Control System Applied to Oil-Water-Sludge Separator

Abstract

The present invention discloses a control system applied to an oil-water-sludge separator, which relates to the technical field of equipment control. A closed-loop control architecture including a liquid level sensor, a physical property sensor, a data processing module, a control decision-making module, and an execution module is constructed. By collecting physical property parameters such as liquid level and conductivity in real time, extracting the liquid level change trend and the conductivity mutation characteristics, constructing an emulsification interference prediction model, dynamically correcting the discharge threshold, and combining the separation time logic to intelligently judge the discharge conditions, the discharge operations of oil, water, and sludge are accurately controlled. This system effectively avoids mis-discharge behaviors under the emulsified state, improves the separation efficiency and the automation level, and significantly enhances the operation stability and environmental protection compliance of the equipment under high-complexity working conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment control, and specifically relates to a control system applied to an oil-water-sludge separator. Background Art

[0002] A control system applied to an oil-water-sludge separator refers to an automated system specifically used to manage and regulate the separation process of oil, water, and solid impurities (three phases). This system monitors the states of each phase of substances in the separator through sensors, and uses the logic set by the program to control the operation of valves, pumps, or stirring devices to achieve efficient and stable oil-water-sludge separation, improve the treatment efficiency and resource recovery rate, while reducing manual intervention and environmental pollution.

[0003] The existing technology has the following deficiencies:

[0004] In the actual application of an oil-water-sludge separator, under complex working conditions such as high temperature, high flow rate, or stirring, it is easy to cause fluctuations in the oil-water interface or form an emulsion layer. It is difficult for the liquid level sensor to accurately identify the true demarcation position, resulting in misjudgment of the liquid level state by the control system. For example, in the treatment of heavy oil wastewater, the sensor misjudges the emulsion layer as the water level, causing the control system to discharge the mixed liquid in advance, resulting in insufficient oil recovery and pollutant discharge, posing a serious environmental compliance risk. Summary of the Invention

[0005] The purpose of the present invention is to provide a control system applied to an oil-water-sludge separator to solve the deficiencies in the background art.

[0006] To achieve the above purpose, the present invention provides the following technical solution: A control system applied to an oil-water-sludge separator, including a liquid level sensor, a physical property sensor, a data processing module, a control decision module, and an execution module;

[0007] The liquid level sensor is used to collect the height data of the oil phase, water phase, and sludge phase in real time;

[0008] The physical property sensor is used to collect information on the conductivity, density, or temperature change of the separation medium;

[0009] The data processing module is used to receive the data from the liquid level sensor and the physical property sensor, extract the liquid level change trend feature and the conductivity mutation feature, construct an emulsion interference prediction model, and is used to judge whether there is emulsion interference at the current liquid surface and correct the discharge threshold;

[0010] The control decision module is connected to the data processing module, and judges whether the discharge condition is reached according to the identified true oil-water interface position and the separation time logic;

[0011] The execution module includes an oil discharge device, a water discharge device, and a slag discharge device connected to the control decision module, and is used to perform oil, water, or slag discharge operations respectively when receiving a discharge instruction.

[0012] Preferably, the liquid level sensor includes liquid level detection units arranged at multiple points, which are arranged along the vertical height direction of the oil-water-slag separator and are used to detect the heights of the oil phase, water phase, and slag phase respectively.

[0013] Preferably, the physical property sensor includes a conductivity sensor, a density sensor, and a temperature sensor. The conductivity sensor is used to detect the change in conductivity at the oil-water interface, the density sensor is used to identify the density distribution of different phase media, and the temperature sensor is used to provide medium temperature information.

[0014] Preferably, the method for extracting the liquid level change trend feature is as follows: Let the liquid level height output by the liquid level sensor at time t be h(t). A continuous data sequence is extracted with a sliding window of length N. For the data sequence within the window, the number of positive and negative changes of the curve per unit time is calculated through the zero-crossing detection method, and the obtained fluctuation frequency f is used as the liquid level change trend feature.

