Intelligent temperature control method and device for precisely regulating microstructure of modal fiber

Abstract

The application discloses an intelligent temperature control method and device for precisely regulating the microstructure of modal fibers, which reconstructs the fiber interface temperature and total thermal disturbance by using an extended state observer through smoothing the reference temperature trajectory and fusing multi-source operation data, and then generates the cold and hot valve duty cycle command according to the enzyme activity out-of-control risk by implementing asymmetric error distribution, nonlinear prospective feedback and disturbance feedforward compensation. Thus, the inhibition capability of heat transfer hysteresis, hydrolysis heat release, friction heat release and environmental thermal load change can be enhanced, the high-temperature side overshoot, cumulative hydrolysis damage and valve frequent commutation can be reduced, so that the fiber interface temperature is stably maintained in the enzyme suitable low temperature window, and the precise and uniform regulation of the microstructure of modal fibers is realized.

Description

Technical Field

[0001] This application relates to the field of intelligent control, and more specifically, to an intelligent temperature control method and device for precisely regulating the microstructure of modal fibers. Background Technology

[0002] During the enzymatic finishing process of Modal fibers, cellulase removes microfibers and fibers from the fiber surface through controlled hydrolysis, altering the surface porosity, fibrillation degree, and interfacial morphology, thereby affecting the fabric's softness, smoothness, and mechanical properties. Because enzyme activity is highly sensitive to temperature changes, especially near the upper limit of the suitable low-temperature window, even a slight increase in temperature can cause a non-linear increase in the enzymatic reaction rate, leading to excessive hydrolysis of the fiber surface and even damage to the main chain. Therefore, it is necessary to develop an intelligent temperature control scheme for precisely regulating the microstructure of Modal fibers.

[0003] Existing temperature control schemes typically use the macroscopic liquid bath temperature within the tank as feedback. They adjust the hot and cold medium valves by setting a temperature threshold and employing on / off control or conventional proportional-integral-derivative (PID) control, while simultaneously controlling the degree of hydrolysis based on process parameters such as enzyme concentration and processing time. However, a dynamic mismatch exists between the macroscopic liquid bath temperature and the microscopic temperature at the fiber-mechanical friction interface. Existing schemes struggle to promptly detect transient thermal disturbances caused by liquid bath heat transfer lag, enzyme hydrolysis heat release, fabric frictional heat release, and changes in environmental heat load. Furthermore, fixed-gain and symmetrical error control do not adequately consider the nonlinear amplification effect of temperature rise on enzyme activity and the cumulative hydrolysis risk from sustained slight overheating, easily leading to temperature overshoot, frequent switching of hot and cold valves, and control lag.

[0004] Therefore, an optimized intelligent temperature control method for precisely regulating the microstructure of modal fibers is desired. Summary of the Invention

[0005] To address the aforementioned technical issues, this application provides an intelligent temperature control method and apparatus for precisely regulating the microstructure of modal fibers.

[0006] According to one aspect of this application, a smart temperature control method for precisely regulating the microstructure of modal fibers is provided, comprising: Step 1: Perform trajectory smoothing and feature fusion on the preset suitable temperature threshold for enzymes and multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate and tank thermal state sequence. Step 2: Using the extended state observer, the thermal state sequence of the tank is expanded and iteratively corrected. The unmodeled temperature change caused by the hysteresis of the liquid bath heat transfer, the heat release of enzyme hydrolysis, the heat release of fabric friction and the change of environmental heat load are equivalent to the total thermal disturbance, so as to obtain the estimated temperature of the fiber interface and the estimated total thermal disturbance. Step 3: Based on the reference temperature trajectory, perform asymmetric error allocation on the estimated temperature of the fiber interface based on the tolerance limit of biochemical enzyme activity runaway to obtain the biochemical risk temperature deviation. Step 4: Perform nonlinear state error look-ahead feedback control that integrates biochemical risk temperature deviation and reference temperature change rate to obtain basic disturbance rejection control quantity. Step 5: Based on the actuator compensation coefficient, feedforward cancellation and amplitude limiting mapping are performed on the basic disturbance rejection control quantity and the estimated total thermal disturbance to obtain the duty cycle command of the hot and cold valves. The duty cycle command of the hot and cold valves is used to adjust the flow rate of the hot and cold medium in the tank so that the interface temperature of the modal fiber is maintained within the low temperature window corresponding to the preset suitable temperature threshold of the enzyme.

[0007] According to another aspect of this application, a smart temperature control device for precisely regulating the microstructure of modal fibers is provided, comprising: The reference trajectory and thermal state generation module is used to perform trajectory smoothing and feature fusion on the preset suitable temperature threshold for enzymes and multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate and tank thermal state sequence. The total thermal disturbance estimation and interface temperature observation module is used to expand and iteratively correct the thermal state sequence of the tank using the expanded state observer. It also equates the unmodeled temperature change caused by the heat transfer lag of the liquid bath, the heat release of enzyme hydrolysis, the heat release of fabric friction and the change of environmental heat load to the total thermal disturbance, so as to obtain the estimated fiber interface temperature and the estimated total thermal disturbance. The asymmetric biochemical risk deviation allocation module is used to perform asymmetric error allocation on the estimated temperature of the fiber interface based on the reference temperature trajectory, according to the lower limit of tolerance for biochemical enzyme activity runaway, to obtain the biochemical risk temperature deviation. The nonlinear state error look-ahead feedback control module is used to perform nonlinear state error look-ahead feedback control that integrates biochemical risk temperature deviation and reference temperature change rate with biochemical constraints to obtain the basic disturbance rejection control quantity. The feedforward compensation and valve duty cycle mapping module is used to perform feedforward cancellation and amplitude limiting mapping on the basic anti-disturbance control quantity and the estimated total thermal disturbance based on the actuator compensation coefficient to obtain the cold and hot valve duty cycle command. The cold and hot valve duty cycle command is used to adjust the flow rate of the cold and hot medium in the tank so that the Modal fiber interface temperature is maintained within the low temperature window corresponding to the preset suitable enzyme temperature threshold.

[0008] Compared with existing technologies, this application provides an intelligent temperature control method and device for precisely regulating the microstructure of modal fibers. It smooths the reference temperature trajectory and integrates multi-source operating data, reconstructs the fiber interface temperature and total thermal disturbance using an expansion state observer, and then implements asymmetric error allocation, nonlinear look-ahead feedback, and disturbance feedforward compensation based on the risk of enzyme activity runaway, generating duty cycle commands for hot and cold valves. This enhances the ability to suppress heat transfer hysteresis, hydrolytic heat release, frictional heat release, and changes in environmental heat load, reduces high-temperature overshoot, accumulated hydrolytic damage, and frequent valve switching, thereby maintaining the fiber interface temperature stably within the suitable low-temperature window for enzymes, achieving precise and uniform regulation of the modal fiber microstructure. Attached Figure Description

[0009] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a flowchart of an intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application; Figure 2 This is a schematic diagram of data flow in an intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Figure 3 This is a flowchart of step three in the intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application; Figure 4 This is a block diagram of an intelligent temperature control device for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Detailed Implementation

[0011] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0012] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0013] While this application makes various references to certain modules of the systems according to embodiments of this application, any number of different modules can be used and run on user terminals and / or servers. The modules described are merely illustrative, and different aspects of the systems and methods may use different modules.

[0014] Flowcharts are used in this application to illustrate the operations performed by the system according to embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed in exact order. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from them.

[0015] The technical solution of this application proposes an intelligent temperature control method for precisely regulating the microstructure of modal fibers. Figure 1 This is a flowchart of an intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Figure 2 This is a schematic diagram of data flow in an intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Figure 1 and Figure 2 As shown, the intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application includes the following steps: S1, performing trajectory smoothing and feature fusion on a preset suitable enzyme temperature threshold and multi-source operating data to obtain a reference temperature trajectory, a reference temperature change rate, and a tank thermal state sequence; S2, using an expanded state observer, performing state expansion and iterative correction on the tank thermal state sequence, and equating the unmodeled temperature change caused by liquid bath heat transfer hysteresis, enzyme hydrolysis heat release, fabric friction heat release, and environmental heat load change to a total thermal disturbance to obtain the estimated fiber interface temperature and the estimated total thermal disturbance; S3, based on the reference temperature... The trajectory is used to perform asymmetric error allocation based on the tolerance lower limit of biochemical enzyme activity runaway to obtain the biochemical risk temperature deviation; S4, the biochemical risk temperature deviation and the reference temperature change rate are combined with nonlinear state error look-ahead feedback control based on biochemical constraints to obtain the basic disturbance rejection control quantity; S5, based on the actuator compensation coefficient, the basic disturbance rejection control quantity and the estimated total thermal disturbance are fed forward to cancel and limit the mapping to obtain the duty cycle command of the hot and cold valves, wherein the duty cycle command of the hot and cold valves is used to adjust the flow rate of the hot and cold medium in the tank so that the temperature of the modal fiber interface is maintained within the low temperature window corresponding to the preset suitable enzyme temperature threshold.

