

[0153] The dynamic degassing efficiency field is compared with the preset target efficiency field in each spatial cell of the virtual degassing space, and the efficiency difference between the two in each spatial cell is calculated, including:

[0154] The preset target efficiency field is an efficiency field template predefined according to the degassing process requirements. It is used to specify the target degassing efficiency value that each spatial unit in the virtual degassing space should achieve when the degassing process is ideally completed. The real-time generated dynamic degassing efficiency field is compared with the preset target efficiency field on a unit-by-unit basis under the same spatial grid index and gas component type index. The efficiency difference of each spatial unit is obtained by the difference operation between the target degassing efficiency value and the predicted degassing efficiency value. If the calculated efficiency difference is positive, it means that the predicted degassing efficiency of the gas component corresponding to the spatial unit at the corresponding time is lower than the target degassing efficiency, and a compensation operation needs to be performed.

[0155] The efficiency differences of all spatial units are summarized to form an efficiency difference matrix, including:

[0156] The efficiency differences calculated from all spatial units are rearranged according to the three-dimensional grid structure of the virtual degassing space to form a three-dimensional matrix that is completely consistent with the grid dimension of the virtual degassing space, namely the efficiency difference matrix. The efficiency difference matrix intuitively shows the distribution of the differences between the current predicted state and the ideal target state in the three dimensions of space, time and gas component types.

[0157] Traverse the efficiency difference matrix and identify all spatial units whose efficiency difference exceeds a set compensation threshold, including:

[0158] Read the entire efficiency difference matrix. Each element of the matrix corresponds to a spatial cell in the virtual degassing space, and its value represents the difference between the predicted degassing efficiency and the target degassing efficiency of that spatial cell. Set a compensation threshold, for example, 0.1, which means that the predicted degassing efficiency is more than 10% lower than the target degassing efficiency. Traverse each element in the efficiency difference matrix and determine whether the value of the element is greater than the compensation threshold. Record the position coordinates of all elements that are greater than the compensation threshold in the virtual degassing space. The spatial cell corresponding to the position coordinates is identified as the problem cell that needs intervention, forming a list of problem cells to be processed.

[0159] Based on the inefficient gas components corresponding to all spatial units, spatial clustering analysis is performed to form several independent regions to be compensated, including:

[0160] From the list of problem units, spatial units belonging to inefficient gas components are selected based on the gas type dimension index. These inefficient gas components are pre-determined based on core correlation parameters, such as the 1-2 gases with the lowest concentration sensitivity factors. For the remaining problem units belonging to inefficient gas components after screening, spatial clustering analysis is performed in a virtual degassing space. Specifically, a density-based clustering algorithm is used to aggregate all problem units that are adjacent or sufficiently close in both time and spatial indices, forming a connected cluster. Each cluster represents a type of inefficient gas component that exhibits persistent inefficiency within a certain continuous time period and / or associated logical region; this cluster is defined as an independent region to be compensated. Each region to be compensated has a clear boundary and a corresponding inefficient gas type attribute, with the boundary defined by the coordinate range of all spatial units it contains.

[0161] For each area to be compensated, based on its spatial geometry and the distribution of internal efficiency differences, plan one or more virtual spray paths that effectively cover the area to be compensated, including:

[0162] For each clustered region to be compensated, its geometry in the virtual degassing space is analyzed. If the region to be compensated is elongated, a straight or curved path is planned that runs through its long axis. If the region to be compensated is planar, a set of parallel scan lines or concentric circle paths is planned. The goal of planning the virtual spray path is to ensure that the spray coverage area can cover as much of the region as possible. At the same time, the direction of the virtual spray path is preferably set to be perpendicular to the predicted concentration gradient of inefficient gas components within the region to be compensated.

