Fireproof glass production and processing parameter monitoring system based on cloud platform

Abstract

The invention relates to the technical field of fireproof glass production and processing control, in particular to a fireproof glass production and processing parameter monitoring system based on a cloud platform, and the system comprises a theoretical model unit which is used for solving a theoretical process temperature representing an ideal process reference; the deviation quantization unit is used for calculating a model mismatch index; wherein the model mismatch index is used for representing the deviation degree of the actual process temperature and the theoretical process temperature; the state judgment unit is used for comparing and analyzing the model mismatch index and a preset trigger threshold value, generating a systematic deviation signal when the model mismatch index is continuously higher than the trigger threshold value, and otherwise, generating a normal noise signal; the model correction unit is used for generating a dynamic energy absorption coefficient; the compensation control unit is used for generating a prospective compensation control instruction and issuing the prospective compensation control instruction to the heating unit; according to the invention, the accuracy of long-term monitoring and control is ensured, and the stability of the production process is remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fireproof glass production and processing control, in particular to a fireproof glass production and processing parameter monitoring system based on a cloud platform. BACKGROUND

[0002] In the production process of high-performance materials such as fireproof glass, accurate control of process parameters is a key link to ensure product quality and consistency. However, the actual production environment is complex and variable, and nonlinear and time-varying disturbance factors such as differences in physical and chemical properties between batches of raw materials and natural aging of production equipment over time continuously interfere with the production process, causing actual process parameters to deviate from ideal process benchmarks. Traditional control methods usually rely on lag compensation for deviations that have already occurred. This passive control mode is difficult to effectively address the above disturbances. When process parameters have deviated, the system only begins to adjust, which may result in insufficient stability of the production process, reduced yield, and difficulty in ensuring the consistency of the final product quality. Therefore, how to achieve early prediction and active compensation of process parameters to overcome the effects of raw material fluctuations and equipment aging and to change lagging passive control to forward-looking active control has become a technical problem to be solved in the field. SUMMARY

[0003] To solve the above technical problems, the present application provides a fireproof glass production and processing parameter monitoring system based on a cloud platform. Specifically, the technical solution of the present application comprises: A theoretical model unit is used to construct a first-principles thermodynamic model based on pre-set material physical and chemical parameters and to calculate a theoretical process temperature representing an ideal process benchmark. A deviation quantification unit is used to obtain an actual process temperature collected by a production line sensor in real time and to calculate a model mismatch index in combination with the theoretical process temperature. The model mismatch index is used to represent the deviation degree of the actual process temperature from the theoretical process temperature. A state discrimination unit is used to compare and analyze the model mismatch index with a pre-set trigger threshold. When the model mismatch index is continuously higher than the trigger threshold, a systematic deviation signal is generated, otherwise a normal noise signal is generated. A model correction unit is used to dynamically correct the theoretical energy absorption coefficient in the first-principles thermodynamic model based on the deviation between the historical actual process temperature and the theoretical process temperature in response to the systematic deviation signal, and to generate a dynamic energy absorption coefficient. A compensation control unit is used to construct a corrected first-principles thermodynamic model based on the dynamic energy absorption coefficient and to generate a forward-looking compensation control instruction according to the corrected first-principles thermodynamic model and to issue the instruction to a heating unit.

[0004] Optionally, the theoretical model unit is specifically used for: acquiring preset material specific heat capacity, density, effective total mass, comprehensive heat exchange coefficient and effective heat dissipation area; determining theoretical energy absorption coefficient and theoretical heat dissipation coefficient in the first-principle thermodynamics model based on the material specific heat capacity, the density, the effective total mass, the comprehensive heat exchange coefficient and the effective heat dissipation area; inputting the heating unit input power at the current moment into the first-principle thermodynamics model containing the theoretical energy absorption coefficient and the theoretical heat dissipation coefficient to solve out the theoretical process temperature.

[0005] Optionally, the theoretical energy absorption coefficient is used for representing a temperature theoretical increment of the system caused by unit input power in unit time; and the theoretical heat dissipation coefficient is used for representing comprehensive heat transfer capacity of the system and the environment.

