Monocrystalline silicon growth state control method based on fuzzy controller
By using a fuzzy controller-based growth state control method in the production process of single crystal silicon, the problem of inaccurate pulling speed control is solved, and a higher product yield is achieved.
Patent Information
- Application Number
- CN202510038408.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-10
- Publication Date
- 2025-03-18
AI Technical Summary
In the process of producing single crystal silicon by direct-pull method, the existing technology lacks effective and accurate lifting speed control methods, resulting in pits or protrusions on the surface of single crystal silicon products, affecting product yield.
The single crystal silicon growth state control method based on the fuzzy controller is adopted. By calculating the corrected acquisition period, the actual diameter of the single crystal silicon is periodically collected, and the speed fuzzy controller and the acceleration fuzzy controller are established. According to the difference between the actual lifting speed and the expected lifting speed, the PID control parameters are adjusted to obtain the corrected lifting speed.
It improves the accuracy of lifting speed control, reduces the surface depression of single crystal silicon products, and improves product yield.
Smart Images
Figure CN119640387A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the field of semiconductor technology, and in particular to a single crystal silicon growth state control method based on a fuzzy controller. Background Art
[0002] Monocrystalline silicon is an important material for the production of solar cells. According to different production methods, it is mainly divided into the Para method, regional recrystallization method, seed method and Czochralski method, among which the Czochralski method is the current mainstream technology for producing monocrystalline silicon. The main process of the Czochralski method is to put polycrystalline silicon in a crucible, heat it to melt it, and then clamp a monocrystalline silicon seed crystal and suspend it above the crucible. When pulling it vertically, one end is inserted into the melt until it melts, and then it is slowly rotated and pulled upward. In this way, the interface between the liquid and the solid will gradually condense to form monocrystalline silicon. The whole process can be regarded as a process of replicating the seed crystal, and the generated silicon crystal is monocrystalline silicon.
[0003] In the process of producing single crystal silicon by the direct pulling method, due to the uneven heat of continuous preparation, the existing technology lacks an effective and accurate method for controlling the pulling speed, which can easily lead to pits or protrusions on the surface of the prepared single crystal silicon, causing depression on the surface of the single crystal silicon product and affecting the product yield. Summary of the invention
[0004] The purpose of the present invention is to provide a method for controlling the growth state of single crystal silicon based on a fuzzy controller, so as to achieve effective and accurate pulling speed control, reduce surface depression of single crystal silicon products, and improve product yield.
[0005] A method for controlling the growth state of single crystal silicon based on a fuzzy controller comprises the following steps: Step S1, calculating the corrected collection period, periodically collecting the actual diameter of the single crystal silicon based on the corrected collection period, and calculating the diameter deviation based on the actual diameter, and then solving the actual pulling speed and the actual pulling acceleration based on the calculated diameter deviation, and dynamically setting the expected pulling speed and the expected pulling acceleration according to actual production needs; Step S2, taking the difference between the actual pulling speed and the expected pulling speed as the input, taking the change of the first PID control parameter as the output, establishing a speed fuzzy controller, selecting a triangular membership function to fuzzify the fuzzy quantity, and then performing fuzzy reasoning according to the formulated speed fuzzy rule, and finally defuzzifying the result through the output membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected first pulling speed; Step S3, taking the difference between the actual pulling acceleration and the expected pulling acceleration as the input, taking the change of the second PID control parameter as the output, establishing an acceleration fuzzy controller, selecting a triangular membership function to fuzzify the fuzzy quantity, and then performing fuzzy reasoning according to the formulated acceleration fuzzy rule, and finally defuzzifying the result by outputting the membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected second pulling speed; Step S4, selecting the corrected first pulling speed and the corrected second pulling speed that is closest to the desired pulling speed as the optimal pulling speed, and controlling the growth state of the single crystal silicon according to the optimal pulling speed.
