Iron tailings based geopolymer panel curing predictive control system

Abstract

The application discloses a kind of iron tailing base polymer board curing prediction control system, belong to the control or regulation technical field of non-electric variable.This system is arranged in the thickness direction of blank multiple-point temperature sensor, and the core temperature and surface temperature of blank are densely measured, and the closed-loop regulation of blank temperature is carried out in digital mode: state prediction unit is according to the recursive estimation of multiple-point temperature feedback blank internal temperature distribution and its variation trend, remaining time of temperature to reach target value is predicted;Temperature regulation unit outputs the adjustment amount of heating and ventilation actuator accordingly, so that each point temperature of blank is controlled according to target temperature trajectory;The system also carries out coordination compensation to humidity and air flow and other auxiliary non-electric variables, and switches conservative regulation strategy when state estimation uncertainty is over limit, to realize the stable digital control of blank temperature.

Description

Technical Field

[0001] This invention belongs to the field of temperature control technology in the solidification process of iron tailings-based polymer boards in the field of physics. Specifically, it relates to a predictive control system for the solidification of iron tailings-based polymer boards, which takes the temperature of the billet as the controlled object and performs predictive control and feedback correction of the temperature field of the billet. Background Technology

[0002] During the curing process of polymer sheets from iron tailings ore deposits, the temperature of the green body changes continuously due to the exothermic reaction of gelation and external heat exchange. The response of the green body temperature field to actuators such as heating, ventilation, and humidification exhibits distributed, time-varying, and multivariate coupling characteristics. Accurately controlling the green body temperature field and ensuring its evolution along the target temperature trajectory is crucial to guaranteeing the curing quality of this type of sheet.

[0003] Greenhouse temperature field control is a type of temperature control problem with distributed parameters. Due to limitations in measurement conditions, existing technologies can only obtain the temperature of a limited number of measuring points on the greenhouse using a finite number of temperature sensors, thus treating temperature field control as a multivariable control problem of measuring point temperatures. The arrangement of temperature sensors depends on experience and is difficult to cover the complete temperature field. Conventional temperature control schemes cannot guarantee that the greenhouse has a transient temperature field that meets the requirements.

[0004] Traditional curing temperature control relies heavily on preset temperature curves or PID controllers, adjusting heating power based on the deviation between the set and measured temperatures. This approach struggles to address nonlinearities, exothermic internal heat sources, and multivariable coupling inherent in the curing process, and also fails to predict the evolution of the preform temperature and the curing endpoint. Therefore, a non-electrical variable control system is needed that uses preform temperature as the primary controlled variable, integrates temperature sensor feedback and a heat transfer physics model, and predictively controls the preform temperature field. Summary of the Invention

[0005] The purpose of this invention is to provide a predictive control system for the curing of polymer sheets in iron tailings bases, addressing the problems of existing curing control methods, such as difficulty in obtaining the internal temperature distribution and curing state of the billet, difficulty in predicting temperature evolution and curing endpoint, and lag in final-stage temperature regulation. This system transforms curing control from a single timetable control to a state-driven predictive control system, with billet temperature as the primary controlled variable and temperature, humidity, and related flow rates as non-electrical variables.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: the system includes a temperature data acquisition unit, a temperature state prediction unit, a temperature prediction control unit, a feedback correction and zoned temperature compensation unit, and a confidence level security unit.

[0007] The temperature data acquisition unit is used to collect data on the core temperature, surface temperature, curing ambient temperature, curing ambient humidity, ventilation frequency, humidification or steam flow rate, and curing characterization features of the polymer sheet during the curing process at the iron tailings base, according to a preset time rhythm. The core temperature and surface temperature of the sheet serve as the main feedback quantities for temperature closed-loop control. After timestamp alignment and outlier processing, the above data forms a synchronous observation vector at the same sampling time, which is used for subsequent state assimilation and temperature control decisions.

