WO2023039846A1

WO2023039846A1 - Set point curve updating method and system, and storage medium

Info

Publication number: WO2023039846A1
Application number: PCT/CN2021/119113
Authority: WO
Inventors: 闻博; 张鹏; 范顺杰
Original assignee: 西门子股份公司; 西门子（中国）有限公司
Priority date: 2021-09-17
Filing date: 2021-09-17
Publication date: 2023-03-23
Also published as: CN117751333A

Abstract

A set point curve updating method and system, and a storage medium. The method comprises: for process control based on a current set point curve, acquiring the actual input value curve of the process control and an error curve between a desired output value and an actual output value (S11); performing interference cancellation as well as inverse operation of the process control on the error curve between the desired output value and the actual output value, obtaining a compensation value curve, superposing the actual input value curve and the compensation value curve, obtaining a current optimization set point curve, and storing the current optimization set point curve as a historical optimization set point curve (S12); and obtaining a next set point curve according to a plurality of historical optimized set point curves the times of storage of which are more recent, replacing the current set point curve with the next set point curve, and obtaining an updated current set point curve (S13). Therefore, the updating accuracy for a set point curve can be improved.

Description

Setting point curve updating method, system and storage medium

technical field

The invention relates to the industrial field, in particular to a set point curve update method, a set point curve update system, a batch process control system and a computer-readable storage medium.

Background technique

In batch process control systems such as monocrystalline silicon growth control systems, set point curves are often used to control controlled variables that lack on-line measurement, such as melt temperature or product quality. In a setpoint curve, the value of the setpoint is a function of time.

The initial set point curve is usually generated from experiments or models, but if the characteristics of the process, such as the crystallization chamber, change, such as model drift, or change from one batch to another, in order to ensure the stability of the control, it needs to be updated periodically Set point curve. But the update of the set point curve usually takes a lot of work and expense.

Therefore, those skilled in the art are devoting themselves to finding an effective scheme for updating the set point curve.

Contents of the invention

In view of this, on the one hand, the embodiment of the present invention proposes a method for updating a set point curve, and on the other hand, proposes a system for updating a set point curve, an intermittent process control system, and a computer-readable storage medium to improve The update accuracy of the setpoint curve.

A method for updating a set point curve proposed in an embodiment of the present invention includes: for process control based on the current set point curve, obtaining the actual input value curve and the corresponding output value error curve of the process control; the output The value error curve is a curve formed by the difference between each actual output value of the process control and a set expected output value; the interference elimination is performed on the output value error curve, and the output value error curve after the interference elimination performing an inverse operation of the process control to obtain a compensation value curve for the actual input value curve, superimposing the actual input value curve and the compensation value curve to obtain a current optimal set point curve, and adding the The current optimized set point curve is stored in chronological order as a historical optimized set point curve; the next set point curve is obtained by estimating the multiple historical optimized set point curves stored with a newer time, and the next set point curve is obtained by using the following The one-time setpoint curve replaces the current setpoint curve, resulting in an updated current setpoint curve.

In one embodiment, the actual input value curve is: the current set point curve; or: the sum of the current set point curve and a control value curve of a proportional-integral-derivative loop.

In one embodiment, each historical optimization set point curve has a corresponding time stamp; the estimation of the next set point curve according to the stored multiple historical optimal set point curves with a newer time includes: For each historical optimization set point curve, assigning a weight value to the historical optimal set point curve according to the time stamp of the historical optimal set point curve; based on the weight value of each historical optimal set point curve, Performing a comprehensive operation on the multiple historical optimized set point curves to obtain the next set point curve; or including: using the multiple historical optimized set point curves together with their time stamps as an input of an estimation model, And the output of the predictive model is used as the next set point curve; the predictive model consists of multiple groups of historical optimization set point curves as input training sets and multiple corresponding forecast optimizations as output training sets The set point curve is trained.

In one embodiment, the storing the current optimized set point curve as a historical optimized set point curve in chronological order includes: storing the current optimized set point curve as a historical optimized set point curve in chronological order in In a memory with a set maximum storage quantity; under the condition that the memory has stored the historical optimal set point curve of the set maximum stored quantity, each time the latest historical optimal set point curve is stored, it will be The oldest historical optimization set point curve is deleted; the multiple historical optimal set point curves with the newest time are all the historical optimal set point curves stored in the memory.

