A green and environment-friendly edible oil pollution reduction and consumption reduction production method

Abstract

The application relates to the edible oil processing technical field and discloses a green and environment-friendly edible oil pollution reduction and consumption reduction production method, which collects material conductivity and circulating pump active power and calculates a difference value, closes a hydration reagent valve when a multivariate joint threshold condition is met, records a critical response time and calculates a conductivity time integral value, constructs a hydration process response characteristic parameter, calculates real-time total heat load, superimposes basic steam target flow and a sine excitation signal to control steam injection, obtains a stripping steam apparent latent heat coefficient through coherent multiplication and low-pass filtering, obtains oil grease volatile net heat load after deducting steam heat load and heat loss, adjusts evaporation latent heat constant by using the characteristic parameter to construct a correction coefficient, calculates real mass evaporation rate in combination with the net heat load and updates the basic steam target flow. The application realizes accurate truncation of the hydration reagent and dynamic adjustment of the stripping steam quantity.

Description

Technical Field

[0001] This invention relates to the field of edible oil processing technology, specifically to a green and environmentally friendly method for reducing pollution and energy consumption in edible oil production. Background Technology

[0002] The hydration and deodorization processes in edible oil processing directly impact oil quality and production energy consumption. In traditional hydration processes, reagents are typically added in a timed or quantitative manner, lacking real-time monitoring of changes in the physicochemical state of the reactants, making it difficult to accurately determine the endpoint of the hydration reaction. This fixed addition pattern easily leads to excessive injection of hydration reagents, thereby increasing the wastewater treatment load of the subsequent deodorization process.

[0003] In the deodorization process, stripping steam is mainly used to remove volatile impurities from the oil. Existing deodorization control systems mostly use fixed steam injection settings, which cannot be dynamically adjusted according to the actual volatilization requirements of impurities. The heat exchange process inside the deodorization tower involves a lot of background heat flow and static heat conduction interference, and conventional temperature or flow monitoring methods are difficult to separate the net heat load generated entirely by the condensation phase change of oil volatiles from the total heat load.

[0004] Furthermore, existing systems typically use a fixed latent heat of vaporization constant for thermodynamic calculations, without correlating and correcting for the actual degree of hydration reaction with thermodynamic parameters. Due to the lack of accurate net heat load data and dynamically corrected latent heat parameters, the system cannot calculate the real-time evaporation rate of volatile oils in the material, resulting in steam supply often exceeding the actual deodorization requirements and increasing energy consumption in the production process. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a green and environmentally friendly method for reducing pollution and energy consumption in edible oil production. It solves the problems of increased wastewater load caused by the fixed addition of hydration reagents in traditional edible oil processing, and increased energy consumption caused by the fixed setting of stripping steam volume in the deodorization process, which cannot be dynamically adjusted according to the actual volatilization of oil.

[0006] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a green and environmentally friendly method for producing edible oil with reduced pollution and energy consumption, comprising the following steps: Collect the conductivity of the material and the active power of the circulating pump, calculate the second-order difference of conductivity and the first-order difference of active power, and output the closing command of the pneumatic regulating valve of the hydration reagent when the multivariate joint threshold condition is met. Record the critical response time from the injection start time to the command cutoff time, calculate the conductivity time integral value within the critical response time, construct the hydration process response characteristic parameters, and store them in the deodorization process data storage area according to batch identification; Collect the condenser cooling water mass flow rate and inlet / outlet water temperature difference, calculate the real-time total heat load, superimpose the basic steam target flow rate and sinusoidal excitation signal, control the injection of stripping steam, collect the actual steam mass flow rate, coherently multiply the total heat load and sinusoidal excitation signal and low-pass filter to obtain the apparent latent heat coefficient of stripping steam, multiply the actual steam mass flow rate and the apparent latent heat coefficient of stripping steam to obtain the stripping steam heat load, subtract the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load to obtain the net heat load of volatile oils; The response characteristic parameters of the hydration process are retrieved in batches, and the phase equilibrium correction coefficient of the non-ideal solution is constructed. The theoretical latent heat of vaporization constant is corrected to the apparent latent heat of vaporization constant. The net heat load of oil volatiles is divided by the apparent latent heat of vaporization constant to obtain the true mass evaporation rate of oil volatiles. The basic steam target flow rate is updated according to the partial pressure ratio constant of the gas-liquid phase equilibrium.

[0007] In the above scheme, the differential changes in conductivity and power reflect the changes in the electrical properties and viscosity of the hydration reaction system, which are used to control reagent injection; the time and conductivity integral of the reaction process are used to quantify the state of the reaction system. Introducing a sinusoidal excitation signal in conjunction with filtering operations allows for the separation of the net heat load generated by volatile condensation from the total heat load. After correcting the latent heat of vaporization constant based on the characteristic parameters of the hydration process, the volatile mass flow rate can be calculated, thereby adjusting the stripping steam quantity to match the reagent and steam supply with the actual reaction requirements.

[0008] Furthermore, the calculation of the second-order difference value of conductivity and the first-order difference value of active power, when satisfying the multivariate joint threshold condition, outputs a closing command for the hydration reagent pneumatic regulating valve, specifically including: Set a positive abrupt change threshold for the second-order difference of conductivity; Set a threshold for negative abrupt changes in the first-order difference value of active power; When the second-order difference value of conductivity is greater than the positive mutation threshold and the first-order difference value of active power is less than the negative mutation threshold, a command to close the pneumatic regulating valve for hydration reagent is output.

[0009] Furthermore, the process of collecting the material conductivity and the active power of the circulating pump, and calculating the second-order difference value of the conductivity and the first-order difference value of the active power, specifically includes: The material conductivity data and circulating pump active power data are time-aligned according to a unified clock. A synchronous sampling sequence is formed according to the same timestamp; A digital low-pass filter is used to filter the synchronous sampling sequence, and the second-order difference value of conductivity and the first-order difference value of active power are calculated based on the filtered synchronous sampling sequence.

[0010] Furthermore, the critical response time from the start of the injection to the instruction truncation time specifically includes: The moment when the pneumatic regulating valve for extracting the hydration reagent receives the opening command and generates an opening action feedback is taken as the injection start moment. The moment when the closing command is issued to the pneumatic regulating valve for the hydration reagent is taken as the command cutoff moment. The critical response time is obtained by calculating the time difference between the instruction truncation time and the injection start time.

[0011] Furthermore, the calculation of the conductivity time integral value within the critical response time specifically includes: The filtered conductivity values ​​from the injection start time to the instruction cutoff time are read according to a unified digital sampling period. The time integral value of conductivity is obtained by multiplying each filtered conductivity value by a uniform digital sampling period and then summing the results. Linear interpolation is used to fill in the missing data caused by communication failures.

[0012] Furthermore, the process of collecting the condenser cooling water mass flow rate and inlet / outlet water temperature difference to calculate the real-time total heat load specifically includes: Read the real-time cooling water mass flow rate of the surface condenser cooling water pipeline; Read the cooling water outlet temperature and cooling water inlet temperature, calculate the difference between the cooling water outlet temperature and the cooling water inlet temperature, and obtain the inlet and outlet water temperature difference; The real-time total heat load is obtained by multiplying the constant specific heat capacity of cooling water at constant pressure, the real-time mass flow rate of cooling water, and the temperature difference between the inlet and outlet water.