[0015] Preferably, the method for extracting the conductivity mutation feature is as follows: Let the conductivity data collected by the sensor in the time series be: ; where: is the conductivity value collected at the t-th moment; Two recursive paths are set: Positive deviation detection: ; Negative deviation detection: ; In the formula, is the expected value of conductivity in the stable state, k is the offset sensitivity constant, h is the determination threshold, and when the cumulative deviation exceeds h, it is regarded as a mutation; , respectively represent the cumulative deviations in the rising and falling directions, and the initial conditions: =0;

[0016] At each sampling moment t, if >h, it is considered that a positive conductivity mutation has occurred; if >h, it is considered that a negative conductivity mutation has occurred; if both are less than h, there is no significant mutation, and the ratio of the number of mutation time points to the total number of detected time points is calculated and used as the conductivity mutation feature.

[0017] Preferably, the emulsification interference prediction model is constructed using an improved decision tree algorithm, specifically including:

[0018] Based on historical operation data and manual annotation results, a training sample set including liquid level change trend features, conductivity mutation features, and emulsification state labels is constructed;

[0019] The information gain ratio is used as the splitting criterion to establish a decision tree structure for emulsification state classification;

[0020] The pruning strategy is adopted to prevent the model from overfitting;

[0021] During the real-time operation of the system, the feature vector of the current cycle is input, and the judgment result of whether there is emulsification interference is output.

[0022] Preferably, during actual operation, sensor data is collected in real time to form a new feature vector Xnew, which is input into the trained decision tree model for judgment. If the model outputs Ynew = 1, it is judged that the current liquid level is in an emulsified state, and the control module pauses the discharge accordingly and starts the emulsification delay processing logic; if the model outputs Ynew = 0, the current discharge process is normally executed.

[0023] Preferably, when emulsification interference is identified, the discharge threshold is dynamically corrected:

[0024] The degree of liquid level fluctuation and the degree of sudden change in conductivity are transformed into fuzzy linguistic variables;

[0025] The input variables are inferred through the fuzzy rule base, and the discharge threshold correction amount is output;

[0026] The centroid method is used to defuzzify the fuzzy set to generate a specific correction value;

[0027] The correction value is applied to the current drainage or oil discharge liquid level threshold to delay or adjust the discharge trigger condition.

[0028] Preferably, the discharge conditions judged by the control decision module include:

[0029] Obtain the current oil-water interface height, water-sludge interface height, separator operation time, emulsification state flag, and conductivity stability index;

[0030] If the oil-water interface height is greater than or equal to the oil discharge threshold, and it is in a non-emulsified state, and the time since the last oil discharge exceeds the minimum oil discharge cycle, then an oil discharge command is triggered;

[0031] If the oil-water interface height is less than or equal to the drainage threshold, and it is in a non-emulsified state, and the conductivity change is stable, and the minimum drainage interval condition is met, then a drainage command is triggered;

[0032] If the slag layer height reaches the set threshold, or the operation time exceeds the slag discharge cycle, and the minimum slag discharge interval is met, then a slag discharge command is triggered.

[0033] In the above technical solution, the technical effects and advantages provided by the present invention:

[0034] 1. In view of complex working conditions such as high temperature, high flow rate, and easy emulsification, the present invention constructs a multi-source data fusion mechanism through the collaborative perception of a liquid level sensor and a physical property sensor, extracts the liquid level change trend and the sudden change characteristics of conductivity, and effectively identifies the emulsification interference state. The system introduces a decision tree recognition model and a fuzzy logic dynamic correction strategy, which not only realizes the intelligent judgment of the discharge decision but also dynamically adjusts the discharge threshold to avoid misdischarging the mixed liquid and ensure the stability and accuracy of the oil-water-sludge separation.