In one embodiment, said set point curve is a power set point curve for controlling the temperature of a melt for monocrystalline silicon production; said actual input value curve is a power actual input value curve; said actual output The value curve is the pulling speed value curve corresponding to the growth rate of single crystal silicon.

The set point curve updating system proposed in the embodiment of the present invention includes: a compensation module, an optimized curve storage module, an optimized curve estimation module and an optimized curve replacement module; process control, and obtain the actual input value curve and the corresponding output value error curve of the process control; the output value error curve is obtained by the difference between each actual output value of the process control and a set expected output value Constituted curve; Carry out interference elimination to described output value error curve, and carry out the inverse operation of described process control to the output value error curve after interference elimination, obtain the compensation value curve for described actual input value curve, described The actual input value curve is superimposed on the compensation value curve to obtain the current optimized set point curve, and the current optimized set point curve is stored in the optimized curve storage module as a historical optimized set point curve in chronological order In; the optimization curve storage module is used to store a plurality of historical optimization set point curves in chronological order; the optimization curve estimation module is used to optimize according to a plurality of newer historical times stored in the optimization curve storage module The set point curve is estimated to obtain the next set point curve; the optimization curve replacement module is used to replace the current set point curve with the next set point curve to obtain an updated current set point curve.

In one embodiment, it further includes: a current input value acquisition module, configured to acquire the current actual input value of the process control for the process control based on the current set point curve, and store the current actual input value in chronological order into a database; the difference acquisition module is used to obtain the current actual output value of the process control for the process control based on the current set point curve, and compare the current actual output value with a set expected output value Comparing to obtain the current output value error, storing the current output value error in the database according to time sequence; the compensation module is used to obtain the actual input value curve and the actual input value curve composed of a series of actual input values from the database Output value error curve composed of a series of output value errors.

In one embodiment, the current actual input value is: the current set point value in the current set point curve; or: the current set point value in the current set point curve and a ratio - Sum of the current control values of the Integral-Derivative loop.

In one embodiment, each historical optimization set point curve has a corresponding time stamp; the optimization curve estimation module is used for each historical optimal set point curve, according to the time stamp of the historical optimal set point curve Time stamp, assigning a weight value to the historical optimization set point curve; based on the weight value of each historical optimal set point curve, perform a comprehensive operation on the multiple historical optimal set point curves to obtain the next set point curve fixed-point curve; or, the optimization curve prediction module uses the plurality of historical optimization set-point curves together with their time stamps as the input of a prediction model, and uses the output of the prediction model as the next setting Fixed-point curves; the predictive model is obtained by training multiple sets of historical optimal set-point curves as input training sets and multiple corresponding estimated optimal set-point curves as output training sets.

In one embodiment, the optimization curve storage module is provided with a set maximum storage quantity, and in the case that the historical optimization set point curves with the set maximum storage quantity have been stored, each time the latest history is stored To optimize a setpoint curve, the oldest historical optimized setpoint curve is deleted.

Another set point curve update system proposed in the embodiment of the present invention includes: at least one memory and at least one processor, wherein: the at least one memory is used to store computer programs; the at least one processor is used to call the The computer program stored in the at least one memory executes the set point curve updating method described in any one of the above embodiments.

The batch process control system proposed in the embodiment of the present invention includes the set point curve update system as described in any one of the above implementation modes.

The computer-readable storage medium proposed in the embodiment of the present invention has a computer program stored thereon; it is characterized in that the computer program can be executed by a processor and realize the update of the set point curve as described in any of the above implementation modes method.

It can be seen from the above scheme that in the embodiment of the present invention, the actual input value is compensated according to the error between the set expected output value and the measured output value of process control, such as batch process control, so as to obtain the The optimized set point curve of the process control is used, and then the set point curve used for the next operation is estimated according to multiple historical optimized set point curves, so that the process drift and other historical change trends can be considered, so that the obtained The accuracy of the set point curve is higher. Further, the function of the process control system applying the update scheme of the set point curve in the embodiment of the present invention can be enhanced.

