Liquid injection flow regulation system for a yarn processing apparatus

Abstract

The present application relates to the field of textile machinery and fluid control technology, in particular to a liquid injection flow regulation system of a yarn processing device; the system comprises a liquid supply pipeline, a flow sensor, a pressure sensor, a temperature sensor, a main regulation actuator, an auxiliary regulation actuator and an injection component, adopts a control architecture based on cross-time domain dynamic decoupling and adaptive compensation, collects yarn linear velocity, tension and liquid injection flow, pressure and temperature, maps the linear velocity and tension into target flow demand, and decomposes the pipeline dead time and inertia time constant into low-frequency reference components and high-frequency compensation components after identification, and the main and auxiliary actuator branches are cooperatively regulated, and prediction compensation and online parameter correction are simultaneously performed, so as to improve the response speed and stability of liquid injection regulation.

Description

Technical Field

[0001] This invention relates to the field of textile machinery and fluid control technology, specifically to a liquid jet flow rate regulation system for yarn processing equipment. Background Technology

[0002] In processes such as yarn sizing, lubrication, or finishing, it is usually necessary to control the liquid output from the spraying component within a range that matches the yarn's operating state through a liquid supply pipeline to ensure stable processing results. Existing liquid spraying adjustment methods generally rely on flow valves, pumps, or proportional actuators to adjust the spraying volume, and combine this with signals such as flow rate, pressure, or temperature to correct the output.

[0003] However, yarn linear speed, yarn tension, and operating parameters such as liquid pressure and temperature are often in a state of continuous change during the production process. Especially during startup, acceleration and deceleration, or tension fluctuations, the liquid spraying demand will change rapidly with the operating conditions. Existing adjustment methods usually cannot distinguish between stable demand and transient disturbances in a timely manner, and are prone to problems such as adjustment lag, overshoot, or slow response.

[0004] In addition, existing systems often centralize the overall flow regulation by a single actuator, making it difficult to balance low-frequency steady-state tracking and high-frequency transient compensation. This results in the flow rate fluctuation exceeding the stability threshold when the operating conditions change rapidly, affecting the consistency and stability of yarn surface treatment. At the same time, the fixed parameter control method has limited adaptability to different batches, different temperatures and different liquid circuit conditions, which can easily lead to uneven spraying or reduced regulation accuracy. Summary of the Invention

[0005] The purpose of this invention is to provide a liquid jet flow rate adjustment system for yarn processing equipment. This system addresses the technical shortcomings of existing technologies, which suffer from adjustment lag, overshoot, slow response, and uneven liquid spraying due to the difficulty in distinguishing between stable demand and transient disturbances when operating conditions change rapidly. Furthermore, this system is more effective in balancing and controlling low-frequency steady-state following and high-frequency transient compensation.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] The liquid jet flow regulation system of the yarn processing equipment includes a liquid supply pipeline, a flow sensor, a pressure sensor, a temperature sensor, a main regulating actuator, an auxiliary regulating actuator, and jet components. The system includes:

[0008] The working condition acquisition unit is used to acquire external working condition parameters including yarn linear velocity and yarn tension, and to acquire physical feedback parameters including liquid flow rate, pipeline pressure and liquid temperature through flow sensor, pressure sensor and temperature sensor.

[0009] The process requirement mapping unit is used to convert yarn linear velocity, yarn tension and its rate of change into target flow requirements through a mapping model containing mapping coefficients;

[0010] The liquid circuit dynamic identification unit is used to identify the step response of the liquid supply pipeline and obtain the dead time and inertial time constant.

[0011] The frequency band decomposition unit is used to determine the smoothing filter parameters based on the inertial time constant and decompose the target flow demand into a low-frequency reference component and a high-frequency compensation component.

[0012] The actuator coordination adjustment unit includes a main execution branch that controls the main adjustment actuator and an auxiliary execution branch that controls the auxiliary adjustment actuator; the main execution branch performs low-frequency steady-state following adjustment based on the low-frequency reference component, and the auxiliary execution branch performs high-frequency transient compensation adjustment based on the high-frequency compensation component.

[0013] The adaptive parameter update unit is used to perform online correction calculations of the mapping coefficients and smoothing filter parameters in the mapping model using a mechanism that combines working condition calibration with recursive least squares method.

[0014] Optionally, the process requirement mapping unit normalizes the yarn linear velocity and yarn tension to construct the target flow requirement. The target flow requirement is calculated by a linear summation rule of the basic flow bias term, the velocity static gain term and velocity dynamic gain term corresponding to the yarn linear velocity, and the tension static gain term and tension dynamic gain term corresponding to the yarn tension.

[0015] Optionally, the liquid path dynamic identification unit identifies the system by applying a known step flow command, determines the dead time by measuring the moment when the flow output first deviates from the baseline state, and calculates the inertial time constant characterizing the pipeline inertia by measuring the difference between the moment when the flow reaches 63.2% of the steady-state change and the dead time.

[0016] Optionally, the frequency band decomposition unit uses a first-order exponential smoothing filter algorithm to generate low-frequency reference components, wherein the smoothing coefficient of the first-order exponential smoothing filter algorithm is determined by the low-frequency equivalent time constant obtained by mapping the inertial time constant to the preset frequency division cutoff frequency.

[0017] The frequency band decomposition unit subtracts the target traffic demand from the low-frequency reference component to separate the high-frequency compensated component.

[0018] Optionally, the system also includes a cross-time domain prediction and compensation module; the cross-time domain prediction and compensation module determines the prediction step size based on the dead time, and calculates the demand compensation amount in the prediction time domain by combining the preset feedforward compensation coefficient, the rate of change of the target flow demand and the currently collected spray flow.

[0019] Optionally, the main execution branch of the actuator coordinated adjustment unit adopts a proportional-integral-derivative control strategy; the main execution branch takes the low-frequency error between the low-frequency reference component and the injection flow rate as input, and calculates the driving quantity of the main branch through proportional control, integral control and derivative control; the tuning process of the proportional, integral and derivative control coefficients in the main execution branch is to perform online optimization with the cost function composed of flow tracking error and control output stability as the optimization objective.

[0020] Optionally, the auxiliary execution branch of the actuator coordination adjustment unit performs segmented threshold adjustment output based on the high-frequency compensation component;

[0021] When the high-frequency compensation component is less than or equal to the low threshold, the auxiliary branch shielding action is output.

[0022] When the high-frequency compensation component is greater than the low threshold and less than or equal to the high threshold, the auxiliary branch driving quantity is generated by combining the high-frequency compensation component with the required compensation quantity using linear amplification compensation.

