Multi-model switching control method and system for temperature of reaction kettle for biofuel preparation

Abstract

This invention belongs to the field of industrial automation control technology, and relates to a multi-model switching control method and system for the temperature of a reactor used in biofuel production. The method includes: acquiring real-time temperature and real-time current of the stirring motor within the reactor, and performing filtering preprocessing to obtain smooth monitoring data; calculating a rheo-thermodynamic coupling characteristic index reflecting the reaction depth and thermodynamic trend based on the ratio of initial current to real-time current and the temperature change rate; calculating the real-time intervention weight of the cooling suppression model using a smoothing switching function based on the deviation of this index from a preset threshold; and finally, weighted fusion of the outputs of the heating-dominant model and the cooling suppression model based on this weight to generate the final control command. This invention achieves adaptive control under all operating conditions by monitoring current changes and feeding forward to predict the reaction stage, effectively preventing temperature runaway during the reaction and significantly improving production safety.

Description

Technical Field

[0001] This invention belongs to the field of industrial automation control technology, specifically relating to a multi-model switching control method and system for the temperature of a reactor used in biofuel preparation. Background Technology

[0002] Biofuels, as an important clean and renewable energy source, are typically produced using waste cooking oil, acidified oil, or vegetable oil residue as raw materials. These materials undergo esterification or transesterification reactions under the action of a catalyst to produce fatty acid methyl esters. This production process is mainly carried out in a batch reactor, and the reaction is accompanied by complex physicochemical changes. It is a typical thermodynamic process characterized by strong nonlinearity, large time hysteresis, and time-varying parameters. As the reaction progresses, the composition of the materials in the reactor changes significantly, leading to substantial dynamic shifts in key physical properties such as viscosity, specific heat capacity, and thermal conductivity of the reaction mixture. The material state gradually evolves from a high-viscosity non-Newtonian fluid in the initial stage to a low-viscosity fluid in the later stages of the reaction.

[0003] In existing industrial production, reactor temperature control mainly employs traditional single-loop PID control strategies or time-based programmed temperature rise methods. These conventional control methods heavily rely on temperature sensors installed inside or on the reactor wall for feedback regulation, attempting to maintain thermal equilibrium by controlling the flow rate of heat transfer oil or cooling water in the jacket. However, conventional PID controllers are linear controllers based on fixed-parameter models, making it difficult to adapt to the drastic fluctuations in the parameters of the controlled object model during biofuel production. In the high-viscosity stage of the initial reaction, poor material flowability leads to low heat transfer efficiency, often requiring a large heating output to maintain the heating rate. Once the reaction starts and enters the period of intense exothermic reaction, the material viscosity rapidly decreases, causing a sudden increase in heat transfer efficiency. If the controller parameters cannot be adjusted in real time, the system is highly susceptible to control overshoot.

[0004] More importantly, due to the significant thermal inertia of large reactors, temperature sensor readings often lag behind the actual thermodynamic state at the reaction center. By the time the temperature sensor detects a noticeable temperature rise, the chemical reaction inside the reactor is often already in a violently exothermic phase. At this point, the heating effect accumulated in traditional control strategies combined with the exothermic effect of the reaction itself can easily lead to uncontrolled heat accumulation and severe temperature spikes. This temperature runaway not only causes high-temperature deactivation of the catalyst, increases side reactions, and reduces product conversion rates, but in extreme cases, it can even lead to serious production safety accidents such as material overflow and explosions. Therefore, there is an urgent need for an adaptive temperature control technology that can adapt to the coupled changes in the rheological and thermodynamic properties of materials. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a multi-model switching control method and system for the temperature of a reactor used in biofuel preparation. The aim is to solve the technical problem of temperature runaway caused by the mismatch of traditional fixed parameter control models due to significant time-varying material properties and lag in temperature detection during existing biofuel preparation processes.