In addition, the embodiment of the present invention also considers the joint control of the set point curve and the proportional-integral-derivative (PID) loop, so that the set point of the application scenario where the set point curve and the PID loop are jointly controlled can be improved. Curve update accuracy.

Description of drawings

Preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, so that those of ordinary skill in the art will be more aware of the above-mentioned and other features and advantages of the present invention. In the accompanying drawings:

Fig. 1 is an exemplary flow chart of a method for updating a set point curve in an embodiment of the present invention.

Fig. 2 is an exemplary structural diagram of a system for updating a set point curve in an embodiment of the present invention.

FIG. 3 is a schematic structural diagram of the compensation module in the system shown in FIG. 2 .

Fig. 4 is a structural schematic diagram of a batch process control system in an example of the present invention.

FIG. 5 is a schematic structural diagram of another system for updating a set point curve in an embodiment of the present application.

Wherein, the reference signs are as follows:

标号label	含义meaning
S11～S13S11～S13		步骤step
210210	补偿模块 compensation module
211211	干扰消除模块 Interference Cancellation Module
212212	逆运算模块 Inverse operation module
213213	相加模块 Addition module
220220	优化曲线存储模块Optimized Curve Storage Module
230230	优化曲线预估模块Optimized Curve Estimation Module
240240	优化曲线替换模块Optimize Curve Replacement Module

250250	当前输入值获取模块Current input value acquisition module
260260	数据库 database
270270	差值获取模块 difference acquisition module
401401	当前设定点曲线模块Current Setpoint Curve Module
402402	PID控制模块 PID control module
403403	被控对象模块 Plant module
5151	存储器 memory
5252	处理器 processor
5353	总线bus

Detailed ways

In the embodiment of the present invention, considering that the system process changes with time, it is considered not only to refer to the current data, but also to refer to the historical data when predicting the next set point curve.

In addition, considering that in some applications, the set point curve is often used together with the proportional-integral-derivative (PID) loop to improve the control performance, therefore, in the embodiment of the present invention, when the next set point curve is estimated , the control value of the PID loop can be taken into consideration together.

In order to make the purpose, technical solution and advantages of the present invention clearer, the following examples are given to further describe the present invention in detail.

Fig. 1 is an exemplary flow chart of a method for updating a set point curve in an embodiment of the present invention. As shown in Figure 1, the method may include the following steps:

Step S11, for the process control based on the current set point curve, the actual input value curve and the corresponding output value error curve of the process control are obtained. The output value error curve is a curve formed by the difference between each actual output value of the process control and a set expected output value. In this embodiment, the process control may be batch process control. Of course, it is also possible for non-batch process control where there is a set point curve in the control recipe.

Wherein, the actual input value curve is the current set point curve. Or when the set point curve is often used together with the PID loop, the actual input value curve may be: the sum of the current set point curve and the control value curve of the PID loop. At this time, the set expected output value may be the set value of the PID loop.

During specific implementation, for the process control based on the current set point curve, the current actual input value and the current actual output value of the process control can be obtained in real time, and the current actual input value can be stored in a database in chronological order; The current actual output value is compared with a set expected output value to obtain a current output value error, and the current output value error is stored in the database in chronological order. Correspondingly, in this step, an actual input value curve composed of a series of actual input values and an output value error curve composed of a series of output value errors can be obtained from the database.

Correspondingly, the current actual input value is: the current set point value in the current set point curve; or: the current set point value in the current set point curve and the current control value of the PID loop Sum.

In this embodiment, if the control of the batch process is the growth process control of single crystal silicon as an example, then the set point curve is a power set point curve for controlling the temperature of the melt used in the production of single crystal silicon; the actual The input value curve is the actual power input value curve; the actual output value curve is the pulling speed value curve corresponding to the growth rate of single crystal silicon.

Step S12, performing interference elimination on the output value error curve, and performing the inverse operation of the process control on the output value error curve after interference elimination, to obtain a compensation value curve for the actual input value curve, and converting the actual The input value curve is superimposed on the compensation value curve to obtain the current optimal set point curve, and the current optimal set point curve is stored as a historical optimal set point curve in chronological order.