[0023] When the high-frequency compensation component is greater than the high threshold, the maximum full-amplitude drive of the auxiliary execution branch is forced to be output through the sign function.

[0024] Optionally, the low threshold is determined based on the standard deviation of the system steady-state flow error and the minimum effective flow rate variation of the process;

[0025] The high threshold is set based on the maximum flow rate change rate of the auxiliary regulating actuator within the unit prediction time domain and the preset safety factor.

[0026] Optionally, the adaptive parameter update unit updates the mapping coefficients using a mechanism combining working condition calibration and recursive least squares method; during the system initialization phase, initial parameter values ​​are determined using typical historical data from start-up, constant speed, and acceleration / deceleration conditions; and during the real-time operation phase, parameters are corrected online based on current working condition data and a set forgetting factor. The parameter estimation matrix of dimension and The system state input eigenvectors contain a constant term, normalized linear velocity, normalized tension, normalized rate of change of linear velocity, and normalized rate of change of tension. Its dimensions.

[0027] Optionally, the actuator coordination adjustment unit algebraically sums the main branch drive quantity and the auxiliary branch drive quantity to obtain the total control command, and after limiting the output upper and lower limits of the total control command with the amplitude of the saturation function, sends the final control command signal to the main adjustment actuator and the auxiliary adjustment actuator respectively.

[0028] The beneficial effects of this invention are:

[0029] 1. This invention acquires yarn linear velocity, yarn tension, liquid flow rate, pipeline pressure, and liquid temperature through a working condition acquisition unit, and maps the linear velocity, tension, and their rate of change to the target flow demand. Then, combined with liquid path dynamic identification, frequency band decomposition, and cross-time domain prediction compensation, the flow demand is split into a low-frequency reference component and a high-frequency compensation component, so that the main regulating actuator and the auxiliary regulating actuator respectively undertake the steady-state following and transient compensation tasks, thereby achieving both stable regulation and rapid response, and reducing lag, overshoot, and flow fluctuation during start-up, acceleration / deceleration, and tension fluctuations. Attached Figure Description

[0030] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0031] Figure 1 This is a schematic diagram of the liquid jet flow rate adjustment system of the yarn processing equipment in the embodiments of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Please see Figure 1 The liquid jet flow regulation system of the yarn processing equipment includes a liquid supply pipeline, a flow sensor, a pressure sensor, a temperature sensor, a main regulating actuator, an auxiliary regulating actuator, and jet components. The system includes:

[0034] The working condition acquisition unit is used to acquire external working condition parameters including yarn linear velocity and yarn tension, and to acquire physical feedback parameters including liquid flow rate, pipeline pressure and liquid temperature through flow sensor, pressure sensor and temperature sensor.

[0035] The process requirement mapping unit is used to convert yarn linear velocity, yarn tension and its rate of change into target flow requirements through a mapping model containing mapping coefficients;

[0036] The liquid circuit dynamic identification unit is used to identify the step response of the liquid supply pipeline and obtain the dead time and inertial time constant.

[0037] The frequency band decomposition unit is used to determine the smoothing filter parameters based on the inertial time constant and decompose the target flow demand into a low-frequency reference component and a high-frequency compensation component.

[0038] The actuator coordination adjustment unit includes a main execution branch that controls the main adjustment actuator and an auxiliary execution branch that controls the auxiliary adjustment actuator; the main execution branch performs low-frequency steady-state following adjustment based on the low-frequency reference component, and the auxiliary execution branch performs high-frequency transient compensation adjustment based on the high-frequency compensation component.

[0039] The adaptive parameter update unit is used to perform online correction calculations of the mapping coefficients and smoothing filter parameters in the mapping model by adopting a mechanism of working condition calibration combined with recursive least squares method;

[0040] The process requirement mapping unit normalizes the yarn linear velocity and yarn tension to construct the target flow requirement. The target flow requirement is specifically calculated by a linear summation rule that includes a basic flow bias term, a speed static gain term and a speed dynamic gain term corresponding to the yarn linear velocity, and a tension static gain term and a tension dynamic gain term corresponding to the yarn tension.

[0041] The fluid dynamic identification unit identifies the system by applying a known step flow command, determines the dead time by measuring the moment when the flow output first deviates from the baseline state, and calculates the inertial time constant characterizing the pipeline inertia by measuring the difference between the moment when the flow reaches 63.2% of the steady-state change and the dead time.

[0042] The frequency band decomposition unit uses a first-order exponential smoothing filter algorithm to generate low-frequency reference components. The smoothing coefficient of the first-order exponential smoothing filter algorithm is determined by the low-frequency equivalent time constant obtained by mapping the inertial time constant to the preset frequency division cutoff frequency. The frequency band decomposition unit subtracts the target flow demand from the low-frequency reference components to separate the high-frequency compensation components.

[0043] The system also includes a cross-time domain prediction and compensation module; the cross-time domain prediction and compensation module determines the prediction step size based on the dead time, and calculates the demand compensation amount in the prediction time domain by combining the preset feedforward compensation coefficient, the rate of change of the target flow demand and the currently collected spray flow.

[0044] This embodiment is used for liquid spraying adjustment in the sizing, lubrication or finishing process of yarn; the liquid supply pipeline connects the liquid storage end and the spraying component; the main adjustment actuator can be a variable frequency pump or a main proportional valve; the auxiliary adjustment actuator can be a fine-tuning valve, bypass pulse valve or piezoelectric micro valve with high frequency response characteristics; the flow sensor is arranged in the liquid line before the spraying component to output the liquid spraying flow signal.

[0045] The pressure sensor outputs pipeline pressure signals; the temperature sensor outputs liquid temperature signals; yarn linear speed and yarn tension are collected by the equipment's existing detection components and then sent to the controller.

[0046] The controller can be implemented by a programmable logic controller in conjunction with a computing module, or by an industrial controller. All the above functional units run in the controller in a program manner, and receive sensor data through the input interface and issue control commands to the main regulating actuator and the auxiliary regulating actuator through the output interface.

[0047] The operating condition acquisition unit first reads the yarn linear velocity, yarn tension, liquid flow rate, pipeline pressure, and liquid temperature in each system cycle;

[0048] Among them, yarn linear velocity reflects the speed at which the yarn passes through the spraying area under the current working conditions, yarn tension reflects the tension state of the yarn, spray flow rate reflects the current output of liquid from the spraying component, and pipeline pressure and liquid temperature are used to characterize whether the liquid circuit working conditions have changed. When the above data enters the controller, the process demand mapping unit uses linear velocity and tension as basic input quantities and further calculates the rate of change of the basic input quantities between adjacent system cycles.