[0016] Specifically, in step S1, trajectory smoothing and feature fusion are performed on the preset suitable enzyme temperature threshold and multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate, and tank thermal state sequence. It should be understood that during the enzyme processing of modal fiber, the activity of cellulase is highly sensitive to temperature, and the process requires the fiber interface temperature to operate stably within the suitable low-temperature window for enzymes. However, the preset suitable enzyme temperature threshold, as a control setpoint, is usually a step signal or a piecewise constant signal. If directly used as a control reference input, it will cause significant initial tracking errors and overshoot in the control system, which is detrimental to the smooth transition of the temperature control system. Furthermore, the multi-source operating data, such as the macroscopic liquid bath temperature, mechanical friction torque, and the opening degree of hot and cold valves within the tank, have different sampling frequencies and dimensions. Without unified time alignment, standardization, and structured fusion, the subsequent extended state observer will be unable to effectively receive input signals with a consistent time reference and characteristic scale. Therefore, this step transforms the step temperature threshold into a continuously differentiable reference temperature trajectory and its rate of change through trajectory smoothing, and fuses the multi-source operating data to form a unified tank thermal state sequence, providing high-quality input signals for subsequent state observation and feedback control.

[0017] In specific implementation, in step 1.1, a second-order tracking differentiator is used to smooth the reference trajectory of the preset suitable enzyme temperature threshold based on the tracking differentiator, and multi-source operating data is collected to obtain the reference temperature trajectory, reference temperature change rate, and operating data set. Here, the multi-source operating data includes the macroscopic liquid bath temperature sequence, the mechanical friction torque sequence, and the hot and cold valve opening sequence, which respectively reflect the thermal environment state of the tank, the heat load generated by the friction between the fabric and the machinery, and the current opening of the hot and cold medium actuator. In this process, the preset suitable enzyme temperature threshold is set as... The second-order tracking differentiator operates in discrete form, and its state update process is as follows: First, the first state of the tracking differentiator is calculated. With input The deviation between them is then calculated using the fastest synthesis function. The control acceleration is calculated, and the first and second states are updated based on the control acceleration. Specifically, the recursive formula for the second-order tracking differentiator is expressed as: ; ; in, For the first The first state output of the tracking differentiator in each control cycle is the reference temperature trajectory. To track the second-state output of the differentiator, i.e., the rate of change of the reference temperature, To control the sampling period, To track the velocity factor, For the filter factor, This is the fastest synthesis function. After the above iterations, This is the reference temperature trajectory. This refers to the reference temperature change rate, for example, when the preset suitable temperature threshold for enzymes is... =45℃, tracking speed factor Filtering factor Control the sampling period At that time, the reference temperature trajectory after system startup Starting from an initial liquid bath temperature of 25 °C, the temperature smoothly rises to around 45 °C via an S-shaped curve after approximately 40 sampling cycles (20 s). The reference temperature change rate... The temperature rises to approximately 1.5℃ / s at the midpoint of the transition, and then continues into steady state. The decay rate is reduced to within ±0.01℃ / s. Simultaneously, multi-source operational data from the current control cycle are collected, including macroscopic liquid bath temperature sequences, mechanical friction torque sequences, and hot and cold valve opening sequences, forming an operational data set.

[0018] In step 1.2, the running data set is interpolated, aligned, and concatenated into vectors, and the reference temperature trajectory and reference temperature change rate are transmitted synchronously to obtain the reference temperature trajectory, reference temperature change rate, and synchronous running feature vector. Since the macroscopic liquid bath temperature sequence, mechanical friction torque sequence, and hot / cold valve opening sequence may have different sampling periods and sampling times, it is necessary to use a control period... To unify the time base, interpolation alignment is performed on each signal channel. Specifically, linear interpolation or zero-order hold interpolation is performed on each channel signal at the control cycle time point to align all channel data to the same time scale. After interpolation alignment, the macroscopic liquid bath temperature value, mechanical friction torque value, and hot / cold valve opening value at the same moment are concatenated into a multi-dimensional vector in a fixed order, called the synchronous operation feature vector. Let the... The macroscopic liquid bath temperature after alignment of each control cycle is: Mechanical friction torque is The opening degree of the hot and cold valves is Then the synchronous running feature vector is represented as: ; in, To synchronize the feature vectors.

[0019] In step 1.3, the synchronous operation feature vector is standardized and time-series framed, and the reference temperature trajectory and reference temperature change rate are transmitted synchronously to obtain the reference temperature trajectory, reference temperature change rate, and tank thermal state sequence. In this process, the standardization process aims to eliminate dimensional and numerical scale differences between the data from different channels, ensuring that each feature has a similar numerical range and preventing any single channel from dominating the subsequent observer calculations due to its large numerical amplitude. Specifically, each component in the synchronous operation feature vector is standardized separately; that is, each channel's operating mean is subtracted and divided by its operating standard deviation to obtain the standardized feature vector. Let the first... The average operating value of each channel is Operating standard deviation is Then the standardized eigenvector of the th The components are: ; in, For the first Synchronous operation feature vectors of each channel For the first The first period of the standardized eigenvector Each component. After standardization, the standardized feature vectors from multiple consecutive control cycles are time-series framed, i.e., with a fixed frame length. As a window, it will start from the first To the The standardized eigenvectors of each period are arranged in chronological order to form a The time series matrix, where The dimension of the feature vector (in this embodiment, =10, =3, meaning the time series matrix is ​​10×3 dimensions, covering the most recent 5-second historical window, containing standardized liquid bath temperature, friction torque, and valve opening data for 10 consecutive sampling periods. This time series matrix represents the thermal state sequence of the tank. , represented as: ; in, For the first The standardized synchronous operation feature vector for each cycle is generated. Therefore, this step ultimately outputs the reference temperature trajectory, reference temperature change rate, and tank thermal state sequence, which are used by the subsequent expansion state observer and feedback control system, respectively.

[0020] Specifically, S2 utilizes an expanded state observer to perform state expansion and iterative correction on the thermal state sequence of the tank. It equates the unmodeled temperature changes caused by liquid bath heat transfer lag, enzyme hydrolysis heat release, fabric frictional heat release, and environmental heat load changes to a total thermal disturbance, thereby obtaining the estimated fiber interface temperature and the estimated total thermal disturbance. It should be understood that in the liquid bath tank for modal fiber enzyme finishing, the actually measurable temperature signal is the macroscopic liquid bath temperature. However, the hydrolysis reaction between cellulase and fiber occurs in the microscopic region of the fiber interface. The temperature in this region is affected by multiple factors, including liquid bath heat transfer lag, enzyme hydrolysis heat release, frictional heat release between the fabric and mechanical parts, and changes in environmental heat load, resulting in a dynamic deviation from the macroscopic liquid bath temperature. Traditional temperature control systems only use the macroscopic liquid bath temperature as feedback, which cannot accurately reflect the true temperature state of the fiber interface or perceive the superposition effect of the aforementioned multi-source thermal disturbances. If the control system cannot obtain timely estimates of the fiber interface temperature and total thermal disturbance, subsequent feedback control will lack an understanding of the true state of the controlled object, making accurate feedforward compensation and disturbance rejection impossible. Therefore, this step utilizes an extended state observer to perform state expansion and iterative correction on the tank's thermal state sequence, estimating the fiber interface temperature as a first-order state, and simultaneously summing all unmodeled thermal disturbances into an extended state for online estimation. This provides estimated fiber interface temperature and estimated total thermal disturbance for subsequent asymmetric error allocation and feedforward cancellation.

[0021] In specific implementation, in step 2.1, an extended state observer is used to calculate the deviation between the macroscopic liquid bath temperature in the tank's thermal state sequence and the observed state, and to perform state iteration to obtain the tank's thermal state sequence, the observation deviation, and the first-order temperature control state value. During this process, the extended state observer uses the current measured value in the macroscopic liquid bath temperature sequence as input and calculates the deviation with the first-order state estimate inside the observer to obtain the observation deviation. Let the... The macroscopic liquid bath temperature measurement value for each control cycle is The first-order state estimate of the extended state observer is Then the observation bias The calculation process can be expressed by the formula: ; in, The first one extracted from the thermal state sequence of the tank The macroscopic liquid bath temperature measurement value for each control cycle. For the extended state observer in the 1st The first-order state estimate for each cycle is the first-order state value of the temperature control system. After obtaining the observation bias, the extended state observer uses this bias to iteratively correct the internal state. The observer adopts a second-order extended structure, where the first-order state... Tracking macroscopic liquid bath temperature dynamics, second-order state Estimate the lumped disturbance. The recursive formula for state iteration is: ; ; in, This is the updated first-order state value for temperature control. The updated expansion state estimate, For the first-order observer gain, For the second-order observer gain, It is a nonlinear function. For the power parameter of the nonlinear function, This is the linear region width parameter for the nonlinear function. In this embodiment, it is taken as the observer bandwidth. ,but This bandwidth setting allows the observer to detect frequencies lower than [a certain value]. The disturbance signal exhibits excellent tracking capability, covering the frequency range of all target disturbance sources, including liquid bath heat transfer hysteresis (time constant approximately 8.5s, corresponding to a frequency of 0.12Hz), enzyme hydrolysis heat release (continuous low-frequency disturbance, <0.05Hz), fabric frictional heat release (periodic variation with drum speed, approximately 0.5–2Hz), and environmental heat load changes (slowly varying disturbance, <0.01Hz). The nonlinear function parameters are taken as... ,in The value of the parameter is matched to the resolution of the temperature sensor (Pt100 platinum resistance four-wire system, resolution ±0.05℃), ensuring that the observer operates in the linear region within the sensor noise level to suppress high-frequency jitter, and automatically switches to a high-gain nonlinear tracking mode when the deviation exceeds the noise level. Specifically, the nonlinear function... The definition of is: ; in, For input deviation, The parameter is a power parameter (its value range is...). It has the characteristics of small error and high gain. The width of the linear region. The sign function is defined. Through the above state iteration, the observer outputs the updated first-order temperature control state value in each control cycle. and expansion state estimates Meanwhile, the thermal state sequence of the tank continues to be passed on to subsequent sub-steps.