[0006] To address the above problems, the technical solution of the multi-model switching control method for the temperature of the biofuel preparation reactor proposed in this invention is as follows: A multi-model switching control method for the temperature of a reactor used in biofuel production includes the following steps: Real-time temperature data of the reactor and real-time current data of the stirring motor are collected, and the real-time current data is filtered to obtain a smooth measurement current, while determining the reference current at the reaction start-up time. Based on the ratio of the reference current to the measured current and the time change rate of the real-time temperature data, the rheological-thermodynamic coupling characteristic index characterizing the material reaction depth and thermodynamic trend is calculated. Based on the deviation between the rheological thermodynamic coupling characteristic index and the preset state switching threshold, the intervention weight coefficient of the cooling and temperature suppression model is calculated using a nonlinear function. Based on the intervention weighting coefficient, the output values ​​of the heating-dominant model and the cooling-suppression model are weighted and fused to generate the final control command for the actuator to adjust the reactor temperature, thereby achieving multi-model switching control of the reactor temperature for biofuel preparation.

[0007] Further, the real-time current data is filtered to obtain a smooth measurement current, including: The real-time current data is processed using a moving average filtering algorithm. By setting the sampling point length of the sliding window, the arithmetic mean of the original current values ​​at multiple consecutive sampling times within the sliding window is calculated, and the arithmetic mean is used as the measured current at the current time.

[0008] By applying a moving average filter to the original current signal, high-frequency noise caused by liquid surface fluctuations and mechanical vibrations can be effectively filtered out, resulting in a smooth current signal that accurately reflects the viscosity characteristics of the material, providing a reliable data basis for subsequent exponential calculations.

[0009] Furthermore, the rheological-thermal coupling characteristic index satisfies the expression:

[0010] In the formula, express The rheological-thermal coupling characteristic index at time t, express Measuring current at any given time The reference current representing the moment the reaction begins. Represents the logarithmic bias constant. The first derivative representing temperature. This represents the safety basis constant.

[0011] By constructing a rheo-thermo-coupling characteristic index that includes the current ratio and the rate of temperature change, the reaction state can be captured in advance by the current change caused by the decrease in viscosity before the temperature changes significantly. This enables an early quantitative assessment of the risk of reaction runaway and solves the hysteresis problem caused by relying solely on temperature feedback.

[0012] Furthermore, the real-time temperature data is collected by a resistance temperature detector (RTD) sensor installed at the bottom of the reactor, and the real-time current data is the effective value of the three-phase current collected by a Hall current sensor installed at the output of the stirrer motor inverter.

[0013] Furthermore, the intervention weighting coefficient satisfies the expression:

[0014] In the formula, express The intervention weighting coefficients of the constant cooling and temperature suppression model Indicates the base of the switching function. Indicates the sensitivity adjustment factor. This represents the rheological-thermal coupling characteristic index. This indicates the set state switching threshold.

[0015] Furthermore, the state switching threshold The base of the switching function is determined based on the average rheological-thermal coupling characteristic index corresponding to the moment when the reaction begins to release heat in historical batch data. Used to control the steepness of the switching curve.

[0016] Furthermore, the method for generating the final control command is as follows: calculate the product of the output value of the heating-dominant model and the first weighting factor, and the product of the output value of the cooling-suppression model and the intervention weighting coefficient, and add the two products together to obtain the final control command; wherein, the sum of the first weighting factor and the intervention weighting coefficient is 1.

[0017] By weighted fusion of the outputs of two different characteristic models, the system can automatically adjust the focus of the control strategy at different stages of the reaction, enabling rapid heating during the heating period and strong temperature suppression during the heat release period, thus achieving precise control under all operating conditions.

[0018] Furthermore, the heating-dominant model is configured with proportional gain and integral time for control during the heating phase when the fluid is highly viscous and endothermic; the cooling-suppression model is configured with differential action and cooling output gain for control during the cooling phase when the fluid is low viscous and exothermic.

[0019] Furthermore, the reference current is determined by calculating the average current value over a preset time period after the system start command is issued and in a cold, mixed state before heating begins.

[0020] The technical solution of the multi-model switching control system for the temperature of the biofuel preparation reactor proposed in this invention is as follows: A multi-model switching control system for the temperature of a biofuel production reactor includes a processor and a memory. The memory stores computer instructions, which, when executed by the processor, implement the multi-model switching control method for the temperature of the biofuel production reactor as described in any of the above technical solutions.