[0006] Optionally, the deviation quantification unit is specifically used for: calculating an absolute deviation of the actual process temperature and the theoretical process temperature; dividing the absolute deviation by a preset sensor system noise standard deviation to generate a model mismatch index.

[0007] Optionally, the model correction unit is specifically used for performing cumulative effect calculation on historical deviations of the actual process temperature and the theoretical process temperature by using an integral controller to generate a dynamic energy absorption coefficient.

[0008] Optionally, the model correction unit is further used for performing judgment by using a gating function, and only when the model mismatch index is greater than a trigger threshold, the generation process of the dynamic energy absorption coefficient is started.

[0009] Optionally, the compensation control unit is specifically used for: substituting the dynamic energy absorption coefficient into the first-principle thermodynamics model and inputting a planned basic power instruction to predict a process temperature after a future time step to obtain a predicted process temperature.

[0010] Optionally, the compensation control unit is further used for: calculating an expected temperature deviation between the predicted process temperature and a theoretical process temperature corresponding to the future time step; calculating a power compensation amount of the basic power instruction based on the expected temperature deviation; combining the basic power instruction and the power compensation amount to generate an optimized actual power instruction, and taking the actual power instruction as a compensation control instruction.

[0011] Compared with the prior art, the present application has the following beneficial effects: 1. The present application converts the traditional lag compensation control into proactive predictive control by constructing the theoretical benchmark of the ideal process and predicting the future process temperature deviation based on the corrected model, which can generate compensation instructions in advance to eliminate the impending deviation, thereby effectively suppressing process fluctuations and significantly improving the stability of the production process; 2. The present application has self-adaptive learning ability and can dynamically correct the key parameters of the model by continuously analyzing the deviation between the actual process and the theoretical model, which enables the system to autonomously recognize and adapt to changes in physical properties caused by raw material batch replacement or equipment aging, ensuring the accuracy of long-term monitoring and control; 3. The present application can intelligently distinguish between systematic process deviation and random sensor noise by statistically standardizing process deviation and setting a trigger threshold, which avoids excessive adjustment of normal fluctuations and greatly enhances the robustness and reliability of system control; 4. The present application can effectively overcome disturbances such as raw material fluctuations and equipment aging through proactive and adaptive closed-loop control, making actual production parameters closely match the ideal process benchmark, which directly ensures the stability of product quality and consistency between batches, thereby significantly improving the yield of fire-resistant glass production. BRIEF DESCRIPTION OF DRAWINGS

[0012] The present application will be further explained in conjunction with the accompanying drawings and examples: Figure 1 is a structural diagram of the system of the present application. DETAILED DESCRIPTION

[0013] To make the purpose, technical scheme and advantages of the present application clearer, the present application will be further explained in conjunction with specific examples.

[0014] Example 1: Please refer to Figure 1 A fire-resistant glass production and processing parameter monitoring system based on a cloud platform, comprising: a theoretical model unit for constructing a first-principle thermodynamic model based on pre-set material physical and chemical parameters and solving a theoretical process temperature representing an ideal process benchmark; a deviation quantification unit for acquiring an actual process temperature collected by a production line sensor in real time and calculating a model mismatch index in combination with the theoretical process temperature, wherein the model mismatch index is used to represent the deviation degree of the actual process temperature from the theoretical process temperature; a state discrimination unit for comparing and analyzing the model mismatch index with a pre-set trigger threshold, generating a systematic deviation signal when the model mismatch index continuously exceeds the trigger threshold, and otherwise generating a normal noise signal; a model correction unit configured to, in response to the systematic deviation signal, dynamically correct a theoretical energy absorption coefficient in the first-principle thermodynamic model based on a deviation between the historical actual process temperature and the theoretical process temperature, to generate a dynamic energy absorption coefficient; a compensation control unit configured to, based on the dynamic energy absorption coefficient, construct a corrected first-principle thermodynamic model, and generate a forward-looking compensation control instruction according to the corrected first-principle thermodynamic model, and send the compensation control instruction to the heating unit.

