Staged self-adaptive control method for drying process of polyvinyl chloride fluidized bed

Abstract

The invention provides a staged self-adaptive control method for a drying process of a polyvinyl chloride fluidized bed, and relates to the technical field of polyvinyl chloride chemical production process control. Wherein in the constant-speed drying stage, model reference adaptive control (MRAC) is adopted, and the hot air temperature is dynamically adjusted to stabilize the drying rate; model predictive control (MPC) is adopted in the speed reduction drying stage, and accurate water content regulation and control and energy consumption optimization are achieved through predictive optimization. According to the invention, the influence of water content fluctuation of fed materials and hot air temperature interference in the PVC drying process can be effectively resisted, so that the water content of PVC is stably reduced, and the fluctuation of the water content in the drying process is reduced.

Description

Technical Field

[0001] This invention relates to the field of polyvinyl chloride (PVC) chemical production process control technology, and in particular to a staged adaptive control method for a PVC fluidized bed drying process. Background Technology

[0002] In polyvinyl chloride (PVC) production, the drying process is a core step in ensuring product quality. It requires precisely reducing the material's moisture content from an initial high value to a target range. Unstable moisture content control directly impacts the product qualification rate in downstream molding and processing. The PVC drying process exhibits significant stage-specific characteristics. In the constant-rate drying stage, surface moisture evaporates rapidly, resulting in a relatively stable drying rate dominated by hot air temperature. In the deceleration drying stage, moisture diffusion shifts from the surface to the interior of the material, gradually decreasing the drying rate and limiting it by the internal diffusion rate. The significant differences in drying kinetics between these two stages pose a challenge to the stable control of the drying process.

[0003] To achieve stable control of moisture content, the most widely used existing technology is the single PID control strategy. This strategy sets fixed proportional, integral, and derivative parameters, and adjusts the opening of the steam regulating valve according to the deviation between the detected moisture content value and the target value, thereby controlling the hot air temperature and maintaining the stability of the drying process.

[0004] However, the single PID control strategy does not fully consider the differences in kinetic characteristics between the constant-rate and falling-rate drying stages of PVC drying. It uses fixed parameters for unified control, failing to dynamically adapt to the different drying mechanisms of the two stages. In actual production, when faced with dynamic conditions such as random fluctuations in feed moisture content and periodic disturbances in hot air temperature, this strategy exhibits weak disturbance rejection capabilities. This not only leads to large fluctuations in moisture content and insufficient control accuracy but also easily results in problems such as fluctuations in drying rate during the constant-rate stage and over-drying or under-drying during the falling-rate stage, ultimately affecting product quality and causing energy waste. Summary of the Invention

[0005] This invention provides a staged adaptive control method for the fluidized bed drying process of polyvinyl chloride, which can solve the problems of lack of staged control, weak anti-disturbance and high energy consumption in existing control systems.

[0006] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions: Firstly, a phased adaptive control method for the fluidized bed drying process of polyvinyl chloride (PVC) is provided. This method includes: real-time acquisition of PVC material moisture content data, calculation of the moisture content derivative to obtain the drying rate, introduction of hysteresis judgment logic to set upper and lower limit switching thresholds, and identification of the constant-rate drying stage or the falling-rate drying stage based on the drying rate and the upper and lower limit switching thresholds; in the constant-rate drying stage, a model reference adaptive control strategy is adopted to establish a constant-rate drying dynamics model describing the relationship between the drying rate and the material moisture content and hot air temperature, designing an adaptive law based on Lyapunov stability theory, dynamically adjusting the hot air temperature control gain, and obtaining the hot air temperature control command corresponding to the constant-rate drying stage, so that the actual drying rate tracks the output of the ideal reference model; in the falling-rate drying stage, a model predictive control strategy is adopted to establish a falling-rate drying dynamics model, and the optimal hot air temperature control quantity is determined based on model prediction, online optimization solution, and rolling optimization; the hot air temperature control command for the constant-rate drying stage or the falling-rate drying stage is output to a steam regulating valve, so that the steam regulating valve dynamically adjusts the hot air temperature of the fluidized bed dryer in response to the hot air temperature control command, forming a closed-loop control.

[0007] In one possible implementation of the first aspect, hysteresis judgment logic is introduced to set upper and lower limit switching thresholds, and the constant-rate drying stage and the falling-rate drying stage of the drying process are identified based on the drying rate and the upper and lower limit switching thresholds. This includes: collecting moisture content data of polyvinyl chloride material at a fixed sampling period, calculating the derivative of the moisture content data using the first-order difference method to obtain the drying rate, and performing first-order low-pass filtering on the drying rate to suppress measurement noise and high-frequency interference; determining a baseline threshold for the drying rate based on the current ambient temperature and process experiments, setting a hysteresis bandwidth in combination with the measured noise level of the moisture content detection signal, and calculating the upper limit switching threshold and the lower limit switching threshold based on the baseline threshold for the drying rate and the hysteresis bandwidth; if the current stage is constant-rate drying, switching to the falling-rate drying stage when the filtered drying rate is ≤ the lower limit switching threshold; and switching back to the constant-rate drying stage when the current stage is falling-rate drying, and the filtered drying rate is ≥ the upper limit switching threshold.

[0008] In one possible implementation of the first aspect, the formula for calculating the upper limit switching threshold is: ; The formula for calculating the lower limit switching threshold is: ; in, The upper limit switching threshold, The lower limit switching threshold, H is the baseline threshold for drying rate, and H is the hysteresis bandwidth.

[0009] In one possible implementation of the first aspect, the specific implementation process of the model reference adaptive control strategy for the constant-rate drying stage for any control cycle includes: obtaining the actual drying rate of the current polyvinyl chloride material, simultaneously reading the ideal drying rate output by the ideal reference model, and calculating the difference between the actual drying rate and the ideal drying rate as the tracking error; updating the hot air temperature control gain for the current control cycle based on a preset discretized adaptive law formula, using the hot air temperature control gain of the previous control cycle, the current tracking error, and the set learning rate as inputs; multiplying the current tracking error by the updated hot air temperature control gain to obtain the hot air temperature adjustment amount, and then superimposing the hot air temperature adjustment amount onto the base hot air temperature to calculate the control output for the current control cycle; outputting the control output as the hot air temperature control command corresponding to the constant-rate drying stage, and the control output includes the final hot air temperature setpoint.