[0021] The beneficial effects of this invention are as follows: By introducing the stirring motor current as a feedforward variable reflecting changes in material viscosity, this invention creatively utilizes the changing trends of physical quantities to infer the progress of chemical reactions, thus solving the hysteresis problem commonly found in traditional temperature control. Before the temperature sensor detects significant heat accumulation, this invention can keenly predict the intensity of the reaction based on the decrease in viscosity and intervene in the cooling control model in advance. This feedforward control mechanism overcomes the limitations of relying solely on temperature feedback, effectively avoiding the runaway boiling phenomenon commonly seen in biofuel production, stabilizing the reaction temperature within the set target range, and improving the safety of chemical production.

[0022] This invention employs a smooth switching strategy based on nonlinear functions, achieving a flexible transition between the heating-dominant model and the cooling-suppression model. Unlike the abrupt changes caused by traditional threshold control, this weighted fusion method based on intervention weight coefficients ensures the continuity and stability of the final control command. It not only effectively avoids frequent oscillations of the actuator at critical operating points and reduces mechanical wear on valves and other hardware, but also extends the service life of the equipment and ensures stable operation throughout the entire reaction process.

[0023] Furthermore, by calculating the ratio of the reference current to the measured current and the rate of temperature change, this invention reduces the dependence on the absolute physical parameters of specific raw materials. The system can achieve high-precision control without repeated manual parameter adjustments for each batch of raw materials. This adaptive capability not only improves the automation level of the production process but also significantly enhances production efficiency and product quality consistency. Attached Figure Description

[0024] Figure 1A flowchart illustrating the steps of the multi-model switching control method for the temperature of a biofuel preparation reactor according to the present invention; Figure 2 This is a schematic diagram comparing the reactor temperature control process provided in this embodiment of the invention with the prior art; Figure 3 A schematic diagram illustrating the principle of key parameter changes and control logic provided in the embodiments of the present invention. Detailed Implementation

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0026] Specific embodiments of the multi-model switching control method for reactor temperature in biofuel preparation proposed in this invention: like Figure 1 As shown, the multi-model switching control method for the temperature of a reactor used in biofuel production includes the following steps: S1. Collect real-time temperature data of the reactor and real-time current data of the stirring motor, and filter the real-time current data to obtain a smooth measurement current, while determining the reference current at the reaction start-up time.

[0027] Specifically, in this embodiment, a PT100 platinum resistance thermometer is installed at the bottom of the reactor to collect real-time temperature data of the materials inside the reactor, denoted as . A Hall effect current sensor is installed at the output terminal of the inverter of the mixing motor to collect the effective value of the three-phase current of the motor, i.e., the real-time current data, denoted as . .

[0028] During reactor operation, the system synchronously acquires raw signals via an analog input module with a sampling period of 100ms. Considering the high-frequency mechanical noise introduced by the shearing action of the agitator blades and liquid surface fluctuations during biofuel preparation, directly using the raw current data could lead to control malfunctions. Therefore, this step employs a moving average filtering algorithm to smooth the real-time current data. Specifically, the sampling point length of the sliding window is set, and the arithmetic mean of the raw current values ​​at multiple consecutive sampling moments within the sliding window is calculated. This arithmetic mean is then used as the measured current at the current moment. That is, the sliding window length is set to... The calculation formula for filtering is as follows:

[0029] In the formula, for Measuring current at any given time Push forward from the current moment The original current values ​​of each sampling point.

[0030] This formula is essentially a low-pass finite impulse response (FIR) filter. The current fluctuations caused by mechanical vibrations, bubble bursts, and liquid surface turbulence within the reactor are zero-mean random high-frequency white noise. By using the moving average of the time series, following the law of large numbers in statistics, it can effectively filter out random high-frequency interference and extract the low-frequency effective signal that truly reflects the macroscopic viscosity changes of the fluid, preventing the control system from malfunctioning due to high-frequency glitches.

[0031] For example, setting the length of the sliding window. The value is 3, at a certain moment Given that the three original current samples within the sliding window are 10.2A, 10.0A, and 9.8A respectively, the smoothed measured current is calculated as follows: .