[0013] Furthermore, the net heat load of oil volatiles is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load, specifically including: By superimposing the target basic steam flow rate and the sinusoidal excitation signal, a stripping steam injection command is generated. Collect the actual steam mass flow rate fed back from the stripping steam mass flow meter; Multiply the actual steam mass flow rate by the apparent latent heat coefficient of the stripping steam to obtain the stripping steam heat load; The net heat load of volatile oils is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load.

[0014] Furthermore, the net heat load of oil volatiles is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load, specifically including: Configure the zero-point clamping protection logic for the net heat load of volatile oils; When the net heat load of volatile oils is less than zero, the net heat load of volatile oils should be set to zero. Synchronous output device external insulation integrity check prompt.

[0015] Furthermore, the step of batch-based retrieval of hydration process response characteristic parameters to construct a non-ideal solution phase equilibrium correction coefficient specifically includes: Divide the time integral of conductivity by the critical response time to obtain the characteristic value of average conductivity; Multiply the characteristic value of the average conductivity by the first system constant and add it to the second system constant to obtain the phase equilibrium correction coefficient for the non-ideal solution; When the critical response time is zero, the phase equilibrium correction factor for the non-ideal solution is set to the default value of 1.0.

[0016] Furthermore, the net heat load of the oil volatiles is divided by the apparent latent heat of vaporization constant to obtain the true mass evaporation rate of the oil volatiles. The basic steam target flow rate is updated based on the gas-liquid phase equilibrium partial pressure ratio constant, specifically including: Multiply the theoretical latent heat of vaporization constant by the phase equilibrium correction factor for the non-ideal solution to obtain the apparent latent heat of vaporization constant; Divide the net heat load of oil volatiles by the apparent latent heat of vaporization constant to obtain the true mass evaporation rate of oil volatiles; The updated basic steam target flow rate is obtained by multiplying the gas-liquid phase equilibrium partial pressure ratio constant by the actual mass evaporation rate of oil volatiles. The updated baseline steam target flow rate is used as the reference input for the sinusoidal excitation signal in the next control cycle.

[0017] This invention provides a green and environmentally friendly method for producing edible oil with reduced pollution and energy consumption. It has the following beneficial effects: 1. This invention controls the hydration reagent valve by performing second-order difference calculations on the fluid conductivity and first-order difference calculations on the active power of the circulating pump, and then combining these two calculations for multi-variable joint determination. This utilizes the physical characteristics of the changes in the electrical properties and fluid viscosity of the reaction system to determine the endpoint of the hydration reaction and promptly cut off reagent injection, avoiding reagent overdose caused by timed or quantitative addition methods. This reduces reagent consumption and the wastewater treatment load of subsequent deodorization processes.

[0018] 2. This invention superimposes a sinusoidal excitation signal of a preset frequency onto the base steam target flow rate and collects the mass flow rate of the cooling water pipeline and the inlet and outlet water temperature difference to calculate the total heat load. The total heat load is processed through coherent multiplication and low-pass filtering to separate static heat transfer interference from the system. Furthermore, the net heat load generated by the condensation of volatiles within the grease is separated from the background heat flow, providing basic data for subsequent closed-loop regulation of the steam flow rate.

[0019] 3. This invention extracts the critical response time and conductivity time integral value during the hydration reagent injection process to construct a correction coefficient for the phase equilibrium of a non-ideal solution. This coefficient is used to dynamically adjust the theoretical latent heat of vaporization constant. Based on the corrected apparent latent heat of vaporization constant and the calculated net heat load, the real-time evaporation rate of volatile oils can be calculated, and the basic steam target flow rate can be updated according to the gas-liquid phase equilibrium relationship. This achieves dynamic adjustment of the stripping steam quantity according to the actual volatilization demand of the material, reducing ineffective steam consumption. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the system architecture of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention; Figure 3 This is a schematic diagram of the degumming joint determination and cutoff control process of the present invention; Figure 4 This is a schematic diagram of the parameter extraction feedforward data flow of the present invention; Figure 5 This is a schematic diagram of the sinusoidal excitation and orthogonal correlation demodulation process of the present invention; Figure 6 This is a schematic diagram of the closed-loop follower control logic flow of the present invention; Figure 7 This is a three-dimensional surface diagram of the integrated operation status joint monitoring of the present invention; Figure 8 This is a multi-signal diagram of heat load stripping and steam control in the deodorization process of the present invention. Detailed Implementation

[0021] The technical solutions in 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.

[0022] See attached document Figure 1 This invention provides a green and environmentally friendly edible oil production system that reduces pollution and energy consumption. The system includes: a degumming joint judgment and cutoff module, a parameter extraction feedforward module, a sinusoidal excitation heat load separation module, and a closed-loop follower control module.

[0023] The above system operates based on a distributed control system hardware platform. The field monitoring instruments include an electrodeless online conductivity meter and a circulating pump frequency converter arranged on the main fluid circulation pipeline of the hydration tank, a reagent flow meter arranged on the hydration reagent injection pipeline, and a cooling water mass flow meter arranged on the cooling water pipeline of the surface condenser of the deodorization tower, as well as a first temperature transmitter and a second temperature transmitter located at the cooling water inlet and cooling water outlet, respectively.

[0024] The on-site actuators include a pneumatic regulating valve for controlling the injection of hydration reagent and a pneumatic regulating valve for controlling the injection of stripping steam.

[0025] A steam mass flow meter is also installed on the stripping steam injection pipeline of the deodorization tower, and the stripping steam pneumatic regulating valve is equipped with a valve position feedback device to obtain the steam valve control command, actual valve position, and actual steam mass flow rate, respectively. All hardware terminal devices are connected to the central processing unit of the distributed control system via an industrial fieldbus.

[0026] The degumming joint determination and cutoff module is used to simultaneously acquire the conductivity of the hydrated mixture and the active power of the circulating pump. The degumming joint determination and cutoff module calculates the second-order difference value of conductivity and the first-order difference value of active power through discrete difference calculation.

[0027] When the multivariate joint judgment rule meets the set threshold condition, the degumming joint judgment cutoff module outputs a hydration reagent pneumatic regulating valve closing command to the hydration reagent pneumatic regulating valve that controls the injection of hydration reagent, so as to stop the addition of hydration reagent; the stripping steam pneumatic regulating valve is independently controlled by the steam flow control loop of the downstream deodorization process according to the basic steam target flow, and does not receive the hydration reagent pneumatic regulating valve closing command.

[0028] The parameter extraction feedforward module is communicatively connected to the degumming joint determination cutoff module. The parameter extraction feedforward module is used to record the critical response time of the hydration endpoint cutoff event and calculate the conductivity time integral within that response time window.

[0029] The parameter extraction feedforward module constructs the response characteristic parameters of the hydration process from the time integral and the critical response time, and writes these parameters into the data storage area of ​​the closed-loop follower control module through distributed bus addressing.

[0030] The sinusoidal excitation heat load separation module is used to calculate the real-time total heat load of the surface condenser by collecting data from the cooling water mass flow meter and the cooling water inlet and outlet temperature transmitters.

[0031] The sinusoidal excitation heat load separation module superimposes a sinusoidal excitation flow signal with a preset frequency and amplitude onto the basic steam target flow rate to control the steam injection at the bottom of the deodorization tower.

[0032] The sinusoidal excitation heat load separation module synchronously acquires the actual steam mass flow rate fed back by the steam mass flow meter, and uses an orthogonal correlation demodulation algorithm to extract the amplitude and phase response of the total heat load and the actual steam mass flow rate at the excitation frequency. After compensating for the transmission delay and dynamic attenuation caused by the steam valve, deodorization tower, condenser and temperature measurement links, the equivalent thermal response corresponding to the stripping steam is identified.