[0049] The rate of change is obtained by difference, that is, by subtracting the normalized value of the previous period from the normalized value of the current period, and then dividing by the system period; the above processing mechanism is configured to achieve dynamic decoupling between the basic liquid application demand under uniform speed and the additional liquid application demand generated by acceleration, deceleration and tension fluctuations.

[0050] In this embodiment, the normalization process uses the speed reference value and tension reference value set by the equipment as references; the controller first compares the current yarn linear speed with the speed reference value to obtain the normalized speed, and then compares the current yarn tension with the tension reference value to obtain the normalized tension.

[0051] To avoid the direct superposition of different dimensions, the normalization process is performed by dividing the difference between the current quantity and the corresponding reference value by the corresponding reference value. When the linear velocity or tension reference value is set to a value lower than or equal to the preset lower limit due to the change of the sensor type, the controller uses 5% of the full scale of the sensor type as the replacement lower limit of the denominator to limit the noise amplification caused by the minimum value of the denominator.

[0052] Normalized linear velocity and tension are preferably limited to the range of -1.5 to +1.5. The portion exceeding this range is truncated according to the boundary value so that the dynamic terms under start-up, emergency stop and abnormal tension peaks remain within a controllable range.

[0053] Target injection flow rate requirement for the current cycle Constructed according to the linear summation rule, the specific formula is as follows:

[0054]

[0055] in, The target spray flow rate requirement for the current cycle; A discrete-time variable representing the system's periodic index;

[0056] This is the basic flow bias term; For velocity static gain;

[0057] The normalized linear velocity; For speed dynamic gain;

[0058] The rate of change of the normalized linear velocity; For tension static gain;

[0059] The tension after normalization; For tension dynamic gain; The normalized rate of change of tension;

[0060] When the yarn runs at a stable speed and the rate of change of tension is lower than the set threshold, the target flow rate requirement is mainly determined by the basic flow rate bias term and the static gain term of speed and tension.

[0061] When the yarn enters the acceleration, deceleration or tension change range, the rate of change increases accordingly. The controller will increase or decrease the target flow rate requirement before the flow rate deviates from the target, thereby reducing the uneven application of liquid caused by liquid path lag.

[0062] The liquid circuit dynamic identification unit is used to identify the dynamic characteristics of the liquid supply pipeline itself; in this embodiment, it can be executed offline during the equipment commissioning stage, or online when production stability conditions are met.

[0063] During identification, the controller applies a known step flow command to the liquid supply line and continuously reads the output of the flow sensor; the step flow command is configured to switch the setpoint of the main regulating actuator from the first steady-state level to the second steady-state level; the controller records the change process of the flow curve after the command is applied;

[0064] When the flow output deviates from the original baseline for the first time, the corresponding number of cycles is converted into dead time. When the flow rate reaches 63.2% of the final steady-state change, record the corresponding number of cycles, and then subtract the dead time. The inertial time constant is obtained. ;

[0065] The first deviation from the original baseline state is configured as follows: the judgment band is defined as the flow rate exceeding the baseline mean for two consecutive sampling periods;

[0066] The larger of three times the standard deviation of the steady-state sampling noise and the minimum resolvable flow rate corresponding to the flow sensor resolution is determined. The final steady-state change is obtained by averaging through a stable window after the step change. The stable window is preferably taken as 5 to 10 consecutive system cycles, and the flow fluctuation within the window does not exceed ±2% of the final average value.

[0067] Dead time reflects the delay required for liquid to travel through the pipeline to the measurement location; the inertial time constant reflects the slowness of response caused by the combined effects of pump, valve, pipeline volumetric elasticity, and fluid compressibility; controller saves... and This is for use in subsequent decomposition and prediction;

[0068] Frequency band decomposition unit based on inertial time constant To determine the smoothing filter parameters; the controller operates according to the preset frequency division cutoff frequency. Combined with low-frequency equivalent time constant Calculate the smoothing coefficient of first-order exponential smoothing. ,in and They correspond to each other according to a preset ratio. It can be rigorously determined according to the exponential mapping relationship of discrete-time systems, i.e., by using the calculation rules:

[0069]

[0070] in, For system period;

[0071] Frequency division cutoff frequency Preferably, the highest frequency of change that the main regulating actuator can stably track, while the auxiliary regulating actuator does not need to intervene, is pre-calibrated and consistent with... Satisfy frequency domain relationship ;

[0072] Low-frequency equivalent time constant Preferred selection The value is 0.8 to 1.5 times that of the fluid circuit body, so that the low-frequency reference component and the fluid circuit body inertia are on the same order of magnitude. If the system control target prioritizes steady-state smoothness, the upper limit of the preset range is taken; if the priority is low-frequency following speed response, the lower limit of the preset range is taken.

[0073] The smoothing coefficient obtained from this The value is limited to between 0.70 and 0.98. If it exceeds this range, the boundary value is used to avoid the low-frequency branch being too slow or the high-frequency branch being incorrectly amplified. In each system cycle, the controller updates the low-frequency reference component of the current cycle according to the following formula. :

[0074]

[0075] in, This is the low-frequency reference component of the previous cycle; The target traffic demand for the current period;

[0076] Traffic demands with a frequency of change lower than the frequency division cutoff frequency will be retained. In the process, the portion of demand exhibiting high-frequency step transient characteristics will be attenuated; the controller calculates the high-frequency compensation component using the following formula. :

[0077]

[0078] in, The high-frequency compensation component is input to the main regulating actuator to perform steady-state following control, while the low-frequency reference component is input to the auxiliary regulating actuator to perform transient compensation control. Based on the decoupled control architecture, the main regulating actuator shields high-frequency fluctuation signals, causing the system flow output to converge to the preset stable range.

[0079] This implementation also includes a cross-time domain prediction and compensation module to convert the current operating condition changes into the output demand after the hydraulic circuit lag; the controller first uses the dead time identified above. and system cycle Determine the prediction step size , usually take and The integer approximation of the ratio;

[0080] The controller calculates the required compensation amount. :

[0081]

[0082] in, This is the amount to compensate for demand. Forward compensation coefficient; For the current target traffic demand; To predict the step size; For system period; The target flow demand change rate; This represents the currently collected spray flow rate;

[0083] The target flow demand change rate introduced in this formula The calculation formula is:

[0084]

[0085] in, This represents the target traffic demand for the previous period. It does not simply replace the original control quantity, but rather represents the compensation needs that should be considered in the current cycle;

[0086] The aforementioned cross-time domain prediction compensation module is configured to perform feedforward correction of the control direction based on the predicted demand value before the yarn speed or tension changes and the liquid circuit has not reflected it in the flow output. With this feedforward correction mechanism, the given amount received by the main and auxiliary execution branches matches the target state of the liquid circuit, thereby improving the response rate of the liquid spray flow to changes in the working conditions.

