Dynamic dry-steaming calcination temperature self-adaptive control method for gypsum production

Abstract

The invention relates to the crossing field of mechanical engineering and process control, and discloses a dynamic dry-steaming calcination temperature self-adaptive control method for gypsum production. The method comprises the following steps: collecting material temperature, furnace body total mass, waste gas humidity and flow and heating power data in real time; the real-time equivalent specific heat capacity and the dehydration degree of the gypsum are estimated on line based on multi-source data; constructing and updating a dynamic thermodynamics prediction model; solving an optimal heating power sequence by adopting a model predictive control algorithm so as to accurately track a target temperature curve; and taking the condition that the real-time dehydration degree reaches a preset threshold value as a calcination end point criterion. According to the technical scheme, self-adaptive response to material thermal characteristic changes is achieved, and the temperature control precision, the energy efficiency and the product quality stability are improved.

Description

A dynamic dry calcination temperature adaptive control method for gypsum production Technical Field

[0001] This invention belongs to the interdisciplinary field of mechanical engineering and process control, and specifically relates to a dynamic dry steaming and calcination temperature adaptive control method for gypsum production. Background Technology

[0002] Gypsum, as an important inorganic cementing material, is widely used in construction, building materials, and industrial molds. Its core production process involves the thermal dehydration of dihydrate gypsum into hemihydrate or anhydrous gypsum. This process is highly sensitive to precise control of the calcination temperature, which directly determines the phase composition, crystal morphology, and final mechanical properties of the product.

[0003] Traditional gypsum calcination systems typically employ PID controllers based on fixed process parameters for temperature control, with control logic relying on preset thermal models and static material property assumptions. However, in actual industrial production, gypsum raw materials often experience changes in heat capacity, thermal conductivity, and dehydration reaction kinetics due to fluctuations in mineral sources and variations in the content of associated impurities. These dynamic disturbances in raw material characteristics cause abrupt changes in the system's thermal inertia, making it difficult for the PID controller with fixed parameters to maintain the optimal temperature trajectory. This not only wastes energy but also easily leads to incomplete dehydration or over-burning, severely impacting product quality stability.

[0004] Dynamic dry calcination is a key technology for improving the quality and energy efficiency of gypsum. Its core lies in adaptively adjusting the heating rate and holding platform based on the real-time thermophysical properties of the material. This process requires the control system to have the ability to quickly sense changes in raw material composition and the ability to reconstruct temperature strategies online, so as to accurately capture the endpoint of the dehydration reaction and avoid side reaction zones.

[0005] In existing technologies, calcination temperature control still mainly relies on offline chemical analysis or empirical settings, lacking online monitoring methods for the composition of impurities in raw materials. Furthermore, control algorithms are mostly limited to classical feedback regulation, unable to proactively optimize control parameters under conditions of sudden changes in heat capacity. Even when some systems introduce feedforward compensation, model mismatch makes it difficult to cope with nonlinear disturbances caused by complex impurity combinations.

[0006] Especially in continuous large-scale production scenarios, where raw material batches are frequently switched, traditional control methods, due to their slow response and rigid adjustment, are prone to increased energy consumption, decreased product purity, and accumulated thermal stress in equipment. There is an urgent need for an adaptive temperature control method that integrates online component sensing and intelligent decision-making mechanisms to achieve high-precision, low-energy consumption, and robust operation of the gypsum calcination process. Summary of the Invention

[0007] The technical problem to be solved by this invention is to overcome the unpredictable changes in the thermodynamic properties, especially the specific heat capacity and dehydration reaction rate, caused by fluctuations in the composition of raw materials in the existing gypsum calcination process. The traditional temperature control method based on a proportional-integral-derivative controller uses fixed control parameters and cannot dynamically adapt to the real-time changes in the thermal properties of the material, which leads to temperature overshoot or lag during calcination, increases unnecessary energy consumption, and the method of judging the dehydration endpoint based on the plateau period of the temperature curve has low accuracy, affecting the uniformity and stability of the final product quality.

[0008] To address the aforementioned technical problems, this invention provides a dynamic dry calcination temperature adaptive control method for gypsum production. This method no longer relies on a fixed control model, but instead constructs a real-time thermodynamic state observation and prediction control system deeply coupled with the physical process to achieve precise adaptive adjustment of the calcination temperature and accurate determination of the dehydration endpoint.