[0035] 2. The present invention significantly improves the operation adaptability and control accuracy of the oil-water-sludge separator under non-ideal working conditions. It not only improves the oil-water separation efficiency and resource recovery rate, reduces the frequency of manual intervention, but also greatly reduces the risk of misdischarging pollutants caused by emulsification and strengthens the discharge compliance. This system has the capabilities of real-time monitoring, self-learning optimization, and feedback regulation, and is widely applicable to fields such as petrochemical industry, kitchen waste water treatment, and oil-containing waste liquid recovery. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0037] Figure 1 It is the system mind map of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention belong to the scope of protection of the present invention.

[0039] Embodiment, please refer to Figure 1 As shown, the control system applied to the oil-water-sludge separator in this embodiment includes a liquid level sensor, a physical property sensor, a data processing module, a control decision module, and an execution module;

[0040] The liquid level sensor is used to collect the height data of the oil phase, water phase, and sludge phase in real time;

[0041] The physical property sensor is used to collect the conductivity, density, or temperature change information of the separation medium;

[0042] A data processing module, which is used to receive the data of the liquid level sensor and the physical property sensor, extract the characteristics of the liquid level change trend and the characteristics of the sudden change of conductivity, construct an emulsification interference prediction model, and is used to judge whether there is emulsification interference on the current liquid surface and correct the discharge threshold;

[0043] A control decision-making module, which is connected to the data processing module, and judges whether the discharge condition is reached according to the identified real oil-water interface position and separation time logic;

[0044] An execution module, including an oil discharge device, a water discharge device and a slag discharge device connected to the control decision-making module, which is used to perform the discharge operations of oil, water or slag respectively when receiving a discharge instruction.

[0045] In the control system of the oil-water-slag separator of the present invention, a liquid level sensor assembly is provided, which is used to monitor the height of the multiphase medium in the separator in real time, so that the system can accurately judge the interface position and thickness of the oil, water and slag layers, thereby providing data support for subsequent discharge control.

[0046] The liquid level sensor can adopt a multi-point installation or a multi-layer recognition structure, and is respectively arranged at different vertical heights of the separator cavity to correspond to the conventional distribution areas of the oil phase layer, the water phase layer and the slag phase layer. Among them, the liquid level sensor adopts a static pressure type liquid level sensor or a capacitance type liquid level sensor, and a radar wave liquid level gauge or an ultrasonic sensor can also be selected according to the actual working conditions to realize non-contact monitoring.

[0047] Each liquid level sensor outputs a signal representing the liquid surface height in real time, and this signal is transmitted to the data acquisition module in the control system through an analog or digital interface. To improve the measurement accuracy, the sensor has an automatic temperature compensation function to adapt to the working conditions in high-temperature and high-oil environments.

[0048] During operation, the liquid level sensor can not only collect the height of each phase layer in the static separation state, but also identify the liquid level fluctuation trend, provide a continuous liquid level change curve, and assist in judging whether there is oil-water mixing, interface fluctuation or emulsification phenomenon.

[0049] Specifically, when the liquid level sensor detects a sharp fluctuation in the height of the oil-water interface within a set time window, or the thickness of the water layer rises significantly within a short time but the physical properties do not change, the system will initially determine that there is emulsification interference. At this time, the liquid level data will be jointly analyzed with other sensors (such as density, conductivity) to further judge whether the real discharge conditions are met.

[0050] Through the implementation of this liquid level sensor system, the multi-phase liquid surface stratification identification and dynamic monitoring of the oil phase, water phase and slag phase can be realized, significantly improving the real-time control ability of the control system for the separation process and avoiding wrong discharge or resource waste caused by misjudgment of the liquid level.

[0051] To improve the accuracy of oil-water-slag separator in identifying the oil-water interface under complex working conditions, the present invention further provides a physical property sensor unit for collecting physical property parameters such as the conductivity, density and temperature of the separation medium, so as to assist in liquid level identification and enhance the multi-source data comprehensive judgment ability of the control system.

[0052] The physical property sensors may include but are not limited to the following types:

[0053] Conductivity sensor: used to measure the change in conductivity between different phase media. Since the conductivity of the oil phase is extremely low and that of the water phase is strong, the oil-water interface can be assisted in identification by detecting the change in conductivity gradient.