[0032] Furthermore, to establish a reference frame for viscosity changes, the system needs to determine a reference current during the cold mixing phase before the reaction begins. Specifically, after the reaction start command is issued, and while in a cold mixing state before heating begins, the system calculates the average current value over a preset time period, such as 30 seconds, and establishes this as the reference current at the reaction start time, denoted as . This serves as a baseline reference value for subsequent calculations of viscosity changes. For example, the average current calculated during the first 30 seconds after system startup... The reference current is 12.0A, which represents the high viscosity state of the material when it has not reacted and is at its initial temperature.

[0033] Thus, through the above-mentioned acquisition and filtering processes, environmental interference can be eliminated, pure current and temperature signals can be obtained, and a benchmark for the initial state of the reaction can be established, providing reliable data support for subsequent accurate evaluation of the reaction process.

[0034] S2. Calculate the rheological-thermal coupling characteristic index, which characterizes the material reaction depth and thermodynamic trend, based on the ratio of the reference current to the measured current and the time change rate of the real-time temperature data.

[0035] To overcome the hysteresis caused by relying solely on temperature feedback, this step constructs a mathematical model that can sensitively reflect the viscosity-temperature coupling state, namely, by introducing a rheo-thermal coupling characteristic index. The rheo-thermal coupling characteristic index integrates the ratio of current change characterizing the reaction depth and the rate of temperature change characterizing the thermodynamic trend, and is used to quantify the current risk of reaction runaway.

[0036] The rheothermal coupling characteristic index satisfies the following expression:

[0037] in, express The rheothermal coupling characteristic index at time t, in units of .

[0038] The reference current representing the moment the reaction begins. express The current is measured at specific times. In the transesterification reaction of biodiesel, as the reaction proceeds, high-viscosity triglycerides are gradually converted into low-viscosity fatty acid methyl esters, leading to a decrease in stirring resistance and a drop in motor current; simultaneously, this reaction is exothermic, causing a temperature increase. Therefore, This reflects the depth of the chemical reaction, i.e., the degree of viscosity reduction. This aligns with the principles of fluid mechanics and electric motor drives: the output electromagnetic torque of the stirring motor is proportional to the fluid viscosity, and the current is positively correlated with the electromagnetic torque. Therefore, the ratio of currents is essentially a physical representation of the macroscopic apparent viscosity reduction of the reactant system. By taking the natural logarithm of this ratio, conforming to the Weber-Fechner law in physical perception, the exponentially decaying viscosity change can be transformed into approximately linear scalar data, preventing the ratio from becoming too large and causing calculation overflow when the viscosity is extremely low in the later stages.

[0039] This is the logarithmic bias constant, with a value of 1; its function is to ensure that the independent variable of the logarithmic function is always greater than 1, preventing bias in the logarithmic function. Mathematical calculation errors or negative values ​​may occur.

[0040] The safety basis constant is 10, and its unit is 10. Its physical meaning is that even if the rate of temperature change is 0, the square root term remains unchanged. To ensure the rheological-thermal coupling characteristic index It is mainly driven by the current ratio, i.e. viscosity change, to prevent the exponent from returning to zero during the isothermal stage and thus losing information about the reaction process.

[0041] Square root term Based on the principles of energy conservation and reaction kinetics in thermodynamics. The first derivative of temperature, i.e., the rate of temperature change, is expressed in °C / min. It directly reflects the rate of net heat accumulation within the reaction system, i.e., the difference between the rate of heat release and the rate of heat dissipation. The squared treatment not only positively measures potential temperature fluctuations but also nonlinearly amplifies the weight of drastic temperature increases. In this embodiment, pass Calculations show that To calculate the step size.

[0042] This formula couples the fluid dynamics characteristics with the thermodynamic characteristics by multiplying them, which conforms to the natural law that the sudden changes in physical properties and exothermic changes in chemical reaction engineering have time consistency.

[0043] To better understand the above formula, a calculation example is provided below.