During specific implementation, in this step, the current optimal set point curve can be stored in a memory with a set maximum storage quantity as a historical optimal set point curve in chronological order; In the case of storing the maximum number of historical optimized set point curves, each time the latest historical optimized set point curve is stored, the oldest historical optimized set point curve is deleted.

Of course, in other implementation manners, the memory may not set the maximum storage quantity, but can store it as long as there is still storage space until the storage space is full. If the fixed-point curve is selected, the operation of deleting the oldest historical optimization set-point curve.

Step S13, estimate and obtain the next set point curve according to multiple historical optimized set point curves stored with a newer time, and use the next set point curve to replace the current set point curve to obtain an update After the current setpoint curve.

In this step, for the memory set with the set maximum storage quantity, the plurality of historical optimization set point curves that are newer in time may be all the historical optimal set point curves stored in the memory. At this time, each time the latest historical optimization set point curve is stored, the execution of step S13 is triggered.

For a memory that is not set with the set maximum storage quantity, the plurality of historical optimization set point curves that are newer in time may be a set number of historical optimization set point curves that are read from the memory in a newer time. fixed-point curve. Similarly, each time the latest historical optimization set point curve is stored, the execution of step S13 can be triggered.

During specific implementation, each historical optimization set point curve has a corresponding time stamp. Correspondingly, in step S13, for each historical optimization set point curve, according to the time stamp of the historical optimal set point curve, assign a weight value to the historical optimal set point curve; The weight value of the set point curve is to carry out comprehensive calculation on the multiple historical optimization set point curves to obtain the next set point curve; or in step S13, the multiple historical optimal set point curves can also be combined with Its time stamp is used as the input of a prediction model, and the output of the prediction model is used as the next set point curve; the prediction model is composed of multiple sets of historical optimization set point curves as the input training set and as A plurality of corresponding estimated optimal setpoint curves for the output training set are trained.

The method for updating the set point curve in the embodiment of the present invention has been described in detail above, and the system for updating the set point curve in the embodiment of the present invention will be described in detail below. The set point curve update system in the embodiment of the present invention can be used to implement the set point curve update method in the embodiment of the present invention. For the details not disclosed in the system embodiment of the present invention, please refer to the corresponding section in the method embodiment of the present invention description, and will not be repeated here.

Fig. 2 is an exemplary structural diagram of a system for updating a set point curve in an embodiment of the present invention. As shown by the solid line in FIG. 2 , the system may include: a compensation module 210 , an optimization curve storage module 220 , an optimization curve estimation module 230 and an optimization curve replacement module 240 .

Wherein, the compensation module 210 is used for obtaining the actual input value curve and the corresponding output value error curve of the process control for the process control based on the current set point curve; the output value error curve is each actual value of the process control The curve formed by the difference between the output value and a set expected output value; the interference elimination is performed on the output value error curve, and the inverse operation of the process control is performed on the output value error curve after the interference elimination, to obtain The compensation value curve of the actual input value curve, the actual input value curve and the compensation value curve are superimposed to obtain the current optimal set point curve, and the current optimized set point curve is used as a historical optimal set point curve The fixed-point curves are stored in the optimization curve storage module 220 in chronological order. In this embodiment, the actual input value curve and the output value error curve can be obtained from a database.

Wherein, the actual input value curve is: the current set point curve; or: the sum of the current set point curve and a control value curve of a PID loop.

During specific implementation, the compensation module 210 may include an interference elimination module 211 , an inverse operation module 212 and an addition module 213 as shown in FIG. 3 .

The interference elimination module 211 is used to perform interference elimination processing such as filtering on the obtained output value error curve, so as to eliminate the influence brought by abnormal behaviors such as man-made or environmental interference.

The inverse operation module 212 is used to perform process control inverse operation on the output value error curve after the interference is eliminated, so as to obtain a compensation value curve.

The addition module 213 is used to superimpose the obtained actual input value curve and the compensation value curve to obtain the current optimal set point curve, and use the current optimized set point curve as a historical optimal set point curve according to The time sequence is stored in the optimization curve storage module 220 .

The optimization curve storage module 220 is used to store multiple historical optimization set point curves in time order.