[0022] In step 2.2, based on the preset work-heat conversion coefficient and liquid bath time delay compensation coefficient, thermal inversion reconstruction is performed on the tank thermal state sequence, observation deviation, and first-order temperature control state value to obtain the estimated fiber interface temperature. Since the fiber interface temperature cannot be directly measured, it is necessary to comprehensively utilize the first-order state estimation of the observer, the local thermal effect generated by mechanical friction torque, and the time compensation for liquid bath heat transfer lag to invert and estimate the fiber interface temperature. Specifically, the mechanical friction torque value for the current period is first extracted from the tank thermal state sequence. Using a preset work-to-heat conversion coefficient This is converted into a contribution to the local temperature rise caused by friction. Then, the liquid bath time delay compensation coefficient is used. A time-delay correction is applied to the observation bias to compensate for the heat transfer delay between the macroscopic liquid bath temperature and the fiber interface temperature. The thermal inversion reconstruction formula for the estimated fiber interface temperature is as follows: ; in, For the first Estimated fiber interface temperature for each control cycle. This is the first-order state value for temperature control. The preset work-to-heat conversion coefficient characterizes the temperature rise effect of a unit frictional torque on the fiber interface region. This represents the mechanical friction torque value extracted from the thermal state sequence of the tank during the current cycle. This is the liquid bath time lag compensation coefficient, which characterizes the proportion of compensation for the fiber interface temperature lead caused by the liquid bath heat transfer lag. This represents the observation bias.

[0023] In this embodiment, the preset work-to-heat conversion coefficient is used. This value was obtained through the following calibration experiment: Under the condition of a constant liquid bath temperature of 45℃, the roller torque was set to 0, 5, 10, 15, and 20 N·m respectively. The difference between the fiber-roller contact surface temperature and the liquid bath temperature was measured using an infrared thermal imager (temperature measurement accuracy ±0.3℃). The slope of the linear fitting of friction temperature rise and torque was found to be 0.012℃ / (N·m). Liquid bath time delay compensation coefficient Its physical meaning is as follows: the pure hysteresis time of heat transfer from the liquid bath to the fiber interface is approximately 3 s (determined through a step response experiment), and the liquid bath thermal time constant is approximately 8.5 s (derived from the liquid bath heat capacity of 3.35 × 10⁻⁶). 6 (Calculated by dividing J / ℃ by the maximum heat transfer coefficient of the hot and cold medium, 395W / ℃), then .

[0024] In step 2.3, a lumped disturbance estimation is performed on the observation bias. The liquid bath heat transfer hysteresis, enzyme hydrolysis heat release, fabric friction heat release, and changes in environmental heat load are equated to the estimated total thermal disturbance to obtain the total estimated thermal disturbance. In the design of the extended state observer, the second-order extended state... The system itself is an online estimate of the total system disturbance, unifying all unmodeled internal dynamics and external disturbances into a single total set. Specifically, the unmodeled temperature change caused by liquid bath heat transfer lag refers to the lag in macroscopic liquid bath temperature change relative to the heating or cooling signal due to the existence of liquid bath heat capacity and heat transfer path; this lag is not explicitly modeled in the simplified model. Enzyme hydrolysis exothermic reaction is the biochemical reaction heat released when cellulase catalyzes the hydrolysis of cellulose molecular chains. Fabric frictional heat release is the heat generated by friction between modal fabric and mechanical parts when moving on rollers or guide rollers. Environmental heat load change is the heat load disturbance caused by convection and radiation heat transfer between the tank and the external environment, as well as environmental temperature fluctuations. The above four types of thermal disturbances are uniformly equivalent to the total thermal disturbance in the extended state observer, and its estimated value is directly obtained from the second-order state output of the observer. ; in, For the first The estimated total thermal disturbance for each control cycle. This is the second-order state output of the extended state observer. The estimated total thermal disturbance includes the combined effects of unmodeled temperature changes caused by liquid bath heat transfer hysteresis, heat release from enzyme hydrolysis, heat release from fabric friction, and changes in environmental heat load. Instead of establishing precise mathematical models for each disturbance source separately, the total disturbance is estimated in real time through online iterative correction of the observer.

[0025] Specifically, in step S3, based on the reference temperature trajectory, an asymmetric error allocation based on the tolerance limit of biochemical enzyme activity runaway is applied to the estimated temperature of the fiber interface to obtain the biochemical risk temperature deviation. It should be understood that during the enzyme finishing process of modal fiber, the activity of cellulase exhibits a highly nonlinear response to temperature changes, especially near the upper limit of the suitable low-temperature window for enzymes. Even a small temperature increase can cause an exponential increase in the rate of enzymatic reactions, leading to excessive hydrolysis of the fiber surface or even main chain breakage. However, when the temperature is below the reference trajectory, the decrease in enzyme activity only causes a decrease in hydrolysis efficiency and does not cause irreversible damage to the fiber structure. This asymmetric effect of temperature deviation on the fiber microstructure requires the control system to not use symmetrical error weights when handling temperature deviations. Instead, a larger gain weight should be applied to the high-temperature deviation (i.e., the fiber interface temperature is higher than the reference temperature) to suppress the deviation in the heating direction with stronger control, while a smaller gain weight should be applied to the low-temperature deviation to avoid unnecessary overheating. Therefore, this step calculates the difference between the estimated temperature of the fiber interface based on the reference temperature trajectory and applies asymmetric gain correction according to the different deviation directions, so that the generated biochemical risk temperature deviation can reflect the high sensitivity of the heating direction to the risk of enzyme activity runaway, thereby driving the subsequent control links to implement a stronger suppression response to the high temperature side deviation.

[0026] Figure 3 This is a flowchart of step three in the intelligent temperature control method for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Figure 3 As shown, in the first embodiment of this application, S3 includes: S3.1, calculating the difference and determining the deviation polarity based on the reference temperature trajectory to obtain the basic temperature deviation amount and the deviation polarity state value; S3.2, performing gain correction based on the basic temperature deviation amount and the deviation polarity state value to obtain the biochemical risk temperature deviation.

[0027] Specifically, in step S3.1, based on the reference temperature trajectory, the estimated temperature at the fiber interface is calculated to determine the difference and the polarity of the deviation, thereby obtaining the basic temperature deviation and the deviation polarity state value. Specifically, firstly, the difference between the estimated fiber interface temperature and the reference temperature trajectory is calculated to obtain the basic temperature deviation. This process can be expressed by the following formula: ; in, For the first The base temperature deviation for each control cycle. For reference, the current value of the temperature trajectory, Estimate the temperature at the fiber interface. When When the estimated temperature at the fiber interface is higher than the reference temperature trajectory, it indicates a temperature rise deviation, and there is a risk of accelerated growth in cellulase activity; when When the estimated temperature at the fiber interface is lower than the reference temperature trajectory, a cooling deviation occurs, resulting in decreased enzyme activity but no fiber damage. After the difference calculation is completed, the polarity of the baseline temperature deviation is determined to identify whether the current deviation belongs to the heating or cooling direction. Specifically, the formula for determining the polarity of the deviation is as follows: ; in, For the deviation polarity state value, when This indicates that the current fiber interface temperature is higher than the reference temperature, placing it on the high-temperature side, and there is a risk of uncontrolled enzyme activity; when This indicates that the current fiber interface temperature is lower than or equal to the reference temperature, and is on the low-temperature side, with a lower risk of enzyme activity.