[0015] The embodiment provides a closed-loop production and processing parameter monitoring system with adaptive cognitive ability; the system aims to solve the problem of deviation of fireproof glass production process caused by nonlinear and time-varying disturbances such as raw material batch fluctuation and equipment aging; the system establishes an ideal process benchmark model based on physical and chemical principles, quantifies the deviation between the actual production process and the ideal model in real time, and intelligently identifies the nature of the deviation; once the systematic deviation is identified, the system can adaptively correct the benchmark model to recognize the real physical properties of the current production batch, and generate a forward-looking compensation control instruction based on the corrected model, thereby changing passive lag control to active predictive control, and realizing accurate and stable monitoring of the processing parameters; The purpose of the theoretical model unit is to construct a first-principle thermodynamic model independent of historical production data and representing an ideal process benchmark, and to calculate a theoretical process temperature; in the embodiment, the unit constructs an energy conservation equation based on preset material physical and chemical parameters, and aggregates key parameters in the equation into a theoretical energy absorption coefficient and a theoretical heat dissipation coefficient, thereby establishing a mathematical model capable of describing the temperature change trajectory in an ideal state; The purpose of the deviation quantification unit is to accurately measure the deviation between the actual production process and the theoretical benchmark; in the embodiment, the unit acquires the actual process temperature in real time through an on-line sensor, compares it with the theoretical process temperature at the same time output by the theoretical model unit, and calculates the absolute deviation therebetween; in order to eliminate the influence of sensor noise and realize standardized evaluation, the unit further introduces a sensor system noise standard deviation to normalize the absolute deviation, and finally generates a dimensionless model mismatch index; the model mismatch index refers to a deviation value that is statistically standardized, and its role is to measure the significance of the current deviation in a statistical sense, and its source is the difference between the actual process temperature and the theoretical process temperature divided by the preset noise standard deviation; The state discrimination unit aims to distinguish random process fluctuations from systematic deviations indicating process abnormalities; in this embodiment, the unit continuously compares the model mismatch index generated by the deviation quantification unit with a preset trigger threshold; the trigger threshold refers to a critical value for determining whether the deviation is significant, which acts as a gating condition for starting the model correction mechanism, and its source can be set according to statistical process control theories such as six sigma criteria; when the model mismatch index is continuously higher than the trigger threshold within a certain time window, the unit determines that a systematic deviation has occurred and generates a systematic deviation signal; otherwise, it is considered to be normal noise disturbance, and a normal noise signal is generated; The purpose of the model correction unit is to respond to the systematic deviation signal and adaptively correct the first-principle thermodynamic model, so that it can recognize and fit the real physical characteristics of the current production batch; in this embodiment, the unit assumes that the observed systematic deviation is mainly due to changes in material energy absorption characteristics, so it uses integral control principle to calculate the cumulative effect of the deviation based on the deviation between the actual process temperature and the theoretical process temperature, thereby dynamically correcting the theoretical energy absorption coefficient in the model to generate a dynamic energy absorption coefficient that reflects the current working condition; The purpose of the compensation control unit is to use the corrected model to predict future states and generate forward-looking compensation control instructions to actively eliminate future process deviations; in this embodiment, the unit constructs a corrected first-principle thermodynamic model based on the dynamic energy absorption coefficient generated by the model correction unit; according to this corrected model, the original planned base power instruction is input to predict the process temperature after a time step; by comparing the predicted temperature with the ideal theoretical process temperature, the expected temperature deviation is calculated, and a power compensation amount is generated accordingly; the compensation amount is combined with the base power instruction to form the final compensation control instruction issued to the heating unit; the present application realizes intelligent monitoring of the production process of fireproof glass by constructing a complete cognitive control closed loop from theoretical modeling to deviation quantification, state discrimination, model correction, and compensation control; it not only can accurately track and compensate for process disturbances caused by raw material batch fluctuations and other unknown factors, but also can learn the physical properties of new materials through the model self-correction mechanism, transforming the traditional passive and lagging control mode into an active and forward-looking predictive control mode, greatly improving the stability, yield, and quality consistency of the production process.

[0016] Embodiment 2: The theoretical model unit is specifically used for: Obtaining the preset material specific heat capacity, density, effective total mass, comprehensive heat exchange coefficient, and effective heat dissipation area; Based on the material's specific heat capacity, density, effective total mass, comprehensive heat transfer coefficient, and effective heat dissipation area, the theoretical energy absorption coefficient and theoretical heat dissipation coefficient in the first-principles thermodynamic model are determined. The current heating unit input power is input into a first-principles thermodynamic model that includes the theoretical energy absorption coefficient and the theoretical heat dissipation coefficient to calculate the theoretical process temperature. The theoretical energy absorption coefficient is used to characterize the theoretical temperature increment of the system caused by a unit input power per unit time; the theoretical heat dissipation coefficient is used to characterize the overall heat transfer capacity of the system and the environment.