[0010] In one possible implementation of the first aspect, the discretization adaptive law formula is: ; in, Indicates the current control cycle. Indicates the previous control cycle. Indicates the first Control gain per control cycle The learning rate is used to adjust the speed at which the adaptive gain is adjusted. The tracking error is calculated at the current sampling time. To preset the ideal error constant, The sampling period; Hot air temperature adjustment The formula for determining it is: ; Final hot air temperature setpoint The formula for determining it is: ; in, This is the base hot air temperature setting.

[0011] In one possible implementation of the first aspect, for any control cycle, the specific implementation process of the model predictive control strategy for the falling-rate drying stage includes: acquiring the current moisture content measurement value of the polyvinyl chloride material and using it as the initial state of the system, substituting it into a pre-established falling-rate drying kinetic model; linearizing and discretizing the falling-rate drying kinetic model near the current operating point to obtain a state-space prediction model; predicting the material moisture content within a preset number of steps based on the state-space prediction model, and obtaining the predicted moisture content corresponding to each step within the preset number of steps; and using each step within the preset number of steps... The weighted sum of the errors between the predicted moisture content and the target moisture content, and the weighted sum of the deviations between the hot air temperature and the base temperature for each step, serve as the optimization objective function. Simultaneously, the upper and lower limits of the hot air temperature, the range of the hot air temperature change rate, and the reasonable moisture content range are used as constraints. The optimization problem is solved online, with the optimization objective being the minimum value of the objective function under the constraints, resulting in a hot air temperature control sequence for the future preset steps. The first control variable in the hot air temperature control sequence is selected as the optimal hot air temperature control variable to obtain the hot air temperature control command corresponding to the falling-rate drying stage.

[0012] In one possible implementation of the first aspect, the state-space prediction model is as follows: ; in, This is a state variable representing the current moisture content of the polyvinyl chloride material, and is a control input. Indicates hot air temperature, output The measured value representing the moisture content of the material. These represent the state transition matrix, input matrix, and output matrix, respectively.

[0013] In one possible implementation of the first aspect, the objective function is optimized as follows: ; in, For the first Predicted moisture content of the step The target moisture content. It is the first hot air temperature of the step, That is the base temperature. and These are weighting coefficients, used to optimize the objective function. Its minimum value needs to be found under the corresponding constraints.

[0014] In one possible implementation of the first aspect, the constraints include hot air temperature constraints, hot air temperature change rate constraints, and moisture content constraints. The upper and lower limits of hot air temperature are constrained as follows: ; Among them, T min T max These are the minimum and maximum limits for hot air temperature, respectively. The constraint on the rate of change of hot air temperature is: ; in, This represents the maximum permissible rate of change in hot air temperature. This represents the minimum permissible rate of change in hot air temperature. Moisture content constraint is: ; Among them, MR min MR is the minimum limit for the moisture content of the material. target This represents the maximum limit for the moisture content of the material.

[0015] In one possible implementation of the first aspect, a polyvinyl chloride fluidized bed drying system comprising hardware components is applied; the hardware components include a controller, a fluidized bed dryer, an online moisture content analyzer, a temperature sensor, and a steam regulating valve; wherein, the online moisture content analyzer is installed at the outlet of the fluidized bed dryer for real-time acquisition of the moisture content data of the polyvinyl chloride material; the temperature sensor is installed at the hot air inlet and inside the bed of the fluidized bed dryer for acquiring hot air temperature and material temperature data; the steam regulating valve is connected to the hot air supply pipeline of the fluidized bed dryer for adjusting the steam flow rate in response to hot air temperature control commands; the controller is used to execute the method described in any one of the first aspects.

[0016] The beneficial effects of this invention are as follows: The method provided by this invention, through the core process of stage identification, staged control, and closed-loop adjustment, firstly, automatically identifies the two stages based on the moisture content derivative and hysteresis logic, solving the problem of missing stages in traditional control; in the constant-rate stage, the MRAC strategy is adopted, which dynamically adjusts the hot air temperature control gain to make the actual drying rate track the ideal model, adapting to the kinetic characteristics where the drying rate is dominated by hot air temperature; in the deceleration drying stage, the MPC strategy is adopted, which determines the optimal temperature through prediction and rolling optimization, adapting to the characteristic where the drying rate is dominated by internal diffusion; finally, a closed-loop control is formed through a steam regulating valve, achieving seamless connection between the two-stage control strategies. This solution fundamentally solves the problem that a single control strategy cannot adapt to the kinetic differences between the two stages, effectively resisting fluctuations in feed moisture content and interference from hot air temperature, ensuring a stable decrease in PVC moisture content, and reducing unnecessary energy consumption through precise staged adjustment, achieving the triple goals of precise humidity control, interference resistance, and energy consumption optimization, meeting the needs of different application scenarios. Attached Figure Description

[0017] Figure 1A phased control overall process framework diagram of a PVC fluidized bed drying system provided in an embodiment of the present invention; Figure 2 A flowchart of a staged adaptive control method for a fluidized bed drying process of polyvinyl chloride provided in an embodiment of the present invention; Figure 3 A logic diagram for identifying the PVC fluidized bed drying stage is provided in this embodiment of the invention. Figure 4 A flowchart illustrating the implementation of a constant-rate stage model reference adaptive control for PVC fluidized bed drying, provided in this embodiment of the invention. Figure 5 This is a flowchart illustrating the implementation of a model-based predictive control method for the falling-rate drying stage of a PVC fluidized bed dryer, as provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the present invention, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. The "or" in the present invention is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A or B can represent: A alone, A and B simultaneously, and B alone. A and B can be singular or plural. Furthermore, in the description of the present invention, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items.

[0019] Furthermore, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" are not necessarily different.

[0020] In this embodiment of the invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in this embodiment of the invention should not be construed as superior or more advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner for ease of understanding.

[0021] In polyvinyl chloride (PVC) production, the drying process is a core step in ensuring product quality. It requires precisely reducing the material's moisture content from an initial high value to a target range. Unstable moisture content control directly impacts the product qualification rate in downstream molding and processing. The PVC drying process exhibits significant stage-specific characteristics. In the constant-rate drying stage, surface moisture evaporates rapidly, resulting in a relatively stable drying rate dominated by hot air temperature. In the deceleration drying stage, moisture diffusion shifts from the surface to the interior of the material, gradually decreasing the drying rate and limiting it by the internal diffusion rate. The significant differences in drying kinetics between these two stages pose a challenge to the stable control of the drying process.

[0022] To achieve stable control of moisture content, the most widely used existing technology is the single PID control strategy. This strategy sets fixed proportional, integral, and derivative parameters, and adjusts the opening of the steam regulating valve according to the deviation between the detected moisture content value and the target value, thereby controlling the hot air temperature and maintaining the stability of the drying process.