[0033] Based on the identified equivalent thermal response, the sinusoidal excitation heat load separation module subtracts the stripping steam heat load and the condensation system baseline background heat load from the total heat load to obtain an estimated value of the net volatile heat load.

[0034] The closed-loop follow-up control module is used to call the hydration process response characteristic parameters transmitted from the upstream degumming process to construct a correction coefficient for the non-ideal solution phase equilibrium. Based on this correction coefficient, the closed-loop follow-up control module updates the apparent latent heat of vaporization constant, estimates the actual mass evaporation rate of volatile oils and fats by combining the net heat load of volatiles, and updates the basic steam target flow rate in real time according to the preset gas-liquid phase equilibrium partial pressure ratio constant.

[0035] The non-ideal solution phase equilibrium correction coefficient, apparent latent heat of vaporization constant, and gas-liquid phase equilibrium partial pressure ratio constant are all stored corresponding to the type of oil, deodorization temperature, and absolute operating pressure. When the corresponding calibration range is exceeded, the adaptive update stops and the preset safety parameters are called.

[0036] In the overall system workflow, the degumming joint judgment and cutoff module continuously monitors the conductivity and circulating pump operating load parameters of the upstream degumming hydration process, and determines the critical state of hydration phase separation to cut off reagent addition.

[0037] Subsequently, the parameter extraction feedforward module performs feature extraction and cross-process feedforward on the dynamic data of the reagent dosing cycle. When the material flows to the deodorization stage, the sinusoidal excitation heat load separation module introduces active excitation at the stripping medium end and analyzes the heat load response.

[0038] The closed-loop follower control module integrates multi-source data to calculate the estimated value of the true mass evaporation rate of volatile oils and oils, and updates the target flow rate of stripping steam accordingly. The steam flow closed-loop control loop adjusts the opening of the pneumatic regulating valve for hydration reagent and the pneumatic regulating valve for deodorization stripping steam, forming an adaptive continuous regulation channel with upper and lower limits of flow, rate of change limitation and abnormal backoff function.

[0039] See attached document Figure 2 This invention provides a green and environmentally friendly method for producing edible oil with reduced pollution and energy consumption, the method comprising: S1. Degumming Joint Judgment Cutoff: Simultaneously collect the conductivity and active power of the circulating pump in the main fluid circulation pipeline of the hydration tank. Perform discrete difference calculation on the collected conductivity and active power to obtain the second-order difference value of conductivity and the first-order difference value of active power. Substitute the aforementioned difference calculation results into the multivariate joint judgment rule for condition comparison. When the preset threshold condition is met, output a shut-off command to the pneumatic regulating valve of the hydration reagent to stop the addition of hydration reagent.

[0040] S2. Parameter extraction feedforward records the critical response time from the start of reagent injection to the moment the hydration reagent valve closes. The conductivity time integral value within this time window is calculated through discrete accumulation. The obtained critical response time and conductivity time integral value are combined to construct the hydration process response characteristic parameter. The system uses a distributed bus addressing mechanism to transmit this characteristic parameter and write it into the data storage area of ​​the downstream deodorization process.

[0041] S3. Harmonic excitation stripping: Mass flow rate and inlet / outlet water temperature difference data of the cooling water pipeline of the surface condenser in the deodorization process are collected to calculate the real-time total heat load of the surface condenser. A sinusoidal excitation flow signal with a preset frequency and amplitude is superimposed on the basic steam target flow rate to control the steam injection operation at the bottom of the deodorization tower. A related demodulation algorithm is used to perform coherent multiplication and low-pass filtering operations on the total heat load and the local reference excitation signal, thereby stripping the heat of the steam itself from the total heat load and calculating the net heat load of volatiles in the deodorization tower.

[0042] S4. Closed-loop follow-up control retrieves the hydration process response characteristic parameters from the upstream degumming process and substitutes them into the algebraic equation to construct a correction coefficient for the non-ideal solution phase equilibrium. This correction coefficient is used to adjust the apparent latent heat of vaporization constant. Combining the calculated net heat load of volatiles with the adjusted latent heat constant, the true mass evaporation rate of volatile oils in the deodorization tower is calculated in reverse. Based on the gas-liquid phase equilibrium partial pressure ratio relationship set in the system, the base steam target flow rate is updated in real time using the calculated true mass evaporation rate of volatile oils.

[0043] See attached document Figure 3 During the degumming and hydration stage, the degumming joint determination and cutoff module determines the phase transition state of the hydration mixture based on the extracted on-site equipment operating parameters, and then performs a reagent cutoff operation. This execution process specifically includes the following sub-steps: S101, the degumming joint judgment and cutoff module obtains the mixture data at the fluid circulation main pipeline of the hydration tank through the field communication bus.

[0044] During this process, the system utilizes an electrodeless online conductivity meter installed inside the pipeline to collect fluid conductivity data characterizing conductivity parameters. This electrodeless online conductivity meter measures the conductivity of a closed loop using the principle of electromagnetic induction, which can reduce polarization measurement errors caused by the adhesion of high-viscosity insulating media.

[0045] The measurement range of the electrodeless online conductivity meter is determined based on the on-site pre-test conductivity range from the normal emulsification stage to the free aqueous phase formation stage. Its resolution and response time should be able to identify the conductivity changes before and after the critical phase transition.

[0046] The conductivity meter is equipped with temperature measurement or temperature compensation functions and is installed according to the minimum pipe diameter, immersion depth, and conductive circuit conditions required by the sensor. When field pre-tests show that the conductivity before the critical phase transition is lower than the effective range of the electrodeless conductivity meter, an electrode-type conductivity meter or impedance measuring device with corresponding low conductivity measurement capabilities should be used instead.

[0047] Simultaneously, the degumming joint judgment and cutoff module obtains the real-time active power driving the circulating pump by reading the internal register of the circulating pump frequency converter. This active power data is used to characterize the changing trend of the apparent viscosity of the hydrated mixture when the circulating pump speed, the opening degree of the circulating pipeline valve, the liquid level in the tank, and the circulating flow rate are kept within the set fluctuation range, and is input into the control system as a characteristic parameter of the circulating pump load.

[0048] During the injection of hydration reagent and the multivariate joint judgment rule, the system will maintain the output frequency of the circulating pump inverter at the preset value and prohibit the adjustment of the circulating pipeline valve position for reasons other than safety. When the circulating pump speed or circulating flow exceeds the set fluctuation range, the active power multivariate joint judgment rule will be suspended and an abnormal operating condition flag will be output.

[0049] For operating modes that cannot maintain a constant speed, the system normalizes the active power based on the speed, circulation flow rate, and pressure difference, and then extracts its trend.

[0050] For analog-to-digital conversion and low-pass filtering anti-interference processing of sensor data in industrial fields, those skilled in the art can configure conventional RC filter circuits or digital smoothing filter algorithms according to the electromagnetic environment of the site. These are well-known technologies in the field and will not be elaborated here.

[0051] In the initial stage of degumming, the added hydration reagent mainly causes the phospholipids in the oil phase to absorb water and swell, and the system is in a water-in-oil emulsion state. During this stage, the change in conductivity is relatively gradual. The hydration reagent refers to the hydration medium used to promote the phospholipids in the oil phase to absorb water and swell, forming a hydration medium with a detectable conductivity response. In this embodiment, water for hydration is used. When using hydration media with different compositions or concentrations, the positive abrupt change threshold of the second-order difference value of conductivity and the negative abrupt change threshold of the first-order difference value of active power should be recalibrated according to the conductivity baseline of the hydration medium, the temperature compensation results, and the field pre-test data.