[0017] This embodiment is a concrete implementation of the theoretical model unit; the construction of this unit does not rely on historical big data, but returns to the physical nature of the material, thereby ensuring the model's universality and interpretability. The theoretical model unit obtains key thermodynamic parameters for specific fire-resistant glass formulations, including the material's specific heat capacity, by consulting material handbooks or conducting experimental measurements. ,density and effective total mass Simultaneously, the comprehensive heat transfer coefficient under the production environment is determined through experimental estimation or theoretical calculation. With effective heat dissipation area ; Based on the law of conservation of energy, this unit establishes a differential equation describing the temperature change of the glass system over time: ; To facilitate system modeling and control, this embodiment rearranges the equation into a lumped parameter form, thereby defining two core theoretical coefficients: the theoretical energy absorption coefficient and the theoretical energy absorption coefficient. The calculation method is as follows: ; in, The energy transfer efficiency from the heater to the glass is a dimensionless parameter; its physical meaning is precisely defined as: the theoretical temperature increment of the entire system per unit input power per unit time, taking into account the material's heat capacity effect, and its dimensions are: Theoretical heat dissipation coefficient The calculation method is as follows: ; This coefficient characterizes the overall heat transfer capacity of the system as a whole with its surrounding environment, and its dimensions are... It reflects the rate at which the system's temperature naturally decreases due to heat dissipation when there is no external energy input. To determine the comprehensive parameters, a calibration experiment was performed during the initial deployment of the system; during the calibration process, a known constant heating power was applied to a standard batch of glass. The actual temperature is continuously monitored until thermal equilibrium is reached, and the steady-state temperature at this point is recorded. Under thermal equilibrium, the rate of temperature change is zero. Substituting this into the energy conservation equation, we can obtain... Furthermore, a cooling experiment can be performed, which involves stopping heating and recording the cooling curve data showing the temperature decay over time; by performing an exponential fit on this cooling curve data, the solution can be obtained separately. and The precise numerical values ​​are obtained; the theoretical model unit substitutes these two coefficients to form the following theoretical process temperature evolution equation: ; in, for The theoretical process temperature at time t is calculated using this equation; for The input power of the heating unit is given by the control system. The ambient temperature is collected by a sensor; by integrating this equation and providing the initial material temperature, the theoretical process temperature at any given time can be calculated. This provides an ideal benchmark for subsequent deviation analysis. By defining and constructing theoretical model units in this way, the process benchmark model is based entirely on measurable and verifiable basic physicochemical parameters. This not only enhances the transparency and interpretability of the model, but also enables the model to be transferred across equipment and formulations, laying a solid foundation for subsequent precise monitoring and control.

[0018] Example 3: The deviation quantization unit is specifically used for: Calculate the absolute deviation between the actual process temperature and the theoretical process temperature; The absolute deviation is divided by the preset standard deviation of the sensor system noise to generate the model mismatch index.

[0019] This embodiment is a concrete implementation of the deviation quantification unit; its design introduces statistical principles to standardize the original deviation, thereby achieving an objective and quantitative assessment of the significance of the deviation. The operation of the deviation quantization unit consists of two steps: Calculate the absolute deviation; this unit acquires the actual temperature measurement value of the glass in real time through temperature sensors deployed on the production line. Simultaneously, it receives the theoretical process temperature from the theoretical model unit at the same time. ; Calculate the absolute deviation between the two ; The model mismatch index is generated; to effectively distinguish between real process disturbances and the random measurement noise of the sensor itself, this embodiment introduces the sensor system noise standard deviation. Standard deviation of sensor system noise This refers to a statistical measure of the fluctuation in the sensor's output signal under stable operating conditions. Its function is to serve as a benchmark for deviation standardization. It is obtained by collecting sensor data over a period of time and performing statistical analysis during the equipment commissioning phase. This unit uses the following formula to generate the model mismatch index. : ; By dividing the absolute deviation by The original deviation value, which has a temperature dimension, is converted into a standardized, dimensionless exponent. The magnitude of this exponent directly reflects how many times the current deviation is a multiple of the sensor's normal fluctuations, thus providing a clear measure of its statistical significance. This design makes deviation quantification no longer a simple numerical comparison. Through normalization, the system can effectively filter out interference from the sensor's inherent noise. The model mismatch exponent provides a unified evaluation standard independent of specific temperature dimensions, allowing subsequent state discrimination units to make decisions based on a stable and statistically significant indicator, greatly improving the robustness and reliability of the monitoring system.