[0008] The temperature state prediction unit receives synchronous observation vectors and runs a state assimilation process based on unscented Kalman filtering or equivalent recursive estimation algorithms. This algorithm recursively fuses the observation data with a coupled model of unsteady heat transfer and reaction exothermics in the thickness direction. Within a preset physical constraint range, it estimates the internal temperature distribution, degree of solidification, equivalent thermal conductivity, and equivalent reaction rate constant of the billet, and provides the remaining solidification time or temperature state deviation at the target age based on the current state.

[0009] The temperature prediction and control unit calculates executable adjustments for heating power, curing environment humidity, ventilation frequency, and humidification valve opening or steam flow rate based on the preform temperature estimate, conservative representative degree of curing, remaining curing time, and target temperature deviation output by the temperature state prediction unit. The primary control objective is to track the target temperature trajectory to the preform core temperature. This unit does not focus on material formulation optimization; instead, it converts the temperature prediction results into non-electrical variable control setpoints.

[0010] The feedback correction and zoned temperature compensation unit performs feedback correction on the physical model based on the deviation between the measured temperature of the temperature measuring point and the predicted temperature of the temperature state prediction unit. On the other hand, it receives the temperature distribution of the blank along the length or thickness direction and compares it with the target temperature distribution to form a temperature deviation distribution. Then, within the equipment capacity boundary of the curing chamber heater, variable frequency fan, humidification valve, atomizing nozzle or steam manifold, it outputs the execution quantity setpoint sequence of each temperature control zone.

[0011] The confidence level safety unit is used to monitor the uncertainty of state assimilation results and control decisions. When the trace of the posterior covariance matrix, the observation residual, or a combination thereof exceeds a preset uncertainty threshold, the system determines that the confidence level of the state estimation is insufficient and switches to a preset conservative solidification process curve; when the uncertainty falls back below the threshold, state-driven predictive control is restored.

[0012] Compared with the prior art, the present invention has at least the following beneficial effects.

[0013] First, this invention takes digital closed-loop control of the core temperature of the billet as the main line, and combines it with the coordinated adjustment of curing environment humidity and humidification, ventilation or steam flow. It unifies multi-source observation, state assimilation and actuator adjustment into a non-electric variable control system, which can improve the foresight and stability of billet temperature regulation.

[0014] Second, the present invention uses a conservative approach to predict the remaining curing time and compensation requirements based on the deviation of the representative degree of curing or the target curing state, thereby avoiding misjudging areas with higher local temperatures or faster reactions as areas where overall curing is complete, thus reducing the risk of under-curing and strength dispersion.

[0015] Third, the present invention performs temperature and humidity decoupling control and zoned temperature compensation in the final stage, which can reduce the mutual disturbance between heating, ventilation, humidification or steam supply, and improve the temperature trajectory, humidity trajectory and curing uniformity of the board along the length direction in the final stage.

[0016] Fourth, the present invention sets up a confidence level safety assessment and a conservative process curve rollback mechanism, which can maintain safe solidification even in the event of sensor malfunction, raw material fluctuations, communication delays or model mismatch, thereby reducing product quality fluctuations caused by erroneous control commands. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall deployment and closed-loop control architecture of the iron tailings base polymer board solidification prediction and control system proposed in this invention in the board production line.

[0018] Figure 2 This is a flowchart illustrating the joint control process between state prediction, temperature prediction control, and execution constraints in this invention.

[0019] Figure 3 This is a sample data analysis diagram showing the correspondence between the curing cycle deviation and the final curing state deviation in the embodiments of the present invention;

[0020] Figure 4 This is a data analysis chart comparing the tracking deviations of the core temperature trajectory of the billet and the humidity trajectory of the curing environment before and after enabling state-driven temperature and humidity control in this embodiment of the invention.

[0021] Figure 5 This is a comparative data analysis diagram of the measured curing state distribution curve of the board along the length direction and the standard curve in the embodiment of the present invention;

[0022] Figure 6 This is a data analysis diagram of the curing uniformity of the entire length of the board before and after zoned temperature compensation in an embodiment of the present invention. Detailed Implementation

[0023] The following examples and Figures 1 to 6The present invention will be further described below. The embodiments described are used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Those skilled in the art can adjust the sampling period, sensor model, controller parameters, and process boundary values ​​without departing from the concept of the present invention.