[0054] Density sensor: based on the principle of differential pressure or float type, it detects the density difference of different liquid layers. The densities of heavy oil, water and sediment are different, and the sensor can be used to distinguish whether the layered structure is clear and whether there is emulsification.

[0055] Temperature sensor: used to detect the overall and local temperature distribution of the separation medium, provide thermal field information, and serve as a temperature compensation reference to improve the accuracy and stability of other sensors.

[0056] During implementation, the physical property sensors are arranged in different height sections of the separator or key areas near the interface. The sensor data is transmitted to the data processing module of the system through the A / D conversion module for digital processing. This module receives the physical property data in real time, cross-verifies it with the height information provided by the liquid level sensor, and forms a complete separation state judgment mechanism.

[0057] For example, when the liquid level sensor detects that the water level height exceeds the discharge threshold, but the conductivity change is not obvious or the density change is extremely gentle, the system judges that the current state is an oil-water mixture or emulsification state, thereby preventing the drain valve from opening in advance and avoiding misdischarging the incompletely separated mixture.

[0058] In addition, the data of the physical property sensors can also be used to identify abnormal operations. For example, a sharp rise in temperature indicates abnormal heating, or a sudden change in density warns of the intrusion of impurities. All sensor data can be recorded and used for subsequent separation efficiency evaluation and control strategy optimization.

[0059] By introducing this physical property sensor system, the present invention effectively overcomes the misjudgment problem existing in traditional methods that only rely on liquid level detection, realizes multi-parameter collaborative identification and dynamic correction control of the oil-water interface, and significantly improves the intelligent level and operation stability of the system.

[0060] In the oil-water-sludge separator control system of the present invention, a data processing module is provided, which is used to receive multi-source data from a liquid level sensor and a physical property sensor, perform feature extraction and intelligent discrimination operations, and construct an emulsification interference prediction model based on the liquid level behavior characteristics and physical property mutation characteristics to achieve intelligent determination and dynamic correction of the discharge conditions.

[0061] The data processing module can be composed of an embedded processor (such as an ARM architecture chip), an industrial computer or an edge computing unit, and internally integrates a set of processing algorithms for emulsification state recognition, mainly including:

[0062] The real-time height signal output by the liquid level sensor, and the conductivity, density, temperature and other data output by the physical property sensor are uniformly collected, and time synchronization, denoising filtering and normalization processing are performed to ensure the consistency of the data source and the processing accuracy.

[0063] Feature extraction of liquid level change trend: By analyzing the fluctuation frequency, amplitude change or rising / falling rate of the liquid level curve through a short-time sliding window, an abnormal fluctuation pattern is extracted. For example, the present invention extracts the liquid level change trend characteristics in the following way:

[0064] Let the liquid level height output by the liquid level sensor at time t be h(t), and a continuous data sequence is extracted with a sliding window of length N: ; where N is the length of the sliding window, in the unit of sampling points (for example, 30 points, corresponding to a 10-second sampling window); is the liquid level data sequence within the current sliding window. For the data sequence within the window, the fluctuation frequency f is extracted, and the number of positive and negative changes of the curve per unit time is calculated by the zero-crossing detection method: ; where: is the average liquid level within the window, Δt is the sampling period, and δ is a conditional function, which is 1 when the condition is satisfied, otherwise 0. The obtained fluctuation frequency f is used as the liquid level change trend characteristic.

[0065] Feature extraction of conductivity mutation: Identify the mutation points of conductivity in the spatial or temporal dimension to reflect whether the oil-water interface is clear, whether mixing or emulsification occurs. For example, the present invention extracts the conductivity mutation characteristics in the following way:

[0066] Let the conductivity data collected by the sensor in the time series be: ; where: is the conductivity value collected at the t-th moment (unit: mS / cm);

[0067] Two recursive paths are set: Forward deviation detection (conductivity suddenly rises): ; Negative deviation detection (conductivity suddenly drops): ; where, is the expected value of the conductivity in the steady state (which can be estimated by the mean value of the historical window); k is the offset sensitivity constant, representing the tolerable normal fluctuation range; h is the determination threshold, and when the cumulative deviation exceeds h, it is regarded as a mutation; and respectively represent the cumulative deviations in the rising and falling directions. Initial condition: = 0.