[0020] Example 4: The model correction unit is specifically used to calculate the cumulative effect of the historical deviation between the actual process temperature and the theoretical process temperature using an integral controller, so as to generate a dynamic energy absorption coefficient. The model correction unit is also used to use a gating function to determine whether the dynamic energy absorption coefficient generation process will be initiated only when the model mismatch index is greater than the trigger threshold.

[0021] This embodiment is a concrete implementation of the model correction unit; it integrates the integral idea and gating logic in classical control theory to construct an intelligent correction mechanism that can effectively correct steady-state errors and avoid overreacting to normal noise. The core of the model correction unit is the theoretical energy absorption coefficient. Dynamic adjustments are made to generate a dynamic energy absorption coefficient. The correction process is defined by the following equation: ; Cumulative effect calculation is performed using the principle of an integral controller; the integral term in the equation It is a continuous accumulation of historical deviations; when the actual energy absorption characteristics deviate continuously due to reasons such as changes in raw material batches, this integral term will continuously accumulate, driving... Make adjustments until the corrected model output is achieved. Able to approach reality again This eliminates systematic biases; model correction rate learning gain which controls the correction speed of the parameters, and its function is to adjust the sensitivity of the model to the deviation, and its source is to set it according to the actual process requirements; selective start-up is performed by using a gating function; the gating function plays a role of a gating mechanism; the gating function has a value logic as follows: when the model mismatch index is greater than the triggering threshold , the value is 1, so that the integral correction is started; otherwise, the value is 0, and the integral correction process is suspended; this function ensures that the integral correction mechanism is activated only when the state discrimination unit determines that a systematic deviation with statistical significance has occurred; the combination of these two features gives the model correction unit high intelligence; the integral effect ensures that the system can learn from the continuous deviation and ultimately eliminate the steady-state error introduced by unknown disturbances; the introduction of the gating function avoids unnecessary correction of normal and random noise, ensuring the stability and convergence of the model correction process.

[0022] Embodiment 5: The compensation control unit is specifically used for: substituting the dynamic energy absorption coefficient into the first-principle thermodynamic model, and inputting the originally planned basic power instruction to predict the process temperature after a future time step, to obtain a predicted process temperature; The compensation control unit is also used for: calculating an expected temperature deviation between the predicted process temperature and a theoretical process temperature corresponding to the future time step; based on the expected temperature deviation, calculating a power compensation amount for the basic power instruction; combining the basic power instruction and the power compensation amount to generate an optimized actual power instruction, and taking the actual power instruction as a compensation control instruction.

[0023] This embodiment is a specific implementation of the compensation control unit; it constructs a forward-looking control framework that predicts future states and compensates in advance; based on a dynamic model to predict future states; this unit uses the dynamic energy absorption coefficient updated in real time by the model correction unit to replace the static coefficient in the theoretical model, thereby obtaining a dynamic model that accurately reflects the current working condition; it takes the basic power instruction originally planned to be issued to the heating unit at the current time as input, and uses the dynamic model to predict the process temperature after a short time step in the future, i.e., the predicted process temperature The prediction process can solve the dynamic model equation by numerical methods such as the first-order Euler method: ; Calculate the expected temperature deviation and the power compensation amount; the unit calculates the future time under ideal conditions ; by comparing the two, the expected temperature deviation is obtained ; this deviation indicates how much process deviation will occur in the future if it is executed as planned; based on this expected deviation, the unit calculates the power compensation amount required for the basic power command; Generate an optimized actual power command; the unit combines the calculated power compensation amount with the basic power command to generate an optimized actual power command , which is issued as the final compensation control command to the heating unit, and its calculation formula is as follows: ; Where the look-ahead control gain is a tunable parameter that converts the expected temperature deviation in units of K to a power compensation amount in units of W, and its source is to be tuned according to the response characteristics of the control system; the present scheme realizes the look-ahead of the control command through the prediction-compensation mechanism; it is no longer a passive adjustment after the deviation has occurred and been measured by the sensor, but anticipates it before it occurs and adjusts the control input in advance to actively eliminate the future deviation; this control strategy can greatly suppress the fluctuation of the process parameters, so that the actual production trajectory can more closely follow the ideal process curve, thereby significantly improving the control precision of the fireproof glass production process and the stability of the product quality.