[0023] However, the single PID control strategy does not fully consider the differences in kinetic characteristics between the constant-rate and falling-rate drying stages of PVC drying. It uses fixed parameters for unified control, failing to dynamically adapt to the different drying mechanisms of the two stages. In actual production, when faced with dynamic conditions such as random fluctuations in feed moisture content and periodic disturbances in hot air temperature, this strategy exhibits weak disturbance rejection capabilities. This not only leads to large fluctuations in moisture content and insufficient control accuracy but also easily results in problems such as fluctuations in drying rate during the constant-rate stage and over-drying or under-drying during the falling-rate stage, ultimately affecting product quality and causing energy waste.

[0024] In view of this, embodiments of the present invention provide a staged adaptive control method for the fluidized bed drying process of polyvinyl chloride (PVC). The method includes: real-time acquisition of PVC material moisture content data, calculation of the moisture content derivative to obtain the drying rate, introduction of hysteresis judgment logic to set upper and lower limit switching thresholds, and identification of the constant-rate drying stage or the falling-rate drying stage of the drying process based on the drying rate and the upper and lower limit switching thresholds; in the constant-rate drying stage, a model reference adaptive control strategy is adopted to establish a constant-rate drying dynamics model describing the relationship between the drying rate and the material moisture content and hot air temperature, design an adaptive law based on Lyapunov stability theory, dynamically adjust the hot air temperature control gain, and obtain the hot air temperature control command corresponding to the constant-rate drying stage so that the actual drying rate tracks the output of the ideal reference model; in the falling-rate drying stage, a model predictive control strategy is adopted to establish a falling-rate drying dynamics model, and determine the optimal hot air temperature control quantity based on model prediction, online optimization solution and rolling optimization; the hot air temperature control command of the constant-rate drying stage or the falling-rate drying stage is output to the steam regulating valve so that the steam regulating valve dynamically adjusts the hot air temperature of the fluidized bed dryer in response to the hot air temperature control command, forming a closed-loop control.

[0025] The method provided by this invention, through a core process of stage identification, staged control, and closed-loop adjustment, firstly automatically identifies the two stages based on the moisture content derivative and hysteresis logic, solving the problem of missing stages in traditional control. In the constant-rate drying stage, a MRAC strategy is adopted, dynamically adjusting the hot air temperature control gain to ensure the actual drying rate tracks the ideal model, adapting to the kinetic characteristics where the drying rate is dominated by hot air temperature. In the deceleration drying stage, an MPC strategy is employed, determining the optimal temperature through prediction and rolling optimization, adapting to the characteristic where the drying rate is dominated by internal diffusion. Finally, a closed-loop control is formed through a steam regulating valve, achieving seamless integration of the two-stage control strategies. This solution fundamentally solves the problem that a single control strategy cannot adapt to the kinetic differences between the two stages, effectively resisting fluctuations in feed moisture content and interference from hot air temperature, ensuring a stable decrease in PVC moisture content. Simultaneously, precise staged adjustment reduces unnecessary energy consumption, achieving the triple goals of precise humidity control, disturbance rejection, and energy optimization, meeting the needs of different application scenarios.

[0026] In some embodiments, the staged adaptive control method for a fluidized bed drying process of polyvinyl chloride (PVC) provided by this invention is applied to a PVC fluidized bed drying system including hardware components. The hardware components include a controller, a fluidized bed dryer, an online moisture content analyzer, a temperature sensor, and a steam regulating valve. The online moisture content analyzer is installed at the outlet of the fluidized bed dryer to collect real-time moisture content data of the PVC material. The temperature sensor is installed at the hot air inlet and inside the bed of the fluidized bed dryer to collect hot air temperature and material temperature data. The steam regulating valve is connected to the hot air supply pipeline of the fluidized bed dryer to adjust the steam flow rate in response to hot air temperature control commands. The controller is used to execute the method of any one of the following embodiments.

[0027] In one possible implementation, the polyvinyl chloride fluidized bed drying system includes: Main drying equipment: fluidized bed dryer. Sensor system: online moisture content analyzer, temperature sensor, flow meter. Actuators: steam regulating valve, feed screw speed-regulating motor. Control system: programmable logic controller (PLC).

[0028] The aforementioned hardware platform provides the necessary physical foundation for the implementation of the control algorithm. Among the hardware components, an online moisture content detector is installed at the dryer outlet to measure the resin moisture content of the dried PVC in real time; a temperature sensor is used to collect the hot air inlet temperature and the bed material temperature; a flow meter is used to detect the feed rate; a steam regulating valve is used to regulate the steam flow rate to control the hot air temperature; a speed-regulating motor is used to control the material feed rate, although in this invention, the main control variable is the hot air temperature, and the feed rate is usually kept constant; and a PLC controller is used to run the control algorithm of this invention.

[0029] In one example, see Figure 1, Figure 1 This invention provides a phased control overall process framework diagram for a PVC fluidized bed drying system. First, the moisture content data of the PVC material is collected in real time using an online moisture content detector and transmitted to a drying stage identification module. This module automatically identifies whether the current drying process is in a constant-rate drying stage or a decreasing-rate drying stage by calculating the moisture content derivative (drying rate), performing first-order low-pass filtering for noise reduction, and comparing hysteresis thresholds, and outputs a stage identification signal. If it is in a constant-rate stage, the Model Reference Adaptive Control (MRAC) module is triggered. This module dynamically adjusts the hot air temperature control gain based on the constant-rate stage drying kinetic model and Lyapunov stability theory, generating a constant-rate stage hot air temperature control command. If it is in a decreasing-rate stage... In the deceleration phase, the Model Predictive Control (MPC) module is triggered. This module determines the optimal hot air temperature control quantity for the deceleration phase by linearizing and discretizing the dynamic model of the deceleration phase, predicting the future moisture content, optimizing the weighted objective function, and performing rolling optimization. Finally, the control commands for both phases are output to the steam regulating valve. The steam regulating valve responds to the commands and dynamically adjusts the hot air temperature of the fluidized bed dryer. At the same time, the temperature sensor collects the hot air temperature in real time, and the online moisture content detector continuously collects the material moisture content, forming a data feedback link to complete the entire phased adaptive closed-loop control, achieving precise and stable control of the PVC material moisture content and energy consumption optimization.

[0030] The following description, in conjunction with the accompanying drawings, illustrates a staged adaptive control method for a fluidized bed drying process of polyvinyl chloride provided by an embodiment of the present invention.