[0028] Specifically, in step S3.2, gain correction is performed based on the baseline temperature deviation and the deviation polarity state value to obtain the biochemical risk temperature deviation. In this process, asymmetric amplification of the baseline temperature deviation is achieved using fixed high-temperature side correction gain and fixed low-temperature side correction gain. The corresponding gain coefficient is selected according to the deviation polarity state value; a larger correction gain is applied to the high-temperature side deviation to enhance the control system's suppression of temperature rise direction deviation, while a smaller correction gain is applied to the low-temperature side deviation to avoid overheating. Specifically, the formula for calculating the biochemical risk temperature deviation is: ; in, For the first Biochemical risk temperature deviation in each control cycle To fix the gain on the high-temperature side, To fix the low-temperature side correction gain, the correction gain for heating deviation is greater than the correction gain for cooling deviation, i.e. This means that, under the constraint of the lower limit of enzyme activity runaway tolerance, the high-temperature side deviation, after being amplified, will produce a larger value of biochemical risk temperature deviation, driving the nonlinear feedback control of the subsequent step four to output a stronger cooling control quantity, thereby pulling the fiber interface temperature back below the reference trajectory more quickly; while the low-temperature side deviation, after being corrected by a small gain, has a relatively small amplitude, and the subsequent control link only uses a gentle heating force to restore the temperature to near the reference trajectory, avoiding heating overshoot that causes the temperature to exceed the high-temperature side boundary.

[0029] In this embodiment, take This value was determined by comprehensively considering the Arrhenius activity-temperature relationship of cellulase and the asymmetry of the consequences of uncontrolled enzyme activity: at a reference temperature of 45℃, the enzyme activity fold increase is approximately [value missing] when the temperature is increased by +1℃ to 46℃. When the temperature is increased by +3℃ to 48℃, the enzyme activity fold is approximately When the temperature is increased by +5℃ to 50℃, the enzyme activity ratio is approximately 1.29; while when the temperature is decreased by -1℃ to 44℃, the enzyme activity ratio is approximately 0.948, and when the temperature is decreased by -3℃ to 42℃, the enzyme activity ratio is approximately 0.85. Although the difference in the rate of change between heating and cooling is approximately 1.1 to 1.4 times based solely on the Arrhenius ratio values, the accelerated enzymatic hydrolysis caused by heating leads to irreversible breakage of the cellulose backbone and permanent loss of strength, while cooling only results in a reversible decrease in hydrolysis efficiency without fiber damage. Therefore, in controlling the gain allocation, a penalty coefficient of approximately 3 times that of the low-temperature side is applied to the high-temperature side bias. This is to reflect the irreversible nature of the heating consequences and the higher requirements for control. Specifically, when the base temperature deviates... When the temperature deviation exceeds 0.5°C on the high-temperature side, the biochemical risk temperature deviation is... ;when When the temperature is 0.5℃ lower on the low-temperature side, .

[0030] Here, when correcting the gain of the baseline temperature deviation, the biochemical risk temperature deviation is calculated based on the generated baseline temperature deviation, the deviation polarity state value, the reference temperature change rate, and the estimated total thermal disturbance. In the first embodiment of this application, S3.2 uses a fixed high-temperature side correction gain and a fixed low-temperature side correction gain to amplify the baseline temperature deviation in segments. That is, when the deviation polarity state value indicates that the fiber interface temperature is higher than the reference temperature, the fixed high-temperature side correction gain is selected; when the deviation polarity state value indicates that the fiber interface temperature is lower than or equal to the reference temperature, the fixed low-temperature side correction gain is selected, thereby generating the biochemical risk temperature deviation. Although this method achieves differentiated responses in the heating and cooling directions, its gain coefficient remains constant throughout the operation. It only performs binary switching selection based on the positive and negative polarities of the baseline temperature deviation at the current moment, failing to incorporate the forward-looking information carried by the reference temperature change rate and the estimated total thermal disturbance into the gain decision.

[0031] However, the thermal dynamics of the modal cellulosic enzyme processing process exhibit significant time-varying and nonlinear characteristics, and the fixed-gain segmented amplification strategy has the following limitations. First, fixed gain does not consider the future temperature change trend caused by the estimated total thermal disturbance. That is, when the estimated total thermal disturbance output by the expansion state observer indicates that a large amount of heat that has not been offset by the cooling system is accumulating in the tank, even if the current base temperature deviation is still small, the fiber interface temperature still has a high probability of exceeding the upper limit of the suitable low temperature window for enzymes in the subsequent control time domain, and fixed gain cannot increase the control intensity in advance based on this trend. Second, fixed gain does not consider the nonlinear amplification effect of temperature changes on cellulase activity. That is, the catalytic reaction rate of cellulase follows an Arrhenius relationship with temperature. The same magnitude of temperature deviation has drastically different effects on enzyme activity in different temperature ranges. When approaching the upper limit of the low temperature window, a small increase in temperature can cause an exponential increase in enzyme activity, but fixed gain applies the same amplification factor to all temperature ranges, which cannot reflect this nonlinear risk difference. Third, the fixed gain does not account for the cumulative hydrolytic damage to the fiber caused by slight overheating. Specifically, when the fiber interface temperature is slightly higher than the reference temperature for several consecutive control cycles but does not trigger a large deviation, the fixed gain only produces a constant, small correction, failing to identify the cumulative hydrolytic damage to the modal fiber backbone caused by continuous slight overheating. After prolonged operation, the fiber strength will still experience an irreversible decrease. Fourth, the fixed gain switching causes frequent reversals of the hot and cold valves. When the fiber interface temperature oscillates slightly around the reference temperature, the deviation polarity value rapidly alternates between the high-temperature and low-temperature sides, causing a step switch between the fixed high-temperature side correction gain and the fixed low-temperature side correction gain. This discontinuous gain jump is transmitted to the duty cycle command of the hot and cold valves through subsequent control loops, resulting in frequent opening and closing reversals of the cooling and heating valves, accelerating actuator wear and introducing additional temperature disturbances.

[0032] To address the aforementioned deficiencies, this application further proposes a second embodiment, which preferably improves the fixed high-temperature side correction gain into a high-temperature side adaptive correction gain that integrates the predicted temperature difference, enzyme activity rate, cumulative thermal risk, and disturbance growth rate. This allows the high-temperature side gain coefficient to be dynamically adjusted based on the future temperature change trend indicated by the estimated total thermal disturbance, adaptively enhanced based on the nonlinear amplification of enzyme activity mapped by the Arrhenius relation, and gradually increased in control strength based on the cumulative thermal risk of continuous overheating. Furthermore, a smooth transition between the high-temperature and low-temperature side gains is achieved through a continuous logic function to eliminate frequent valve switching caused by step switching.

[0033] Specifically, firstly, based on the estimated total thermal disturbance and equivalent heat capacity, the equivalent temperature rise rate of the disturbance is calculated. Then, by combining the baseline temperature deviation and the reference temperature change rate, the predicted end-time temperature deviation is obtained. In this process, the disturbance dimensions are first unified. That is, when the estimated total thermal disturbance output by the extended state observer is in the form of thermal power, it is converted into an equivalent temperature rise rate using the equivalent heat capacity of the liquid bath and the fabric, expressed as: ; in, For the first The disturbance equivalent temperature rise rate per control cycle To estimate the thermal power corresponding to the total thermal disturbance, The equivalent mass of the liquid bath, modal fabric, and tank involved in the heat exchange is... This is the equivalent specific heat capacity.

[0034] In this embodiment, equivalent mass The combined mass of the bath water (800 kg) and the modal fabric (50 kg) is approximately 850 kg (ignoring the small heat capacity contribution from the tank walls). The equivalent specific heat capacity is... The specific heat of the liquid bath water (4186 J / (kg·℃)) and the specific heat of the fabric (1340 J / (kg·℃)) are weighted by a mass average of approximately 4018 J / (kg·℃). Predicted time domain length This predicts the temperature trend over the next 5 seconds, covering the pure lag time (3 seconds) of heat transfer in the liquid bath and allowing a 2-second margin. For example, if the estimated total thermal disturbance corresponds to a heat power of 5000W (equivalent to approximately 471W from roller friction 15 N·m × 2π × 5 r / s plus the heat dissipation from hydrolysis and the ambient heat load, totaling 5000W), then the equivalent temperature rise rate of the disturbance is... If the current base temperature deviation (Temperature is precisely tracked), reference temperature change rate (Steady state), then predict the temperature deviation at the end of the time domain. This indicates that the fiber interface temperature will be 0.0073℃ higher than the reference temperature after 5 seconds, and the system will begin slight cooling ahead of schedule.

[0035] This completes the dimensional conversion from heat power to temperature change rate, thus unifying the exothermic reactions of enzyme hydrolysis, fabric friction, environmental heat transfer, and heat transfer hysteresis into a single representation of the temperature drift that may occur per unit time. Secondly, based on the baseline temperature deviation, the reference temperature change rate, and the disturbance equivalent temperature rise rate, the predicted temperature deviation at the end of the control time domain is expressed as: ; in, To predict the temperature deviation at the end of the time domain, The base temperature deviation, To predict the length of the time domain, For reference temperature change rate, The equivalent temperature rise rate is the disturbance. Therefore, it is determined whether the fiber interface temperature might exceed the suitable low-temperature window for enzymes before the tank temperature sensor detects significant overheating. Specifically, when... When the temperature is above the reference temperature, it indicates that the predicted fiber interface temperature will be higher than the reference temperature.