During specific implementation, the optimization curve storage module 220 may be provided with a set maximum storage quantity, and in the case that the historical optimization set point curves with the set maximum storage quantity have been stored, each time the latest historical optimization set point curve is stored. setpoint curve, the oldest historical optimization setpoint curve is deleted. In addition, every time a newest historical optimal set point curve is stored, the optimal curve estimation module 230 is triggered to read all stored historical optimal set point curves, no matter whether they reach the maximum storage quantity or not. Or, the optimization curve storage module 220 can also not set the maximum storage quantity of the setting, but as long as there is still storage space, it can be stored all the time until the storage space is full, and then the latest historical optimization setting for each storage time is executed. If the fixed-point curve is selected, the operation of deleting the oldest historical optimization set-point curve. Correspondingly, each time the latest historical optimization set point curve is stored, the optimization curve estimation module 230 may be triggered to read a set number of stored historical optimal set point curves.

The optimization curve estimation module 230 is used for estimating and obtaining the next set point curve according to a plurality of newer historical optimization set point curves stored in the optimization curve storage module.

During specific implementation, each historical optimization set point curve has a corresponding time stamp. Correspondingly, the optimal curve estimation module 230 may assign a weight value to each historical optimized set point curve according to the time stamp of the historical optimized set point curve; based on each The weight value of the historical optimization set point curve, and comprehensive calculation is performed on the multiple historical optimal set point curves to obtain the next set point curve; or, the multiple historical optimal set point curves together with their time Stamp is used as the input of a prediction model, and the output of the prediction model is used as the next set point curve; the prediction model is trained by multiple sets of historical optimization set point curves as input training set and as output Multiple corresponding estimated optimal setpoint curves for the set are trained.

The optimization curve replacement module 240 is used to replace the current set point curve with the next set point curve to obtain an updated current set point curve.

In addition, in some embodiments, if a process control system such as a batch process control system does not record the actual input value curve and the output value error curve, the set point curve update system in the batch process control in this example As shown by the dotted line in FIG. 2 , it further includes: a current input value acquisition module 250 , a database 260 and a difference value acquisition module 270 .

Wherein, the current input value obtaining module 250 is used for obtaining the current actual input value of the process control for the process control based on the current set point curve, and storing the current actual input value in the database 260 in chronological order. The current actual input value is the current set point value in the current set point curve; or the sum of the current set point value in the current set point curve and a current control value of a PID loop.

The difference acquisition module 270 is used for obtaining the current actual output value of the process control for the process control based on the current set point curve, and comparing the current actual output value with a set expected output value to obtain the current output value error, and store the current output value error in the database 260 in chronological order.

Correspondingly, the compensation module 210 is configured to acquire an actual input value curve formed by a series of actual input values and an output value error curve formed by a series of output value errors from the database 260 .

Fig. 4 is a structural schematic diagram of a batch process control system in an example of the present invention. As shown in Figure 4, the intermittent process control system includes: current set point curve module 401, PID control module 402, controlled object module 403, compensation module 210 in Figure 2, optimization curve storage module 220, optimization curve preset Estimation module 230, optimization curve replacement module 240, current input value acquisition module 250, database 260 and difference acquisition module 270.

Wherein, the current set point curve module 401 is used to sequentially output the current set point value according to the current set point curve.

The PID control module 402 is used to determine the current control value of the PID loop according to the current output value error obtained by the difference value acquisition module 270 .

Correspondingly, the current input value acquisition module 250 is used to add the current set point value and the current control value to obtain the current input value.

The controlled object module 403 is used to obtain the corresponding current output value according to the current input value.

For example, for the process control of monocrystalline silicon production, the controlled object module 403 can be a melt used for monocrystalline silicon production; the current input value is a power value, and the melt is controlled to reach the corresponding temperature by this power value value, and then obtain the pulling speed corresponding to the production speed of the single crystal silicon, that is, the current output value of the melt is a pulling speed value. Correspondingly, when the inverse operation of the process control is performed for the error of the pulling speed value, the corresponding melt temperature error can be deduced from the error of the pulling speed value, and then the corresponding input power error can be deduced, and the input power error is for The compensation power of the actual input power.