[0163] Based on the average value of the efficiency difference within each region to be compensated, compensation levels are set, including:

[0164] Calculate the arithmetic mean of the efficiency differences of all spatial units within each compensation region. Based on the magnitude of the arithmetic mean, three compensation levels are established: primary compensation, intermediate compensation, and advanced compensation. For compensation regions with an arithmetic mean between 0.1 and 0.2, primary compensation is initiated by adjusting the spray pump frequency to increase or decrease the total spray flow rate. For compensation regions with an arithmetic mean between 0.2 and 0.3, intermediate compensation is initiated by further adjusting the nozzle orifice diameter or pressure to change the atomized particle size and increase the gas-liquid contact area, in addition to adjusting the flow rate. For compensation regions with an arithmetic mean greater than 0.3, advanced compensation is initiated by synergistically adjusting the spray incident angle (by driving a rotating nozzle) and providing localized auxiliary heating to the corresponding area of ​​the oil chamber, building upon the first two levels.

[0165] Based on the virtual spray path planned for each area to be compensated, and in conjunction with the set compensation level, corresponding spray control commands and temperature control commands are generated. These commands drive the actuators to sequentially perform graded area compensation operations along the virtual spray path, including:

[0166] The controller converts the virtual spray path into a sequence of motion control commands for the real spray mechanism, such as controlling the multi-degree-of-freedom spray arm to move to a specified position and angle. At the same time, it generates specific parameter commands such as flow rate setpoint, atomization pressure setpoint, and local heater power setpoint according to the compensation level. The generated commands are synchronously sent to the actuators such as the spray pump, proportional valve, atomization controller, rotary drive motor, and local heater. The actuators work together according to the commands to make the spray head move along the planned physical path and perform enhanced spray degassing operation on the area to be compensated under the set flow rate, atomization effect, and temperature conditions.

[0167] It should be further explained that, based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, the determination of whether the degassing termination condition is met is made. The specific implementation is as follows:

[0168] After performing hierarchical regional compensation, the dynamic degassing efficiency field is updated; the root mean square error of all spatial units between the updated dynamic degassing efficiency field and the preset target efficiency field is calculated to determine whether the convergence of the dynamic degassing efficiency field meets the standard.

[0169] Understandably, the updated dynamic degassing efficiency field is compared with the same preset target efficiency field again on a cell-by-cell basis. The sum of squares of the efficiency differences of all corresponding spatial cells is calculated and divided by the total number of spatial cells to obtain the mean square error. The square root of this mean square error is then taken to obtain the root mean square error. This error value comprehensively quantifies the overall gap between the current prediction state and the ideal target state. Among them, the convergence degree of the dynamic degassing efficiency field refers to the degree to which the overall difference between the updated predicted degassing efficiency field and the preset target efficiency field is reduced after the compensation operation.

[0170] Extract the newly generated pressure-time data after the compensation operation begins, and construct a new pressure-time curve; calculate the fluctuation characteristic vector of the new pressure-time curve to determine whether the morphological recovery of the pressure-time curve meets the standard; whereby the morphological recovery of the pressure-time curve refers to the degree to which the fluctuation characteristic vector (such as energy distribution gradient and cumulative intensity of abrupt change points) of the newly acquired pressure-time curve recovers to the stable morphology of the normal stable degassing state after the compensation operation is performed.

[0171] If the efficiency field convergence and morphological restoration both meet the standards, and the real-time degassing efficiency of all monitored gas components has entered their respective preset qualified ranges, then the degassing termination condition is met. The qualified range refers to the range of degassing efficiency values ​​that are pre-set for each monitored gas component to characterize that its degassing effect has met the standards. Otherwise, return to S4 for iterative optimization.