[0006] According to the single crystal silicon growth state control method based on fuzzy controller provided by the present invention, the actual diameter of the single crystal silicon is periodically collected based on the revised collection period, which can improve the accuracy of data collection, thereby improving the calculation accuracy of the actual pulling speed and the actual pulling acceleration. On the basis of the actual pulling speed and the actual pulling acceleration, the present invention sets up two fuzzy controllers. In the speed fuzzy controller, fuzzy reasoning is performed according to the formulated speed fuzzy rules to obtain the revised first pulling speed. In the acceleration fuzzy controller, fuzzy reasoning is performed according to the formulated acceleration fuzzy rules to obtain the revised second pulling speed. Then, the pulling speed closest to the expected pulling speed is selected as the optimal pulling speed, which can effectively improve the accuracy of pulling speed control, thereby reducing surface depressions of single crystal silicon products and improving product yield. BRIEF DESCRIPTION OF THE DRAWINGS
[0007] Figure 1 It is a schematic flow chart of the single crystal silicon growth state control method based on fuzzy controller of the present invention. DETAILED DESCRIPTION
[0008] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0009] See also Figure 1 , a method for controlling the growth state of single crystal silicon based on a fuzzy controller, comprising steps S1 to S4: Step S1, calculate the corrected collection cycle, based on the corrected collection cycle, periodically collect the actual diameter of the single crystal silicon, and calculate the diameter deviation of the single crystal silicon according to the actual diameter, and then based on the calculated diameter deviation, solve the actual pulling speed and the actual pulling acceleration, and dynamically set the expected pulling speed and the expected pulling acceleration according to actual production needs.
[0010] Among them, the corrected collection cycle satisfies the following conditional formula: ; in, Indicates the corrected acquisition period, Indicates the preset collection period. It represents the difference between the actual pulling speed and the expected pulling speed. V e Indicates the expected pulling speed.
[0011] Specifically, ,in, V T Indicates the actual pulling speed.
[0012] Then, the actual pulling speed and actual pulling acceleration are solved using the following formula: ;
[0013] in, D represents the diameter of the crucible, d Indicates the actual diameter of single crystal silicon, f s represents the macroscopic growth rate of the crystal, represents the actual pulling acceleration, Indicates the diameter deviation of single crystal silicon, k is the proportionality constant, is the set time interval.
[0014] Then, the actual pulling speed and actual pulling acceleration are solved using the following formula: ;
[0015] in, V T Indicates the actual pulling speed. D represents the diameter of the crucible, d Indicates the actual diameter of single crystal silicon, f s represents the macroscopic growth rate of the crystal, represents the actual pulling acceleration, Indicates the diameter deviation of single crystal silicon, k is the proportionality constant, is the set time interval.
[0016] Step S2, taking the difference between the actual pulling speed and the expected pulling speed as the input, and the change of the first PID control parameter as the output, establish a speed fuzzy controller, select the triangular membership function to fuzzy the fuzzy quantity, and then perform fuzzy reasoning according to the established speed fuzzy rules, and finally defuzzify the result through the output membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected first pulling speed.
[0017] Wherein, step S2 specifically includes: In the speed fuzzy controller, the input quantity is defined as and output , is the fuzzy amount, where , They are the changes of PID proportional and integral control parameters of speed fuzzy controller respectively; The above fuzzy quantities are fuzzified by the triangular membership function, the fuzzy domain is unified to [-2, +2], and the fuzzy subset is defined as [NB, NM, NS, Z, PS, PM, PB], where NB is negative large, NM is negative medium, NS is negative small, Z is zero, PS is positive small, PM is positive medium, and PB is positive large; Formulate separately and The rule base, The rule base is shown in Table 1. The rule base is shown in Table 2. E 1 is the error in the speed fuzzy controller, EC 1 is the error change rate in the speed fuzzy controller, according to and The rule base is used to perform fuzzy reasoning, and the fuzzy reasoning method is the Mamdani method; Table 1
[0018] Table 2
[0019] After completing the fuzzy reasoning, the centroid method is used to defuzzify the result and obtain and ; Based on the obtained and , calculate the first pulling force and the first rotational force of the pulling device, the expression is as follows: ;
[0020] in, represents the first pulling force of the pulling device, represents the first rotational force of the lifting device, E 1 is the error in the speed fuzzy controller; Based on the first lifting force and the first rotational force of the lifting device, the corrected first lifting speed is calculated, and the expression is: ; in, V 1 represents the first pulling speed after correction, It indicates the difference in weight of single crystal silicon measured at two adjacent collection moments according to the corrected collection cycle.