[0052] As the amount of reagent injected continues to increase, when the critical saturation state of hydration is reached, a free water phase begins to precipitate within the system, forming a conductive path and causing a rapid increase in conductivity, with an increase in its second-order conductivity difference value. Simultaneously, the disruption of the emulsion structure alters the rheological properties of the fluid, resulting in a decrease in the apparent viscosity, manifested as a negative change in the load on the circulating pump.

[0053] S102, the degumming joint judgment and cutoff module performs discrete difference calculation on the data sequence to be processed according to the set discrete sampling period, so as to calculate the first-order difference value of active power and the second-order difference value of conductivity.

[0054] Before performing differential calculations, the degumming joint determination and cutoff module aligns the conductivity data, active power data, circulating pump speed data, and reagent flow rate data with a unified clock, and uses the same data timestamp to form a synchronous sampling sequence.

[0055] The data sequence is first passed through a digital low-pass filter or a sliding fit filter with a defined cutoff frequency to suppress the amplification of high-frequency measurement noise by differential calculation.

[0056] The digital sampling period can be set according to the controller's computing power and the sensor's response characteristics. In practical applications, it is usually selected within the range of 0.1 seconds to 2 seconds, and it should be ensured that the expected critical change process can be characterized by at least five consecutive sampling points. The group delay caused by filtering should be compensated in the subsequent critical moment calculation.

[0057] At any discrete time The conductivity of the filtered hydrated mixture obtained by the degumming joint determination and cutoff module is The active power of the circulating pump is The debonding joint judgment cutoff module calculates the second-order difference value of conductivity according to the following formula. : ; In the formula, Represents any discrete time. The calculated second-order difference value of conductivity; Indicates the current discrete time. The obtained conductivity values ​​of the filtered hydrated mixture; Indicates the digital sampling period of the system; The conductivity value was measured at the previous sampling time. These are the conductivity values ​​measured at the first two sampling times. This second derivative characterizes the rate of change of conductivity and is used to capture the numerical jump characteristics during the formation of the free water phase.

[0058] The degumming joint judgment and cutoff module synchronously calculates the first-order difference value of the active power of the circulating pump according to the following formula. : ; In the formula, This represents the first-order difference value of the calculated active power of the circulating pump; Indicates the current discrete time. The obtained active power value of the circulating pump; Indicates the digital sampling period of the system; This represents the active power value measured at the previous sampling time. This first derivative characterizes the instantaneous change trend of the apparent viscosity of the hydrated mixture.

[0059] S103, the degumming joint determination and truncation module substitutes the above differential calculation results into the multivariate joint determination rule to perform threshold comparison and trigger the subsequent truncation mechanism.

[0060] During the continuous injection of hydration reagent into the hydration tank, the hydration mixture gradually evolves from a homogeneous fluid into a water-in-oil microemulsion, and the overall apparent viscosity of the fluid increases accordingly. When the hydration reaction reaches the critical saturation point and a free water phase is formed, the free water phase generated inside the fluid causes the system conductivity to rise rapidly, manifested as a positive abrupt change in the second-order difference value of conductivity; at the same time, the structural disintegration of the emulsion state leads to a decrease in fluid shear resistance and apparent viscosity, manifested as a decrease in the output load of the circulating pump motor, i.e., a negative abrupt change in the first-order difference value of active power.

[0061] Set the threshold for the positive abrupt change in the second-order difference of conductivity as follows: ,and The threshold for negative abrupt change in the first-order difference of active power is set as follows: ,and The above mutation threshold and The determination method can be achieved by collecting the fluctuation amplitude of relevant variables during the normal emulsification stage and setting it in advance in combination with an empirical safety factor, or by obtaining it through offline statistical calibration based on the batch historical operation data of the process, so as to reduce the probability of false triggering caused by background fluctuations.

[0062] During the offline statistical calibration process, the distribution of changes in the normal stage was established according to the type of oil, processing temperature, reagent concentration, and operating conditions of the circulating pump. The statistical upper limit of the changes in the normal stage was compared with the changes in the phase change batch confirmed by manual sampling, and a threshold that could distinguish between the two states was selected.

[0063] Thresholds, duration of judgment, and applicable operating conditions are stored as the same parameter group. When the operating conditions exceed the applicable range of the corresponding parameter group, the normal multivariate joint judgment rule is not executed.

[0064] The debonding joint determination and cutoff module is internally configured with an interlock trigger function. : ; In the formula, This indicates the interlock trigger function configured inside the debonding joint determination and cutoff module; This represents the first-order difference value of the calculated active power of the circulating pump; This indicates the set positive mutation threshold; This represents the set positive mutation threshold. The second-order difference value of conductivity is given if and only if... Greater than the set positive mutation threshold And the first-order difference of active power At the same time, it is less than the set negative mutation threshold. At that time, the interlock trigger function The output state is set to 1; otherwise, the output is set to 0.

[0065] The cross-multivariate joint judgment rule eliminates misjudgments caused by univariate fluctuations due to local material inhomogeneity or local signal fluctuations in the fluid within the hydration tank. (When the interlock trigger function...) When the output state remains at 1 for a period exceeding the preset judgment time window, the degumming joint judgment cutoff module determines that the critical state of hydration phase separation has been reached. It then outputs a closing command to the positioner of the hydration reagent pneumatic regulating valve, driving the valve to close and stopping the addition of hydration reagent. The stripping steam pneumatic regulating valve does not receive the hydration reagent closing command.

[0066] Specifically, the critical detection moment is defined as the moment when the second-order change in conductivity and the first-order change in active power simultaneously meet the threshold conditions for the first time. The interlock confirmation moment is defined as the moment when the threshold conditions are continuously met until the end of the judgment time window. The actual cutoff moment is defined as the moment when the closing feedback of the pneumatic regulating valve for hydration reagent and the pneumatic regulating valve for deodorization stripping steam reaches the preset closing position and the reagent flow rate decreases below the preset cutoff flow rate. Subsequent critical response times are calculated with the interlock confirmation moment as the endpoint, and the actual reagent dosage is accumulated up to the actual cutoff moment.

[0067] To ensure the integrity of the control logic and avoid risks caused by the failure of normal judgment logic to be triggered, this embodiment is supplemented with a timeout interlock protection mechanism.

[0068] The degumming and debinding joint determination and cutoff module has a preset maximum reagent injection time limit. Timing starts from the start of reagent injection; if the injection time reaches this maximum limit, and the interlocking trigger function... If the condition for setting the output to 1 is still not met, the degumming joint judgment cutoff module will forcibly output a hydration reagent valve closing command to both the hydration reagent pneumatic regulating valve and the deodorization stripping steam pneumatic regulating valve, and simultaneously send a status alarm signal to the central control room. This reduces the risk of continuous water addition caused by sensor malfunctions or deviations in data from special operating conditions.

[0069] For batches triggered by timeout protection, the degumming joint judgment cutoff module simultaneously generates an abnormal batch mark, records the timeout reason, sensor status, cumulative reagent dosage, actual cutoff time and corresponding timestamp, and transmits the abnormal batch mark along with the hydration process response characteristic parameters to the closed-loop follow control module.