[0031] Figure 2 A flowchart illustrating a staged adaptive control method for a fluidized bed drying process of polyvinyl chloride (PVC) provided in this embodiment of the invention. The method may include the following steps: S1. Real-time acquisition of moisture content data of polyvinyl chloride material, calculation of moisture content derivative to obtain drying rate, introduction of hysteresis judgment logic to set upper and lower limit switching thresholds, and identification of constant-rate drying stage or falling-rate drying stage of drying process based on drying rate and upper and lower limit switching thresholds.

[0032] In one possible implementation, hysteresis judgment logic is introduced to set upper and lower limit switching thresholds, and the constant-rate drying stage and the falling-rate drying stage of the drying process are identified based on the drying rate and the upper and lower limit switching thresholds, including: Moisture content data of polyvinyl chloride material is collected at a fixed sampling period. The derivative of the moisture content data is calculated using the first-order difference method to obtain the drying rate. The drying rate is then processed by a first-order low-pass filter to suppress measurement noise and high-frequency interference. The drying rate baseline threshold is determined based on the current ambient temperature and process test results. The hysteresis bandwidth is set by combining the measured noise level of the moisture content detection signal. The upper limit switching threshold and the lower limit switching threshold are calculated based on the drying rate baseline threshold and the hysteresis bandwidth. If the current drying stage is constant speed, switch to the deceleration drying stage when the filtered drying rate is less than or equal to the lower limit switching threshold; if the current drying stage is deceleration, switch back to the constant speed drying stage when the filtered drying rate is greater than or equal to the upper limit switching threshold.

[0033] Furthermore, the formula for calculating the upper limit switching threshold is as follows: ; The formula for calculating the lower limit switching threshold is: ; in, The upper limit switching threshold, The lower limit switching threshold, H is the baseline threshold for drying rate, and H is the hysteresis bandwidth.

[0034] For a better understanding of this solution, please refer to [link / reference]. Figure 3 The S1 provided in this embodiment of the invention specifically includes: Signal acquisition and preprocessing. This step involves using an online moisture content analyzer to read the real-time moisture content signal of the PVC material at a fixed sampling period. Data acquisition was completed. Subsequently, the real-time signals acquired by the online moisture content analyzer were analyzed. Pretreatment: First, the instantaneous change rate of PVC moisture content, i.e., the drying rate, is calculated using the first-order difference method. Subsequently, the obtained drying rate signal is subjected to first-order low-pass filtering to suppress measurement noise and high-frequency interference, providing a stable and reliable data foundation for subsequent stage identification.

[0035] Determine the PVC drying stage based on real-time PVC moisture content. With moisture content derivative The threshold judgment logic is designed by comparing the absolute values ​​of the derivatives of the moisture content. Based on the threshold value, the system automatically identifies the "constant-rate drying stage" and the "falling-rate drying stage," and outputs a stage identification signal. When the absolute value of the derivative of moisture content When the output is greater than or equal to a preset threshold, it represents the constant-rate stage of the drying process. When the absolute value of the derivative of moisture content When the output is less than the preset threshold, it enters the slow-down drying stage, and the output... The above process provides trigger signals for phased control.

[0036] Hysteresis detection logic is introduced. To eliminate frequent stage switching caused by signal fluctuations, this invention introduces a hysteresis range for anti-jitter processing. Specifically, a hysteresis value is set. This establishes upper and lower limit switching thresholds. When the system is in a constant-speed phase, only when... The system switches to the slow-down drying stage only when the temperature falls below the lower switching threshold; when the system is in the slow-down drying stage, it only switches to the slow-down drying stage when the temperature falls below the lower switching threshold. It only switches back to the constant speed stage when the speed exceeds the upper limit switching threshold.

[0037] This can also be understood as follows: In one example, S1 specifically includes the following steps: S11. Data Acquisition and Preprocessing. The specific process for this step is as follows: a1. Collect the moisture content of PVC material. The sensor reads the real-time moisture content of the PVC material from the online moisture content analyzer at a fixed sampling period. .

[0038] a2. Calculate the derivative of the material's moisture content (drying rate). Use the first-order difference method to calculate the instantaneous rate of change of moisture content. The calculation formula is as follows: In the formula, , Representing respectively in the and the The moisture content value of PVC material measured at each sampling time. The sampling period.

[0039] a3. First-order low-pass filtering. Since the difference calculation in step a2 amplifies high-frequency noise in the signal, it is necessary to adjust the derivative of the moisture content. A first-order low-pass filter is applied. The cutoff frequency of the filter is determined based on the noise spectrum measured on-site. Its purpose is to filter out high-frequency noise and obtain a smooth drying rate signal that accurately reflects the drying process for subsequent judgment.

[0040] S12. Hysteresis Comparison and Stage Judgment. Setting a drying rate threshold. With a hysteresis bandwidth The drying rate threshold is related to the current ambient temperature and can be determined through preliminary process experiments; hysteresis bandwidth. The selection should be based on the drying rate signal. The measured noise level ensures The value is greater than its peak fluctuation range; at the same time, it is necessary to combine process characteristics to achieve a balance between preventing switching delays and avoiding sluggish response, and finally determine the final fine-tuning through online operation and debugging.

[0041] When the PVC fluidized bed drying system is started, according to and , The relationship determines the initial stage; if the current stage is the constant-rate drying stage of the drying process, then the signal is identified. At that time, only when The process will only switch to the falling-rate drying stage when certain conditions are met; if the current process is in the falling-rate drying stage, an indicator signal will be displayed. At that time, only when Only when the conditions are met will it return to the constant-rate drying stage. In short, as long as the fluctuation of the drying rate is limited to the hysteresis range... Within this process, the drying stage remains unchanged. This effectively filters out signal noise and minor disturbances, preventing the drying stage indicator from oscillating back and forth between "1" and "2" at high frequencies, thus ensuring the stable and smooth operation of the subsequent MRAC and MPC controllers.

[0042] For a better understanding of this solution, please refer to [link / reference]. Figure 3 The S1 provided in this embodiment of the invention specifically includes: Signal acquisition and preprocessing. This step involves using an online moisture content analyzer to read the real-time moisture content signal of the PVC material at a fixed sampling period. Data acquisition was completed. Subsequently, the real-time signals acquired by the online moisture content analyzer were analyzed. Pretreatment: First, the instantaneous change rate of PVC moisture content, i.e., the drying rate, is calculated using the first-order difference method. Subsequently, the obtained drying rate signal is subjected to first-order low-pass filtering to suppress measurement noise and high-frequency interference, providing a stable and reliable data foundation for subsequent stage identification.