[0024] This embodiment provides a predictive control system for the curing of polymer sheets in an iron tailings ore production base. The system is located between the curing chamber, the post-curing chamber, and the control cabinet. The curing chamber houses heating actuators, variable frequency ventilation actuators, atomizing humidification actuators, and optional steam manifolds. The control cabinet contains an industrial controller or edge controller. All sensors, actuators, and controllers communicate via an industrial fieldbus or real-time Ethernet to ensure the timing consistency of acquired data and control commands. Its overall deployment and control closed-loop within the sheet production line correspond to… Figure 1 .

[0025] The temperature data acquisition unit constitutes the system's sensing layer. This unit collects the core temperature of the billet, the surface temperature of the billet, the curing environment temperature, the curing environment humidity, the ventilation frequency, the humidification or steam flow rate, and curing characterization features according to a preset time interval. The sampling period can be set according to the thickness of the board, heating inertia, and sensor response time, for example, from 1 second to 30 seconds. For operating conditions with a long sampling period for curing characterization features, the sampling period can be set to an integer multiple of the temperature and humidity sampling period, and the most recent valid feature is timestamped during state assimilation.

[0026] The core temperature of the preform can be obtained using alkali- and heat-resistant encapsulated thermocouples or fiber optic temperature sensors. These sensors are positioned at representative locations along the thickness direction during the preform forming or molding stage, preferably at the geometric center or in areas with large temperature gradients. The surface temperature of the preform can be obtained using shallow-embedded temperature sensors, infrared thermometers, or a combination of both. The curing environment temperature and humidity are obtained using a combined temperature and humidity sensor positioned at representative locations within the curing chamber, serving as boundary conditions and control feedback.

[0027] Curing characterization features can be acquired using an interdigital capacitive sensor array, a conductivity probe, an infrared temperature field scanning device, or a combination thereof. The interdigital capacitive sensors can be attached to the inside of the mold or mounted on an insulating support near the sidewall of the preform; the infrared temperature field scanning device can be positioned at the press outlet, the inlet of the initial curing unit, or the inlet of the later curing chamber. The system does not use a single testing instrument as the final protection target, but rather uses the curing characterization features as auxiliary observations to correct the curing state estimate and generate control setpoints for temperature, humidity, and related flow rates.

[0028] This unit also includes a data quality verification module. This module adds a timestamp to each set of observation data, removes temperature, humidity, flow rate, and solidification characteristics that are significantly outside the physical range, compensates for short-term missing data using the previous valid value, trend extrapolation, or adjacent sensor data, and sends a sensor anomaly flag to the confidence security unit for continuously abnormal data. The verified data is encapsulated into a synchronous observation vector and sent to the temperature state prediction unit.

[0029] The temperature state prediction unit is used to estimate the internal state of the billet, which is difficult to measure directly. Considering that the thermal gradient along the thickness direction of the sheet is usually greater than that along the surface direction, this embodiment uses the thickness direction as the main heat transfer direction and establishes a one-dimensional nonlinear transient heat conduction model. This model is used to describe the coupling relationship between the temperature field, degree of cure, and reaction exothermics. The controller uses the following energy balance relationship when predicting the state.

[0030]

[0031] In the formula, Indicates the position of the billet in the thickness direction and time Temperature at that location; Indicates degree of curing; Apparent density as a function of degree of cure; This indicates the specific heat capacity as a function of degree of cure. Indicates equivalent thermal conductivity; This represents the exothermic reaction term per unit volume generated by the gelation reaction. All the above variables appear in the energy balance relationship, and no further undefined symbols will be introduced. Their values ​​can be determined or corrected through offline thermophysical property experiments and online state assimilation.

[0032] To link the gelation reaction progress with the temperature field, the exothermic reaction term is adopted in the form related to the heat of reaction per unit volume, the equivalent reaction rate constant, and the curing kinetic function, as follows.