[0070] See attached document Figure 4 To achieve parametric coupling between upstream and downstream processes, after triggering the joint degumming determination truncation, the parameter extraction feedforward module extracts features from the process data of the degumming process and performs cross-process data feedforward operations. This execution process specifically includes the following sub-steps: S201, the parameter extraction feedforward module acquires and records the key time nodes of the reagent injection cycle in order to calculate the critical response time for the formation of the free aqueous phase.

[0071] The parameter extraction feedforward module uses the system's internal clock synchronization mechanism to extract the moment when the hydration reagent pneumatic regulating valve receives the opening command and generates an opening action feedback as the start time. .

[0072] Simultaneously, the moment when the interlock trigger function meets the set conditions and issues a closing command to the hydration reagent pneumatic regulating valve is defined as the command cutoff moment. The command cutoff moment is used to calculate the critical response time; the moment when the closing feedback of the hydration reagent pneumatic regulating valve reaches the preset closing position and the reagent flow rate decreases below the preset cutoff flow rate is defined as the actual cutoff moment. The actual hydration reagent dosage is accumulated to the actual cutoff moment. The parameter extraction feedforward module calculates the critical response time of this batch of materials according to the following formula. : ; In the formula, Indicates the critical response time of the current batch of materials; This indicates the actual cutoff time, i.e., the moment when the interlock trigger function meets the set conditions and sends a hydration reagent valve closing command to the regulating valve; This indicates the start time, which is the moment when the pneumatic regulating valve for hydration to hydration reagent and the pneumatic regulating valve for deodorization to stripping steam receive the opening command and generate action feedback.

[0073] To ensure the integrity of the control logic, computational stability, and data validity, if the actual truncation time... This is triggered by the aforementioned timeout interlock protection mechanism. The parameter extraction feedforward module marks the confidence level of the current batch of data as abnormal and assigns it a preset baseline response time value calibrated offline to prevent abnormally long-duration data from being substituted into subsequent algorithms and causing subsequent calculation anomalies. The preset baseline response time value can be determined by extracting the average value of the critical response time in historical normal production batches.

[0074] S202, the parameter extraction feedforward module extracts the discrete data sequence within the aforementioned time window by discretely accumulating the time integral value of conductivity. For discretely sampled industrial field data, analytical integration of continuous functions is typically difficult to employ. The parameter extraction feedforward module uses discrete summation over the initial time... up to the actual cutoff time Real-time conductivity over a time period The integral calculation is performed, and the specific calculation relationship is as follows: ; In the formula, This is the calculated time integral value of the conductivity; This is an incrementing index variable for the sampling period, and its value is an integer. The digital sampling period of the system; The total number of sampling points is calculated as follows: When the calculation result contains a decimal, it is rounded down. In the first The filtered conductivity values ​​read at each sampling point.

[0075] In actual calculations, if individual sampling points are affected due to sensor communication failures... If data is missing, the parameter extraction feedforward module can use linear interpolation to fill in the missing data to ensure the continuity of the integration results.

[0076] Conductivity time integral value This comprehensively reflects the evolution of the system's conductive medium accumulation throughout the entire dynamic cycle, from the injection of the hydration reagent to the formation of the free aqueous phase. From the perspective of colloid chemistry principles, this integral value is numerically correlated with the fluid's ability to encapsulate the hydration reagent.

[0077] A higher integral value indicates a greater tolerance for the colloidal network within the fluid to maintain an emulsified state. Therefore, in the control algorithm architecture, this conductivity time integral value is crucial. It is explicitly defined as the cumulative characteristic value of hydration conductivity of the current batch of oils, used to characterize the changing trend of emulsification stability at the interface of non-hydrated phospholipids and colloidal particles.

[0078] S203, the parameter extraction feedforward module combines the calculated data results to construct the hydration process response characteristic parameters and performs cross-module feedforward transmission.

[0079] The parameter extraction feedforward module will extract the critical response time. The time integral value of conductivity Pack them into arrays to construct a set of characteristic parameters that characterize the reaction properties of this batch. .

[0080] Considering the time differences in process flow and the buffering capacity of storage tanks in continuous industrial production, this set of characteristic parameters is uniquely bound to the crude oil tracking identification code of the current batch in the control program or the data management layer of the distributed control system. The crude oil tracking identification code can be automatically generated by the manufacturing execution system based on the production work order number and batch unloading time.

[0081] The parameter extraction feedforward module writes the feature parameter group bound to the batch identification code into the data storage area of ​​the closed-loop follower control module through the distributed communication bus at the bottom of the system.

[0082] In subsequent production stages, when the material corresponding to this batch enters the deodorization tower after passing through the dehydration and decolorization intermediate processes, the deodorization-related control unit can retrieve the characteristic parameter group in the corresponding register based on the batch identification code. When different batches of material exist simultaneously in the storage tank or when batches are mixed, the control unit determines the current batch identifier entering the deodorization tower based on the material tracking information recorded by the manufacturing execution system; if a unique corresponding batch cannot be determined, the adaptive retrieval of the hydration process response characteristic parameters is suspended, and preset safety parameters are used for control.

[0083] For data transmission and verification between controllers according to the adopted industrial communication protocol, those skilled in the art can configure it according to the specific industrial communication protocol, which is a well-known technology in the field and will not be described in detail here.

[0084] See attached document Figure 5 In the deodorization process, to overcome the challenge of directly identifying and extracting changes in volatile matter heat load due to the large steam heat load and the relatively small volatile matter heat load, a sinusoidal excitation heat load separation module performs heat load separation calculations for the total heat load. This execution process specifically includes the following sub-steps: S301, the sinusoidal excitation heat load separation module collects the operating status parameters of the surface condenser through field instruments to calculate the real-time total heat load of the system.

[0085] The sinusoidal excitation heat load separation module reads the real-time cooling water mass flow rate fed back by the cooling water mass flow meter on the surface condenser cooling water pipeline. Synchronously read the cooling water inlet and outlet temperature difference fed back by the temperature transmitter. Cooling water inlet and outlet temperature difference The inlet and outlet temperatures of the cooling water are the inlet and outlet temperatures, respectively, and the outlet temperature is the inlet temperature of the cooling water minus the outlet temperature.

[0086] Based on the principle of thermodynamic energy conservation, and neglecting the slight heat dissipation from the tube wall, the heat removed by the condenser through the cooling water is equivalent to the total heat released by the condensation of the mixed escaping gas at the top of the deodorization tower. The sinusoidal excitation heat load separation module calculates the real-time total heat load according to the following thermodynamic balance formula. : ; In the formula, This represents the calculated real-time total heat load of the surface condenser; This indicates the real-time mass flow rate of cooling water on the surface condenser cooling water pipeline, as read at the current moment. This indicates the temperature difference between the inlet and outlet cooling water of the surface condenser read at the current moment; This is the constant specific heat capacity of cooling water at constant pressure.

[0087] For the range calibration and pipeline installation specifications of field temperature and flow transmitters, those skilled in the art can select and configure them according to the relevant instrumentation and control engineering design manuals. These are well-known technologies in the field and will not be elaborated here.

[0088] S302, the sinusoidal excitation heat load separation module constructs a sinusoidal excitation flow signal and superimposes it onto the basic steam target flow.

[0089] In industrial measurements, if the steady-state total heat is directly subtracted from the theoretical steam heat, the net heat generated by small amounts of volatiles can be masked by the large proportion of steam heat load estimation error due to the measurement deviation of the instrument range.