[0087] In engineering implementation, the feedforward compensation coefficient The initial value is preferably set to 0.3 to 0.8 based on the offline step test calibration, and then adaptively updated according to the following state branching rules during real-time operation:

[0088] The controller calculates the average error of the flow rate relative to the target flow rate demand within each correction cycle of 10 consecutive system cycles. The steady-state tolerance band of the process is set as the target flow rate requirement. 2% and the standard deviation of the system steady-state flow error The larger of the absolute values ​​will Subtract the preferred fixed correction step size of 0.05. ;

[0089] If within two consecutive correction cycles If the value is negative and its absolute value is greater than the steady-state tolerance band, it is judged to have significant hysteresis. Add correction step size ; Corrected The amplitude is forcibly limited within the boundary of 0.3 to 0.8 to avoid the compensation term exciting the system in the opposite direction;

[0090] In a typical operating condition, the yarn transitions from a constant speed state to an accelerating state. The operating condition acquisition unit detects an increase in the rate of change of linear velocity within an adjacent system cycle, and the process demand mapping unit subsequently increases the target flow rate requirement. ;

[0091] The fluid dynamic identification unit previously saved and This allows the controller to be aware of the existing delay and inertia in the current fluid path; the frequency band decomposition unit divides the increased flow demand into slower-changing segments. and those that change rapidly ;

[0092] The cross-time domain prediction compensation module then relies on The current demand is converted into the output time domain after the liquid circuit delay, forming Based on the above time-domain conversion, the controller generates the basic given data for subsequent control within the same system cycle, and the main branch and auxiliary branch execute independent adjustment of the corresponding frequency band in parallel.

[0093] In a preferred embodiment of the present invention, the main execution branch of the actuator coordinated adjustment unit adopts a proportional-integral-derivative control strategy; the main execution branch takes the low-frequency error between the low-frequency reference component and the injection flow rate as input, and calculates the driving quantity of the main branch through proportional control, integral control and derivative control; the tuning process of the proportional, integral and derivative control coefficients in the main execution branch is to perform online optimization with the cost function composed of flow tracking error and control output stability as the optimization objective;

[0094] The auxiliary execution branch of the actuator coordination adjustment unit performs segmented threshold adjustment output based on the high-frequency compensation component;

[0095] When the high-frequency compensation component is less than or equal to the low threshold, the auxiliary branch shielding action is output.

[0096] When the high-frequency compensation component is greater than the low threshold and less than or equal to the high threshold, the auxiliary branch driving quantity is generated by combining the high-frequency compensation component with the required compensation quantity using linear amplification compensation.

[0097] When the high-frequency compensation component is greater than the high threshold, the maximum full-amplitude drive of the auxiliary execution branch is forced to be output through the sign function;

[0098] The low threshold is determined based on the standard deviation of the system steady-state flow error and the minimum effective flow change of the process; the high threshold is set based on the maximum flow change rate of the auxiliary regulating actuator in the unit prediction time domain and the preset safety factor.

[0099] This embodiment describes the specific working process of the main execution branch and the auxiliary execution branch in the actuator coordinated adjustment unit; the main execution branch corresponds to the main adjustment execution mechanism and undertakes the basic flow regulation below the set cutoff frequency; the auxiliary execution branch corresponds to the auxiliary adjustment execution mechanism and undertakes the transient compensation regulation above the set cutoff frequency; the two branches share the same controller, and the controller calculates the driving quantity of the main branch and the driving quantity of the auxiliary branch in each system cycle respectively;

[0100] The main execution branch adopts a proportional-integral-derivative control strategy; the controller first reads the low-frequency reference component obtained in the previous embodiment. and the injection flow rate of the current cycle ;

[0101] Since the main execution branch only handles low-frequency steady-state following, the controller first checks the injection flow rate. The low-frequency measurement component is extracted using the same smoothing coefficient and system period rule as the low-frequency reference component. Then, this low-frequency measurement component is compared with... Calculate low-frequency error ;

[0102] The error signal of the main execution branch is in the same low-frequency band as its control target, thereby isolating the high-frequency transient action introduced by the auxiliary branch from the disturbance of the main branch parameter calculation; the controller performs proportional, integral and derivative operations respectively; the proportional part directly forms the adjustment amount according to the current error, which is used to make immediate correction to the current deviation;

[0103] Main branch drive quantity Combining the integral state of the main branch It can be calculated using the discrete proportional-integral-differential method as follows:

[0104]

[0105] in, The driving quantity of the main branch; This represents the low-frequency error in the current period. This represents the low-frequency error from the previous cycle. The integral state quantity of the main branch; For system period; This is the proportionality coefficient; The integral coefficient; These are the differential coefficients;

[0106] When the actuator is not at the saturation boundary, or although it is at the boundary but the error direction is conducive to its exit from saturation, the controller updates the integral state quantity according to the following formula. :

[0107]

[0108] When the main regulating actuator reaches its upper or lower output limit and the error direction continues to drive it to further saturation, the current period's integral increment is temporarily frozen; the main regulating actuator receives... Then, change the pump speed or the opening of the main proportional valve to make the base flow of the liquid supply line closer to the low-frequency reference component.

[0109] By using the above integral state update rules, the integral term is prevented from being accumulated multiple times in each cycle due to repeated superposition of historical sums, so that the discrete control calculation relationship of the main branch remains consistent and the integral saturation phenomenon during start-up, emergency stop or major type change is suppressed.

[0110] While maintaining low-frequency tracking accuracy, the main execution branch optimizes the proportional, integral, and derivative coefficients online to limit the overshoot of the actuator's action amplitude.

[0111] The cost function on which the tuning is based for:

[0112]

[0113] in, This is the current statistical window length; This is a low-frequency error; This represents the change in the driving quantity of the main branch between adjacent system cycles. ; For precision weights; As a stationarity weight; For discrete-time dummy variables used for historical cycle traversal summation; in the frequency band decoupling architecture, if the cost function uses the total traffic error of the entire frequency band, the high-frequency transient compensation action and broadband disturbance of the auxiliary branch will pollute the optimization gradient of the main branch, causing the optimization direction to oscillate in a contradictory manner.