[0009] The core of this invention lies in using multimodal sensing technology to monitor multidimensional physical quantities within the calcining furnace in real time. Based on these observations, the equivalent thermodynamic parameters of the gypsum material are estimated online, establishing a dynamic prediction model that reflects the real-time state of the material. Using this model, a model predictive control strategy is employed to derive the optimal heating power sequence for a future period, thereby achieving precise tracking of the temperature curve and optimization of total energy consumption. Simultaneously, this invention introduces real-time dehydration degree based on direct material balance as the core criterion for determining the calcination endpoint, completely eliminating the indirect and unreliable reliance on temperature plateaus in traditional methods.

[0010] This invention provides a dynamic dry calcination temperature adaptive control method for gypsum production. The method includes the following steps: real-time acquisition of multi-dimensional process status data, including real-time material temperature data in the calcination furnace, real-time total mass data of the calcination furnace, exhaust gas humidity data and exhaust gas flow data in the exhaust gas discharge pipe of the calcination furnace, and real-time heating power data of the heating system input to the calcination furnace.

[0011] Based on the multi-dimensional process state data, real-time thermodynamic state estimation of gypsum material is performed. The real-time thermodynamic state estimation step specifically includes calculating the real-time equivalent specific heat capacity and real-time dehydration degree of gypsum material.

[0012] Based on the real-time equivalent specific heat capacity and the real-time dehydration degree, a dynamic thermodynamic prediction model for the calcination process is constructed and updated. The dynamic thermodynamic prediction model is used to describe the trajectory of the material temperature in the calcination furnace in the future prediction time domain under a given heating power input.

[0013] Based on the dynamic thermodynamic prediction model, and combined with the preset target temperature curve and the preset control objective function, the model predictive control algorithm is used to solve for the optimal heating power sequence in the future control time domain.

[0014] The first control variable in the optimal heating power sequence is used as the heating power setpoint at the current moment and output to the heating system of the calcining furnace to control the heating power of the heating system.

[0015] The above steps are repeated until the real-time dehydration degree reaches the preset dehydration endpoint threshold. At this point, the calcination process is determined to be over, and the heating operation of the heating system is stopped.

[0016] As one embodiment of the present invention, the real-time acquisition of multi-dimensional process status data specifically includes: acquiring real-time temperature data of the material through a sheathed thermocouple located at the geometric center of the calcining furnace; acquiring real-time total mass data of the calcining furnace through an array of four high-precision piezoelectric force sensors located below the main structure of the calcining furnace, and summing and digitally filtering the measured values ​​of the four sensors; acquiring exhaust gas humidity data and exhaust gas flow data through an exhaust gas analysis unit located at the exhaust gas main pipe of the calcining furnace, which integrates a zirconium oxide humidity sensor and a thermal gas mass flow meter; and acquiring real-time heating power data by monitoring the control input signal of the heating system or by direct measurement through an external power meter.

[0017] All acquired data is stamped with a high-precision clock synchronization timestamp to ensure data consistency in the time dimension.

[0018] As one embodiment of the present invention, in the real-time thermodynamic state estimation step of the gypsum material, the method for calculating the real-time dehydration degree is as follows: First, before calcination begins, the initial total mass of the gypsum material is acquired and stored; the theoretical total water loss mass is pre-calculated based on the chemical composition analysis of the gypsum; during calcination, the real-time mass of the material at the current moment is obtained by subtracting the preset empty mass of the calcining furnace from the real-time total mass data of the calcining furnace at the current moment; the mass of water removed is obtained by subtracting the real-time mass of the material from the initial total mass; finally, the real-time dehydration degree is obtained by dividing the mass of water removed by the theoretical total water loss mass.

[0019] The calculation process is performed in real time during each control cycle.

[0020] In one embodiment of the present invention, the method for calculating the real-time equivalent specific heat capacity in the real-time thermodynamic state estimation step of the gypsum material is as follows: within a very short time step, the heat balance equation is solved simultaneously. The input terms of the heat balance equation are the real-time heating power data multiplied by the time step. The output terms of the heat balance equation include the increase in sensible heat of the material, the latent heat absorbed by the dehydration reaction, and the heat loss of the system. The increase in sensible heat of the material is determined by the real-time mass of the material, the real-time equivalent specific heat capacity, and the temperature change within the time step. The latent heat absorbed by the dehydration reaction is determined by the product of the mass of water removed within the time step and the latent heat of vaporization of water, wherein the mass of water removed is calculated by the rate of change of the real-time dehydration degree. The heat loss of the system is pre-calibrated using a heat dissipation model related to the furnace surface temperature and the ambient temperature.