[0090] This technical solution introduces a sinusoidal excitation method in system identification, causing the injection volume of the stripping medium to fluctuate periodically according to a known pattern. The sinusoidal excitation heat load separation module generates the actual injection commands issued to the pneumatic regulating valves for hydration reagent conversion and deodorization stripping steam conversion, based on the following formula. : ; In the formula, This indicates the actual injection command issued to the pneumatic regulating valve for the hydration reagent and the pneumatic regulating valve for the stripping steam; Represents the sine trigonometric function; The basic steam target flow rate calculated for the current cycle of the system; The set sinusoidal harmonic excitation amplitude; The set low-frequency angular velocity; This refers to the system uptime.

[0091] To ensure the fluid dynamic stability of the vacuum system inside the deodorization tower and avoid abnormal surge of the steam jet vacuum pump, the excitation amplitude is... The value can be set between 3% and 5% of the basic command range, for low-frequency angular velocity. The corresponding oscillation period can be set to a value greater than the system's thermal inertia time delay. In practical engineering applications, this value can be taken in the range of 30 to 120 seconds. Angular velocity... With oscillation period The conversion relationship satisfies .

[0092] The S303 sinusoidal excitation heat load separation module uses a relevant demodulation algorithm to perform coherent multiplication and low-pass filtering calculations on the total heat load data to obtain the real-time apparent latent heat coefficient of the steam. Combined with the first law of thermodynamics, the real-time total heat load... It consists of the heat release from basic steam, the heat release from induced steam, the heat release from volatiles, and the static heat loss of the equipment. The expanded physical model is represented as follows: ; In the formula, This indicates the real-time total heat load; The base steam target flow rate updated for the current calculation cycle; The set sinusoidal harmonic excitation amplitude; Represents the sine trigonometric function; The set low-frequency angular velocity; This refers to system uptime. The apparent latent heat coefficient of the stripping steam under the current operating conditions; The net heat load of the actual volatile oils extracted for the target; This is the calibrated value for the heat loss of the equipment itself.

[0093] To facilitate understanding of the technical logic of this separation algorithm, a principle explanation is provided here. Analyzing the mass transfer process of volatiles, the volatilization rate of free fatty acids and odor substances in oils is limited by the high-temperature oil phase viscosity and molecular diffusion resistance, constituting a slow and gradual process, with the corresponding heat load... In the frequency domain, it basically does not include angular velocity. The fluctuation component.

[0094] Therefore, the frequency in the total heat load signal is the same. The fluctuation component can be considered as a response caused purely by superimposed excitation steam.

[0095] The sinusoidal excitation heat load separation module deploys relevant demodulation logic at the algorithm's underlying layer to separate the measured total heat load. Locally generated co-frequency reference signal Coherent multiplication is performed. The multiplication results in a mixed term containing high-order harmonic noise and the target DC component. Subsequently, the sinusoidal excitation heat load separation module uses digital low-pass filter logic to filter out high-frequency components and extract the DC component characterizing the current latent heat state, thereby calculating... : ; In the formula, The apparent latent heat coefficient of the stripping steam under the current operating conditions; The set sinusoidal harmonic excitation amplitude; This indicates the real-time total heat load; Represents the sine trigonometric function; The set low-frequency angular velocity; This refers to system uptime. This represents a digital low-pass filter function, and its specific algorithm implementation can be achieved by discretizing the first-order inertial filter difference equation.

[0096] To effectively filter out high-frequency interference from the second harmonic while preserving the low-frequency latent heat variation trend, the cutoff frequency of the low-pass filter should be set to be less than the harmonic excitation angular velocity. One-tenth of the original value. This effectively suppressed complex broadband non-co-frequency thermal noise interference in the production site and enabled dynamic identification of the latent heat coefficient of steam.

[0097] S304, the sinusoidal excitation heat load separation module eliminates the steam heat load based on the identified apparent latent heat coefficient, and calculates the true net volatile heat load. This is achieved by obtaining high-confidence... Subsequently, the sinusoidal excitation heat load separation module calculates the dynamic heat load in the deodorization tower contributed purely by the stripping medium according to the following formula. : ; In the formula, This indicates the dynamic heat load within the deodorization tower that is purely contributed by the stripping medium; This indicates the actual injection command issued to the pneumatic regulating valve for the hydration reagent and the pneumatic regulating valve for the stripping steam; This represents the apparent latent heat coefficient of the stripping steam under the current operating conditions.

[0098] Subsequently, the sinusoidal excitation heat load separation module performs an algebraic subtraction operation on the total heat load to calculate the net heat load of the volatile oils. : ; In the formula, This represents the calculated net heat load of volatile oils; This indicates the real-time total heat load; This indicates the dynamic heat load within the deodorization tower that is purely contributed by the stripping medium; This is the calibrated value for the heat loss of the equipment itself.

[0099] The above solution process avoids direct subtraction of static large datasets, and uses dynamic excitation response to separate the weak target variable. The heat loss calibration value in the formula... During the no-load hot-state commissioning phase of the equipment, offline measurements can be performed by maintaining the same physical vacuum and operating temperature conditions through energy balance calculations, and these results can be fixed as constants in the control system.

[0100] Considering that drastic changes in the external environment, such as changes in ambient temperature, may lead to... The increased actual deviation leads to abnormal heat load estimation. The sinusoidal excitation heat load separation module has a net heat load zero-point clamping protection logic.

[0101] When the calculated net heat load When a negative value is obtained, the sinusoidal excitation heat load separation module forces the value to zero and simultaneously triggers a system maintenance prompt for the equipment external insulation integrity check.

[0102] See attached document Figure 6 After obtaining the net heat load of volatiles, the closed-loop follower control module combines the hydration process response characteristic parameters from the cross-process feedforward to perform equivalent latent heat of volatiles parameter correction and command update. This execution process specifically includes the following sub-steps: S401, the closed-loop follower control module retrieves the characteristic parameter set extracted from the upstream degumming process to construct a non-ideal solution phase equilibrium correction coefficient. The closed-loop follower control module reads the hydration process response characteristic parameters addressed and written to the batch by the parameter extraction feedforward module from the register via the system's underlying distributed communication bus. These characteristic parameters include the critical response time calculated previously. The time integral value of conductivity .

[0103] In actual production, tiny colloidal particles and impurities that may not have been fully reacted or separated in the upstream degumming process may enter the deodorization tower of the downstream deodorization process along with the material.

[0104] These residual substances alter the intermolecular forces between the bulk oil and volatile components, causing the physicochemical behavior of the actual deodorized oil to deviate from the ideal solution model. Thermodynamically, this deviation manifests as a discrepancy between the actual latent heat of vaporization of the volatiles and the theoretical value under standard conditions. Therefore, this embodiment introduces a non-ideal solution phase equilibrium correction coefficient to compensate for this deviation.

[0105] The closed-loop follower control module calculates the phase equilibrium correction coefficient for the non-ideal solution according to the following formula. : ; In the formula, This refers to the calculated correction factor for the phase equilibrium of the non-ideal solution. This is the first system constant; This is the second system constant; Indicates the critical response time of the current batch of materials; This is the calculated time integral value of the conductivity. (Constant) and All of these are dimensionless empirical coefficients, and their values ​​can be obtained by offline multivariate regression identification algorithms during the equipment commissioning phase, based on the internal topology and fluid dynamics characteristics of the deodorization tower.