[0114] Therefore, the system employs strict alignment of low-frequency errors in the low-frequency band. As an evaluation benchmark, the coupling interference between the two branches at the parameter optimization level is cut off by combining the stability penalty of the main branch driving quantity change.

[0115] like The error term accounts for a large proportion, such as the sum of squared errors. If the value is above 70%, the controller will trigger a tentative increase in the proportional or integral coefficient; if The output variation terms of the main branch account for a large proportion, such as the penalty term. If the value is more than 50%, the controller will trigger a tentative reduction in the proportional or derivative coefficient.

[0116] With parameter iteration constraints, the main execution branch is configured to perform stable low-frequency steady-state follow-up adjustment; during online optimization, the controller does not simultaneously and unconstrainedly change all coefficients, but instead uses a fixed-step coordinate alternating optimization algorithm for small-step, phased adjustments: first maintaining... Without changing, respectively to and Add or subtract 5% of its current value as a trial parameter; if the current statistical window cost function after the trial... If the value decreases, the proposed direction will be accepted and the coefficients will be updated.

[0117] If both positive and negative directions are made If the value increases, the trial step size will be halved and continued; after the peak error within a consecutive statistical window decreases by 10% compared to before the adjustment, then the process will be switched to adjusting... Similarly, make trial corrections; the total change in the coefficients after each correction is limited to no more than 10% of the original baseline coefficient, and , , Each is limited to the stable range obtained during the debugging phase;

[0118] If a certain environmental disturbance causes If the increase exceeds 5% compared to the previous window, the correction for this and the previous statistical window will be immediately revoked and the system will be restored to the basic stability coefficient, thereby ensuring that the online tuning has a definite algorithm convergence boundary and clear backoff conditions.

[0119] The auxiliary execution branch is based on the high-frequency compensation component. Perform segmented threshold adjustment; the segmented threshold adjustment is configured so that the controller first adjusts the threshold. The absolute value is compared with the low and high thresholds, and the corresponding output mode is called based on the comparison result; the controller first determines within the current system cycle... Is it less than or equal to the low threshold? ;

[0120] If this condition is met, it indicates that the current high-frequency compensation amount is still within the noise band or below the minimum effective resolution threshold of the process, and the auxiliary execution branch output... The auxiliary regulating actuator remains inactive; by shielding low threshold fluctuations, the auxiliary regulating actuator is prevented from frequently opening and closing due to minor disturbances, thus reducing the system's spray stability.

[0121] when Greater than the low threshold And less than or equal to the high threshold At this time, the controller enters the linear amplification compensation region; at this time, the auxiliary branch drive quantity is combined with the high-frequency compensation component and the demand compensation quantity output by the cross-time domain prediction compensation module for comprehensive calculation, and the calculation formula is as follows:

[0122]

[0123] in, For auxiliary branch drive quantity; For high-frequency compensation components; This is the amount to compensate for demand. To increase the gain of the auxiliary branch; To predict feedforward gain;

[0124] This generates the final auxiliary branch linear drive output command;

[0125] By introducing the predicted demand compensation into this linear regulation process, the auxiliary branch can output feedforward compensation in advance before the flow rate deviates, thus overcoming the dynamic error caused by the hysteresis of the liquid circuit.

[0126] When the high-frequency compensation amount within this range is higher than the noise identification threshold but lower than the limit flow amplitude of the auxiliary regulating actuator, the system drives the auxiliary regulating actuator to adjust the liquid flow rate through linear amplification compensation.

[0127] The optimal auxiliary branch gain and predictive feedforward gain are obtained through calibration tests of the auxiliary regulating actuator driven separately: first, with the main branch unchanged, the driving amount of the auxiliary actuator is gradually increased, the transient flow change corresponding to the unit driving change is recorded, and then the gain that makes no obvious oscillation after single-cycle compensation is selected as the benchmark working value.

[0128] Simultaneously, the controller establishes a linkage correction mechanism for environmental disturbances. Specifically, the controller monitors the current liquid temperature in real time through a temperature sensor. When it detects that the current liquid temperature has dropped by more than a preset threshold of, for example, 5 degrees Celsius relative to the reference temperature recorded during the initial system calibration, it determines that the increased fluid viscosity has led to an increase in the flow resistance of the auxiliary actuator. At this time, the controller updates the gain using a linear temperature compensation function, specifically according to the formula:

[0129]

[0130] in, The current liquid temperature, Reference temperature, temperature compensation coefficient The value range is from 0.004 to 0.02 degrees Celsius. The specific compensation multiplier is calculated based on this value to increase the gain of the auxiliary branch and the predictive feedforward gain. This cross-physical dimension gain adaptive scaling is used to overcome the nonlinear decay of the transient response. Conversely, if the fluid temperature rises by more than the set threshold, the system will perform the corresponding gain reduction to limit the amplitude overshoot of the dynamic compensation.

[0131] when Exceeding the high threshold When this occurs, it indicates that the current operating condition change has exceeded the tracking range of the auxiliary regulating actuator under normal proportional output. The controller switches the auxiliary branch to the maximum full-amplitude drive state. The direction of the auxiliary branch drive quantity is determined by the sign function, and the polarity of the control output is determined by the sign of the high-frequency compensation component to indicate the compensation direction. Its amplitude is forcibly set to the maximum drive quantity of the auxiliary branch. If If positive, the auxiliary regulating actuator will rapidly increase the opening degree or increase the pulse duty cycle;

[0132] like When the value is negative, the auxiliary regulating actuator quickly reduces the opening or the pulse duty cycle; the above execution logic is configured such that when the yarn starts, stops suddenly, accelerates or decelerates abruptly, or the tension changes stepwise, the auxiliary regulating actuator is given priority to perform high-frequency flow compensation, and the main regulating actuator smoothly takes over the basic flow regulation.

[0133] To avoid high-frequency oscillations caused by the high-frequency compensation component crossing back and forth near the threshold, the controller preferably sets a hysteresis band to exit the fast compensation mode. That is, after entering full-amplitude drive, the hysteresis band only appears when... Fall back to no higher than The auxiliary branch exits full-amplitude drive and returns to the linear amplification compensation region only after maintaining this state for two consecutive system cycles.