[0021] By solving the above equations, the only unknown quantity, namely the real-time equivalent specific heat capacity, can be obtained.

[0022] Furthermore, in order to improve the robustness and accuracy of the state estimation, the real-time thermodynamic state estimation step of the gypsum material adopts the Kalman filter algorithm to fuse and estimate the real-time equivalent specific heat capacity and the real-time dehydration degree.

[0023] The Kalman filter algorithm establishes a state-space model with real-time equivalent specific heat capacity and real-time dehydration degree as state vectors. The process model describes the evolution of these two state variables over time, while the observation model establishes a nonlinear mapping relationship between the state vector and the real-time temperature data of the material, the humidity data of the exhaust gas, and the flow rate data of the exhaust gas.

[0024] By using the two steps of prediction and update in Kalman filtering, the observation information from multiple sensors is optimally fused to obtain the posterior optimal estimate of the two state variables, thereby suppressing the influence of measurement noise from a single sensor and model uncertainty.

[0025] As one embodiment of the present invention, the construction and updating of the dynamic thermodynamic prediction model of the calcination process specifically involves: establishing a first-order nonlinear differential equation based on physical mechanisms to describe the rate of change of material temperature over time.

[0026] The expression for this equation is: ; For real-time heating power, The power that absorbs latent heat for the dehydration reaction; The system heat dissipation power loss due to the average temperature of the material; for Real-time material quality at any given moment; This is the real-time equivalent specific heat capacity.

[0027] As one embodiment of the present invention, the steps of using the model predictive control algorithm to solve for the optimal heating power sequence are as follows: First, a prediction time domain and a control time domain are defined, wherein the length of the prediction time domain is greater than or equal to that of the control time domain; second, a multi-objective optimization function is defined, which includes two parts: the first part is the integral of the square of the deviation between the predicted temperature trajectory calculated by the dynamic thermodynamic prediction model and the preset target temperature curve within the prediction time domain, used to ensure temperature tracking accuracy; the second part is the square integral of the heating power sequence to be solved within the control time domain, multiplied by an adjustable weighting coefficient, used to limit the consumption of control energy.

[0028] Then, in each control cycle, based on the current material state and the updated dynamic thermodynamic prediction model, a numerical optimization algorithm is used to solve for the heating power sequence in the control time domain that minimizes the value of the multi-objective optimization function. This is the optimal heating power sequence.

[0029] The numerical optimization algorithm is either a sequential quadratic programming algorithm or an interior point method.

[0030] Furthermore, the preset target temperature curve is a piecewise function curve, specifically including: the first segment is a rapid heating segment, the goal of which is to heat the material from the initial temperature to the dehydration reaction initiation temperature as quickly as possible; the second segment is a constant-rate dehydration segment, the goal of which is to maintain the material within the most suitable dehydration temperature range to achieve a stable dehydration rate and excellent crystal transformation, and the temperature slope of this segment is close to 0; the third segment is a ripening and heat preservation segment, in which the temperature is maintained at a relatively high temperature for a certain period of time after the main dehydration process is completed to complete the crystal transformation and remove residual free water.

[0031] The model predictive control algorithm employs different weighting coefficients at different stages: in the first stage, it focuses on the heating rate; in the second stage, it focuses on temperature stability and energy consumption; and in the third stage, it focuses on precise temperature maintenance.

[0032] As one embodiment of the present invention, the method for setting the dehydration endpoint threshold is as follows: according to the requirements of the target product, the ideal chemical ratio of dihydrate gypsum, hemihydrate gypsum and anhydrous gypsum in the final product is determined.

[0033] Based on stoichiometry, this ideal chemical ratio is converted into the corresponding theoretical total water loss percentage, which is the dehydration endpoint threshold.

[0034] The typical range of the dehydration endpoint threshold is 0.75 to 0.85, depending on the type of building gypsum or high-strength gypsum to be produced.