[0033]

[0034] In the formula, Indicates the equivalent density of the system participating in the reaction; It represents the heat of reaction per unit mass; Represents the temperature-dependent equivalent reaction rate constant; This represents the solidification kinetics function. For different iron tailings-based polymer board formulations, and It can be pre-calibrated by differential scanning calorimetry or equivalent thermal analysis.

[0035] The curing kinetics function can be expressed using a Sesták-Berggren type expression to describe the combined trends of accelerated early reaction, development of mid-stage reaction, and limited diffusion in the later stage, as detailed below.

[0036]

[0037] In the formula, , and Preset model parameters can be identified through offline isothermal curing experiments. To avoid numerical calculations... Instability occurs when the value is close to 0 or 1; the controller will [perform certain actions] in actual calculations. It is limited to a preset effective range, and the target curing degree boundary value is set to a process value less than 1.

[0038] The equivalent reaction rate constant is described in Arrhenius form to account for the effect of temperature, as follows.

[0039]

[0040] In the formula, Pre-exponential factors; It is the apparent activation energy; It is the ideal gas constant; This represents absolute temperature and uses the same dimensions as the aforementioned temperature variables. To improve the stability of online control, this embodiment does not require complete re-identification in each sampling period. and Instead, it recursively corrects the equivalent reaction rate constant within the constraints of the offline calibration values, thus balancing physical interpretability and online feasibility.

[0041] During state assimilation, the thickness direction of the plate is divided into several discrete nodes. The temperature state prediction unit constructs augmented state variables from the discrete node temperatures, discrete node curing degrees, equivalent thermal conductivity, and equivalent reaction rate constant, and estimates them using unscented Kalman filtering or equivalent recursive estimation algorithms. Within each time cycle, the unit first generates unscented sampling points based on the posterior mean and posterior covariance of the previous time step, and then substitutes each unscented sampling point into the discretized physical model for state prediction. The spatial discretization of the physical model can use the finite difference method or the finite volume method, and the time progression can use an implicit scheme to improve computational stability. This recursive estimation corresponds to the joint process between temperature predictive control and execution constraints. Figure 2 .

[0042] After the state prediction is completed, the temperature state prediction unit reconstructs the observations based on the predicted state. The reconstructed observations include the predicted values ​​of the core temperature of the billet, the surface temperature of the billet, the humidity of the curing environment, and the curing characterization features. Subsequently, the controller compares the reconstructed observations with the synchronous observation vector output by the temperature data acquisition unit, calculates the gain, and corrects the predicted state and prediction covariance. This multi-observation update method can reduce the risk of unidentifiable factors that may occur when relying solely on the core temperature to estimate the degree of curing.

[0043] To verify the supporting role of the state estimation and curing endpoint prediction stages in the calculation of control quantities, this embodiment performs offline playback verification of historical curing records. During playback, the temperature state prediction unit is driven by the sensor observation sequence of the same batch of green bodies. The estimated core temperature, degree of curing, and remaining curing time of the green bodies are compared with the results of offline thermophysical tests and actual breakage measurements. Two baseline methods are set up for comparison: surface temperature feedback only and a fixed curing time schedule. This verification is only used to confirm that the state estimation and prediction stages can output reliable state deviations, remaining curing times, and compensation requirements, and does not take data fitting results that deviate from the physical model as independent protection priorities. Among them, the sample relationship and residual distribution of curing cycle deviation and final curing state deviation are corresponding. Figure 3 .