[0134] low threshold and high threshold The setting in this embodiment has a clear physical basis; low threshold Based on the standard deviation of the system steady-state flow error and process minimum effective flow rate variation It is jointly determined that the larger of the two values ​​can be taken, that is, when a high-frequency compensation amount neither exceeds the steady-state error fluctuation band nor reaches the minimum change amount sufficient to affect the liquid application effect, the auxiliary execution branch does not operate;

[0135] The low threshold parameter is determined by the measurement fluctuation index of the current equipment and the effective resolution capability of the process. The preferred method is to calculate the flow error by continuously collecting no less than 50 system cycles of samples under uniform steady-state conditions. Preferably, the result is obtained by calibrating the minimum flow rate increment in the process sample that can cause a discernible difference in the applied solution;

[0136] High threshold Based on the maximum flow change rate of the auxiliary regulating actuator within the unit prediction time domain and safety factor The prediction time domain is determined comprehensively by the system period. With prediction step size Correspondence can be represented as Therefore, high thresholds can be set according to set up;

[0137] Safety factor This is used to reserve a certain margin for the auxiliary regulating actuator, usually taking a value close to 1; based on this, rules are set that when the high-frequency compensation exceeds... The system determines that ordinary linear compensation cannot meet the real-time requirements and directly switches to full-amplitude output mode;

[0138] In this embodiment, The preferred value is between 0.75 and 0.95; when the auxiliary actuator ages or the liquid viscosity increases, causing a decrease in transient capability, a smaller value in this range should be used; when the auxiliary actuator is in good condition and it is desirable to enhance the speed, a larger value in this range should be used.

[0139] Under a typical operating condition, the yarn tension suddenly increases within a short period of time. The target flow rate demand obtained by the process requirement mapping unit will increase accordingly, and the high-frequency compensation component calculated by the frequency band decomposition unit will... It will also increase; if Only slightly above the low threshold The auxiliary execution branch outputs a non-full-amplitude linear corresponding drive quantity and superimposes predictive feedforward compensation to assist the regulating actuator in performing linear range regulation.

[0140] if Continue to increase and exceed the high threshold The auxiliary execution branch forces the maximum full-amplitude drive to increase the liquid injection volume rapidly, while the main execution branch maintains low-frequency steady-state following adjustment based on the low-frequency reference component. The control system achieves high-frequency following of transient operating conditions and smooth control during steady-state operation through the independent control mechanism of the above two branches.

[0141] In a preferred embodiment of the present invention, the adaptive parameter update unit updates the mapping coefficients using a mechanism combining condition-specific calibration and recursive least squares method; during the system initialization phase, initial parameter values ​​are determined using typical historical data from startup, constant speed, and acceleration / deceleration conditions; and during the real-time operation phase, online corrections are made based on current condition data and a set forgetting factor. The parameter estimation matrix of dimension and The covariance matrix is ​​dimensional, where the system state input eigenvectors include constant terms, normalized linear velocity, normalized tension, normalized rate of change of linear velocity, and normalized rate of change of tension. Its dimensions;

[0142] The actuator coordination adjustment unit algebraically sums the main branch drive quantity and the auxiliary branch drive quantity to obtain the total control command. After limiting the output upper and lower limits of the total control command with the amplitude of the saturation function, the final control command signal is sent to the main adjustment actuator and the auxiliary adjustment actuator respectively.

[0143] This embodiment describes the implementation process of the adaptive parameter update unit and the output of the overall control command. The adaptive parameter update unit is located inside the controller. Its update objects include the mapping coefficients in the process requirement mapping unit. It can also be linked to correct the smoothing filter parameters in the frequency band decomposition unit and the control gain and threshold in the actuator coordination adjustment unit. However, this embodiment focuses on describing the update process of the mapping coefficients.

[0144] During the system initialization phase, the controller first calls the typical operating condition data saved from the previous operation of the equipment; the typical operating conditions include start-up condition, constant speed condition and acceleration / deceleration condition; the start-up condition is used to reflect the change in liquid application demand when the yarn enters normal operation from a standstill or low speed.

[0145] The constant speed condition is used to reflect the correspondence between the base flow rate and the stable linear velocity and stable tension; the acceleration and deceleration condition is used to reflect the impact of the rate of change of velocity and tension on the liquid injection demand.

[0146] The controller extracts the relationships between linear velocity, tension, injection flow rate, and adjacent system cycles from these three operating conditions through a step-by-step decoupled statistical calculation method, and processes them according to a normalized benchmark consistent with real-time operation to determine the mapping coefficients. , , , , The initial value;

[0147] The specific determination rule is as follows: Extract segments from the uniform operating condition data where both linear velocity and tension remain steady. Treat the rate of change of linear velocity and the rate of change of tension as zero. Perform a multiple linear regression with the injection flow rate as the dependent variable and linear velocity and tension as independent variables. The intercept term of the regression equation is used as the basic flow rate bias term. The initial values, linear velocity, and regression coefficients of tension are respectively used as the static velocity gain. and tension static gain The initial value;

[0148] Extracting the aforementioned components from the starting and acceleration / deceleration data. , and The static baseline flow rate is calculated, and the remaining flow fluctuation value is used as a new observed dependent variable. A multiple linear regression is then performed again with the linear velocity change rate and tension change rate as independent variables. The resulting coefficients are used as the velocity dynamic gain. and tension dynamic gain The initial value;

[0149] The normalization process maintains the same reference benchmark as the real-time operation phase, ensuring that the initial values ​​obtained from the regression of historical data are consistent with the dimensions of subsequent real-time input variables;

[0150] The above initial value determination mechanism is configured to account for the deviation in the target flow demand calculation when the convergence system enters the real-time control stage, and to ensure the decoupling characteristics of each physical parameter;

[0151] To ensure the comparability of the contributions of the three types of operating conditions to the initial values ​​of the parameters, the controller preferably selects historical segments from each type of operating condition that have a duration of no less than 20 system cycles and where the pressure and temperature signals do not show obvious abnormal drift. If the historical samples for a certain type of operating condition are insufficient, the most recent manual calibration result is used as the source of the initial values ​​for that type of operating condition.

[0152] After entering the real-time operation phase, the adaptive parameter update unit updates the mapping coefficients based on the current operating condition data when the system is in a steady-state following condition.

[0153] Here, the preferred steady-state following condition is that the linear velocity change rate and tension change rate do not exceed 10% of the corresponding allowable upper limit within 10 consecutive system cycles, neither the main regulating actuator nor the auxiliary regulating actuator is in output saturation state, and the deviation of pipeline pressure and liquid temperature from their respective most recent steady-state average values ​​does not exceed 5%.