[0172] Example 2

[0173] Please see Figure 2 This invention provides an adaptive degassing control system for spray oil-gas separation based on air pressure monitoring, comprising:

[0174] Data synchronization acquisition module: acquires the time-series data of air pressure in the degassing chamber during the spray degassing process, and synchronously records the time-series data of spray flow rate and oil temperature at the corresponding time points;

[0175] Pressure fluctuation extraction module: Constructs a pressure-time curve by combining pressure time series data, and extracts the fluctuation feature vector from the pressure-time curve;

[0176] Multivariate coupling analysis module: Performs multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and outputs the multidimensional stability coefficient of the current degassing process;

[0177] Core parameter determination module: Quantifies the mapping relationship between multidimensional stability coefficient and gas component degassing, thereby determining the core correlation parameters affecting degassing efficiency;

[0178] Degassing efficiency prediction module: Based on the core correlation parameters, construct a coupled correlation map between the concentration of degassing gas components and the rate of change of gas pressure to predict the dynamic degassing efficiency field of each gas component;

[0179] Spray path planning module: Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, it plans the virtual spray path and performs hierarchical regional compensation;

[0180] Degassing termination judgment module: Based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, it determines whether the degassing termination condition is met.

[0181] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. The focus of each embodiment is on its differences from other embodiments. In particular, the apparatus embodiments are described simply because they are fundamentally based on the method embodiments; relevant details can be found in the descriptions of the method embodiments.

[0182] For ease of description, the above devices are described separately by function as various units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0183] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0184] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0185] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0186] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0187] Secondly: The accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other.

[0188] In conclusion, the above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. An adaptive degassing control method for spray oil-gas separation based on air pressure monitoring, characterized in that, include: S1. Acquire the gas pressure time series data in the degassing chamber during the spray degassing process, and simultaneously record the spray flow time series data and oil temperature time series data at the corresponding time points. S2. Construct a pressure-time curve by combining pressure time series data, and extract the fluctuation feature vector in the pressure-time curve; wherein, the fluctuation feature vector is composed of the pressure fluctuation energy change trend component obtained by wavelet packet decomposition and the pressure sudden event intensity component obtained by derivative analysis. S3. Perform multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and output the multidimensional stability coefficient of the current degassing process; where the multidimensional stability coefficient is a comprehensive scalar obtained by weighted summation of the components based on four dimensions: pressure fluctuation stability, event sudden stability, flow stability and temperature stability, after determining the weights by the information entropy method. S4. Quantify the mapping relationship between the multidimensional stability coefficient and the gas component extraction to determine the core correlation parameters affecting the degassing efficiency. Among them, the core correlation parameters include the concentration-sensitive factor used to characterize the influence of different stability dimensions on the extraction rate of each gas component, and the pressure change rate influence weight used to characterize the influence of the pressure change rate on the extraction rate of each gas component. S5. Based on the core correlation parameters, construct a coupled correlation map between the concentration of the extracted gas components and the rate of change of gas pressure, and predict the dynamic degassing efficiency field of each gas component; wherein, the dynamic degassing efficiency field is a data field generated by mapping the predicted degassing efficiency values ​​of each gas component at each future time point to a virtual degassing space with time, gas type and spatial unit as dimensions. S6. Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, plan the virtual spray path and perform hierarchical regional compensation. S7. Determine whether the degassing termination condition is met based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve; wherein, the morphological recovery is determined by calculating the fluctuation characteristic vector corresponding to the newly generated pressure time series data after the compensation operation.

2. The pneumatic pressure monitoring based spray oil and gas separation self-adaptive degassing control method according to claim 1, characterized in that, S1 includes: Several high-precision pressure sensors are distributed and deployed inside the degassing chamber to synchronously collect time-series pressure data; The total flow rate and branch flow ratio of the spray pipeline are monitored in real time by electromagnetic flowmeter to obtain spray flow time series data; The real-time temperature of the oil is collected by a platinum resistance temperature sensor attached to the wall of the oil chamber, and oil temperature time series data is obtained.

3. The pneumatic pressure monitoring based spray oil and gas separation self-adaptive degassing control method according to claim 2, characterized in that, S2 includes: Wavelet packet decomposition is performed on the barometric pressure time series data to extract wavelet coefficients within a preset frequency band, and the energy of the wavelet packet coefficients within each time window is calculated. The energy ratio is calculated based on the wavelet packet coefficients of adjacent time windows, and the energy ratios are sorted by time to form an energy ratio sequence. A linear fit is performed on the energy ratio sequence, and the slope obtained from the linear fit is used as the energy distribution gradient characterizing the trend of energy change in air pressure fluctuations during that period.