[0044] In the initial stage of the reaction, i.e., the heating stage, the viscosity of the material has not yet changed significantly, and the temperature rises slowly. Assuming... It is close to the reference current; Substituting into the formula, we get: At this point, the rheological-thermal coupling characteristic index is low, indicating that the system is in a stable heating region.

[0045] During the most intense, exothermic phase of the reaction, the viscosity drops sharply and is accompanied by heat release. Down to The rate of temperature rise at this time Substituting into the formula, we get: At this point, the rheological-thermal coupling characteristic index increases significantly, and the system keenly detects the sudden change in operating conditions.

[0046] By constructing a rheological-thermodynamic coupling characteristic index, the system combines the physical trend of current change with the thermodynamic characteristics of temperature change, which plays a key role in signal amplification. This effectively overcomes the hysteresis of single temperature feedback, enabling the system to capture sudden changes in operating conditions earlier and more sensitively than relying solely on temperature signals, thereby achieving early feedforward warning of overheating risks.

[0047] S3. Based on the deviation between the rheological thermodynamic coupling characteristic index and the preset state switching threshold, the intervention weight coefficient of the cooling and temperature suppression model is calculated using a nonlinear function.

[0048] To overcome the problems of abrupt changes in control quantities and system oscillations caused by hard threshold judgments in traditional control logic, this invention does not employ hard threshold switching, but instead adopts a smooth weighted switching strategy based on the Sigmoid function. The core of this strategy lies in calculating the intervention weight coefficients of the cooling and temperature suppression model in real time according to the operating conditions, thereby achieving a flexible transition of the control mode. Specifically, the cooling and temperature suppression model is specially configured with strong differential action parameters and cooling output gain, aiming to rapidly and powerfully suppress the exothermic reaction.

[0049] In this embodiment, the intervention weighting coefficient satisfies the expression:

[0050] in, express The intervention weighting coefficients of the constant cooling and temperature suppression model; The base of the switching function determines the basic curvature of the function curve. In this embodiment, the value is set to 10 to construct a suitable exponential decay or growth trend.

[0051] This is the sensitivity adjustment factor, in units of The value 1.5 is used to adjust the response speed of the weight as the characteristic exponent changes, i.e. the steepness of the curve. This value ensures that when the operating condition approaches the state switching threshold, the weight can be increased quickly but without a step, thus ensuring the smoothness of the transition.

[0052] The set state switching threshold is a critical point calibrated based on a large amount of historical batch data; in this embodiment, it is set to 3.5. The state switching threshold represents the boundary between the stable period and the high-risk exothermic period of the reaction, and it is also the benchmark point at which the weighting function undergoes a center reversal.

[0053] This formula is based on fuzzy logic and the principle of perturbation-free switching of multiple models in advanced control theory. Essentially, it is a smoothed version of the step function. When the response is in a plateau phase, i.e. At that time, the weight approaches 0; when entering the period of intense heat release, that is... At this point, the weight rapidly approaches 1. This S-shaped transition law not only conforms to the nonlinear evolution law of state phase transition in nature, but also avoids the frequent jittering phenomenon caused by traditional relay logic near the critical point, protecting mechanical actuators such as valves from wear.

[0054] Following step S2, the following numerical examples illustrate the calculation process and physical meaning of the weighting coefficients in the cooling and temperature suppression model: When rheothermal coupling characteristic index When it rises to 4.786, The results show that the intervention weight coefficient is close to 1 at this time, and the system control smoothly transitions to the cooling and temperature suppression model to cope with the upcoming intense heat release.

[0055] Conversely, in the initial cold mixing stage of the reaction, when At a lower level, such as 2.21, the calculated value is... This extremely low value means that the system is still controlled by the heating-dominant model, with the cooling mechanism hardly intervening, thus ensuring heating efficiency.

[0056] The above calculation logic utilizes the characteristics of nonlinear functions to achieve dynamic and precise allocation of control weights: during periods of mild reaction, the weights approach 0 to ensure normal temperature rise; once an intensified reaction is detected, the weights rapidly increase to 1 to implement strong temperature suppression. This mechanism not only ensures adaptive control under all operating conditions but also achieves a smooth and seamless switching from heating to cooling modes, effectively avoiding system oscillations caused by hard switching.