[0036] Next, the absolute temperature of the fiber interface is predicted based on the predicted time-domain end-temperature deviation, and then substituted into the Arrhenius equation to solve for the predicted enzyme activity ratio. The cumulative enzyme activity heat risk is then updated using a recursive mechanism. In this process, the absolute temperature of the fiber interface is first predicted based on the predicted time-domain end-temperature deviation, expressed as: ; in, To predict fiber interface temperature using Kelvin, For reference, the current value of the temperature trajectory, This is the conversion constant for converting Celsius temperature to Kelvin temperature. The predicted temperature is obtained by inverting the prediction bias to provide the absolute temperature input for subsequent calculations of cellulase activity based on the Arrhenius relation. Furthermore, within the low-temperature sand washing operating range where the cellulase has not yet entered a significant thermal inactivation zone, the enzyme activity ratio of the predicted temperature relative to the reference temperature is calculated using the Arrhenius relation, expressed as: ; in, To predict enzyme activity fold, This represents the apparent activation energy obtained by standardization for the current batch of cellulase. The molar gas constant, The temperature is in Celsius. To predict fiber interface temperature using Kelvin, It is a natural exponential function, in this embodiment, (The reported epiactivation energy of neutral cellulase derived from Trichodermareesei ranges from 40 to 55 kJ / mol, with a typical value of 45 kJ / mol.) .by For example, when predicting the temperature deviation at the end of the time domain... (That is, when the predicted fiber interface temperature is 47℃) Predicting enzyme activity fold When the predicted fiber interface temperature rises to 48℃, When the predicted temperature reaches 50℃, Enzyme activity increased by nearly 30%, but the risk of hydrolysis also increased significantly. By mapping linear temperature changes to exponential enzyme activity changes, it is possible to distinguish different hydrolysis risks corresponding to the same temperature difference but different temperature ranges, where... When the predicted enzyme activity is higher than that at the reference temperature, it indicates that the enzyme activity is higher than that at the reference temperature.

[0037] Furthermore, to identify the cumulative hydrolytic damage caused by continuous slight overheating, a recursive method with decay memory was used to calculate the cumulative risk of enzyme activity heat: ; in, This represents the cumulative risk of enzyme activity over the current period. This represents the cumulative risk of enzyme activity heat from the previous cycle. To control the sampling period, For risk memory decay time constant, The normalized time constant for risk accumulation. This is an index indicating the risk of enzyme activity exceeding the limit. To predict enzyme activity fold, This is a function that maximizes the value. In this embodiment, it takes... This indicates that the historical overheat risk decays to approximately 37% of its original value after about 60 seconds, meaning that the impact of an overheat event about 1 minute ago on the current risk assessment becomes insignificant. This value matches the timescale at which the modal fiber cellulose backbone is sensitive to sustained overheating (literature reports that sustained overheating for 1–3 minutes in the 45–50°C range can lead to a measurable decrease in strength). As a normalization benchmark, make The dimensionless risk index is denoted by . This indicates that the risk of enzyme activity exceeding the limit accumulates in a squared manner, i.e., slight exceeding the limit ( Contribution per cycle The risk of increased risk, and serious over-limit ( Each cycle contributes approximately 0.25 / 30 ≈ 8.3 × 10⁻⁶. -3 The risk increment varies by as much as 25 times. For example, if The temperature remained at 1.2 (meaning the fiber interface temperature was consistently about 3.3°C higher than the reference temperature) for 60 seconds. The cumulative value is approximately (The actual value is approximately 0.12 after considering attenuation). By performing non-linear time integration on the enzyme activity fold exceeding the baseline, minor overheating over multiple consecutive cycles can be accumulated into a higher risk, avoiding the failure to detect continuous hydrolytic damage to modal fibers based solely on instantaneous temperature differences.

[0038] Furthermore, the predicted end-time temperature deviation, the predicted enzyme activity fold, the cumulative enzyme activity heat risk, and the perturbation growth rate are integrated to calculate the high-temperature side adaptive correction gain, which is then smoothly switched with the low-temperature side correction gain to generate the biochemical risk temperature deviation. In this process, firstly, the normalized positive perturbation growth rate is calculated using the perturbation equivalent temperature rise rates of the current and previous periods, expressed as: ; in, The normalized positive disturbance growth rate, and These are the perturbation equivalent temperature rise rates for the current period and the previous period, respectively. To control the sampling period, This serves as the benchmark for normalizing the perturbation growth rate. By extracting the first-order positive component of the perturbation change, we can identify rapid temperature rise trends caused by sudden increases in friction, fabric agglomeration, or accelerated enzyme reactions, without imposing additional high-temperature penalties on the perturbation reduction process.

[0039] Therefore, to calculate the adaptive correction gain on the high-temperature side, a continuous logic function is first used to map the degree to which the predicted temperature approaches the upper limit of the low-temperature window to a risk range of zero to one: ; in, For continuous logic functions, For the variable to be mapped, It is a natural exponential function. This is to avoid sudden jumps in the fixed piecewise gain near the critical temperature, thereby reducing the high-frequency switching of hot and cold valves.

[0040] Then, by fusing the predicted overheat risk, enzyme activity fold, cumulative heat risk, disturbance growth rate, and current bias polarity, the adaptive correction gain on the high-temperature side is calculated as follows: ; in, For adaptive gain correction on the high-temperature side, Based on high-temperature gain, This represents the upper limit of the suitable low-temperature window for enzymes. As a smoothing scale for temperature risk, The weights are respectively the predicted temperature, enzyme activity rate, cumulative risk, disturbance growth rate, and current high temperature state. Determined by the deviation polarity state value, it is set to 1 if the current fiber interface temperature is higher than the reference temperature, and 0 otherwise. It is a saturation limiting function. and These are the lower and upper limits of the gain, respectively. In other words, the adaptive correction gain on the high-temperature side is a multi-factor dynamic gain constrained by the upper and lower limits, which makes the high-temperature suppression strength increase simultaneously with the future overheating probability, the degree of enzyme activity enhancement, the duration of overheating, and the growth rate of thermal disturbance.

[0041] In this embodiment, the values ​​of each parameter are: The logic for this weight allocation is as follows: The highest weight (0.4) is given because the predicted temperature being close to the upper limit of the low temperature window is the most direct indicator of overheating threat. Secondly (0.3), because the enzyme activity ratio reflects the degree of nonlinear amplification of the hydrolysis rate in the current temperature range; A value of 0.15 is used to reflect the cumulative effect of the duration of overheating; A value of 0.1 is used to capture dynamic early warnings of rapidly increasing disturbances; A value of 0.05 is used as an immediate indication that the temperature is currently at the high-temperature side. Gain upper limit. =5.0 ensures that the control strength on the high-temperature side is 5 / 0.8 = 6.25 times that on the low-temperature side under the most extreme conditions, which is the lower limit of the gain. =1.5 ensures that even when all risk factors are zero, the high-temperature side maintains a fundamental asymmetry of 1.5 / 0.8 ≈ 1.875 times. For example, in a specific scenario: when predicting temperature... ,but ; , ; ; ; ,but .

[0042] Furthermore, a continuous logic function is used to smoothly switch between gain correction on the low-temperature side and adaptive gain correction on the high-temperature side to generate a biochemical risk temperature bias: ; in, Temperature deviation due to biochemical risks To correct the gain on the low-temperature side, For adaptive gain correction on the high-temperature side, To predict the temperature deviation at the end of the time domain, For gain switching smoothing scale, It is a continuous logic function. In this embodiment, it is taken as... That is, when the predicted time domain end temperature deviation Within the temperature range of -0.3℃ to +0.3℃, the gain is... and A smooth and continuous transition between them. The value changes from 0.27 to 0.73 without any abrupt jumps, avoiding frequent switching of the hot and cold valves near the reference temperature. In other words, when the predicted future temperature exceeds the reference temperature, a larger negative temperature control deviation is generated in advance, driving subsequent steps to open earlier or increase the cooling medium flow rate; when the predicted temperature is lower than the reference temperature, a smaller low-temperature side gain is used to avoid overheating.

[0043] Thus, even if the current baseline temperature deviation does not yet indicate overheating, as long as the estimated total thermal perturbation shows that the prediction deviation will be in the time domain... When the internal value is turned negative, the system can output the biochemical risk temperature deviation in the cooling direction in advance. Compared with a fixed asymmetric gain, it can reduce temperature overshoot near the upper limit of the low temperature window, shorten the cumulative time of the fiber interface temperature in the high-risk range of enzyme activity, reduce the hydrolysis of the modal main chain caused by continuous slight overheating, and reduce the frequent switching of the hot and cold valves by smoothing the gain switching.