[0024] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A cloud platform-based monitoring system for fireproof glass production and processing parameters, characterized in that, The method comprises the following steps: a theoretical model unit is configured to construct a first-principle thermodynamic model based on preset material physical and chemical parameters, and to calculate a theoretical process temperature representing an ideal process reference; a deviation quantification unit is configured to obtain an actual process temperature collected by a production line sensor in real time, and to calculate a model mismatch index in combination with the theoretical process temperature; wherein the model mismatch index is used to represent a deviation degree of the actual process temperature from the theoretical process temperature; a state discrimination unit is configured to compare and analyze the model mismatch index with a preset trigger threshold; when the model mismatch index continuously exceeds the trigger threshold, a systematic deviation signal is generated; otherwise, a normal noise signal is generated; a model correction unit is configured to, in response to the systematic deviation signal, dynamically correct a theoretical energy absorption coefficient in the first-principle thermodynamic model based on historical deviations of the actual process temperature from the theoretical process temperature, and to generate a dynamic energy absorption coefficient; a compensation control unit is configured to construct a corrected first-principle thermodynamic model based on the dynamic energy absorption coefficient, and to generate a forward-looking compensation control instruction according to the corrected first-principle thermodynamic model, and to issue the compensation control instruction to a heating unit.

2. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 1, characterized in that, The theoretical model unit is specifically configured to: obtain preset material specific heat capacity, density, effective total mass, comprehensive heat exchange coefficient and effective heat dissipation area; determine a theoretical energy absorption coefficient and a theoretical heat dissipation coefficient in the first-principle thermodynamic model based on the material specific heat capacity, density, effective total mass, comprehensive heat exchange coefficient and effective heat dissipation area; input a current heating unit input power into the first-principle thermodynamic model containing the theoretical energy absorption coefficient and the theoretical heat dissipation coefficient to calculate the theoretical process temperature.

3. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 2, characterized in that, The theoretical energy absorption coefficient is used to represent a temperature theoretical increment of the system caused by unit input power in unit time; The theoretical heat dissipation coefficient is used to represent the comprehensive heat transfer capacity of the system and the environment.

4. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 1, characterized in that, The deviation quantification unit is specifically configured to: calculate an absolute deviation of the actual process temperature from the theoretical process temperature; divide the absolute deviation by a preset sensor system noise standard deviation to generate the model mismatch index.

5. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 1, characterized in that, The model correction unit is specifically configured to use an integral controller to calculate the cumulative effect of the historical deviation of the actual process temperature from the theoretical process temperature to generate the dynamic energy absorption coefficient.

6. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 5, characterized in that, The model correction unit is also configured to use a gating function to determine whether to start the generation process of the dynamic energy absorption coefficient only when the model mismatch index is greater than the trigger threshold.

7. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 1, characterized in that, The compensation control unit is specifically configured to: substitute the dynamic energy absorption coefficient into the first-principle thermodynamic model, and input a planned basic power instruction to predict a process temperature after a future time step, to obtain a predicted process temperature.

8. The cloud platform-based monitoring system for production and processing parameters of fireproof glass according to claim 7, characterized in that, The compensation control unit is also configured to: calculate an expected temperature deviation between the predicted process temperature and a theoretical process temperature corresponding to the future time step; calculate a power compensation amount of the basic power instruction based on the expected temperature deviation; combine the basic power instruction and the power compensation amount to generate an optimized actual power instruction, and use the actual power instruction as the compensation control instruction.

Images