[0057] S4. Based on the intervention weight coefficient, the output values ​​of the heating-dominant model and the cooling-suppression model are weighted and fused to generate the final control command for the actuator to adjust the reactor temperature, so as to realize multi-model switching control of the reactor temperature for biofuel preparation.

[0058] Based on the calculated intervention weight coefficients, the system performs weighted fusion of the output values ​​of the two models to generate the final control command. The signal is then sent to the actuator. The actuator uses a three-way regulating valve with three ports: one connects to a heat source, such as high-temperature heat transfer oil; one connects to a cold source, such as cooling water; and the third connects to the reactor jacket inlet, enabling control of heating and cooling.

[0059] The final control command is generated using the following formula:

[0060] in, The output value is the heating-dominant model, which is configured with a strong proportional gain and integral time to control the heating stage of the fluid when it is highly viscous and endothermic. The output values ​​are from the cooling and temperature suppression model, which is configured with differential action and cooling output gain to control the fluid during its low viscosity and exothermic phase, aiming to suppress sudden temperature changes. The output values ​​of both models are valve opening percentages, ranging from 0% to 100%; 100% corresponds to the maximum heating opening, and 0% corresponds to the maximum cooling opening or complete heating cutoff. Therefore, when the cooling and temperature suppression model determines that strong cooling is required, its output value... It will tend towards 0%.

[0061] This formula is based on the dynamic control power allocation mechanism in control engineering, because... and The sum of these values ​​is always 1, meaning that the total control command issued by the system to the actuators at any given time is a linear interpolation of the heating-dominant model and the cooling-suppression model. This architecture allows the heating PID to dominate the temperature rise during low-risk conditions and the cooling PID to suppress the temperature during high-risk conditions, thus meeting the control requirements of complex time-varying systems under multiple operating conditions and ensuring energy balance during transition periods.

[0062] Taking the period of intense reaction as an example, the cooling weight calculated at this time The system status is as follows: Assuming that the heating-dominant model, based on error integration, fails to detect the sudden change in reaction rate because the temperature has not yet fully exceeded the set value, it still determines that heating needs to be maintained and outputs... The cooling and temperature suppression model detected a sudden drop in current and an extremely high rate of temperature change, indicating a risk of overheating. It immediately requested powerful cooling and output... .

[0063] Final output: The calculation results indicate that the system effectively shut down heating, thereby suppressing potential temperature runaway.

[0064] This process clearly demonstrates the core advantages of this invention: although the traditional PID model still issues 60% of the heating command, the system successfully blocks the hysteretic heating command and enforces the cooling strategy of the cooling suppression model because the rheological-thermal coupling index rapidly increases the cooling weight. Through this weighted fusion mechanism, the system ensures efficiency during the heating phase while minimizing overheating risks by utilizing the smooth transition characteristics of multi-model switching, thus ensuring stable operation of the reaction process under all operating conditions.

[0065] The following combination Figure 2 and Figure 3 The technical solution and technical effects of the present invention will be further explained.

[0066] Figure 2 The figure visually demonstrates the significant difference in temperature control performance between this invention and existing technologies. As can be seen from the figure, the existing technology exhibits drastic temperature fluctuations after approximately 60 minutes of reaction, with peak values ​​even reaching 150 degrees Celsius, exhibiting severe overshoot. In contrast, the temperature trajectory of this invention consistently closely adheres to the target value of 120 degrees Celsius, without any significant temperature overshoot throughout the entire process, maintaining extremely high control precision.

[0067] Figure 3 This further reveals the control mechanism behind this excellent performance. As the reaction time progresses, the current curve of the stirring motor, which characterizes the material viscosity, shows a step-down trend from a high level. At the same time, the cooling model intervention weight curve rapidly rises from 0 to 1 during the critical period when the current decreases significantly and is accompanied by a temperature rise. This change process strongly verifies that the present invention can keenly capture viscosity change signals and automatically enhance cooling control before the risk of reaction runaway occurs, thereby effectively suppressing the temperature runaway phenomenon.