[0044] Specifically, S4 involves using nonlinear state error look-ahead feedback control that integrates biochemical risk temperature deviation and reference temperature change rate with biochemical constraints to obtain the basic disturbance rejection control quantity. This reflects the equivalent severity of the fiber interface temperature deviation from the reference trajectory in terms of enzyme activity risk. However, simply obtaining the deviation quantity is insufficient to generate effective control; it is also necessary to convert this deviation quantity into a control output signal with appropriate dynamic characteristics. In traditional linear proportional-derivative control, the proportional term linearly amplifies the current deviation, and the derivative term linearly amplifies the deviation change rate. However, linear control laws are insufficiently responsive to large deviations and over-responsive to small deviations, and lack look-ahead compensation capability for the reference temperature trajectory change trend. In the temperature control scenario of modal fiber enzyme processing, the biochemical risk temperature deviation, after asymmetric correction, already contains enzyme activity risk weight information. A nonlinear feedback control law is needed to achieve variable gain response characteristics that allow rapid suppression of large deviations and fine adjustment of small deviations. Simultaneously, the reference temperature change rate carries information about the future change trend of the temperature target trajectory. Introducing it as a look-ahead compensation component into the control law allows the control system to respond to the change direction of the target trajectory in advance, reducing tracking lag. Therefore, this step performs nonlinear proportional feedback calculation on the temperature deviation of biochemical risks, performs nonlinear trend compensation on the reference temperature change rate, and superimposes the two to form the basic disturbance rejection control quantity, providing a control reference quantity with biochemical constraint characteristics and look-ahead compensation capability for the feedforward cancellation and valve mapping in the subsequent step five.

[0045] In specific implementation, step 4.1 involves performing a proportional feedback calculation on the biochemical risk temperature deviation to obtain a nonlinear proportional term output. During this process, the biochemical risk temperature deviation is used as input, and a nonlinear function is employed... Perform variable gain proportional feedback calculations. Nonlinear function. When the absolute value of the deviation is large, the output is close to the product of the sign function and the power function, achieving an approximately linear large deviation response; when the absolute value of the deviation is small, the output is a high-gain amplification within the linear region, achieving fine adjustment. Specifically, the calculation formula for the nonlinear proportional term output is: ; in, For the first The nonlinear proportional term output for each control cycle, This is the proportional feedback gain coefficient. Temperature deviation due to biochemical risks For the power parameter of the proportional channel nonlinear function, This is the linear region width parameter for the proportional channel nonlinear function. It is a nonlinear function.

[0046] In this embodiment, take . The critical gain is obtained by tuning using the critical oscillation method: without ESO feedforward compensation, the proportional gain is gradually increased until the system exhibits constant-amplitude oscillations, and the critical gain is recorded. Based on ADRC experience . Make the deviation Within the fine adjustment region, the equivalent gain is Approximately the gain in the large deviation region It is 1.5 times faster, enabling fine-grained locking with small deviations. For example, when hour, ;when hour, .

[0047] In step 4.2, trend compensation is applied to the reference temperature change rate to obtain a nonlinear differential term output. Specifically, the reference temperature change rate is used as input, and a nonlinear function is employed... Variable gain trend compensation calculation is performed. The rate of change of reference temperature characterizes the trend of the reference temperature trajectory at the current moment. The rate of change is positive when the reference temperature is in the heating phase, close to zero when the reference temperature is in a steady state, and negative when the reference temperature is in the cooling phase. Processing the rate of change of reference temperature through a nonlinear function and using it as a look-ahead compensation component allows the control system to apply control action in the corresponding direction in advance when the reference trajectory is about to change, reducing tracking delay. The formula for calculating the output of the nonlinear differential term is: ; in, For the first The nonlinear differential term output for each control cycle, The differential compensation gain coefficient is... For reference temperature change rate, Let be the power parameter of the differential channel nonlinear function. This is the linear region width parameter of the differential channel nonlinear function. When the absolute value of the reference temperature change rate is large (i.e., the reference trajectory is in a rapid heating or cooling transition phase), the nonlinear function output provides bounded trend compensation, avoiding overshoot caused by excessive look-ahead compensation. When the absolute value of the reference temperature change rate is small (i.e., the reference trajectory is close to steady state or changing slowly), the nonlinear function provides a high-gain response in the linear region, enabling the control system to still have sensitive look-ahead compensation capability for small trend changes.

[0048] In this embodiment, take For example, when the reference temperature trajectory is in a transition phase, hour, When the reference temperature is in a steady state, hour, Differential channel taking (less than 0.75 of the proportional channel) makes the nonlinear compression of the rate of change signal stronger, and avoids overshoot caused by excessive look-ahead compensation when the reference trajectory changes rapidly.

[0049] In step 4.3, the outputs of the nonlinear proportional term and the nonlinear derivative term are superimposed to obtain the basic disturbance rejection control quantity. Specifically, the outputs of the nonlinear proportional term and the nonlinear derivative term are linearly superimposed to form the basic disturbance rejection control quantity that integrates deviation feedback and trend compensation. Specifically, the formula for superimposing the control components is: ; in, For the first The basic disturbance rejection control quantity for each control cycle. Output is a non-linear proportional term. This is the output of the nonlinear differential term. The basic disturbance rejection control quantity. It combines the functions of two control components: the nonlinear proportional term output provides an immediate feedback correction force based on the current biochemical risk temperature deviation, the magnitude and direction of which are determined by the magnitude and polarity of the deviation; the nonlinear derivative term output provides a look-forward trend compensation force based on the reference temperature change rate, the magnitude and direction of which are determined by the changing trend of the reference trajectory. When these two are superimposed, the basic disturbance rejection control quantity can effectively suppress the current deviation and respond in advance to future changes in the reference trajectory, possessing the characteristics of nonlinear state error look-forward feedback control that integrates biochemical constraints.

[0050] Specifically, in step S5, based on the actuator compensation coefficient, feedforward cancellation and amplitude limiting mapping are performed on the basic disturbance rejection control quantity and the estimated total thermal disturbance to obtain the duty cycle command for the hot and cold valves. This duty cycle command is used to adjust the flow rate of the hot and cold media in the tank, maintaining the Modal fiber interface temperature within the low-temperature window corresponding to the preset suitable enzyme temperature threshold. It should be understood that the basic disturbance rejection control quantity only reflects the output of the feedback control loop and has not yet actively cancelled the estimated total thermal disturbance in step two. If feedforward compensation is not performed on the estimated total thermal disturbance, the feedback control loop can only produce a corrective effect after the disturbance has already caused a temperature deviation, resulting in an inherent response lag. Furthermore, the dimensions and numerical ranges of the basic disturbance rejection control quantity and the estimated total thermal disturbance do not match the physical drive range of the hot and cold valves; the numerical values ​​of the control quantity need to be scaled to be converted into an acceptable duty cycle signal for the valves. Meanwhile, the actual hot and cold valves have physical opening limitations, and their duty cycles cannot exceed the range of zero to full opening. Furthermore, the cooling and heating valves are driven by different polarities of the control quantity, requiring polarity splitting to distribute the unified control signal to the corresponding cooling or heating channel. Therefore, this step performs feedforward cancellation and scaling on the basic disturbance rejection control quantity and the estimated total thermal disturbance based on the actuator compensation coefficient. Then, it utilizes the driving range of the hot and cold valves for polarity splitting and amplitude limiting mapping, ultimately generating the hot and cold valve duty cycle command. This command is used to adjust the flow rate of the hot and cold media in the tank, maintaining the Modal fiber interface temperature within the low-temperature window corresponding to the preset suitable enzyme temperature threshold.

[0051] In specific implementation, in step 5.1, based on the actuator compensation coefficient, feedforward cancellation and scaling are performed on the basic disturbance rejection control quantity and the estimated total thermal disturbance to obtain the dynamic compensation control value. Specifically, the basic disturbance rejection control quantity is fedforward-cancelled with the estimated total thermal disturbance output in step two, that is, the equivalent control contribution of the estimated total thermal disturbance is subtracted from the basic disturbance rejection control quantity, so that the control output actively cancels the estimated disturbance component in the system. Here, the core idea of ​​feedforward cancellation is: since the extended state observer has already estimated the total thermal disturbance online, a compensation quantity equal to and opposite to the disturbance is applied in advance at the control output, so that the influence of the disturbance on the controlled variable in the closed-loop system is greatly weakened, thereby reducing the burden on the feedback control loop and accelerating the disturbance suppression speed. After the feedforward cancellation is completed, the result is multiplied by the actuator compensation coefficient for scaling, converting the control quantity from the abstract temperature control domain to the valve drive domain. Specifically, the calculation formula for the dynamic compensation control value is: ; in, For the first Dynamic compensation control value for each control cycle For the compensation coefficient of the implementing agency, Basic disturbance rejection control quantity To estimate the total thermal disturbance, here, the actuator compensation coefficient... The value of is determined by the flow-temperature gain characteristics of the hot and cold valves, that is, the reciprocal of the temperature change per unit time caused by a unit change in valve opening, which is obtained through static calibration of the actuator and calculation of the tank heat capacity parameters. Feedforward cancellation term. The physical meaning is as follows: when the estimated total thermal disturbance is positive (indicating a net temperature rise disturbance in the system), the feedforward term is negative, driving the control output to shift towards cooling to actively counteract the temperature rise disturbance; when the estimated total thermal disturbance is negative (indicating a net temperature drop disturbance in the system), the feedforward term is positive, driving the control output to shift towards heating to actively counteract the temperature drop disturbance. Therefore, the dynamic compensation control value... It combines the dual functions of feedback control and feedforward compensation, and its numerical scale has been converted to a range that matches the valve drive range through the actuator compensation coefficient.