[0068] At each sampling moment t, the system judges: If > h, it is considered that a positive mutation of the conductivity occurs (which may be caused by the entry of water layer or the dissipation of the mixed layer); if > h, it is considered that a negative mutation of the conductivity occurs (which may be caused by the mixing of oil and water or the aggravation of emulsification); if both are less than h, the system considers that the conductivity is stable and there is no significant mutation. Calculate the ratio of the number of mutation time points to the total number of detected time points, and use it as the conductivity mutation feature.

[0069] Based on the training samples (which can be historical data or preset rules), construct an emulsification state recognition model, and use the improved decision tree model to classify and judge the feature combination to judge whether the current liquid level is in the emulsification interference state. Specifically:

[0070] Take the obtained liquid level change trend feature and conductivity mutation feature as the input feature vector X of the decision tree model, and the output item Y is: Y = 1, the liquid level is in the emulsification interference state; Y = 0, the liquid level is in the non-emulsification state;

[0071] Combine the historical operation data and the manual annotation results to construct a training sample data set including input features and emulsification state labels. The emulsification state labels in the samples include two categories: emulsification and non-emulsification, which are used to train the classification model.

[0072] The present invention constructs an emulsification state recognition tree by using an improved C4.5 algorithm optimized based on the information gain ratio, specifically including:

[0073] Calculate the information gain ratio of each feature to the output Y;

[0074] Select the feature with the largest information gain ratio for splitting to prevent overfitting;

[0075] If the information gain is less than the set threshold, no further splitting is performed (pruning process).

[0076] During actual operation, the system collects sensor data in real time, forms a new feature vector Xnew, and inputs it into the trained decision tree model for judgment. If Ynew = 1, the system determines that the current liquid level is in an emulsified state, and the control module pauses the discharge accordingly and starts the emulsification delay processing logic; if Ynew = 0, the current discharge process is normally executed.

[0077] If the system determines that there is emulsification interference, the current discharge instruction is postponed, and the liquid level discharge threshold is dynamically corrected according to the characteristics of the liquid level change trend and the sudden change of conductivity. The present invention introduces a fuzzy logic control algorithm to dynamically correct the liquid level discharge threshold, specifically:

[0078] The characteristics of the liquid level change trend and the sudden change of conductivity are used as the input items of fuzzy logic, and the correction amount of the liquid level discharge threshold is used as the output item of fuzzy logic, and they are transformed into fuzzy language variables.

[0079] For example, the degree of liquid level fluctuation can be described as levels such as "low" and "medium", and the sudden change of conductivity can be described as "stable", "slight mutation" or "severe mutation". The system calculates its membership degree on different language values through the membership function according to the position of the current value in the set interval, so as to complete the transformation from the numerical value to the fuzzy set.

[0080] The fuzzy control system presets a set of control rules in the form of "if... then..." to describe the relationship between the input variable and the output variable. For example:

[0081] If the liquid level fluctuation is "high" and the sudden change of conductivity is "severe", the discharge threshold should be "significantly increased";

[0082] If the liquid level fluctuation is "medium" and the sudden change of conductivity is "weak", the discharge threshold is "slightly increased".

[0083] In this step, the system matches the fuzzy levels corresponding to the current input with the preconditions in all rules to find all the fuzzy rules that meet the conditions.

[0084] For all the rules that match successfully, the system determines the activation strength of each rule according to the membership degree of the input variable. The activation strength reflects the influence degree of the current input condition on the output conclusion of the rule. If multiple rules are activated simultaneously, each rule will contribute to the final control result to different degrees according to its activation degree.

[0085] The output parts of all the activated rules are synthesized to form a comprehensive fuzzy output set. This set contains multiple output language variables (such as "slightly increased", "moderately increased", "significantly increased"), and each variable has a corresponding fuzzy membership degree.