[0106] To ensure the integrity of the control logic and avoid division-by-zero failures, when the critical response time is retrieved... When the value is zero or exceeds the set judgment range, the anomaly handling program inside the closed-loop follow control module will adjust the correction coefficient. Force the value to the default value of 1.0 to ensure the stability of subsequent thermodynamic calculations.

[0107] The above-mentioned judgment interval can be set according to the statistical distribution of historical normal batch data, for example, it can be set to the range of 50% to 150% of the historical baseline response time.

[0108] S402, the closed-loop follower control module uses the corrected apparent latent heat of vaporization constant to back-calculate the true mass evaporation rate of the volatile oils. The closed-loop follower control module receives the net heat load of the volatiles calculated and output by the sinusoidal excitation heat load separation module and protected by zero-point clamping. .

[0109] Subsequently, the closed-loop follow-up control module calls the theoretical latent heat of vaporization constant of the deodorized oil volatiles under standard conditions. And combined with the calculated correction coefficient Perform multiplication to obtain the apparent latent heat of vaporization constant, which characterizes the actual physical state of the current batch of materials. .

[0110] Based on this, the closed-loop follower control module calculates the true mass evaporation rate of the volatile oils using the following algebraic equation of heat. : ; In the formula, This represents the actual mass evaporation rate of volatile oils and fats in the deodorization tower at the current moment, calculated in reverse. The net heat load of the actual volatile oils extracted for the target; This refers to the calculated correction factor for the phase equilibrium of the non-ideal solution. This represents the theoretical latent heat of vaporization of deodorized oil volatiles under standard conditions.

[0111] Theoretical latent heat of vaporization The specific values ​​can be obtained by consulting relevant oleochemical thermodynamics handbooks or by measuring with a laboratory calorimeter. This calculation step utilizes the net heat load of the stripped volatiles divided by the latent heat parameter after feedforward correction to achieve an indirect online calculation of the volatile mass escaped.

[0112] S403, the closed-loop follower control module updates the target flow rate of basic steam in real time based on the gas-liquid phase equilibrium relationship. In the high-vacuum operating environment of the deodorization process, the main function of injecting stripping steam is to reduce the partial pressure of the target volatile components in the gas phase space, so as to maintain the mass transfer driving force at the gas-liquid interface.

[0113] Based on the phase equilibrium theory of gas-liquid phase equilibrium and the principle of steam stripping, the mass flow rate of stripping steam required to maintain stable gas-liquid phase equilibrium shows a positively correlated linear relationship with the mass evaporation rate of volatiles.

[0114] The closed-loop follower control module updates the control commands according to the following formula: ; In the formula, The base steam target flow rate updated for the current calculation cycle; This is a preset gas-liquid phase equilibrium partial pressure ratio constant; This represents the evaporation rate of the actual mass of volatile oils and fats in the deodorization tower at the current moment, calculated in reverse.

[0115] proportionality constant The value of this constant is determined by the system's absolute operating pressure, the working temperature of the stripping steam, and the deodorization qualification standard of the target product. In actual engineering, it can be configured and solidified by process technicians according to specific production formula specifications. Usually, the value of this constant can be set between 0.5 and 3.0.

[0116] After completing the calculation, the closed-loop follower control module will update the basic steam target flow rate. The feedback is sent to the sinusoidal excitation thermal load separation module, which serves as the reference input for the harmonic excitation superposition equation in the next control cycle.

[0117] At the same time, the update command is converted into a corresponding analog or digital output signal, which acts on the underlying control loops of the hydration reagent pneumatic regulating valve and the deodorization stripping steam pneumatic regulating valve, forming a closed-loop regulation channel for the stripping steam consumption.

[0118] This adaptive steam target flow rate update method allows the steam medium input to dynamically adapt to changes in volatile matter removal load. The specific parameter settings for the proportional-integral controller of the steam flow rate in the regulating loop, as well as the compensation method for the valve mechanical dead zone, can be configured by those skilled in the art based on the specific model of the field actuator and the pipeline pressure characteristics. These are well-known technologies in the field and will not be elaborated upon here.

[0119] Specific application examples: The pollution reduction and energy saving production system of this invention was introduced into a continuous edible oil refining production line of an oil processing plant. The plant processes crude soybean oil and has a designed daily processing capacity of 500 tons.

[0120] During the degumming and hydration stage, the system collects the conductivity of the main circulation pipeline of the hydration tank and the active power of the circulation pump in real time with a sampling period of 0.5 seconds. As the hydration reagent is continuously injected, the crude soybean oil changes from a homogeneous state to a water-in-oil emulsion state. After 15 minutes of reagent injection, the fluid reaches hydration saturation and the free water phase precipitates. The system detects a positive abrupt change in the second-order difference value of conductivity, which exceeds the preset threshold. The microemulsion structure disintegrates, causing a decrease in the apparent viscosity of the fluid. The first-order difference value of the active power of the circulation pump also experiences a negative abrupt change, exceeding the preset threshold. When both conditions are met simultaneously and last for more than 5 seconds, the joint judgment module identifies the phase separation critical point, and the system outputs a command to close the hydration reagent regulating valve.

[0121] After the cutoff operation is completed, the parameter extraction feedforward module records the critical response time of the batch from reagent injection cutoff. The system calculates the conductivity time integral value within this time period by discrete accumulation. These two values ​​are packaged into the hydration process response characteristic parameters of the batch and assigned a batch tracking code. The system writes the above characteristic parameters into the storage area of ​​the deodorization process through the distributed control system bus.

[0122] When the batch of soybean oil is transferred to the deodorization tower, the sinusoidal excitation heat load separation module reads the mass flow rate of the cooling water in the surface condenser and the temperature difference between the inlet and outlet water to calculate the real-time total heat load of the condenser. The system superimposes a sinusoidal excitation flow signal with an amplitude of 3% of the base range and a period of 60 seconds into the base steam target flow rate. The system uses a phase-sensitive detection demodulation algorithm to coherently multiply the total heat load signal with a sinusoidal reference signal of the same frequency and then performs low-pass filtering. The system identifies the real-time apparent latent heat coefficient of the stripping steam. The system deducts the base heat load and excitation fluctuation heat load generated by the steam and the fixed heat loss of the equipment from the total heat load. The system calculates the net heat load of the volatiles.

[0123] The closed-loop follow control module calls the hydration characteristic parameters of the previous feedforward to calculate the correction coefficient of the non-ideal solution phase equilibrium. The system uses the corrected apparent latent heat of vaporization constant and the calculated net heat load to back-calculate the true mass evaporation rate of volatile oils in the deodorization tower. The system calculates and updates the basic steam target flow rate in real time according to the gas-liquid phase equilibrium partial pressure ratio constant required by the process. The system adjusts the valve opening through closed-loop control to make the steam input follow the actual removal requirements of volatiles.

[0124] Appendix Figure 7 The X-axis in the figure represents the critical hydration response time. Figure 7 The Y-axis in the figure represents the time integral characteristic of conductivity. Figure 7 The Z-axis in the figure represents the equivalent latent heat correction coefficient, and the numerical mapping relationship between the characteristic parameters and the correction coefficient is reflected by the three-dimensional surface.

[0125] Appendix Figure 8 The X-axis in the diagram represents the continuous refining operation time. Figure 8 The Y-axis in the graph represents the heat and flow rate amplitude parameters. Figure 8 The light gray area in the image represents the total heat load collected by the system. Figure 8 The dark gray solid line in the diagram represents the net heat load calculated from the separation. Figure 8 The gray dashed line in the text represents the adaptive steam target command.