[0035] When the real-time dehydration degree first reaches and exceeds this threshold, the control system immediately determines that the calcination process is complete.

[0036] Compared with existing technologies, the advantages of this invention are as follows: 1. It achieves real-time adaptive response to changes in the thermal properties of materials. By estimating the equivalent specific heat capacity online, this method can respond instantly to changes in the thermodynamic properties of materials caused by fluctuations in raw material impurities, particle size distribution, or moisture content, and dynamically adjust the control model and strategy, fundamentally solving the mismatch problem of traditional fixed-parameter controllers when facing changes in operating conditions.

[0037] 2. Improved accuracy of temperature control and energy efficiency. Based on model predictive control algorithms, this method proactively plans future heating power, suppressing temperature overshoot and fluctuations, ensuring that the calcination temperature strictly follows the optimal process curve, and avoiding energy waste caused by repeated adjustments and overheating.

[0038] 3. This method significantly improves the accuracy and reliability of calcination endpoint determination. It uses real-time dehydration degree based on mass balance as the endpoint criterion, directly measuring the progress of the chemical reaction. Its physical meaning is clear and unaffected by factors such as uneven temperature distribution within the furnace or thermocouple measurement point deviations. This ensures that each batch of product completes the reaction at the same degree of chemical transformation, thereby guaranteeing a high degree of consistency and stability in product quality.

[0039] 4. Enhanced robustness of the entire control system. By fusing multi-source sensor information such as temperature, mass, exhaust gas humidity, and flow rate through the Kalman filter algorithm, the system can filter out measurement noise and perform state compensation based on information from other sensors when a sensor drifts or fails, thereby improving the overall system reliability and environmental adaptability. Attached Figure Description

[0040] Figure 1 is a schematic diagram of the overall technical architecture of the dynamic dry steaming and calcination temperature adaptive control method for gypsum production proposed in this invention; Figure 2 is a schematic diagram of the core principle framework of real-time material state observation and predictive control based on multimodal sensing and thermodynamic state estimation in this invention; Figure 3 is a logical flow diagram of multi-dimensional process state data acquisition and synchronous processing in this invention; Figure 4 is a schematic diagram of the fusion calculation framework of real-time thermodynamic state estimation (including equivalent specific heat capacity and dehydration degree) of gypsum material in this invention; Figure 5 is a logical flow diagram of the model predictive control strategy based on dynamic thermodynamic prediction model to solve the optimal heating power sequence in this invention; Figure 6 is a schematic diagram of the multi-level interaction relationship and data flow between the calcining furnace, multi-source sensors and cloud control unit in this invention. Detailed Implementation

[0041] Please refer to Figures 1 to 6. This invention provides a dynamic adaptive control method for dry steaming and calcination temperature in gypsum production. It aims to solve the problem that traditional proportional-integral-derivative controllers, due to parameter rigidity, cannot adjust the calcination temperature or determine the dehydration endpoint when unpredictable changes in the thermodynamic properties of gypsum materials occur due to fluctuations in raw material composition. This method achieves precise adaptive adjustment of the calcination temperature and accurate determination of the dehydration endpoint by constructing a real-time thermodynamic state observation and predictive control system deeply coupled with the physical process.

[0042] The method includes the following steps: S1, acquiring multi-dimensional process state data in real time; S2, estimating the real-time thermodynamic state of gypsum material based on the multi-dimensional process state data; S3, constructing and updating a dynamic thermodynamic prediction model for the calcination process based on the real-time equivalent specific heat capacity and the real-time dehydration degree; S4, using a model predictive control algorithm to solve for the optimal heating power sequence in the future control time domain, based on the dynamic thermodynamic prediction model and combined with a preset target temperature curve and a preset control objective function; S5, using the first control variable in the optimal heating power sequence as the heating power setpoint at the current moment and outputting it to the heating system of the calcination furnace to control the heating power of the heating system; S6, repeating the above steps until the real-time dehydration degree reaches a preset dehydration endpoint threshold, at which point the calcination process is determined to be over, and the heating operation of the heating system is stopped.