[0154] Triggering parameter estimation updates only when the above conditions are met can reduce the erroneous writing of data from disturbance transients, sensor anomalies, or actuator saturation phases into the model.

[0155] To avoid the covariance matrix from diverging due to the lack of dynamic excitation in the recursive least squares method under pure steady-state conditions, the controller performs decoupling control based on excitation conditions on the state input vector and its update dimension.

[0156] The controller constructs a system state input feature vector based on the system state of the current period. When, it is split into a static subvector containing the constant term 1, the current normalized linear velocity, and the current normalized tension, and a dynamic subvector containing the current normalized linear velocity change rate and the current normalized tension change rate.

[0157] During the aforementioned steady-state following period, as the dynamic sub-vectors approach zero, the controller reduces the system state input vector to a form containing only static sub-vectors, and updates only the estimated values ​​of the basic flow bias term, velocity static gain, and tension static gain, while simultaneously updating the covariance matrix. Recursive calculation of the rows and columns of elements corresponding to the dynamic gain of forced freezing speed and the dynamic gain of tension;

[0158] Only in the initial stage after the system has just finished accelerating, decelerating or experiencing a sudden change in tension and re-entering the steady-state window. Within a single cycle, the complete input vector containing historical transient changes is extracted to activate the joint update of all-dimensional parameters; to avoid incorrectly using flow deviations caused by temporary disturbances such as pipeline anomalies as parameter estimation targets, the controller needs to determine the expected model correction. ;

[0159] Specifically, when the system is determined to be under the aforementioned steady-state following condition, the controller calculates the arithmetic average of the injection flow rate collected in the previous 10 consecutive cycles of the current system cycle, and directly uses the calculated average value as the expected model correction for the current cycle. This rule is used to eliminate the influence of high-frequency measurement noise and obtain the true steady-state reference flow rate.

[0160] The controller corrects the parameter estimation matrix for parameters with activated update permissions using the recursive least squares method. and the corresponding covariance matrix The standard update format for the block is as follows:

[0161]

[0162]

[0163]

[0164] in, The expected value of the model is corrected according to the aforementioned mean rule; This is the gain matrix; The forgetting factor; obtained through the above calculations. The corresponding activation update mapping coefficient estimate, and the covariance matrix The update reflects the degree of uncertainty in the current estimate;

[0165] The forgetting factor is used to balance the weights of new and old data. When the frequency of changes in operating conditions is higher than the first judgment threshold, the controller can use a forgetting factor that is biased towards the lower limit of the set range, so that the influence of recent operating conditions on the mapping coefficient is more obvious.

[0166] When the frequency of operating condition changes is lower than the second judgment threshold, the controller can use a forgetting factor that is biased towards the upper limit of the set range to make the changes in the mapping coefficients smoother; in this embodiment, Dynamically switch within the range of 0.95 to 0.995;

[0167] Specifically, the controller maintains a sliding time window with a width of 50 system cycles and calculates the standard deviation of the liquid temperature data within this window;

[0168] If the standard deviation exceeds, for example, a set temperature fluctuation threshold of 2 degrees Celsius, or the aforementioned model correction expectation... If the residual moving average relative to the prediction baseline suddenly increases and exceeds the warning limit for strong physical disturbances characterizing the system, such as fluid change or viscosity abrupt changes, the controller triggers a state machine switch, incorporating the forgetting factor. The weight of the new data is forcibly reduced to 0.95 to accelerate its convergence to the new operating condition.

[0169] Once the standard deviation falls back to within the corresponding threshold and the flow residual stabilizes again, the controller gradually increases the flow rate by a preset ramp rate of 0.005 every 50 cycles. The value has been adjusted upwards to 0.995.

[0170] After this dynamic forgetting update with strict condition triggering and slope recovery, the mapping coefficients can adapt more quickly and accurately to changes in liquid temperature, viscosity, and operating conditions in different operating ranges, thereby achieving the target flow rate requirement. Always stay close to the boundary;

[0171] In this embodiment, the overall control command is generated by the actuator coordination adjustment unit. After the controller obtains the main branch drive quantity and the auxiliary branch drive quantity in the same system cycle, it first converts the two into equivalent physical flow prediction values ​​through the flow-drive mapping coefficient of their respective actuators and then algebraically adds them to obtain the total expected output flow. On this dimension, the amplitude limit judgment is performed, and then the control command is allocated to each of the issued controllers according to the original ratio. The overall control command represents the comprehensive adjustment amount that the liquid supply system needs to achieve in the current cycle.

[0172] Since both the main regulating actuator and the auxiliary regulating actuator have permissible output ranges, the controller cannot directly issue unrestricted total control commands, so amplitude limiting is also required;

[0173] The specific procedure is as follows: If the total control command is greater than or equal to the output limit... Then take the master control command equal to If the total control command is less than or equal to the lower output limit Then take the master control command equal to If the overall control command is between the upper and lower limits, then retain the original value; this is used to generate the final control command processed by the saturation function.

[0174] After the final control command is generated, the controller sends it to the main regulating actuator and the auxiliary regulating actuator through the output interface respectively; the output configuration sent is as follows: the controller first calculates the driving quantity of the main branch. and auxiliary branch drive quantity Total algebraic sum ;

[0175] when Output limit not reached Or output lower limit When, the two branches output their original calculated values ​​respectively; when When the saturation boundary is reached or exceeded, the controller calculates the compression ratio coefficient according to the following rules. If the output exceeds the upper limit, equal Divide by ;

[0176] If it is below the lower limit of output equal Divide by The controller will respectively with the original and Multiply them to generate the restricted main branch output command and auxiliary branch output command, and send the commands to the corresponding actuators respectively;

[0177] The two branches are compressed to the allowable range according to the same scaling factor determined above. This satisfies the control logic of the total control command first performing algebraic summation and unifying the amplitude limit, while also not losing the division of labor between the main regulating actuator, which undertakes basic regulation, and the auxiliary regulating actuator, which undertakes rapid compensation. For the main regulating actuator, the controller can convert the scaled main branch control quantity into a speed command signal or a valve opening command signal.

[0178] For auxiliary control actuators, the controller can convert the scaled auxiliary branch control quantity into a fast opening / closing signal, a duty cycle signal, or a small opening adjustment signal; since the total control command has been limited by the upper and lower limits, it can prevent the actuator from exceeding the available adjustment range;

[0179] The equipment is initially in a constant speed operating condition. The adaptive parameter update unit slowly corrects the mapping coefficients based on the data of this stage to keep the target flow demand stable.