[0086] The result of this step is a fuzzy set that contains all possible output emission threshold correction amounts and their corresponding degrees of importance, reflecting the current system's tolerance for uncertainties in control actions.

[0087] To enable the control system to make specific adjustments to the emission threshold, it is necessary to convert the fuzzy output set into a definite numerical result. The system uses the centroid method (or "center of mass method") to process the fuzzy set and calculates a representative numerical result, namely the emission threshold correction amount.

[0088] This correction amount directly acts on the liquid level threshold set in the current emission control logic, appropriately adjusting the trigger conditions for oil or water discharge operations in this cycle and avoiding misjudgment of the emission timing in the emulsified state.

[0089] The data processing module also has a model self-learning function, which can reverse-correct the parameters of the prediction model according to the subsequent emission results and the actual separation efficiency, improving the system's long-term adaptability and accuracy. By introducing this data processing module, the present invention can achieve intelligent identification and intervention control of emulsification interference under complex working conditions, thereby significantly improving the stability of oil-water-sludge separation and emission compliance, and overcoming the problem of emission failure caused by misjudgment of emulsification in traditional control systems.

[0090] The control decision-making module receives the following core parameters in real time: the current recognized height of the oil-water interface How; the current recognized height of the water-sludge interface Hws; the cumulative running time Trun of the separator; whether there is emulsification interference in the current separation state (Boolean flag); the liquid level change trend and the conductivity stability identifier; the last emission time Tlast and the minimum interval period Tmin.

[0091] The control decision-making module first determines whether the current conditions for oil-phase emission are met. The basis for this determination is as follows:

[0092] The position of the current oil-water interface is higher than or equal to the preset oil discharge height threshold, that is, the liquid level has reached the upper limit for oil discharge; the current system recognition result indicates that it is not in the state of emulsification interference to ensure a clear interface and sufficient separation; since the last oil discharge operation, the system running time has exceeded the set minimum oil discharge time interval.

[0093] When the above three conditions are simultaneously met, the control system determines that the oil-phase emission conditions have been reached, issues an oil discharge command, and starts the corresponding oil discharge device for discharge.

[0094] Subsequently, the control decision-making module determines whether water-phase discharge needs to be performed. Its judgment logic includes:

[0095] The currently recognized position of the oil-water interface is lower than or equal to the preset water discharge liquid level threshold, indicating that the water level has risen to the set discharge level;

[0096] The current system state is in a non-emulsified state, ensuring that the discharged water layer does not contain any oil.

[0097] The conductivity monitoring result is in a stable state, that is, there is no sharp decline or mutation. Usually, the change amount of the negative CUSUM value is less than the preset tolerance as the standard; the system operation time since the last drainage operation has reached the set minimum drainage interval period.

[0098] When all the above conditions are met, the control decision module issues a drainage instruction to execute the water phase discharge.

[0099] Finally, the control system determines whether to execute the slag phase discharge. This determination is based on any one of the following main conditions and a time condition:

[0100] The currently detected slag layer height is greater than or equal to the preset slag discharge threshold, indicating that the bottom sediment has accumulated to a state where it can be discharged; or the system has been running for more than the preset maximum allowable running period, and even if no slag layer overrun is detected, periodic slag discharge maintenance should be carried out; at the same time, the system confirms that the time since the last slag discharge operation has exceeded the minimum slag discharge interval time to avoid over-frequent operation.

[0101] As long as one of the above conditions and the time condition are met simultaneously, the slag discharge instruction is triggered, and the control system drives the slag discharge device to operate to complete the removal of bottom impurities.

[0102] In the oil-water-slag separator control system of the present invention, an execution module is provided for receiving the discharge instructions issued by the control decision module and respectively executing the oil phase, water phase or slag phase discharge operations according to the instruction content. This module is the terminal execution unit for realizing automatic separation control, and its composition and functions are as follows:

[0103] The execution module includes the following sub-devices:

[0104] Oil discharge device: Usually set at the upper part of the separator for discharging the oil layer floating on the top. The oil discharge device can include actuators such as electric ball valves, solenoid valves, explosion-proof oil pumps, etc., for opening the oil phase discharge port and starting pumping or gravity discharge when receiving the oil discharge instruction.