[0068] Specific embodiments of the multi-model switching control system for the temperature of the biofuel production reactor proposed in this invention are as follows: A multi-model switching control system for the temperature of a biofuel production reactor includes a processor and a memory. The memory stores computer instructions, which, when executed by the processor, implement the multi-model switching control method for the temperature of the biofuel production reactor in the above embodiments.

[0069] The multi-model switching control system for the temperature of the reactor used in biofuel production also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0070] While various embodiments of the invention have been shown and described in this specification, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention.

Claims

1. A multi-model switching control method for the temperature of a reactor used in biofuel production, characterized in that, Includes the following steps: Real-time temperature data of the reactor and real-time current data of the stirring motor are collected, and the real-time current data is filtered to obtain a smooth measurement current, while determining the reference current at the reaction start-up time. Based on the ratio of the reference current to the measured current and the time change rate of the real-time temperature data, the rheological-thermodynamic coupling characteristic index characterizing the material reaction depth and thermodynamic trend is calculated. Based on the deviation between the rheological thermodynamic coupling characteristic index and the preset state switching threshold, the intervention weight coefficient of the cooling and temperature suppression model is calculated using a nonlinear function. Based on the intervention weighting coefficient, the output values ​​of the heating-dominant model and the cooling-suppression model are weighted and fused to generate the final control command for the actuator to adjust the reactor temperature, thereby achieving multi-model switching control of the reactor temperature for biofuel preparation.

2. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 1, characterized in that, Filtering the real-time current data to obtain a smooth measurement current includes: The real-time current data is processed using a moving average filtering algorithm. By setting the sampling point length of the sliding window, the arithmetic mean of the original current values ​​at multiple consecutive sampling times within the sliding window is calculated, and the arithmetic mean is used as the measured current at the current time.

3. The multi-model switching control method for the temperature of the biofuel preparation reactor according to claim 1, characterized in that, The rheo-thermal coupling characteristic index satisfies the expression: In the formula, express The rheological-thermal coupling characteristic index at time t, express Measuring current at any given time The reference current representing the moment the reaction begins. Represents the logarithmic bias constant. The first derivative representing temperature. This represents the safety basis constant.

4. The multi-model switching control method for the temperature of the biofuel preparation reactor according to claim 1, characterized in that, The real-time temperature data is collected by a resistance temperature detector (RTD) sensor installed at the bottom of the reactor, and the real-time current data is the effective value of the three-phase current collected by a Hall current sensor installed at the output of the stirrer motor inverter.

5. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 3, characterized in that, The intervention weighting coefficient satisfies the expression: In the formula, express The intervention weighting coefficients of the constant cooling and temperature suppression model Indicates the base of the switching function. Indicates the sensitivity adjustment factor. This represents the rheological-thermal coupling characteristic index. This indicates the set state switching threshold.

6. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 5, characterized in that, The state switching threshold The base of the switching function is determined based on the average rheological-thermal coupling characteristic index corresponding to the moment when the reaction begins to release heat in historical batch data. Used to control the steepness of the switching curve.

7. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 1, characterized in that, The method for generating the final control command is as follows: calculate the product of the output value of the heating-dominant model and the first weighting factor, and the product of the output value of the cooling-suppression model and the intervention weighting coefficient, and add the two products together to obtain the final control command; wherein, the sum of the first weighting factor and the intervention weighting coefficient is 1.

8. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 7, characterized in that, The heating-dominant model is configured with proportional gain and integral time for control during the heating phase when the fluid is highly viscous and endothermic; the cooling-suppression model is configured with differential action and cooling output gain for control during the cooling phase when the fluid is low viscous and exothermic.

9. The multi-model switching control method for the temperature of a biofuel preparation reactor according to claim 1, characterized in that, The reference current is determined by calculating the average current value over a preset time period after the system start command is issued and before heating begins, in a cold, mixed state.

10. A multi-model switching control system for the temperature of a reactor used in biofuel production, characterized in that, It includes a processor and a memory, the memory storing computer instructions, which, when executed by the processor, implement the multi-model switching control method for the temperature of the biofuel preparation reactor as described in any one of claims 1-9.

Images