[0021] Step S3, taking the difference between the actual pulling acceleration and the expected pulling acceleration as the input, and the change of the second PID control parameter as the output, establish an acceleration fuzzy controller, select the triangular membership function to fuzzy the fuzzy quantity, and then perform fuzzy reasoning according to the formulated acceleration fuzzy rules, and finally defuzzify the result through the output membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected second pulling speed.
[0022] Wherein, step S3 specifically includes: In the acceleration fuzzy controller, the input quantity is defined as and output , , is the fuzzy amount, where is the difference between the actual pulling acceleration and the expected pulling acceleration, , is the expected pulling acceleration, , , is the change of PID proportional, integral and differential control parameters of the acceleration fuzzy controller; The above fuzzy quantities are fuzzified by the triangular membership function, the fuzzy domain is unified to [-1, +1], and the fuzzy subset is defined as [NB, NM, NS, Z, PS, PM, PB]; Formulate separately , and The rule base, The rule base is shown in Table 3. The rule base is shown in Table 4. The rule base of is shown in Table 5. E 2 is the error in the acceleration fuzzy controller, EC 2 is the error change rate in the acceleration fuzzy controller, according to , and The rule base is used to perform fuzzy reasoning, and the fuzzy reasoning method is the Mamdani method; Table 3
[0023] Table 4
[0024] Table 5
[0025] After completing the fuzzy reasoning, the centroid method is used to defuzzify the result and obtain , and ; Based on the obtained , and , calculate the second pulling force, second rotational force and lateral force of the pulling device, the expressions are as follows: ; ; ; in, represents the second lifting force of the lifting device, represents the second rotational force of the lifting device, represents the lateral force of the lifting device, E 2 is the error in the acceleration fuzzy controller, represents the differential of the error in the acceleration fuzzy controller, Represents the differential of time; Based on the second lifting force, the second rotational force and the lateral force of the lifting device, the corrected second lifting speed is calculated, and the expression is: ; in, V 2 represents the second pulling speed after correction.
[0026] Step S4, selecting the corrected first pulling speed and the corrected second pulling speed that is closest to the desired pulling speed as the optimal pulling speed, and controlling the growth state of the single crystal silicon according to the optimal pulling speed.
[0027] Among them, the expression of the optimal pulling speed is: ; in, V b Indicates the optimal pulling speed.
[0028] In summary, according to the single crystal silicon growth state control method based on fuzzy controller provided by the present invention, the actual diameter of the single crystal silicon is periodically collected based on the revised collection period, which can improve the accuracy of data collection, thereby improving the calculation accuracy of the actual pulling speed and the actual pulling acceleration. On the basis of the actual pulling speed and the actual pulling acceleration, the present invention sets up two fuzzy controllers. In the speed fuzzy controller, fuzzy reasoning is performed according to the formulated speed fuzzy rules to obtain the revised first pulling speed. In the acceleration fuzzy controller, fuzzy reasoning is performed according to the formulated acceleration fuzzy rules to obtain the revised second pulling speed. Then, the pulling speed closest to the expected pulling speed is selected as the optimal pulling speed, which can effectively improve the accuracy of pulling speed control, thereby reducing surface depressions of single crystal silicon products and improving product yield.
[0029] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0030] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.
Claims
1. A method for controlling the growth state of single crystal silicon based on a fuzzy controller, characterized in that: The following steps are involved: Step S1, calculating the corrected collection period, periodically collecting the actual diameter of the single crystal silicon based on the corrected collection period, and calculating the diameter deviation of the single crystal silicon based on the actual diameter, and then solving the actual pulling speed and the actual pulling acceleration based on the calculated diameter deviation, and dynamically setting the expected pulling speed and the expected pulling acceleration according to actual production needs; Step S2, taking the difference between the actual pulling speed and the expected pulling speed as the input, taking the change of the first PID control parameter as the output, establishing a speed fuzzy controller, selecting a triangular membership function to fuzzify the fuzzy quantity, and then performing fuzzy reasoning according to the formulated speed fuzzy rule, and finally defuzzifying the result through the output membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected first pulling speed; Step S3, taking the difference between the actual pulling acceleration and the expected pulling acceleration as the input, taking the change of the second PID control parameter as the output, establishing an acceleration fuzzy controller, selecting a triangular membership function to fuzzify the fuzzy quantity, and then performing fuzzy reasoning according to the formulated acceleration fuzzy rule, and finally defuzzifying the result by outputting the membership function to output the adjustment value of the PID control parameter, thereby obtaining the corrected second pulling speed; Step S4, selecting the corrected first pulling speed and the corrected second pulling speed that is closest to the desired pulling speed as the optimal pulling speed, and controlling the growth state of the single crystal silicon according to the optimal pulling speed.