[0105] Water discharge device: Usually set in the middle part of the separator or at an intermediate position near the bottom for discharging the water phase. The water discharge device generally uses an electric valve in combination with a variable frequency water pump to accurately control the drainage flow rate and time according to the instructions of the control decision module, ensuring that the water layer is fully discharged without entraining oil or slag.

[0106] Sludge discharge device: It is set at the bottom of the separator and is mainly used to discharge solid impurities or granular waste slag deposited at the bottom. The sludge discharge device can adopt an automatic sludge discharge valve, a screw slag discharge mechanism or a pulsed pneumatic device. When receiving the sludge discharge instruction, it starts the sludge discharge program to quickly discharge the sediment and avoid accumulation.

[0107] The execution module is connected to the control decision-making module through wired or wireless communication. The response logic is as follows:

[0108] Receive the discharge instruction: When the control decision-making module determines that the conditions for oil discharge, water discharge or slag discharge are met, it sends a control instruction of the corresponding discharge type to the execution module.

[0109] Start the corresponding device: The execution module drives the corresponding execution device according to the instruction type:

[0110] Receive the oil discharge instruction → Start the oil discharge pump / valve;

[0111] Receive the water discharge instruction → Start the water discharge pump / valve;

[0112] Receive the slag discharge instruction → Open the slag discharge mechanism or valve.

[0113] Execution status feedback: Each execution device is equipped with operation status detection sensors (such as switch position feedback, flow sensors, current monitors, etc.), which feedback the actual execution situation to the control decision-making module for verifying whether the instruction execution is successful and updating the operation record.

[0114] If situations such as the valve not opening, abnormal flow or unsmooth discharge are detected during the execution process, the system can automatically trigger the alarm logic and enter the safety mode to avoid misoperation or equipment damage.

[0115] Through the setting of this execution module, the rapid response and closed-loop feedback between the control system and physical execution can be achieved, greatly improving the automatic operation ability, processing efficiency and safety of the oil-water-sludge separator. At the same time, each sub-device can be independently controlled to adapt to various operation strategies (such as discharging oil first and then water, timed slag discharge, etc.) to meet the intelligent separation control requirements under complex working conditions.

[0116] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by technicians in this field according to the actual situation.

[0117] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0118] As described above, the specific implementation manners of the present application are only described, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. A control system applied to an oil-water-sludge separator, characterized in that: It includes a liquid level sensor, a physical property sensor, a data processing module, a control decision-making module, and an execution module; The liquid level sensor is used to collect the height data of the oil phase, water phase, and slag phase in real time; The physical property sensor is used to collect the conductivity, density, or temperature change information of the separation medium; The data processing module is used to receive the data from the liquid level sensor and the physical property sensor, extract the liquid level change trend feature and the conductivity mutation feature, construct an emulsification interference prediction model, and is used to judge whether there is emulsification interference on the current liquid surface and correct the discharge threshold. Specifically: When the liquid level sensor detects a sharp fluctuation in the height of the oil-water interface within a set time window, or the thickness of the water layer rises significantly within a short time but the physical properties do not change, it is initially determined that there is emulsification interference. At this time, the liquid level data will be jointly analyzed with the density and conductivity data to further determine whether the real discharge conditions are met; when the liquid level sensor detects that the water level height exceeds the discharge threshold but the conductivity change is not significant, it is judged that the current state is an oil-water mixture or emulsification state, thus preventing the drain valve from opening in advance; The method for extracting the liquid level change trend feature is: assuming that the liquid level height output by the liquid level sensor at time t is h(t), a continuous data sequence is extracted with a sliding window of length N. For the data sequence within the window, the number of positive and negative changes of the curve per unit time is calculated through the zero-crossing detection method, and the fluctuation frequency f is calculated. The obtained fluctuation frequency f is used as the liquid level change trend feature; The method for extracting the conductivity mutation characteristics is as follows: Let the conductivity data collected by the sensor in the time series be: ; where: is the conductivity value collected at the t-th moment; Set two recursive paths: Forward deviation detection: ; Negative deviation detection: ; In the formula, is the expected value of the conductivity in the stable state, k is the offset sensitivity constant, h is the determination threshold, and when the cumulative deviation exceeds h, it is regarded as a mutation; and respectively represent the cumulative deviations in the rising and falling directions, and the initial conditions are: = 0; At each sampling moment t, if > h, it is considered that a positive conductivity mutation has occurred; if > h, it is considered that a negative conductivity mutation has occurred; if both are less than h, there is no significant mutation, calculate the ratio of the number of mutation time points to the total number of detected time points, and use it as the conductivity mutation characteristic; The control decision-making module is connected to the data processing module and judges whether the discharge condition is reached according to the identified real oil-water interface position and separation time logic; The execution module includes an oil discharge device, a water discharge device, and a slag discharge device connected to the control decision-making module, and is used to perform the oil, water, or slag discharge operation respectively when receiving the discharge instruction.