4. The pneumatic pressure monitoring based spray oil-gas separation self-adaptive degassing control method according to claim 2, wherein, S2 further includes: The derivative calculation and analysis of the pressure-time curve are performed to automatically identify and mark candidate abrupt change points; Within the set analysis time window, the absolute values ​​of the first derivatives of all candidate mutation points are accumulated, and this accumulated value is used as the cumulative intensity of mutation points in that analysis time window. The energy distribution gradient calculation process is repeatedly performed for the continuous analysis time window to generate an energy distribution gradient sequence; at the same time, the cumulative intensity calculation process of abrupt change points is repeatedly performed for the continuous analysis time window to generate a cumulative intensity sequence of abrupt change points. The energy distribution gradient sequence and the cumulative intensity sequence of abrupt change points were normalized respectively. The normalized energy distribution gradient sequence and the cumulative intensity sequence of abrupt change points are spliced ​​together according to the same time index to form a fluctuation feature vector that comprehensively characterizes the dynamic characteristics of the current degassing process.

5. The pneumatic pressure monitoring based spray oil-gas separation self-adaptive degassing control method according to claim 4, characterized in that, S3 includes: Based on the fluctuation feature vector, the data components corresponding to the energy distribution gradient and the data components of the cumulative intensity of the mutation point are extracted and used as the pressure fluctuation stability component and the event sudden stability component of the current degassing process, respectively. Acquire the current spray flow rate time series data and oil temperature time series data, and calculate their relative deviation from the corresponding preset benchmark values, which are used as the flow rate stability component and temperature stability component, respectively. Obtain the stability components of N consecutive backtracking time windows during the current degassing process, and construct an N×4 comprehensive stability component matrix; Using the information entropy method, by analyzing the uncertainty of each column of the comprehensive stability component matrix, a stability contribution entropy weight representing its relative importance is assigned to each stability component. The multidimensional stability coefficient of the current degassing process is obtained by multiplying each stability component value of the current backtracking time window by its corresponding stability contribution entropy weight and summing the results.

6. The pneumatic pressure monitoring based spray oil-gas separation self-adaptive degassing control method according to claim 5, characterized in that, S4 includes: Real-time acquisition of concentration time-series data of various gaseous components dissolved in oil during the degassing process; Based on the concentration time series data of multiple gas components, the concentration change of each gas component per unit time is calculated to obtain the instantaneous extraction rate of the gas component; the calculation is repeated for continuous time points to form a sequence of instantaneous extraction rates of each gas component. A multivariate nonlinear regression model was constructed using the multidimensional stability coefficient as the independent variable and the instantaneous extraction rate sequence of each gas component as the dependent variable. By fitting a multivariate nonlinear regression model, the influence of different stability dimensions contained in the multidimensional stability coefficient on the desorption rate of each gas component is analyzed. The regression coefficients corresponding to the stability dimensions with the highest influence are extracted and normalized to obtain the concentration sensitivity factor of the gas component. Obtain time-series air pressure data and calculate its first-order difference to obtain the first-order rate of change sequence of air pressure. The mutual information value between the instantaneous extraction rate sequence of each gas component and the first-order rate of change of gas pressure is calculated separately. After normalizing the mutual information value, it is defined as the influence weight of the rate of change of gas pressure. The gas component concentration sensitivity factor and the influence weight of the gas pressure change rate together constitute the core correlation parameter.