[0044] Table 1. Offline playback validation data for state estimation and remaining curing time prediction

[0045] Validate Project This method (data assimilation) Surface temperature feedback baseline only Fixed maintenance schedule baseline Root mean square error (°C) of core temperature estimation for billet 0.84 2.97 4.15 Mean absolute error of curing degree estimation 0.031 0.092 0.137 Relative error in remaining curing time prediction (%) 6.8 18.4 27.5

[0046] The curing endpoint prediction function in the temperature state prediction unit receives a conservative representative degree of cure and an equivalent reaction rate constant. Within a single prediction step, the controller estimates the equivalent reaction rate constant at the current representative temperature and makes rolling corrections in subsequent control cycles. The curing kinetics progress model uses the first-order relationship between the degree of cure and time, expressed as:

[0047]

[0048] In this relationship, This indicates the rate at which the degree of curing increases over time. and Consistent with the aforementioned definition, the curing endpoint prediction function uses the current conservative representative degree of curing as the lower limit of integration and the target degree of curing boundary value as the upper limit of integration to calculate the remaining curing time:

[0049]

[0050] In the formula, Indicates the remaining curing time; This indicates the current level of conservatism and solidification. Indicates the target degree of cure boundary value; This represents the equivalent reaction rate constant at the current representative temperature; This represents the solidification kinetic function. The above integral can be solved using an adaptive step-size numerical integration method. Since the system is updated in each control cycle... and Therefore, even if the external temperature or the state of the raw materials changes, the remaining curing time can be adjusted according to the real-time status.

[0051] The temperature prediction control unit determines the control mode based on the remaining curing time. The preset final stage control switching threshold can be set according to the board thickness, target strength, and curing regime, for example, from 600 seconds to 1800 seconds. When the remaining curing time exceeds the final stage control switching threshold, the preform is in the main curing stage, and the system prioritizes stable heating and humidity maintenance. At this time, the heating power is determined by constant temperature control or duty cycle control with a narrow dead zone, the ventilation actuator maintains the basic ventilation frequency, and the humidification actuator maintains the basic humidification state or is turned off to ensure a stable reaction environment.

[0052] When the remaining curing time is less than or equal to the final stage control switching threshold, the preform enters the final stage of fine-tuning temperature and humidity. The temperature prediction and control unit generates the final stage target temperature trajectory and the final stage target humidity trajectory. The final stage target temperature trajectory may include a platform insulation section and a slow cooling section to reduce thermal stress; the final stage target humidity trajectory can be set to slowly decrease or maintain in stages according to the moisture content and shrinkage sensitivity of the board to suppress excessively rapid surface water loss. The corresponding graphical results of the relevant comparative data are shown below. Figure 4 .

[0053] Table 2 Comparison of control performance before and after enabling state-driven temperature and humidity control

[0054] index unit State-driven control is not enabled. Enable state-driven control Change Root mean square value of core temperature trajectory tracking deviation of billet ℃ 3.92 0.95 -2.97 Root mean square value of humidity trajectory tracking deviation in curing environment %RH 8.74 2.31 -6.43 Final curing degree along the thickness direction has a wide range — 0.142 0.046 -0.096 under-cured batch ratio % 18.6 3.2 -15.4

[0055] Since heating power affects the humidity of the curing environment, and ventilation frequency, humidification valve opening, and steam flow rate also affect the temperature of the curing chamber, the final controlled object is a bivariate or multivariate coupled system. Therefore, a decoupling compensator is installed in the temperature prediction control unit. Based on the coupling relationship obtained from process identification experiments, this decoupling compensator estimates the disturbance of humidity caused by heating and the disturbance of temperature caused by humidity regulation, and adds feedforward compensation to the controller output. This ensures that the temperature control loop primarily acts on the core temperature of the preform, and the humidity control loop primarily acts on the humidity of the curing environment.

[0056] After decoupling compensation is completed, the temperature controller outputs heating power commands based on the core temperature tracking deviation of the preform, and the humidity controller outputs ventilation frequency commands, humidification valve opening commands, atomizing nozzle flow commands, steam manifold flow commands, or combinations thereof, based on the humidity tracking deviation of the curing environment. Each executed command, after being subjected to amplitude limiting, speed limiting, and interlock protection, acts on the heater, variable frequency fan, and humidification device. If the curing environment humidity is higher than the target trajectory, the system prioritizes increasing the ventilation frequency; if the curing environment humidity is lower than the target trajectory, the system reduces the ventilation frequency and activates the humidification or steam actuator.