[0043] Determine the PVC drying stage based on real-time PVC moisture content. With moisture content derivative The threshold judgment logic is designed by comparing the absolute values ​​of the derivatives of the moisture content. Based on the threshold value, the system automatically identifies the "constant-rate drying stage" and the "falling-rate drying stage," and outputs a stage identification signal. When the absolute value of the derivative of moisture content When the output is greater than or equal to a preset threshold, it represents the constant-rate stage of the drying process. When the absolute value of the derivative of moisture content When the output is less than the preset threshold, it enters the slow-down drying stage, and the output... The above process provides trigger signals for phased control.

[0044] Hysteresis detection logic is introduced. To eliminate frequent stage switching caused by signal fluctuations, this invention introduces a hysteresis range for anti-jitter processing. Specifically, a hysteresis value is set. This establishes upper and lower limit switching thresholds. When the system is in a constant-speed phase, only when... The system switches to the slow-down drying stage only when the temperature falls below the lower switching threshold; when the system is in the slow-down drying stage, it only switches to the slow-down drying stage when the temperature falls below the lower switching threshold. It only switches back to the constant speed stage when the speed exceeds the upper limit switching threshold.

[0045] S2. In the constant-rate drying stage, a model reference adaptive control strategy is adopted to establish a constant-rate drying dynamics model that describes the relationship between drying rate and material moisture content and hot air temperature. An adaptive law based on Lyapunov stability theory is designed to dynamically adjust the hot air temperature control gain and obtain the hot air temperature control command corresponding to the constant-rate drying stage, so that the actual drying rate tracks the output of the ideal reference model.

[0046] In some embodiments, for any control cycle, the specific implementation process of the model reference adaptive control strategy in the constant-rate drying stage includes: The actual drying rate of the current PVC material is obtained, and the ideal drying rate output by the ideal reference model is read. The difference between the actual drying rate and the ideal drying rate is calculated as the tracking error. Based on the preset discretization adaptive law formula, the hot air temperature control gain of the previous control cycle, the current tracking error, and the set learning rate are used as inputs to update the hot air temperature control gain of the current control cycle. The current tracking error is multiplied by the updated hot air temperature control gain to obtain the hot air temperature adjustment amount. The hot air temperature adjustment amount is then added to the base hot air temperature to calculate the control output of the current control cycle. The output control output serves as the hot air temperature control command corresponding to the constant-rate drying stage. The control output includes the final hot air temperature setpoint.

[0047] In one possible implementation, the discretization adaptive law formula is: ; in, Indicates the current control cycle. Indicates the previous control cycle. Indicates the first Control gain per control cycle The learning rate is used to adjust the speed at which the adaptive gain is adjusted. The tracking error is calculated at the current sampling time. To preset the ideal error constant, The sampling period; Hot air temperature adjustment The formula for determining it is: ; Final hot air temperature setpoint The formula for determining it is: ; in, This is the base hot air temperature setting.

[0048] Specifically, S2 mentioned above includes: A kinetic model for the constant-rate stage of the PVC fluidized bed drying process was established. Based on empirical data and simulation models of the PVC drying process, the most suitable model for the constant-rate stage was selected, as follows: in, and These are model constants. This refers to the drying time. This represents the moisture content of the PVC material. A constant was obtained by fitting experimental data under different temperature conditions. and This model can describe the drying rate of PVC in the constant-rate stage quite well.

[0049] Based on the constant-rate drying process of PVC, an adaptive law of MRAC is selected. This invention employs Model Reference Adaptive Control (MRAC) to dynamically adjust the hot air temperature during the constant-rate drying stage of PVC, ensuring that the system can effectively regulate and maintain an ideal drying rate. To achieve precise adaptive control of the hot air temperature, this invention designs an adaptive law of the following form: in, The learning rate is used to adjust the speed at which the adaptive gain is adjusted. The error between the actual system and the reference model, Let be a set ideal error constant. To control the gain. This adaptive law is one of the core elements of the control strategy in the constant-speed phase of this invention; its structure ensures that the system error can be driven and stabilized at the ideal error. This allows for real-time, adaptive adjustment of the hot air temperature.

[0050] To ensure the stability and convergence of the control system under this adaptive law, it can be demonstrated using Lyapunov stability theory: The MRAC algorithm requires designing a reference model and an actual system model, and adjusting the control input based on the error between these two models to achieve precise control of the actual system. The actual system's drying rate... It is usually affected by a variety of factors, such as hot air temperature and material moisture content. Therefore, the actual drying rate... There will be some deviation from the ideal drying rate of the reference model. : in, This represents the actual PVC drying rate of the system. Indicates in Always refer to the ideal drying rate output by the model. This represents the error between the actual system and the reference model. The goal of the MRAC algorithm is to adjust the control gain to minimize this error. Approaching 0.

[0051] To ensure the stability of the adaptive controller, it is necessary to prove the error. It tends to zero in time. The Lyapunov method is used for analysis.

[0052] First, define a Lyapunov function. The Lyapunov function is chosen as the error to describe the system's energy. Square form: Lyapunov function Non-negative and exist When the value reaches its minimum, it indicates that the system state has stabilized and the equilibrium point is reached. According to Lyapunov's second method, the stability of a system can be determined through analysis. The derivative is used to determine the derivative of the derivative. Differentiation yields: Errors in the actual system Substituting the dynamic equation into it, we assume that the error dynamic equation of the system can be expressed as: in, is a constant, representing the inherent decay rate of the system. Let be the relationship constant between the system and the control gain. Substitute the system's dynamic equations into... From the derivative formula, we get: To ensure system stability, it is necessary to guarantee If it is negative, then the Lyapunov function is... Only then will it decrease, thus reducing the error. It tends to 0. To satisfy this condition, an adaptive gain is selected. The update pattern is as follows: Based on this adaptive law, the control gain is... Based on the current error Continuous adjustments are made to gradually reduce the error. The value eventually tends to 0, proving the stability of the system.

[0053] It should be noted that the MRAC controller is implemented during the constant-rate drying stage. Its core objective is to dynamically adjust the controller gain so that the output of the actual drying system can quickly and stably track the output of an ideal reference model.

[0054] For example, see S2 above. Figure 4 Specifically, it includes the following steps: S21. Establish a reference model. Based on the drying kinetics model of the constant-rate stage. Differentiating this equation yields the drying rate. Therefore, the reference model can be set as a first-order inertial element: in, The time constant is used as a reference model; its value can be set manually according to the desired response speed. The smaller the value, the faster the response. Set a value for the desired drying rate. The ideal drying rate is the reference model at the input... The ideal output trajectory is then generated.