[0043] In step S1, multi-dimensional process status data is acquired in real time. The multi-dimensional process status data includes real-time material temperature data in the calcining furnace, real-time total mass data of the calcining furnace, exhaust gas humidity data and exhaust gas flow data in the exhaust gas emission pipe of the calcining furnace, and real-time heating power data of the heating system input to the calcining furnace.

[0044] To ensure the accuracy and consistency of data acquisition, all sensors are equipped with high-precision time synchronization modules, enabling all data streams to be sampled and recorded under a unified time reference.

[0045] Real-time material temperature data is obtained through a sheathed thermocouple positioned at the geometric center inside the calcining furnace.

[0046] This thermocouple features high temperature resistance, corrosion resistance, and fast response. Its temperature measurement range covers room temperature to 900 degrees Celsius, with a resolution greater than 0.1 degrees Celsius and a sampling frequency set to 10 times per second.

[0047] The real-time total mass data of the calcining furnace is obtained through an array of four high-precision piezoelectric force sensors located below the main structure of the calcining furnace.

[0048] The four sensors are arranged in a rectangle and are located at the support feet of the furnace body. Each sensor has a range of 0 to 5 tons and a resolution of 100 grams.

[0049] The raw output signals from the four sensors are converted from analog to digital and then sent to the central processing unit for summation. A low-pass digital filter is applied to eliminate high-frequency noise caused by mechanical vibration, ultimately yielding a stable value reflecting the total mass of the material inside the furnace.

[0050] The exhaust gas humidity and exhaust gas flow data are obtained by the exhaust gas analysis unit located at the exhaust gas discharge main of the calcining furnace.

[0051] This unit integrates a zirconium oxide humidity sensor and a thermal gas mass flow meter.

[0052] The zirconia humidity sensor measures the partial pressure of water vapor in exhaust gas based on the principle of oxygen ion conductivity, and then converts it into relative humidity or absolute humidity. The measurement range is 0 to 100% relative humidity, with an accuracy of ±1.5%.

[0053] The thermal gas mass flow meter measures the mass flow rate of exhaust gas using the constant temperature difference method. The range is 0 to 2000 standard cubic meters per hour, and the repeatability error is less than 0.5%.

[0054] Real-time heating power data is obtained by monitoring the control input signal of the heating system, or by directly measuring the instantaneous values ​​of voltage and current in the three-phase AC power supply line through an external high-precision power meter. The active power is then calculated after Fourier transform and power factor correction.

[0055] All the above data were immediately stamped with a timestamp generated synchronously by an atomic clock after collection. The timestamp accuracy was better than 1 millisecond to ensure time alignment during subsequent multi-source data fusion processing.

[0056] In step S2, the real-time thermodynamic state of the gypsum material is estimated based on the multi-dimensional process state data.

[0057] The core of this estimation process is calculating the real-time equivalent specific heat capacity and real-time dehydration degree of the gypsum material. The calculation of the real-time dehydration degree is based on the principle of direct material balance.

[0058] Before calcination begins, the control system records and stores the initial total mass M0 of the gypsum material.

[0059] The initial total mass The real-time total mass data obtained in step S1 Subtract the pre-calibrated no-load mass of the calcining furnace To obtain, that is: Simultaneously, based on the chemical composition analysis report of the incoming batch, the theoretical total water loss of this batch of gypsum material was calculated in advance. This theoretical value is determined based on the difference between the molecular weight of gypsum (CaSO4·2H2O) and that of anhydrous gypsum (CaSO4), combined with the actual percentage of dihydrate gypsum in the material. During the calcination process... Real-time material quality : ; the quality of the removed moisture Real-time dehydration Defined as the ratio of the mass of water removed to the theoretical total water loss, i.e.; This calculation is performed in real time within each control cycle, which is set to two seconds.

[0060] Real-time equivalent specific heat capacity The calculation is based on an extremely short time step. The internal heat balance equation. This equation states that the net heat input equals the sum of the sensible heat increment of the material, the latent heat consumption of the dehydration reaction, and the system's heat dissipation loss. The mathematical expression is: ; This refers to the real-time heating power. For time step The temperature change of the material inside is calculated from the data of the armored thermocouples. The latent heat of vaporization of water under the current pressure is taken as 2260 kJ per kilogram; For time step The mass of water removed internally is calculated by the rate of change of the real-time degree of dehydration, i.e.: ; for Dehydration level at all times; The power loss due to heat dissipation of the system is the furnace temperature. The function is established by using a lookup table function or a fitting polynomial through prior cold and hot calibration experiments. A typical form is: ;in , For calibration coefficients, Let be the ambient temperature. In this equation, except... Apart from the above, all other variables are either known or can be calculated from step S1 and the aforementioned degree of dehydration. Therefore, the real-time equivalent specific heat capacity can be directly solved by algebraic operations.