[0180] When the yarn enters an accelerated operating condition, the linear velocity and tension change rate increase, and the system state input vector of the current cycle changes accordingly. The new mapping coefficients obtained by recursive least squares calculation will enhance the role of the dynamic term, making the target flow demand more sensitive to changes in operating conditions.

[0181] Meanwhile, the main branch drive quantity and the auxiliary branch drive quantity given by the actuator coordination adjustment unit are algebraically added to form a total control command. The controller then limits the amplitude according to the upper and lower limits and sends it to the two actuators. The system gradually corrects the demand calculation relationship as the working conditions change, while ensuring that the output command is within the range that the actuators can realize.

[0182] With this implementation method, the mapping coefficient is no longer fixed, but is continuously corrected according to the operating conditions, which can reduce the demand deviation caused by changes in liquid temperature, liquid characteristics or operating rhythm; the amplitude limit is applied before the main control command is issued, which can also reduce the excessive movement of the main regulating actuator and the auxiliary regulating actuator, making the liquid injection flow regulation more stable.

Claims

1. A liquid jet flow rate regulation system for yarn processing equipment, comprising a liquid supply pipeline, a flow sensor, a pressure sensor, a temperature sensor, a main regulating actuator, an auxiliary regulating actuator, and a jetting component, characterized in that, The system includes: The working condition acquisition unit is used to acquire external working condition status quantities including yarn linear velocity and yarn tension, and to acquire physical feedback quantities including liquid spray flow rate, pipeline pressure and liquid temperature through the flow sensor, the pressure sensor and the temperature sensor. The process requirement mapping unit is used to convert the yarn linear velocity, yarn tension and its rate of change into target flow requirements through a mapping model containing mapping coefficients; The liquid circuit dynamic identification unit is used to identify the step response of the liquid supply pipeline and obtain the dead time and inertial time constant. The frequency band decomposition unit is used to determine the smoothing filter parameters based on the inertial time constant and decompose the target flow demand into a low-frequency reference component and a high-frequency compensation component. The actuator coordination adjustment unit includes a main execution branch that controls the main adjustment actuator and an auxiliary execution branch that controls the auxiliary adjustment actuator; the main execution branch performs low-frequency steady-state following adjustment based on the low-frequency reference component, and the auxiliary execution branch performs high-frequency transient compensation adjustment based on the high-frequency compensation component; An adaptive parameter update unit is used to perform online correction calculations on the mapping coefficients and smoothing filter parameters in the mapping model using a mechanism that combines working condition calibration with recursive least squares method. The fluid dynamic identification unit identifies the system by applying a known step flow command, determines the dead zone time by measuring the moment when the flow output first deviates from the baseline state, and calculates the inertial time constant characterizing the pipeline inertia by measuring the difference between the moment when the flow reaches 63.2% of the steady-state change and the dead zone time. The frequency band decomposition unit uses a first-order exponential smoothing filter algorithm to generate the low-frequency reference component, wherein the smoothing coefficient of the first-order exponential smoothing filter algorithm is determined by the low-frequency equivalent time constant obtained by mapping the inertial time constant to the preset frequency division cutoff frequency. The frequency band decomposition unit calculates the difference between the target traffic demand and the low-frequency reference component to separate the high-frequency compensation component. The auxiliary execution branch of the actuator coordination adjustment unit performs segmented threshold adjustment output based on the high-frequency compensation component; When the high-frequency compensation component is less than or equal to the low threshold, the auxiliary branch shielding action is output. When the high-frequency compensation component is greater than the low threshold and less than or equal to the high threshold, the auxiliary branch driving quantity is generated by combining the high-frequency compensation component with the required compensation quantity using linear amplification compensation. When the high-frequency compensation component is greater than the high threshold, the maximum full-amplitude drive of the auxiliary execution branch is forcibly output through the sign function; The low threshold is determined based on the standard deviation of the system steady-state flow error and the minimum effective flow rate variation of the process. The high threshold is set based on the maximum flow rate change rate of the auxiliary adjustment actuator within the unit prediction time domain and a preset safety factor.

2. The liquid jet flow rate adjustment system of the yarn processing equipment according to claim 1, characterized in that, The process requirement mapping unit normalizes the yarn linear velocity and the yarn tension to construct a target flow requirement. The target flow requirement is specifically calculated by a linear summation rule that includes a basic flow bias term, a velocity static gain term and a velocity dynamic gain term corresponding to the yarn linear velocity, and a tension static gain term and a tension dynamic gain term corresponding to the yarn tension.

3. The liquid jet flow rate adjustment system of the yarn processing equipment according to claim 1, characterized in that, The system also includes a cross-time domain prediction compensation module; The cross-time domain prediction compensation module determines the prediction step size based on the dead time, and calculates the demand compensation amount in the prediction time domain by combining the preset feedforward compensation coefficient, the rate of change of the target flow demand, and the currently collected spray flow rate.

4. The liquid jet flow rate adjustment system of the yarn processing equipment according to claim 1, characterized in that, The main execution branch of the actuator coordinated adjustment unit adopts a proportional-integral-derivative control strategy. The main execution branch takes the low-frequency error between the low-frequency reference component and the liquid injection flow rate as input, and calculates the main branch drive quantity through proportional control, integral control and derivative control. The tuning process of the proportional, integral, and derivative control coefficients in the main execution branch is performed online with the cost function consisting of flow tracking error and control output stability as the optimization objective.

5. The liquid jet flow rate adjustment system of the yarn processing equipment according to claim 4, characterized in that, The adaptive parameter update unit updates the mapping coefficients using a mechanism that combines working condition calibration with recursive least squares method. During the system initialization phase, initial parameter values ​​are determined using typical historical data from startup, constant speed, and acceleration / deceleration conditions. During real-time operation, these values ​​are corrected online based on current operating data and a predefined forgetting factor. The parameter estimation matrix of dimension and The system state input eigenvectors are covariance matrices of dimension N, where N is a constant term, normalized linear velocity, normalized tension, normalized rate of change of linear velocity, and normalized rate of change of tension.

6. The liquid jet flow rate adjustment system of the yarn processing equipment according to claim 5, characterized in that, The actuator coordination adjustment unit algebraically sums the main branch drive quantity and the auxiliary branch drive quantity to obtain the total control command. After limiting the amplitude of the saturation function of the upper and lower output limits of the total control command, the final control command signal is sent to the main adjustment actuator and the auxiliary adjustment actuator respectively.

Images