[0055] S22. Design an adjustable controller. The above reference model provides a clear performance tracking target for the entire control system. By comparing the ideal output of the reference model with the measured output of the actual system, the system tracking error is defined: The controller output is the adjustment amount for the hot air temperature. From the current adjustable gain With system tracking error A joint decision was made to adopt the proportional control law: In the above formula, To control the gain, it varies over time. This represents the system tracking error, specifically the difference between the ideal drying rate and the actual drying rate. Adjustment amount. This will be superimposed on a base hot air temperature setpoint. The final control command is then generated. S23. Update the control gain according to the adaptive law. To achieve adaptive adjustment of the control gain, based on Lyapunov stability theory, the following adaptive law is adopted to ensure the stability and convergence of the control system: Since a PLC is a digital control system, it operates at discrete points in time and performs cycle calculations. Therefore, the above formula needs to be discretized during implementation. The discretized implementation of the adaptive law formula in a digital controller is as follows: In the formula, Indicates the current control cycle. Indicates the previous control cycle. The execution cycle of the control algorithm, i.e., the sampling cycle. This is the error value calculated at the current sampling time.

[0056] At the beginning of each control cycle, the system first samples and calculates the error at the current moment. Then based on this error and the Control gain per control cycle Calculate the first Control gain per control cycle .

[0057] Based on all the above steps, the implementation method of the MRAC controller for the entire constant-rate drying stage can be described as follows: The system samples and obtains the current drying rate. And calculate the error ; Based on the discretized adaptive law formula above, update the adaptive gain. ; Based on error With control gain Calculate control output ; Output final hot air temperature setpoint ; Waiting for a control cycle Then jump to S21 and start a new round of control cycle.

[0058] S3. In the falling-rate drying stage, a model predictive control strategy is adopted to establish a drying dynamics model for the falling-rate drying stage. Based on model prediction, online optimization solution and rolling optimization, the optimal hot air temperature control quantity is determined.

[0059] In some embodiments, for any control cycle, the specific implementation process of the model predictive control strategy in the falling-rate drying stage includes: The current moisture content of the polyvinyl chloride (PVC) material is measured and used as the initial state of the system. This value is then substituted into a pre-established drying kinetic model for the falling-rate drying stage. The drying kinetic model for the falling-rate drying stage is linearized and discretized near the current operating point to obtain a state-space prediction model. Based on this model, the moisture content of the material within a preset number of steps is predicted, resulting in the predicted moisture content for each step within that preset number of steps. The objective function is the weighted sum of the errors between the predicted and target moisture contents for each step, and the weighted sum of the deviations between the hot air temperature and the base temperature for each step. Simultaneously, the upper and lower limits of the hot air temperature, the range of the hot air temperature change rate, and the reasonable range of moisture content are used as constraints. The optimization problem is solved online, with the objective being the minimum value of the objective function under the constraints. This yields the hot air temperature control sequence for the preset number of steps. The first control variable in the hot air temperature control sequence is selected as the optimal hot air temperature control variable to obtain the hot air temperature control command for the falling-rate drying stage.

[0060] In one possible implementation, the state-space prediction model is as follows: ; in, This is a state variable representing the current moisture content of the polyvinyl chloride material, and is a control input. Indicates hot air temperature, output The measured value representing the moisture content of the material. These represent the state transition matrix, input matrix, and output matrix, respectively.

[0061] Furthermore, the objective function is optimized as follows: ; in, For the first Predicted moisture content of the step The target moisture content. It is the first hot air temperature of the step, That is the base temperature. and These are weighting coefficients, used to optimize the objective function. Its minimum value needs to be found under the corresponding constraints.

[0062] In some embodiments, the constraints include hot air temperature constraints, hot air temperature change rate constraints, and moisture content constraints. The upper and lower limits of hot air temperature are constrained as follows: ; Wherein, Tmin and Tmax are the minimum and maximum limits of hot air temperature, respectively; The constraint on the rate of change of hot air temperature is: ; in, This represents the maximum permissible rate of change in hot air temperature. This represents the minimum permissible rate of change in hot air temperature. Moisture content constraint is: ; Among them, MR min MR is the minimum limit for the moisture content of the material. target This represents the maximum limit for the moisture content of the material.

[0063] It should be noted that, considering the characteristic that the drying rate in the falling-rate drying stage is dominated by internal diffusion, a predictive controller is designed based on the kinetic model of the PVC falling-rate drying stage. Through the MPC control algorithm, the PVC drying process in the falling-rate drying stage can be precisely controlled, ensuring the achievement of the target moisture content and optimizing energy consumption. The specific MPC control process in the falling-rate drying stage is as follows: A kinetic model for the falling-rate drying stage of the PVC fluidized bed drying process was established. Based on the simulation results of the PVC drying process, the drying kinetic model for this stage can be expressed as: This formula describes the moisture content of the material during the falling-rate drying stage. Over time The change model. Here, It represents the water content at a certain moment. It is the diffusion coefficient, which is related to the hot air temperature. As the temperature increases, the diffusion coefficient increases and the rate of moisture diffusion accelerates. Indicates the duration of the drying process. It is a constant reflecting the rate of moisture diffusion. A key characteristic of this model is that the moisture content increases over time. The rate of moisture decreases exponentially, reflecting the diffusion of moisture from the interior of the material to the surface. During the falling-rate drying stage, the rate of moisture diffusion gradually slows down, leading to a decrease in the drying rate. Therefore, temperature control is crucial. This is crucial for ensuring the effective execution of the drying process.

[0064] Application of Model Predictive Control (MPC) in the Falling Drying Stage. When using Model Predictive Control (MPC) for the falling drying stage, we will make predictions based on the kinetic model of this stage. MPC will use this model to predict the moisture content over a future period. And calculate the corresponding optimal hot air temperature. This is to achieve control over the drying process.

[0065] Specifically, the MPC controller will adjust the current moisture content... and hot air temperature This model is used to predict the moisture content of PVC at future times. By optimizing the objective function, MPC will dynamically adjust the hot air temperature. To minimize the target error and satisfy the constraints, the process for this step is as follows: a1. System State Prediction. In MPC, the first step is to predict the system state based on the current moisture content. and hot air temperature Predicting moisture content in the near future Set the time step to The prediction within each time step can be calculated using the following equation: in, It is the first Moisture content of the step, It is the diffusion coefficient. For hot air temperature, It is the diffusion index. Through iterative calculation, MPC can predict the future. The moisture content value at this step. This prediction will serve as the basis for subsequent control decisions.