2. The control system applied to the oil-water-sludge separator according to claim 1, characterized in that: The liquid level sensor includes a liquid level detection unit arranged at multiple points, which is arranged along the vertical height direction of the oil-water-slag separator and is used to detect the heights of the oil phase, water phase, and slag phase respectively.

3. The control system applied to the oil-water-sludge separator according to claim 1, characterized in that: The physical property sensor includes a conductivity sensor, a density sensor, and a temperature sensor. The conductivity sensor is used to detect the change in the conductivity of the oil-water interface, the density sensor is used to identify the density distribution of different phase media, and the temperature sensor is used to provide the medium temperature information.

4. The control system applied to the oil-water-sludge separator according to claim 1, wherein: The emulsification interference prediction model is constructed by using an improved decision tree algorithm, specifically including: Based on historical operation data and manual annotation results, a training sample set including liquid level change trend features, conductivity mutation features, and emulsification state labels is constructed; Using the information gain ratio as the splitting criterion, a decision tree structure for emulsification state classification is established; Prevent model overfitting through pruning strategies; During the real-time operation of the system, the feature vector of the current cycle is input, and the judgment result of whether there is emulsification interference is output.

5. The control system applied to the oil-water-sludge separator according to claim 4, wherein: During actual operation, sensor data is collected in real time to form a new feature vector Xnew, which is input into the trained decision tree model for judgment. If the model outputs Ynew = 1, it is judged that the current liquid level is in an emulsified state, and the control module accordingly pauses the discharge and starts the emulsification delay processing logic; if the model outputs Ynew = 0, the current discharge process is normally executed.

6. The control system applied to the oil-water-sludge separator according to claim 5, characterized in that: When emulsification interference is identified, the discharge threshold is dynamically corrected: The degree of liquid level fluctuation and the degree of sudden change in conductivity are converted into fuzzy linguistic variables; The input variables are inferred through the fuzzy rule base, and the discharge threshold correction amount is output; The centroid method is used to defuzzify the fuzzy set to generate specific correction values; The correction value is applied to the current drainage or oil discharge liquid level threshold to delay or adjust the discharge trigger condition.

7. The control system applied to the oil-water-sludge separator according to claim 6, characterized in that: The control decision module judges the discharge conditions including: Obtain the current oil-water interface height, water-sludge interface height, separator operation time, emulsification state flag, and conductivity stability index; If the oil-water interface height is greater than or equal to the oil discharge threshold, and it is in a non-emulsified state, and the time since the last oil discharge exceeds the minimum oil discharge cycle, then an oil discharge command is triggered; If the oil-water interface height is less than or equal to the drainage threshold, and it is in a non-emulsified state, and the conductivity change is stable, and the minimum drainage interval condition is met, then a drainage command is triggered; If the slag layer height reaches the set threshold, or the operation time exceeds the slag discharge cycle, and the minimum slag discharge interval is met, then a slag discharge command is triggered.

Images