FIG. 5 is a schematic structural diagram of another system for updating a set point curve in an embodiment of the present application. The system can be used to implement the method shown in FIG. 1 , or implement the systems shown in FIGS. 2 to 4 . As shown in FIG. 5 , the system may include: at least one memory 51 and at least one processor 52 . In addition, some other components may also be included, such as communication ports and the like. These components communicate via bus 53 .

Among them, at least one memory 51 is used to store computer programs. In one embodiment, the computer program can be understood as including various modules of the set point curve updating system shown in FIGS. 2 to 4 . In addition, at least one memory 51 can also store an operating system and the like. The operating system includes but is not limited to: Android operating system, Symbian operating system, Windows operating system, Linux operating system and so on.

At least one processor 52 is used to call the computer program stored in at least one memory 51 to execute the method for updating the set point curve described in the embodiment of the present application. The processor 52 may be a CPU, a processing unit/module, an ASIC, a logic module or a programmable gate array, and the like. It can receive and send data through the communication port.

It should be noted that not all steps and modules in the above-mentioned processes and structure diagrams are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The division of each module is only to facilitate the description of the functional division adopted. In actual implementation, one module can be divided into multiple modules, and the functions of multiple modules can also be realized by the same module. These modules can be located in the same device. , or on a different device.

It can be understood that the hardware modules in the foregoing embodiments may be implemented mechanically or electronically. For example, a hardware module may include specially designed permanent circuits or logic devices (such as special-purpose processors, such as FPGAs or ASICs) to perform specific operations. Hardware modules may also include programmable logic devices or circuits (eg, including general-purpose processors or other programmable processors) temporarily configured by software to perform particular operations. As for implementing the hardware module in a mechanical way, using a dedicated permanent circuit, or using a temporarily configured circuit (such as configured by software) to realize the hardware module, it can be decided according to cost and time considerations.

In addition, an embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored, the computer program can be executed by a processor and implement the method for updating the set point curve described in the embodiment of the present application . Specifically, a system or device equipped with a storage medium may be provided, on which the software program code for realizing the functions of any implementation manner in the above-mentioned embodiments is stored, and the computer (or CPU or MPU of the system or device) ) to read and execute the program code stored in the storage medium. In addition, an operating system or the like operated on a computer may also complete part or all of the actual operations through instructions based on program codes. It is also possible to write the program code read from the storage medium into the memory set in the expansion board inserted into the computer or into the memory set in the expansion unit connected to the computer, and then based on the instructions of the program code, the memory installed in the Some or all of the actual operations are performed by the CPU on the expansion board or the expansion unit, so as to realize the functions of any of the above-mentioned implementation manners. Embodiments of storage media for providing program codes include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), Tape, non-volatile memory card, and ROM. Alternatively, the program code can be downloaded from a server computer via a communication network.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of the present invention. within the scope of protection.

Claims

The set point curve update method is characterized in that it includes:

For the process control based on the current set point curve, the actual input value curve and the corresponding output value error curve of the process control are obtained; the output value error curve is each actual output value of the process control and a set The curve formed by the difference of the expected output value;

Perform interference elimination on the output value error curve, and perform the inverse operation of the process control on the output value error curve after interference elimination to obtain a compensation value curve for the actual input value curve, and convert the actual input value curve to superimposing with the compensation value curve to obtain the current optimal set point curve, and storing the current optimized set point curve as a historical optimized set point curve in chronological order;

The next set point curve is obtained by estimating the stored multiple historical optimized set point curves, and the current set point curve is replaced by the next set point curve to obtain the updated current set point curve. Set point curve.
The method according to claim 1, wherein the actual input value curve is: the current set point curve; or is:

The sum of the current set point curve and the control value curve of a proportional-integral-derivative loop.
The method according to claim 1, wherein each historical optimization set point curve has a corresponding time stamp;

The estimation of the set point curve for the next time according to the stored multiple historical optimization set point curves with newer time includes:

For each historical optimization set point curve, assigning a weight value to the historical optimal set point curve according to the time stamp of the historical optimal set point curve; based on the weight value of each historical optimal set point curve, Performing a comprehensive operation on the multiple historical optimization set point curves to obtain the next set point curve; or include:

Using the plurality of historical optimization set point curves together with their time stamps as an input to a predictive model, and using the output of the predictive model as the next set point curve; the predictive model is trained as input by Multiple sets of historical optimal set-point curves for the set and multiple corresponding estimated optimal set-point curves for the output training set are trained.
The method according to claim 1, wherein storing the current optimal set point curve as a historical optimized set point curve in chronological order comprises:

storing the current optimal set point curve as a historical optimized set point curve in a memory with a set maximum storage quantity in chronological order; In the case of point curves, each time the latest historical optimization set point curve is stored, the oldest historical optimal set point curve is deleted;

The plurality of historical optimization set point curves that are newer in time are all the historical optimal set point curves stored in the memory.
The method according to any one of claims 1 to 4, characterized in that said set point curve is a power set point curve for controlling the temperature of a melt for monocrystalline silicon production; said actual input The value curve is the actual input value curve of power; the actual output value curve is the pulling speed value curve corresponding to the growth rate of single crystal silicon.
The set point curve update system is characterized in that it includes: a compensation module, an optimization curve storage module, an optimization curve estimation module and an optimization curve replacement module; wherein,

The compensation module is used for obtaining the actual input value curve and the corresponding output value error curve of the process control for the process control based on the current set point curve; the output value error curve is each actual output of the process control The curve formed by the difference between the value and a set expected output value; the interference elimination is performed on the output value error curve, and the inverse operation of the process control is performed on the output value error curve after the interference elimination, to obtain The compensation value curve of the actual input value curve, the actual input value curve and the compensation value curve are superimposed to obtain the current optimal set point curve, and the current optimized set point curve is used as a historical optimal setting Point curves are stored in the optimization curve storage module in chronological order;

The optimization curve storage module is used to store a plurality of historical optimization set point curves in chronological order;

The optimization curve estimation module is used to estimate and obtain the next set point curve according to multiple historical optimization set point curves stored in the optimization curve storage module with a newer time;

The optimization curve replacement module is used to replace the current set point curve with the next set point curve to obtain an updated current set point curve.
The system of claim 6, further comprising:

A current input value acquisition module, configured to acquire the current actual input value of the process control for the process control based on the current set point curve, and store the current actual input value in a database in chronological order;

The difference acquisition module is used to obtain the current actual output value of the process control for the process control based on the current set point curve, and compare the current actual output value with a set expected output value to obtain the current output value error, storing the current output value error in the database in chronological order;

The compensation module is used to obtain an actual input value curve composed of a series of actual input values and an output value error curve composed of a series of output value errors from the database.
The system according to claim 7, wherein the current actual input value is: a current set point value in the current set point curve; or is:

The sum of the current setpoint value in the current setpoint curve and the current control value of a proportional-integral-derivative loop.
The system according to claim 6, wherein each historical optimization set point curve has a corresponding time stamp;

The optimization curve estimation module is used for each historical optimization set point curve, according to the time stamp of the historical optimal set point curve, assigning a weight value to the historical optimal set point curve; based on each historical Optimizing the weight value of the set point curve, and performing a comprehensive operation on the multiple historical optimized set point curves to obtain the next set point curve; or,

The optimization curve estimation module uses the plurality of historical optimization set point curves together with their time stamps as an input of an estimation model, and uses the output of the estimation model as the next set point curve; The prediction model is trained by multiple sets of historical optimal set point curves as input training set and multiple corresponding estimated optimal set point curves as output training set.
The system according to claim 6, wherein the optimization curve storage module is provided with a set maximum storage quantity, and when the historical optimization set point curve with the set maximum storage quantity has been stored, Each time the newest historical optimized set point curve is stored, the oldest historical optimized set point curve is deleted.
A setpoint profile update system, comprising: at least one memory and at least one processor, wherein:

said at least one memory is for storing computer programs;

The at least one processor is used for invoking a computer program stored in the at least one memory to execute the set point curve updating method according to any one of claims 1 to 5.
The batch process control system is characterized by comprising the set point curve updating system according to any one of claims 6 to 11.
A computer-readable storage medium, on which a computer program is stored; it is characterized in that the computer program can be executed by a processor and realize the set point curve updating method according to any one of claims 1 to 5.