[0061] To further improve the robustness and accuracy of state estimation, this embodiment employs a Kalman filter algorithm to fuse and estimate the real-time equivalent specific heat capacity and real-time dehydration degree. The state vector of this Kalman filter... : The process model describes the evolution of the state vector. It is assumed that within the control period, the changes in state variables are mainly affected by the slow evolution of the material's own physicochemical properties; therefore, a first-order Markov process model is adopted. ,in The noise is a process noise that follows a zero-mean Gaussian distribution.

[0062] Observation model State vectors and multi-sensor observations were established. The nonlinear mapping relationship between them.

[0063] Including real-time material temperature Exhaust gas humidity and exhaust gas flow rate .

[0064] The relationship with the state vector is implicit in the heat balance equation; Exhaust gas flow rate It works in conjunction with the pressure, temperature and dehydration rate inside the furnace.

[0065] The Kalman filter performs a prediction step (predicting prior states based on the process model) and an update step (utilizing...) within each control cycle. The prediction is revised to obtain the optimal posterior estimate.

[0066] This multi-source information fusion mechanism suppresses estimation bias caused by single sensor drift, such as thermocouple drift or zero-point drift of mass sensor.

[0067] In step S3, a dynamic thermodynamic prediction model for the calcination process is constructed and updated based on the latest estimated real-time equivalent specific heat capacity and real-time dehydration degree. This model uses a first-order nonlinear differential equation based on physical mechanisms to describe the average temperature of the material. rate of change over time Its expression is: ; The power for absorbing latent heat in a dehydration reaction is calculated using the following formula: ; The system heat dissipation loss power is the average temperature of the material.

[0068] This differential equation fully describes the dynamic equilibrium relationship between heating, heat dissipation, phase change heat absorption, and material temperature rise.

[0069] At the beginning of each control cycle, the control system uses the latest state estimate of equivalent specific heat capacity and dehydration degree output by the Kalman filter in step S2, substitutes it into the above equation, and completes the online update of the prediction model parameters to ensure that the model can closely track the actual physical process.

[0070] In step S4, the optimal heating power sequence is solved using a model predictive control algorithm based on the updated dynamic thermodynamic prediction model. First, the prediction time domain is defined. 600 seconds, controlling the time domain The timeframe is 60 seconds, meaning the algorithm plans the temperature trajectory for the next 10 minutes, but only optimizes the heating power for the next minute. Secondly, a multi-objective optimization function is defined. : ; yes The timing is predicted by a dynamic thermodynamic model for a given future heating power sequence. The predicted temperature trajectory obtained from the simulation; The preset target temperature curve; It is an adjustable weighting coefficient used to balance temperature tracking accuracy and energy consumption.

[0071] Target temperature curve It is a piecewise function: the first segment is a rapid heating segment from 0 to 300 seconds, linearly increasing the temperature from room temperature to 170 degrees Celsius; the second segment is a constant-rate dehydration segment from 300 to 1500 seconds, maintaining the temperature at 170 degrees Celsius; the third segment is a curing and holding segment after 1500 seconds, increasing the temperature to 190 degrees Celsius and holding it thereafter.

[0072] At different stages, Take different values: rapid heating section The smaller value is selected to prioritize the heating rate; a larger value is selected for the constant-rate dehydration section to suppress power fluctuations and save energy; the curing and heat preservation section... A middle value is chosen to balance accuracy and energy consumption.

[0073] Then, in each control cycle, using the current material temperature, mass, estimated equivalent specific heat capacity, and degree of dehydration as initial values ​​for the model, the dynamic thermodynamic prediction model is discretized, and a sequential quadratic programming algorithm is used to solve it. The smallest future Heating power sequence within seconds This sequence is the optimal heating power sequence. They represent time respectively The control quantity.

[0074] In step S5, the first control quantity in the optimal heating power sequence is used as the heating power setpoint at the current moment and sent to the heating system of the calcining furnace through the analog output module or digital communication interface.