[0057] The feedback correction and zoned temperature compensation unit operates in conjunction with the temperature prediction and control unit. Before the board completes its initial setting and enters the later curing chamber, the system determines the curing state deviations at the head, middle, and tail sections based on the measured curing state distribution curve along the board's length. These deviations are then mapped to a sequence of heating, ventilation, humidification, or steam setpoints for each control zone. This process allows for differentiated temperature and humidity compensation at different locations on the same board; the measured curing state distribution curve, standard curve, and deviation curve along the length correspond to... Figure 5 .

[0058] Table 3 represents the uniformity of the plate samples before and after temperature compensation in different zones.

[0059] Board number Standard deviation before reprogramming (MPa) Standard deviation after reprogramming (MPa) Before and after replanning (MPa) Head-to-tail difference after replanning (MPa) 1 2.183 1.431 1.718 0.794 2 2.371 1.609 0.212 0.115 3 2.071 1.620 1.336 0.959 4 2.290 1.673 0.242 0.383

[0060] The confidence level safety unit operates synchronously with the temperature state prediction unit. Within each time cycle, this unit reads the posterior covariance matrix and calculates its trace, or combines it with the mean square value of the observed residuals to form a comprehensive uncertainty index. This uncertainty index can serve as a scalar indicator of the overall confidence level of the current state estimate. The preset uncertainty threshold can be determined through simulation, historical production data, and field tests. The graphical results of the curing uniformity of the entire length of the board before and after zoned temperature compensation are shown. Figure 6 .

[0061] When the trace or overall uncertainty index of the posterior covariance matrix exceeds a preset threshold, it may indicate sensor malfunction, model mismatch, raw material batch fluctuations, communication delays, or filter divergence. In this case, the confidence safety unit triggers a safety control command, overriding the normal output of the temperature prediction control unit and switching the system to a preset conservative curing process curve. This conservative curing process curve may include a gentler heating rate, a longer holding time, and a higher final humidity limit to avoid under-curing or excessively rapid water loss when the state estimate is unreliable. Once the uncertainty index falls below the threshold and remains below the preset time, the system resumes state-driven predictive control.

[0062] Through the above structure, the control system of this embodiment can convert billet temperature, ambient humidity, ventilation frequency, humidification or steam flow rate, and curing characterization features into state information that can be used for temperature control decisions. It then switches between the main curing stage, the final fine-tuning stage, and the zoned temperature compensation stage based on billet temperature prediction, remaining curing time, and state deviation along the length direction. The main control objective of this system is the predictive control of non-electrical variables such as billet temperature. Material formulation, finished sheet structure, and individual testing instruments are only used as application objects or observation sources and are not considered as the main improvement points of this invention. Therefore, the technical subject of this invention is consistent with non-electrical variable control or regulation systems.

Claims

1. A predictive control system for the solidification of polymer boards in iron tailings ore deposits, characterized in that, The system uses the billet temperature as the controlled object to perform closed-loop predictive control of the billet temperature field during the solidification process of polymer plates from iron tailings, including: The temperature data acquisition unit is configured to acquire the core temperature and surface temperature of the billet according to a preset sampling period using temperature sensors arranged at multiple measuring points in the thickness direction of the billet, and to acquire the operating parameters of the heating, ventilation, humidification or steam actuators and the curing environment temperature as auxiliary quantities. The temperature state prediction unit is configured to recursively estimate the internal temperature field distribution, equivalent thermal conductivity, and equivalent reaction rate constant of the billet based on a physical model describing unsteady heat transfer and reaction exothermic in the thickness direction of the billet. It also predicts the evolution trajectory of the billet temperature according to the current control quantity and the remaining time for the billet temperature to reach the target value in the prediction time domain. The temperature prediction control unit is configured to minimize the deviation of the billet temperature from the target temperature trajectory. Under the constraint of the actuator capacity, it continuously calculates and outputs the control increments of heating power, ventilation frequency, humidification valve opening degree or steam flow rate, forming a closed-loop predictive control of the billet temperature. Among them, the ventilation, humidification or steam execution quantities coordinate the control of the curing environment humidity while adjusting the billet temperature. The feedback correction and zoned temperature compensation unit is configured to perform feedback correction on the physical model based on the deviation between the measured temperature of the temperature measuring point and the predicted temperature of the temperature state prediction unit; and when the temperature distribution of the billet is uneven along the length or thickness direction, the billet is divided into multiple temperature control zones, and heating, ventilation, humidification or steam execution amount set values ​​are generated for each zone respectively. The confidence level safety unit is configured to determine the reliability of temperature control based on the uncertainty index obtained from the feedback correction. When the uncertainty index exceeds a preset threshold, the billet temperature control is switched to a conservative heating and cooling curve, and the rate of change of heating power, ventilation frequency, humidification valve opening, or steam flow is limited. The system takes the predictive control of billet temperature as the main control objective, and the degree of curing, remaining curing time, and curing characterization features of the billet are only used to correct the prediction of the billet temperature state and generate control commands for temperature, humidity, and related flow rates.