[0052] In this embodiment, the actuator compensation coefficient This value was obtained through the following calibration process: With the liquid bath temperature stable at 45℃, the cooling valve was stepped from 0% to 100% fully open, and the initial rate of temperature decrease in the liquid bath was recorded as approximately -6.7℃ / min (i.e., the maximum cooling capacity of the cooling system causes the liquid bath temperature to decrease at approximately 6.7℃ / min). The gain of the unitized rate of temperature change corresponding to the valve being fully open is then calculated. , (Here, the control quantity is normalized using per minute as the unit). Similarly, the heating rate corresponding to the full opening of the heating valve is calibrated to be approximately 5.8℃ / min. Specific calculation example: when the basic disturbance rejection control quantity... (Indicating the need for cooling), estimate the total thermal disturbance. When the contribution of the corresponding equivalent control domain is +1.2, the dynamic compensation control value This refers to 93% of the driving force in the cooling direction. After polarity splitting, the duty cycle of the cooling valve... (93% opening), heating valve (closure).

[0053] In step 5.2, the polarity of the dynamic compensation control value is split and limited using the driving range of the hot and cold valves to obtain the duty cycle command for the hot and cold valves. Specifically, the dynamic compensation control value is first split according to its polarity: when the dynamic compensation control value is negative, it indicates that the control system needs cooling to reduce the fiber interface temperature, and its absolute value is allocated to the cooling valve channel; when the dynamic compensation control value is positive, it indicates that the control system needs heating to increase the fiber interface temperature, and its value is allocated to the heating valve channel. Then, the control values ​​of each channel after splitting are limited and mapped to constrain them within the physically allowed duty cycle range of the valves, i.e., between zero (fully closed) and fully open. Specifically, the formula for calculating the duty cycle command of the cooling valve is: ; in, For the first Cooling valve duty cycle command for each control cycle This is the maximum allowable duty cycle (fully open) for the cooling valve. This is the saturation limiting function. The definition of the saturation limiting function is: ; in, Input to be limited As the lower limit, The upper limit is used. For cooling valve channels, the negative value of the dynamic compensation control value is taken. As input, when hour This indicates that cooling is required, and after limiting, the value is zero to zero. Cooling duty cycle within the range; when hour After the lower limit is applied, the cooling duty cycle becomes zero, indicating that cooling is not required. The formula for calculating the heating valve duty cycle command is: ; in, For the first The duty cycle command for the heating valve in each control cycle. This refers to the maximum permissible duty cycle (fully open) of the heating valve. When, it indicates that heating is required; after limiting, the value is obtained from zero to... Heating duty cycle within the range; when When the lower limit is applied, the heating duty cycle becomes zero, indicating that heating is not required. Through the above polarity splitting and limiting mapping, the dynamic compensation control value is converted into a cooling valve duty cycle command. and heating valve duty cycle command The combination refers to the duty cycle commands for the cooling and heating valves. Within any control cycle, the cooling and heating valves will not be simultaneously at their maximum opening, because the polarity of the dynamic compensation control value determines that the control action is only distributed to the valve channel in one direction. The duty cycle commands for the cooling and heating valves are used to adjust the flow rates of the cooling and heating media in the tank. When the opening of the cooling valve increases, the flow rate of the cooling media increases, causing the tank temperature to decrease. When the opening of the heating valve increases, the flow rate of the heating media increases, causing the tank temperature to rise. This maintains the Modal fiber interface temperature within the low-temperature window corresponding to the preset suitable enzyme temperature threshold.

[0054] To verify the effectiveness of the intelligent temperature control method described in this application, a comparative experiment of enzyme sand washing of modal fibers was conducted on a certain type of airflow dyeing machine (effective tank capacity 800 L). The experimental conditions are as follows: Fabric: 100% Modal knitted fabric, with a surface density of 180 g / m² 2 Each cylinder has a capacity of 50 kg. Enzyme: Neutral cellulase Cellusoft L, 3% owf, pH 5.5 (acetic acid / sodium acetate buffer). Set temperature: Processing time: 60 min; Hot and cold media: cooling water 12 ℃, steam 0.3 MPa.

[0055] Three sets of comparison schemes were set up: Option A (Traditional PID Control): Using the macroscopic liquid bath temperature as the feedback variable, conventional incremental PID control is employed. Symmetric error processing; Option B (First Embodiment of this Application): Employing ESO + Fixed Asymmetric Gain ( ) + nonlinear feedback + feedforward compensation; Scheme C (Second Embodiment of this Application): Based on Scheme B, an adaptive high-temperature side correction gain and smooth switching mechanism are adopted.

[0056] Each experiment was repeated three times, and the average value was taken. The results are shown in Table 1.

[0057] Table 1

[0058] Table 2

[0059] As can be seen from Tables 1 and 2: (1) In terms of temperature control accuracy, the fiber interface temperature of Scheme A has a maximum overshoot of +2.8 ℃, which has significantly exceeded the upper limit of the low temperature window of 48 ℃, and the cumulative over-temperature time accounts for 8.3% (about 300 s), resulting in excessive hydrolysis of local fibers; Scheme B reduces the overshoot to +1.1 ℃ and the over-temperature time accounts for 1.2%; Scheme C further controls the overshoot to within +0.6 ℃, and the over-temperature time accounts for only 0.15% (about 5.4 s), which basically eliminates the risk of exceeding the upper limit of the low temperature window.

[0060] (2) Regarding fiber protection, Scheme A suffered a 16.4% decrease in breaking strength and a CV value as high as 8.7% due to frequent overheating (indicating uneven hydrolysis), a weight loss rate of 5.8%, and an inter-cylinder CV of 12.5%; Scheme C maintained a strength of 94.5% (only a 5.5% decrease), a CV value of 3.8% close to the original level, a weight loss rate of 3.6%, and an inter-cylinder CV of only 3.9%, indicating that the degree of enzyme hydrolysis was uniform and controllable, and the fiber microstructure was precisely regulated.

[0061] (3) Regarding the lifespan of the actuator, the valve in Scheme A changed direction 1820 times within 60 minutes (average 0.51 Hz), while Scheme C changed it only 380 times (0.11 Hz), a reduction of 79%, effectively reducing the mechanical wear of the solenoid valve and the pneumatic valve.

[0062] (4) In terms of finishing effect, the pilling level and surface smoothness of the three schemes all meet the process requirements (≥3.5 grade). However, while maintaining the same finishing effect, scheme C achieved the same polishing effect with a lower weight loss rate (3.6% vs 5.8%) and a higher strength retention rate. This indicates that precise temperature control avoids unnecessary excessive hydrolysis and better protects the main fiber structure under the premise of achieving the same appearance finishing effect.

[0063] In addition, a special experiment on disturbance suppression was conducted: at the 30th minute of steady-state operation (45 ℃), an artificial step disturbance was injected (the drum speed was suddenly increased from 30 r / min to 60 r / min to simulate a sudden increase in friction torque), and the temperature response of the three schemes was recorded. The results are shown in Table 3.

[0064] Table 3

[0065] Scheme C starts increasing the opening of the cooling valve 0.5s after the frictional disturbance is injected (thanks to the online real-time estimation and feedforward compensation of the disturbance by ESO). The maximum temperature rise is only 0.4℃, and it recovers to the steady-state zone of ±0.3℃ within 7s, which verifies the ability of the extended state observer combined with the feedforward cancellation mechanism to quickly suppress sudden thermal disturbances.

[0066] In summary, the intelligent temperature control method for precisely regulating the microstructure of modal fibers according to the embodiments of this application is explained. It smooths the reference temperature trajectory and integrates multi-source operating data, reconstructs the fiber interface temperature and total thermal disturbance using an expansion state observer, and then implements asymmetric error allocation, nonlinear look-ahead feedback, and disturbance feedforward compensation based on the risk of enzyme activity runaway, generating duty cycle commands for hot and cold valves. This enhances the ability to suppress heat transfer hysteresis, hydrolytic heat release, frictional heat release, and changes in environmental heat load, reduces high-temperature overshoot, accumulated hydrolytic damage, and frequent valve switching, thereby maintaining the fiber interface temperature stably within the suitable low-temperature window for enzymes, achieving precise and uniform regulation of the modal fiber microstructure.

[0067] Furthermore, an intelligent temperature control device for precisely regulating the microstructure of modal fibers is also provided.