[0066] a2. Optimize the objective function design. In MPC, the optimization objective is achieved by adjusting the control input, namely the hot air temperature. The goal is to make the system behave as expected. The optimization objective function includes the following parts: Minimize moisture content error: Requires a minimum moisture content of [missing information]. Try to get as close as possible to the desired target value ,Right now: in, For the first Predicted moisture content of the step The target moisture content.

[0067] Minimize hot air temperature: Avoid excessive temperature fluctuations, thereby reducing energy consumption and equipment wear. In the formula, It is the first hot air temperature of the step, That is the base temperature.

[0068] Comprehensive optimization objective function: The final optimization objective function can be a weighted sum of water content error and energy consumption. Right now: in, and These are weighting coefficients used to balance the priorities of the two objective terms: moisture content error and energy consumption. (Optimization objective function) It is necessary to find its minimum value under constraints.

[0069] a3. Set the constraints for the MPC control algorithm. In practical applications, the system's control input and state variables are subject to certain physical constraints, and the MPC algorithm needs to be optimized under these constraints. For the drying process in the falling-rate drying stage, some common constraints include: Temperature constraint: The hot air temperature must not exceed the equipment's maximum withstand temperature. At the same time, it must not be lower than the minimum operating temperature. Therefore: Temperature change rate constraint: To prevent excessively rapid operation of the steam regulating valve during the drying process, which could lead to equipment wear and energy consumption fluctuations, the hot air temperature change rate is constrained. It must be limited to the permitted range: Moisture content constraint: To ensure drying effect and prevent over-drying, the moisture content is... It should be kept within a reasonable range: These constraints ensure that the drying process is carried out under multiple safeguards, including equipment safety, stable operation, and compliance with process requirements.

[0070] a4. Optimization and Rolling Optimization. MPC uses optimization algorithms to solve for the optimal hot air temperature sequence. And select the first control variable. This serves as the current control input. Over time, the system recalculates the control input based on the new state, ensuring optimization at every moment based on the latest predictions. This "rolling optimization" strategy can cope with uncertainties and external disturbances during the drying process, enabling the system to adapt to dynamic changes and remain stable.

[0071] For example, see Figure 5 The above S3 specifically includes the following steps: S31. Establish and discretize the prediction model. Develop a kinetic model for the PVC drying falling-rate drying stage. By linearizing and discretizing the model around its current operating point, we obtain the state-space model for MPC: In the formula This is a state variable, representing the current moisture content of the PVC material, and is a control input. Indicates hot air temperature, output The measured value representing the moisture content of the material. These three matrices, representing the state transition matrix, input matrix, and output matrix respectively, together constitute the state-space model of a linear discrete system, describing the dynamic characteristics of the system, the influence of input on the state, and the relationship between the state and the output.

[0072] Online solution for S32 and MPC optimization problems. In each control cycle... The MPC controller will perform the following steps: a1. State Initialization. The MPC controller will use the measured current moisture content of the PVC material as the initial value of the system state. .

[0073] a2. System Prediction. Using the discrete state-space model described above, predict the future. The system output of the step, i.e., moisture content ,in .

[0074] a3. Optimization Solution. In this patent, the objective function is optimized. This is a weighted sum of moisture content error and energy consumption. Solve the following optimization problem to obtain the future... Optimal control sequence of steps , : The above formula optimizes the objective function It directly balances control precision and energy consumption, among which, For the first Predicted moisture content of the step For the first Predicting the hot air temperature step by step, The base hot air temperature is usually set as the minimum economical temperature to meet drying requirements. Weighting coefficients ( This is used to balance the priorities of moisture content tracking accuracy and energy conservation. Increase Emphasizing control precision and increasing They place greater emphasis on reducing energy consumption.

[0075] The first two constraints in the above formula are model constraints, meaning that the MPC controller's prediction of the future system behavior must be based on the state-space model we have established; the last three constraints are about hot air temperature, hot air temperature change rate, and PVC material moisture content. These constraints are determined by the process and equipment to ensure that the PVC drying process does not exceed the capability range of the actuator, the safe operating limits of the equipment, and the allowable range of the process.

[0076] S33, MPC controller rolling optimization solution. After solving the above constrained optimization problem, the MPC controller will select the first element in the optimal control sequence. The actual control command for the current moment is output to the hot air temperature regulating valve. This continues until the next control cycle. The controller receives new moisture content measurements, updates the initial state, and re-optimizes based on the new predictions. This rolling optimization strategy enables the MPC to continuously cope with external disturbances in the system under complex operating conditions, maintaining excellent control performance at all times.

[0077] S4. Output the hot air temperature control command for the constant speed drying stage or the falling speed drying stage to the steam regulating valve, so that the steam regulating valve can dynamically adjust the hot air temperature of the fluidized bed dryer in response to the hot air temperature control command, forming a closed-loop control.

[0078] As described in S1-S4 above, the method provided by this embodiment of the invention, through the core process of stage identification, staged control, and closed-loop adjustment, firstly, automatically identifies the two stages based on the derivative of moisture content and hysteresis logic, solving the problem of missing stages in traditional control. In the constant-rate stage, the MRAC strategy is adopted, dynamically adjusting the hot air temperature control gain to ensure the actual drying rate tracks the ideal model, adapting to the kinetic characteristics where the drying rate is dominated by hot air temperature. In the deceleration drying stage, the MPC strategy is adopted, determining the optimal temperature through prediction and rolling optimization, adapting to the characteristic where the drying rate is dominated by internal diffusion. Finally, a closed-loop control is formed through a steam regulating valve, achieving seamless integration of the two-stage control strategies. This solution fundamentally solves the problem that a single control strategy cannot adapt to the kinetic differences between the two stages, effectively resisting fluctuations in feed moisture content and interference from hot air temperature, ensuring a stable decrease in PVC moisture content. Simultaneously, precise staged adjustment reduces unnecessary energy consumption, achieving the triple goals of precise humidity control, interference resistance, and energy optimization, meeting the needs of different application scenarios.