2. The predictive control system for the solidification of polymer plates in iron tailings ore deposits according to claim 1, characterized in that, The physical model is a one-dimensional nonlinear transient heat conduction model established along the thickness direction of the billet. The model includes reaction exothermic terms related to the billet temperature, the degree of solidification of the billet, and the equivalent reaction rate constant. The temperature state prediction unit uses the temperature, degree of solidification, equivalent thermal conductivity, and equivalent reaction rate constant of the discrete nodes in the thickness direction as augmented state variables, and takes the billet temperature as the main estimated and controlled variable, and performs joint recursive estimation within the preset physical constraints.

3. The predictive control system for the solidification of polymer boards in iron tailings ore bases according to claim 2, characterized in that, The temperature state prediction unit employs unscented Kalman filtering, extended Kalman filtering, or a state estimation algorithm with equivalent recursive update functionality. Within each sampling period, it generates the predicted state and the predicted billet temperature based on the posterior state estimate of the previous period, and updates the current set of state variables based on the latest measured temperature at the temperature measurement point.

4. The predictive control system for the solidification of polymer plates in iron tailings ore deposits according to claim 1, characterized in that, The temperature state prediction unit uses the conservative representative degree of curing as the target endpoint criterion for the temperature control of the billet. The conservative representative degree of curing is determined by selecting the minimum value, low quantile value, or weighted value that can reflect the under-cured area from the state distribution in the thickness or length direction of the billet. The remaining curing time is calculated with the conservative representative degree of curing as the starting point and the target degree of curing boundary value as the endpoint, and is used to determine the timing of the heat preservation and cooling sections of the target temperature trajectory of the billet.

5. The predictive control system for the solidification of polymer plates in iron tailings ore deposits according to claim 1, characterized in that, The temperature prediction and control unit constructs a dual-variable control structure with the tracking deviation of the core temperature of the billet and the tracking deviation of the curing environment humidity as inputs. It performs feedforward decoupling compensation based on the coupling effect of heating power on curing environment humidity and the coupling effect of ventilation, humidification or steam execution on core temperature of the billet, so that the temperature control loop mainly acts on the core temperature of the billet.

6. The predictive control system for the solidification of polymer plates in iron tailings ore deposits according to claim 1, characterized in that, The feedback correction and zone temperature compensation unit generates a zone temperature compensation amount based on the deviation between the measured temperature distribution and the target temperature distribution along the length of the billet, and converts the zone temperature compensation amount into a sequence of execution amount setpoints for the heater, variable frequency fan, humidification valve, atomizing nozzle or steam manifold.

7. A predictive control system for the solidification of polymer plates in iron tailings ore deposits according to any one of claims 1 to 6, characterized in that, The confidence level safety unit calculates the trace of the posterior covariance matrix of the state estimation, the mean square value of the temperature observation residual, or a combination of the two as the uncertainty index, and compares it with a preset threshold. When the comparison result shows that the billet temperature state estimation is unreliable, the system switches to a preset conservative heating and cooling curve until the uncertainty index falls back below the preset threshold and continues to meet the preset time.

Images