[0075] After receiving the set value, the heating system adjusts the electrical energy supplied to the heating element through its internal power regulation unit (such as a thyristor voltage regulator or frequency converter), thereby precisely controlling the actual heating power.

[0076] In step S6, the control system continuously executes steps S1 to S5 in a loop. At the end of each cycle, it checks whether the real-time dehydration degree has reached the preset dehydration endpoint threshold. .

[0077] This threshold is set based on the chemical composition requirements of the target product. For example, if the target product is building gypsum (its main component is hemihydrate gypsum CaSO4·0.5H2O), the theoretical degree of dehydration is 75%; if it is high-strength gypsum, a higher degree of dehydration, such as 82%, may be required. Therefore, The typical value range is 0.75 to 0.85.

[0078] When the real-time dehydration degree is first greater than or equal to At that time, the control system immediately issues a command to set the heating power setting to 0, shut down the heating system, and simultaneously start the cooling program, thus announcing the end of this calcination batch.

[0079] Through the closed-loop execution of the above six steps, this invention achieves adaptive control of the entire lifecycle of the gypsum dry steaming and calcination process. The entire control logic completely abandons the dependence on fixed PID parameters and instead adopts an advanced control strategy based on real-time physical state estimation and model prediction, fundamentally solving the control problems caused by raw material fluctuations and ensuring the stability of product quality and the high efficiency of energy utilization.

[0080] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0081] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dynamic dry steaming and calcination temperature adaptive control method for gypsum production, characterized in that, include: Real-time acquisition of multi-dimensional process status data, including real-time material temperature data in the calcining furnace, real-time total mass data of the calcining furnace, exhaust gas humidity data and exhaust gas flow data in the exhaust gas emission pipe of the calcining furnace, and real-time heating power data of the heating system input to the calcining furnace; based on the multi-dimensional process status data, real-time thermodynamic state estimation of gypsum material is performed, including calculation of real-time equivalent specific heat capacity and real-time dehydration degree of gypsum material; Based on the real-time equivalent specific heat capacity and the real-time dehydration degree, a dynamic thermodynamic prediction model for the calcination process is constructed and updated. This model describes the trajectory of material temperature change in the calcination furnace over a given heating power input in the future prediction time domain. According to the dynamic thermodynamic prediction model, and combined with a preset target temperature curve and a preset control objective function, a model predictive control algorithm is used to solve for the optimal heating power sequence in the future control time domain. The first control variable in the optimal heating power sequence is used as the heating power setpoint for the current moment and output to the heating system of the calcination furnace to control the heating power of the heating system. The above steps are repeated until the real-time dehydration degree reaches a preset dehydration endpoint threshold. At this point, the calcination process is considered complete, and the heating operation of the heating system is stopped.

2. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 1, characterized in that, The real-time acquisition of multi-dimensional process status data includes: acquiring real-time material temperature data through a sheathed thermocouple positioned at the geometric center of the calcining furnace; acquiring real-time total mass data of the calcining furnace through an array of four high-precision piezoelectric force sensors located below the main structure of the calcining furnace, and summing and digitally filtering the measured values ​​of the four sensors; acquiring exhaust gas humidity and flow rate data through an exhaust gas analysis unit integrated with a zirconia humidity sensor and a thermal gas mass flow meter located at the exhaust gas main pipe of the calcining furnace; acquiring real-time heating power data by monitoring the control input signal of the heating system or by direct measurement through an external power meter; all acquired data are stamped with a high-precision clock synchronization timestamp to ensure data consistency in the time dimension.

3. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 2, characterized in that, The calculation of the real-time dehydration degree includes: acquiring and storing the initial total mass of the gypsum material before calcination begins; pre-calculating the theoretical total water loss mass based on the chemical composition analysis of the gypsum; during calcination, subtracting the preset empty mass of the calcining furnace from the current real-time total mass data of the calcining furnace to obtain the current real-time mass of the material; subtracting the real-time mass of the material from the initial total mass to obtain the mass of water removed; and dividing the mass of water removed by the theoretical total water loss mass to obtain the real-time dehydration degree.

4. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 3, characterized in that, The calculation of the real-time equivalent specific heat capacity includes: solving a heat balance equation simultaneously within a very short time step. The input term of the heat balance equation is the real-time heating power data multiplied by the time step. The output term of the heat balance equation includes the increase in sensible heat of the material, the latent heat absorbed by the dehydration reaction, and the system heat loss. The increase in sensible heat of the material is determined by the real-time mass of the material, the real-time equivalent specific heat capacity, and the temperature change within the time step. The latent heat absorbed by the dehydration reaction is determined by the product of the mass of water removed within the time step and the latent heat of vaporization of water, wherein the mass of water removed is calculated by the rate of change of the real-time dehydration degree. The system heat loss is pre-calibrated using a heat dissipation model related to the furnace surface temperature and ambient temperature. Through the above equations, the unique unknown quantity, namely the real-time equivalent specific heat capacity, is solved.

5. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 4, characterized in that, The real-time thermodynamic state estimation of the gypsum material employs a Kalman filter algorithm to fuse and estimate the real-time equivalent specific heat capacity and the real-time dehydration degree. The Kalman filter algorithm establishes a state-space model with the real-time equivalent specific heat capacity and the real-time dehydration degree as state vectors. The process model describes the evolution of these two state variables over time, while the observation model establishes a nonlinear mapping relationship between the state vectors and the real-time temperature data of the material, the humidity data of the exhaust gas, and the flow rate data of the exhaust gas. Through the prediction and update steps of the Kalman filter, the observation information from multiple sensors is optimally fused to obtain the posterior optimal estimate of the two state variables.

6. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 5, characterized in that, The construction and updating of the dynamic thermodynamic prediction model for the calcination process includes: establishing a first-order nonlinear differential equation based on physical mechanisms to describe the rate of change of material temperature over time. The expression for this equation is: ； For real-time heating power, The power that absorbs latent heat for the dehydration reaction; The system heat dissipation power loss due to the average temperature of the material; for Real-time material quality at any given moment; This is the real-time equivalent specific heat capacity.

7. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 6, characterized in that, The method of using model predictive control algorithm to solve for the optimal heating power sequence in the future control time domain includes: defining a prediction time domain and a control time domain, wherein the length of the prediction time domain is greater than or equal to that of the control time domain; defining a multi-objective optimization function, which includes: the integral of the square of the deviation between the predicted temperature trajectory calculated by the dynamic thermodynamic prediction model and the preset target temperature curve in the prediction time domain; and the square integral of the heating power sequence to be solved in the control time domain, multiplied by an adjustable weighting coefficient; and in each control cycle, based on the current material state and the updated dynamic thermodynamic prediction model, using a numerical optimization algorithm to solve for the heating power sequence in the control time domain that minimizes the value of the multi-objective optimization function, which is the optimal heating power sequence.

8. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 7, characterized in that, The preset target temperature curve is a piecewise function curve, which includes: a first segment is a rapid heating segment, the goal of which is to heat the material from the initial temperature to the dehydration reaction initiation temperature as quickly as possible; a second segment is a constant-rate dehydration segment, the goal of which is to maintain the material within the most suitable dehydration temperature range to achieve a stable dehydration rate and excellent crystal transformation; a third segment is a ripening and heat preservation segment, in which the temperature is maintained at a higher temperature for a certain period of time after the main dehydration process is completed to complete the crystal transformation and remove residual free water; the model predictive control algorithm uses different weighting coefficients at different stages, focusing on the heating rate in the first segment, temperature stability and energy consumption in the second segment, and precise temperature maintenance in the third segment.

9. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 8, characterized in that, The method for setting the dehydration endpoint threshold is as follows: Based on the requirements of the target product, determine the ideal chemical ratio of dihydrate gypsum, hemihydrate gypsum, and anhydrous gypsum in the final product; based on the stoichiometric relationship, convert this ideal chemical ratio into the corresponding theoretical total water loss percentage, which is the dehydration endpoint threshold; the typical range of the dehydration endpoint threshold is 0.75 to 0.85; when the real-time dehydration degree first reaches and exceeds this threshold, the control system immediately determines that the calcination process is complete.

10. The dynamic dry steaming and calcination temperature adaptive control method for gypsum production according to claim 9, characterized in that, The numerical optimization algorithm is either a sequential quadratic programming algorithm or an interior point method.

Images