[0079] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A staged adaptive control method for a fluidized bed drying process of polyvinyl chloride, characterized in that, The method includes: Real-time collection of moisture content data of polyvinyl chloride material, calculation of moisture content derivative to obtain drying rate, introduction of hysteresis judgment logic to set upper and lower limit switching thresholds, and identification of constant rate drying stage or falling rate drying stage of drying process based on drying rate and the upper and lower limit switching thresholds. In the constant-rate drying stage, a model reference adaptive control strategy is adopted to establish a constant-rate drying dynamics model that describes the relationship between drying rate and material moisture content and hot air temperature. An adaptive law based on Lyapunov stability theory is designed to dynamically adjust the hot air temperature control gain and obtain the hot air temperature control command corresponding to the constant-rate drying stage, so that the actual drying rate tracks the output of the ideal reference model. In the falling-rate drying stage, a model predictive control strategy is adopted to establish a drying dynamics model for the falling-rate drying stage. The optimal hot air temperature control quantity is determined based on model prediction, online optimization solution and rolling optimization. The hot air temperature control command for the constant-speed drying stage or the deceleration drying stage is output to the steam regulating valve, so that the steam regulating valve dynamically adjusts the hot air temperature of the fluidized bed dryer in response to the hot air temperature control command, forming a closed-loop control.

2. The method according to claim 1, characterized in that, The introduction of hysteresis judgment logic sets upper and lower limit switching thresholds, and identifies the constant-rate drying stage and the falling-rate drying stage of the drying process based on the drying rate and the upper and lower limit switching thresholds, including: Moisture content data of polyvinyl chloride material is collected at a fixed sampling period. The derivative of the moisture content data is calculated using the first-order difference method to obtain the drying rate. The drying rate is then subjected to first-order low-pass filtering to suppress measurement noise and high-frequency interference. The drying rate baseline threshold is determined based on the current ambient temperature and process test results. The hysteresis bandwidth is set by combining the measured noise level of the moisture content detection signal. The upper limit switching threshold and the lower limit switching threshold are calculated based on the drying rate baseline threshold and the hysteresis bandwidth. If the current drying stage is constant speed, switch to the deceleration drying stage when the filtered drying rate is less than or equal to the lower limit switching threshold; if the current drying stage is deceleration, switch back to the constant speed drying stage when the filtered drying rate is greater than or equal to the upper limit switching threshold.

3. The method according to claim 2, characterized in that, The formula for calculating the upper limit switching threshold is: ； The formula for calculating the lower limit switching threshold is: ； in, The upper limit switching threshold, The lower limit switching threshold, H is the baseline threshold for drying rate, and H is the hysteresis bandwidth.

4. The method according to claim 1, characterized in that, For any given control cycle, the specific implementation process of the model reference adaptive control strategy in the constant-rate drying stage includes: Obtain the actual drying rate of the current polyvinyl chloride material, and at the same time read the ideal drying rate output by the ideal reference model. Calculate the difference between the actual drying rate and the ideal drying rate as the tracking error. Based on the preset discretization adaptive law formula, the hot air temperature control gain of the previous control cycle, the current tracking error, and the set learning rate are used as inputs to update the hot air temperature control gain of the current control cycle. Multiply the current tracking error by the updated hot air temperature control gain to obtain the hot air temperature adjustment amount, and then add the hot air temperature adjustment amount to the base hot air temperature to calculate the control output of the current control cycle. The control output is used as a hot air temperature control command for the constant speed drying stage, and the control output includes the final hot air temperature setpoint.

5. The method according to claim 4, characterized in that, The formula for the discretization adaptive law is: ； in, Indicates the current control cycle. Indicates the previous control cycle. Indicates the first Control gain per control cycle The learning rate is used to adjust the speed at which the adaptive gain is adjusted. The tracking error is calculated at the current sampling time. To preset the ideal error constant, The sampling period; The hot air temperature adjustment amount The formula for determining it is: ； The final hot air temperature setpoint The formula for determining it is: ； in, This is the base hot air temperature setting.

6. The method according to claim 1, characterized in that, For any given control cycle, the specific implementation process of the model predictive control strategy used in the rate-deceleration drying stage includes: The current moisture content of the polyvinyl chloride material is measured and used as the initial state of the system, and then substituted into the pre-established drying kinetics model of the falling rate drying stage. The drying kinetics model of the falling-rate drying stage is linearized and discretized near the current working point to obtain a state-space prediction model. Based on the state-space prediction model, the material moisture content within a preset number of steps is predicted to obtain the predicted moisture content corresponding to each step within the preset number of steps. The optimization objective function is to use the weighted sum of the errors between the predicted moisture content and the target moisture content corresponding to each step within the preset number of steps, and the weighted sum of the deviations between the hot air temperature and the base temperature corresponding to each step, as the optimization objective function. At the same time, the upper and lower limits of hot air temperature, the range of hot air temperature change rate, and the reasonable range of moisture content are used as constraints. The optimization problem is solved online. The optimization objective of the optimization problem is the minimum value of the optimization objective function under the constraints, so as to obtain the hot air temperature control sequence for the future preset number of steps. The first control variable in the hot air temperature control sequence is selected as the optimal hot air temperature control variable to obtain the hot air temperature control command corresponding to the falling-rate drying stage.

7. The method according to claim 6, characterized in that, The state-space prediction model is as follows: ； in, This is a state variable representing the current moisture content of the polyvinyl chloride material, and is a control input. Indicates hot air temperature, output The measured value representing the moisture content of the material. These represent the state transition matrix, input matrix, and output matrix, respectively.

8. The method according to claim 7, characterized in that, The optimization objective function is: ； in, For the first Predicted moisture content of the step For the target moisture content, It is the first hot air temperature of the step, It is the base temperature. and These are the weighting coefficients, and the optimization objective function... Its minimum value needs to be found under the corresponding constraints.

9. The method according to claim 8, characterized in that, The constraints include hot air temperature constraints, hot air temperature change rate constraints, and moisture content constraints. The upper and lower limits of the hot air temperature are constrained as follows: ； Among them, T min T max These are the minimum and maximum limits for hot air temperature, respectively. The constraint on the hot air temperature change rate is: ； in, This represents the maximum permissible rate of change in hot air temperature. This represents the minimum permissible rate of change in hot air temperature. The moisture content constraint is: ； Among them, MR min MR is the minimum limit for the moisture content of the material. target This represents the maximum limit for the moisture content of the material.

10. The method according to claim 9, characterized in that, Applied to polyvinyl chloride fluidized bed drying systems, including hardware components; The hardware components include a controller, a fluidized bed dryer, an online moisture content detector, a temperature sensor, and a steam regulating valve. The online moisture content detector is installed at the outlet of the fluidized bed dryer to collect real-time moisture content data of the polyvinyl chloride material; the temperature sensor is installed at the hot air inlet and inside the bed of the fluidized bed dryer to collect hot air temperature and material temperature data; the steam regulating valve is connected to the hot air supply pipeline of the fluidized bed dryer to regulate the steam flow rate in response to hot air temperature control commands; and the controller is used to execute the method of any one of claims 1-9.

Images