7. The adaptive degassing control method for spray oil-gas separation based on air pressure monitoring according to claim 6, characterized in that, S5 includes: A coupled correlation map was constructed by using time as the horizontal axis, the rate of change of air pressure as the vertical axis, and the influence weights of the concentration sensitivity factor and the rate of change of air pressure of each gas component as coupling modulation parameters. In the state space represented by the coupling correlation map, the concentration of each gas component is predicted and analyzed in the next K time steps, forming the concentration prediction trajectory of each gas component. Based on the current oil temperature time series data, the concentration prediction trajectory of each gas component is converted into the corresponding oil-gas two-phase equilibrium partial pressure prediction value. Based on the relationship between the predicted partial pressure of the oil and gas two-phase equilibrium and the current total pressure of the degassing chamber, the predicted degassing efficiency of each gas component at each future time point is calculated. The predicted degassing efficiency values ​​of different gas components are mapped to a preset virtual degassing space, thereby generating a dynamic degassing efficiency field.

8. The pneumatic pressure monitoring based spray oil-gas separation self-adaptive degassing control method according to claim 7, characterized in that, S6 includes: The dynamic degassing efficiency field and the preset target efficiency field are compared in each spatial unit of the virtual degassing space, and the efficiency difference between the two in each spatial unit is calculated. Summarize the efficiency differences of all spatial units to form an efficiency difference matrix; Traverse the efficiency difference matrix and identify all spatial units in which the efficiency difference exceeds the set compensation threshold; Based on the inefficient gas components corresponding to all spatial units, spatial clustering analysis is performed to form several independent regions to be compensated. For each area to be compensated, one or more virtual spray paths that effectively cover the area to be compensated are planned based on its spatial geometry and the distribution of internal efficiency differences. The compensation level is set based on the average value of the efficiency difference in each region to be compensated; Based on the virtual spray path planned for each area to be compensated, and in conjunction with the set compensation level, corresponding spray control commands and temperature control commands are generated, driving the actuator to perform graded area compensation operations sequentially along the virtual spray path.

9. The adaptive degassing control method for spray oil-gas separation based on air pressure monitoring according to claim 8, characterized in that, S7 includes: After performing hierarchical regional compensation, the dynamic degassing efficiency field is updated; the root mean square error of all spatial units between the updated dynamic degassing efficiency field and the preset target efficiency field is calculated to determine whether the convergence of the dynamic degassing efficiency field meets the standard. Extract the newly generated air pressure time series data after the compensation operation begins, and construct a new air pressure-time curve; calculate the fluctuation characteristic vector of the new air pressure-time curve to determine whether the morphological recovery of the air pressure-time curve meets the standard. If the efficiency field convergence and morphological restoration both meet the standards, and the real-time degassing efficiency of all monitored gas components has entered their respective preset qualified ranges, then the degassing termination condition is met; otherwise, return to S4 for iterative optimization.

10. An adaptive degassing control system for spray oil-gas separation based on air pressure monitoring, characterized in that, The system, applied to the adaptive degassing control method for spray oil-gas separation based on air pressure monitoring as described in any one of claims 1-9, comprises: Data synchronization acquisition module: acquires the time-series data of air pressure in the degassing chamber during the spray degassing process, and synchronously records the time-series data of spray flow rate and oil temperature at the corresponding time points; Pressure fluctuation extraction module: Constructs a pressure-time curve by combining pressure time series data, and extracts the fluctuation feature vector from the pressure-time curve; Multivariate coupling analysis module: Performs multivariate coupling analysis on the fluctuation feature vector with the acquired spray flow time series data and oil temperature time series data, and outputs the multidimensional stability coefficient of the current degassing process; Core parameter determination module: Quantifies the mapping relationship between multidimensional stability coefficient and gas component degassing, thereby determining the core correlation parameters affecting degassing efficiency; Degassing efficiency prediction module: Based on the core correlation parameters, construct a coupled correlation map between the concentration of degassing gas components and the rate of change of gas pressure to predict the dynamic degassing efficiency field of each gas component; Spray path planning module: Based on the spatial difference distribution between the dynamic degassing efficiency field and the preset target efficiency field, it plans the virtual spray path and performs hierarchical regional compensation; Degassing termination judgment module: Based on the convergence of the compensated dynamic degassing efficiency field and the morphological recovery of the pressure-time curve, it determines whether the degassing termination condition is met.

Images