2. The method for controlling the growth state of single crystal silicon based on a fuzzy controller according to claim 1, characterized in that: In step S1, the corrected acquisition period satisfies the following conditional formula: ； in, Indicates the corrected acquisition period, Indicates the preset collection period. It represents the difference between the actual pulling speed and the expected pulling speed. V e Indicates the expected pulling speed.
3. The method for controlling the growth state of single crystal silicon based on a fuzzy controller according to claim 2, characterized in that: In step S1, the actual pulling speed and actual pulling acceleration are solved using the following formula: ； in, V T Indicates the actual pulling speed. D represents the diameter of the crucible, d Indicates the actual diameter of single crystal silicon, f s represents the macroscopic growth rate of the crystal, Indicates the actual pulling acceleration, Indicates the diameter deviation of single crystal silicon, k is the proportionality constant, is the set time interval.
4. The method for controlling the growth state of single crystal silicon based on a fuzzy controller according to claim 3, characterized in that: Step S2 specifically includes: In the speed fuzzy controller, the input quantity is defined as and output , is the fuzzy amount, where , They are the changes of PID proportional and integral control parameters of speed fuzzy controller respectively; The above fuzzy quantities are fuzzified by the triangular membership function, the fuzzy domain is unified to [-2, +2], and the fuzzy subset is defined as [NB, NM, NS, Z, PS, PM, PB], where NB is negative large, NM is negative medium, NS is negative small, Z is zero, PS is positive small, PM is positive medium, and PB is positive large; Formulate separately and The rule base is based on and The rule base is used to perform fuzzy reasoning, and the fuzzy reasoning method is the Mamdani method; After completing the fuzzy reasoning, the centroid method is used to defuzzify the result and obtain and ; Based on the obtained and , calculate the first pulling force and the first rotational force of the pulling device, the expression is as follows: ； in, represents the first pulling force of the pulling device, represents the first rotational force of the lifting device, E 1 is the error in the speed fuzzy controller; Based on the first lifting force and the first rotational force of the lifting device, the corrected first lifting speed is calculated, and the expression is: ； in, V 1 represents the first pulling speed after correction, It indicates the difference in weight of single crystal silicon measured at two adjacent collection moments according to the corrected collection cycle.
5. The method for controlling the growth state of single crystal silicon based on a fuzzy controller according to claim 4, characterized in that: Step S3 specifically includes: In the acceleration fuzzy controller, the input quantity is defined as and output , , is the fuzzy amount, where is the difference between the actual pulling acceleration and the expected pulling acceleration, , , is the change of PID proportional, integral and differential control parameters of the acceleration fuzzy controller; The above fuzzy quantities are fuzzified by the triangular membership function, the fuzzy domain is unified to [-1, +1], and the fuzzy subset is defined as [NB, NM, NS, Z, PS, PM, PB]; Formulate separately , and The rule base is based on , and The rule base is used to perform fuzzy reasoning, and the fuzzy reasoning method is the Mamdani method; After completing the fuzzy reasoning, the centroid method is used to defuzzify the result and obtain , and ; Based on the obtained , and , calculate the second pulling force, second rotational force and lateral force of the pulling device, the expressions are as follows: ； ； ； in, represents the second lifting force of the lifting device, represents the second rotational force of the lifting device, represents the lateral force of the lifting device, E 2 is the error in the acceleration fuzzy controller, represents the differential of the error in the acceleration fuzzy controller, Represents the differential of time; Based on the second lifting force, the second rotational force and the lateral force of the lifting device, the corrected second lifting speed is calculated, and the expression is: ； in, V 2 represents the second pulling speed after correction.
6. The method for controlling the growth state of single crystal silicon based on a fuzzy controller according to claim 5, characterized in that: In step S4, the expression of the optimal pulling speed is: ； in, V b Indicates the optimal pulling speed.