[0126] 6,000 tons of crude soybean oil from the same production area and the same batch were divided into two groups for processing, with 3,000 tons in each group.

[0127] The first batch of 3,000 tons of crude oil was put into the traditional refining control mode. The amount of water added for degumming hydration was manually set and turned off at a fixed ratio of 2% of the oil weight. The steam flow rate of the deodorization tower was set to a fixed value of 200 kg per hour. The system recorded the total consumption of hydration reagents in the traditional mode, the average steam consumption per ton of oil in the deodorization tower, and the entrainment loss rate of neutral oil in the deodorization distillate.

[0128] After cleaning the pipeline, the second batch of 3,000 tons of crude oil was put into the adaptive control mode of the pollution reduction and consumption reduction production system for processing. The system automatically identifies the critical point of hydration and cuts off the water supply. During the deodorization stage, it uses sinusoidal excitation and phase balance correction algorithm to dynamically adjust the stripping steam valve in real time. The system records the process energy consumption and quality data throughout the process and takes samples to be sent to a third-party laboratory for testing and comparison.

[0129] Both groups of final products met the national first-grade soybean oil standards in terms of phospholipid content and acid value. The consumption of hydration reagent in the adaptive control mode group was reduced by 18% compared with the traditional manual fixed ratio mode. The steam consumption per ton of oil stripping in the adaptive control mode group was reduced by 25%, and the loss rate of neutral oil in the distillate was reduced by 30%.

[0130] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A green and environmentally friendly method for producing edible oil with reduced pollution and energy consumption, characterized in that, include: Collect the conductivity of the material and the active power of the circulating pump, calculate the second-order difference of conductivity and the first-order difference of active power, and output the closing command of the pneumatic regulating valve of the hydration reagent when the multivariate joint threshold condition is met. Record the critical response time from the injection start time to the command cutoff time, calculate the conductivity time integral value within the critical response time, construct the hydration process response characteristic parameters, and store them in the deodorization process data storage area according to batch identification; Collect the condenser cooling water mass flow rate and inlet / outlet water temperature difference, calculate the real-time total heat load, superimpose the basic steam target flow rate and sinusoidal excitation signal, control the injection of stripping steam, collect the actual steam mass flow rate, coherently multiply the total heat load and sinusoidal excitation signal and low-pass filter to obtain the apparent latent heat coefficient of stripping steam, multiply the actual steam mass flow rate and the apparent latent heat coefficient of stripping steam to obtain the stripping steam heat load, subtract the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load to obtain the net heat load of volatile oils; The response characteristic parameters of the hydration process are retrieved in batches, and the phase equilibrium correction coefficient of the non-ideal solution is constructed. The theoretical latent heat of vaporization constant is corrected to the apparent latent heat of vaporization constant. The net heat load of oil volatiles is divided by the apparent latent heat of vaporization constant to obtain the true mass evaporation rate of oil volatiles. The basic steam target flow rate is updated according to the partial pressure ratio constant of the gas-liquid phase equilibrium.

2. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The calculation of the second-order difference value of conductivity and the first-order difference value of active power, when satisfying the multivariate joint threshold condition, outputs a closing command for the pneumatic regulating valve of the hydration reagent, specifically including: Set a positive abrupt change threshold for the second-order difference of conductivity; Set a threshold for negative abrupt changes in the first-order difference value of active power; When the second-order difference value of conductivity is greater than the positive mutation threshold and the first-order difference value of active power is less than the negative mutation threshold, a command to close the pneumatic regulating valve for hydration reagent is output.

3. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The process of collecting the conductivity of the material and the active power of the circulating pump, and calculating the second-order difference of conductivity and the first-order difference of active power, specifically includes: The material conductivity data and circulating pump active power data are time-aligned according to a unified clock. A synchronous sampling sequence is formed according to the same timestamp; A digital low-pass filter is used to filter the synchronous sampling sequence, and the second-order difference value of conductivity and the first-order difference value of active power are calculated based on the filtered synchronous sampling sequence.

4. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The critical response time from the start of the injection to the instruction truncation time specifically includes: The moment when the pneumatic regulating valve for extracting the hydration reagent receives the opening command and generates an opening action feedback is taken as the injection start moment. The moment when the closing command is issued to the pneumatic regulating valve for the hydration reagent is taken as the command cutoff moment. The critical response time is obtained by calculating the time difference between the instruction truncation time and the injection start time.

5. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The calculation of the conductivity time integral value within the critical response time specifically includes: The filtered conductivity values ​​from the injection start time to the instruction cutoff time are read according to a unified digital sampling period. The time integral value of conductivity is obtained by multiplying each filtered conductivity value by a uniform digital sampling period and then summing the results. Linear interpolation is used to fill in the missing data caused by communication failures.

6. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The process of collecting the condenser cooling water mass flow rate and inlet / outlet water temperature difference to calculate the real-time total heat load specifically includes: Read the real-time cooling water mass flow rate of the surface condenser cooling water pipeline; Read the cooling water outlet temperature and cooling water inlet temperature, calculate the difference between the cooling water outlet temperature and the cooling water inlet temperature, and obtain the inlet and outlet water temperature difference; The real-time total heat load is obtained by multiplying the constant specific heat capacity of cooling water at constant pressure, the real-time mass flow rate of cooling water, and the temperature difference between the inlet and outlet water.

7. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The net heat load of volatile oils is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load, specifically including: By superimposing the target basic steam flow rate and the sinusoidal excitation signal, a stripping steam injection command is generated. Collect the actual steam mass flow rate fed back from the stripping steam mass flow meter; Multiply the actual steam mass flow rate by the apparent latent heat coefficient of the stripping steam to obtain the stripping steam heat load; The net heat load of volatile oils is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load.

8. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The net heat load of volatile oils is obtained by subtracting the stripping steam heat load and the equipment heat loss calibration value from the real-time total heat load, specifically including: Configure the zero-point clamping protection logic for the net heat load of volatile oils; When the net heat load of volatile oils is less than zero, the net heat load of volatile oils should be set to zero. Synchronous output device external insulation integrity check prompt.

9. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The step of batch-based hydration process response characteristic parameters extraction and construction of non-ideal solution phase equilibrium correction coefficients specifically includes: Divide the time integral of conductivity by the critical response time to obtain the characteristic value of average conductivity; Multiply the characteristic value of the average conductivity by the first system constant and add it to the second system constant to obtain the phase equilibrium correction coefficient for the non-ideal solution; When the critical response time is zero, the phase equilibrium correction factor for the non-ideal solution is set to the default value of 1.

0.

10. The green and environmentally friendly edible oil production method with reduced pollution and energy consumption according to claim 1, characterized in that, The true mass evaporation rate of the oil volatiles is obtained by dividing the net heat load of the oil volatiles by the apparent latent heat of vaporization constant. The base steam target flow rate is updated based on the gas-liquid phase equilibrium partial pressure ratio constant, specifically including: Multiply the theoretical latent heat of vaporization constant by the phase equilibrium correction factor for the non-ideal solution to obtain the apparent latent heat of vaporization constant; Divide the net heat load of oil volatiles by the apparent latent heat of vaporization constant to obtain the true mass evaporation rate of oil volatiles; The updated basic steam target flow rate is obtained by multiplying the gas-liquid phase equilibrium partial pressure ratio constant by the actual mass evaporation rate of oil volatiles. The updated baseline steam target flow rate is used as the reference input for the sinusoidal excitation signal in the next control cycle.

Images