[0068] Figure 4 This is a block diagram of an intelligent temperature control device for precisely regulating the microstructure of modal fibers according to an embodiment of this application. Figure 4 As shown, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers according to an embodiment of this application includes: a reference trajectory and thermal state generation module 310, used to perform trajectory smoothing and feature fusion on a preset suitable enzyme temperature threshold and multi-source operating data to obtain a reference temperature trajectory, a reference temperature change rate, and a tank thermal state sequence; a total thermal disturbance estimation and interface temperature observation module 320, used to perform state expansion and iterative correction on the tank thermal state sequence using an expanded state observer, and to equate the unmodeled temperature change caused by liquid bath heat transfer lag, enzyme hydrolysis heat release, fabric friction heat release, and environmental heat load change as total thermal disturbance to obtain the fiber interface estimated temperature and estimated total thermal disturbance; and an asymmetric biochemical risk deviation allocation module 330, used to... Based on the reference temperature trajectory, the estimated temperature of the fiber interface is subjected to asymmetric error allocation based on the tolerance lower limit of biochemical enzyme activity runaway to obtain the biochemical risk temperature deviation; the nonlinear state error look-ahead feedback control module 340 is used to perform nonlinear state error look-ahead feedback control that integrates biochemical constraints on the biochemical risk temperature deviation and the reference temperature change rate to obtain the basic disturbance rejection control quantity; the feedforward compensation and valve duty cycle mapping module 350 is used to perform feedforward cancellation and amplitude limiting mapping on the basic disturbance rejection control quantity and the estimated total thermal disturbance based on the actuator compensation coefficient to obtain the hot and cold valve duty cycle command, wherein the hot and cold valve duty cycle command is used to adjust the flow rate of the hot and cold medium in the tank so that the modal fiber interface temperature is maintained within the low temperature window corresponding to the preset suitable enzyme temperature threshold.

[0069] As described above, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers according to the embodiments of this application can be implemented in various wireless terminals, such as servers with intelligent temperature control algorithms for precisely regulating the microstructure of modal fibers. In one possible implementation, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers according to the embodiments of this application can be integrated into the wireless terminal as a software module and / or a hardware module. For example, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers can be a software module in the operating system of the wireless terminal, or it can be an application developed for the wireless terminal; of course, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers can also be one of many hardware modules of the wireless terminal.

[0070] Alternatively, in another example, the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers and the wireless terminal can also be separate devices, and the intelligent temperature control device 300 for precisely regulating the microstructure of modal fibers can be connected to the wireless terminal via wired and / or wireless networks, and transmit interactive information in accordance with an agreed data format.

[0071] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A smart temperature control method for precisely regulating the microstructure of modal fibers, characterized in that, include: Step 1: Perform trajectory smoothing and feature fusion on the preset suitable temperature threshold for enzymes and multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate and tank thermal state sequence. Step 2: Using the extended state observer, the thermal state sequence of the tank is expanded and iteratively corrected. The unmodeled temperature change caused by the hysteresis of the liquid bath heat transfer, the heat release of enzyme hydrolysis, the heat release of fabric friction and the change of environmental heat load are equivalent to the total thermal disturbance, so as to obtain the estimated temperature of the fiber interface and the estimated total thermal disturbance. Step 3: Based on the reference temperature trajectory, perform asymmetric error allocation on the estimated temperature of the fiber interface based on the tolerance limit of biochemical enzyme activity runaway to obtain the biochemical risk temperature deviation. Step 4: Perform nonlinear state error look-ahead feedback control that integrates biochemical risk temperature deviation and reference temperature change rate to obtain basic disturbance rejection control quantity. Step 5: Based on the actuator compensation coefficient, feedforward cancellation and amplitude limiting mapping are performed on the basic disturbance rejection control quantity and the estimated total thermal disturbance to obtain the duty cycle command of the hot and cold valves. The duty cycle command of the hot and cold valves is used to adjust the flow rate of the hot and cold medium in the tank so that the interface temperature of the modal fiber is maintained within the low temperature window corresponding to the preset suitable temperature threshold of the enzyme.

2. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 1, characterized in that, The multi-source operating data includes macroscopic liquid bath temperature sequences, mechanical friction torque sequences, and hot and cold valve opening sequences.

3. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 1, characterized in that, Step one includes: Step 1.1: Using a second-order tracking differentiator, smooth the reference trajectory based on the tracking differentiator for the preset suitable temperature threshold of the enzyme, and collect multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate and operating data set. Step 1.2: Interpolate and align the running data set and concatenate the vectors, and synchronously transmit the reference temperature trajectory and reference temperature change rate to obtain the reference temperature trajectory, reference temperature change rate and synchronous running feature vector; Step 1.3: Standardize and time-series frame the synchronous operation feature vector, and synchronously transmit the reference temperature trajectory and reference temperature change rate to obtain the reference temperature trajectory, reference temperature change rate and tank thermal state sequence.

4. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 1, characterized in that, Step two includes: Step 2.1: Using the extended state observer, the deviation between the macroscopic liquid bath temperature and the observed state in the thermal state sequence of the tank is calculated and the state is iterated to obtain the thermal state sequence of the tank, the observation deviation, and the first-order state value of the temperature control. Step 2.2: Based on the preset work-heat conversion coefficient and liquid bath time delay compensation coefficient, perform thermal inversion reconstruction on the tank thermal state sequence, observation deviation and temperature control first-order state value to obtain the estimated temperature of the fiber interface. Step 2.3: Perform lumped disturbance estimation on the observation deviation, and equate the liquid bath heat transfer lag, enzyme hydrolysis heat release, fabric friction heat release and environmental heat load change to estimate the total thermal disturbance to obtain the estimated total thermal disturbance.

5. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 1, characterized in that, Step three includes: Step 3.1: Based on the reference temperature trajectory, calculate the difference and determine the polarity of the estimated temperature at the fiber interface to obtain the basic temperature deviation and the deviation polarity state value. Step 3.2: Perform gain correction based on the baseline temperature deviation and the deviation polarity state value to obtain the biochemical risk temperature deviation.

6. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 5, characterized in that, The correction gain for heating deviation is greater than the correction gain for cooling deviation.

7. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 6, characterized in that, Step four includes: Step 4.1: Perform proportional feedback calculation on the temperature deviation of biochemical risks to obtain the nonlinear proportional term output; Step 4.2: Perform trend compensation on the reference temperature change rate to obtain the nonlinear differential term output; Step 4.3: Superimpose the control components of the nonlinear proportional term output and the nonlinear derivative term output to obtain the basic disturbance rejection control quantity.

8. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 1, characterized in that, Step five includes: Step 5.1: Based on the actuator compensation coefficient, feedforward cancellation and scaling are performed on the basic disturbance rejection control quantity and the estimated total thermal disturbance to obtain the dynamic compensation control value; Step 5.2: Using the driving range of the hot and cold valves, polarity splitting and amplitude limiting mapping are performed on the dynamic compensation control value to obtain the duty cycle command of the hot and cold valves.

9. The intelligent temperature control method for precisely regulating the microstructure of modal fibers according to claim 5, characterized in that, Step 3.2 includes: Step 3.2.1: Based on the estimated total thermal disturbance and equivalent heat capacity, calculate the equivalent temperature rise rate of the disturbance, and combine the basic temperature deviation and the reference temperature change rate to calculate and obtain the predicted end-time temperature deviation. Step 3.2.2: Based on the predicted time-domain end temperature deviation, the absolute temperature of the fiber interface is inverted and predicted. The Arrhenius equation is substituted to solve for the predicted enzyme activity ratio, and the cumulative amount of enzyme activity heat risk is updated through a recursive mechanism. Step 3.2.3: The predicted end-time temperature deviation, the predicted enzyme activity ratio, the cumulative amount of enzyme activity heat risk and the perturbation growth rate are fused to calculate the high-temperature side adaptive correction gain, and the high-temperature side correction gain is smoothly switched to generate the biochemical risk temperature deviation.

10. An intelligent temperature control device for precisely regulating the microstructure of modal fibers, characterized in that, include: The reference trajectory and thermal state generation module is used to perform trajectory smoothing and feature fusion on the preset suitable temperature threshold for enzymes and multi-source operating data to obtain the reference temperature trajectory, reference temperature change rate and tank thermal state sequence. The total thermal disturbance estimation and interface temperature observation module is used to expand and iteratively correct the thermal state sequence of the tank using the expanded state observer. It also equates the unmodeled temperature change caused by the heat transfer lag of the liquid bath, the heat release of enzyme hydrolysis, the heat release of fabric friction and the change of environmental heat load to the total thermal disturbance, so as to obtain the estimated fiber interface temperature and the estimated total thermal disturbance. The asymmetric biochemical risk deviation allocation module is used to perform asymmetric error allocation on the estimated temperature of the fiber interface based on the reference temperature trajectory, according to the lower limit of tolerance for biochemical enzyme activity runaway, to obtain the biochemical risk temperature deviation. The nonlinear state error look-ahead feedback control module is used to perform nonlinear state error look-ahead feedback control that integrates biochemical risk temperature deviation and reference temperature change rate with biochemical constraints to obtain the basic disturbance rejection control quantity. The feedforward compensation and valve duty cycle mapping module is used to perform feedforward cancellation and amplitude limiting mapping on the basic anti-disturbance control quantity and the estimated total thermal disturbance based on the actuator compensation coefficient to obtain the cold and hot valve duty cycle command. The cold and hot valve duty cycle command is used to adjust the flow rate of the cold and hot medium in the tank so that the Modal fiber interface temperature is maintained within the low temperature window corresponding to the preset suitable enzyme temperature threshold.

Images