A grinding pressure dressing method, device, computer device and medium

Abstract

The present invention provides a grinding pressure trimming method, device, computer device and medium, which are applied to a gas grinding head, and include: determining a plurality of characteristic parameters according to the pressure set value and the actual pressure value of each region; calculating the pressure set value of each region that minimizes the function output value according to a prediction function, the prediction function is established according to each characteristic parameter and the weight of each characteristic parameter, the function output value is used to characterize the difference between the pressure set value and the actual pressure value, and the weight of each characteristic parameter is obtained by solving based on an existing data set; controlling the gas grinding head according to the pressure set value of each region. The present invention can make the pressure set value of each region of the grinding head more accurate and the precision control higher.

Description

Technical Field

[0001] The present invention relates to the technical field of wafer grinding, and particularly to a grinding pressure trimming method, device, computer device and medium. Background Art

[0002] As a key device in integrated circuit manufacturing, the CMP device is the core and foundation of the semiconductor industry, and the surface topography of the wafer processed and polished by it directly affects the overall performance of the integrated circuit.

[0003] At present, during the polishing process of the CMP device, the rotating and oscillating gas type grinding head can apply gas pressure to the wafer surface and run a complex polishing trajectory for grinding, so as to achieve multi-region control of the grinding of the wafer surface. During the grinding process, the actual amount of each pressure of the grinding head is monitored in real time and compared with the set value of the pressure. If it exceeds a certain threshold, an alarm will be processed.

[0004] However, due to the instability of the aerodynamic control system, the pressure set value is not accurate enough and the precision control is not high. Summary of the Invention

[0005] In order to solve the deficiencies in the prior art, the present invention provides a grinding pressure trimming method, device, computer device and medium.

[0006] The first aspect of the present invention provides a grinding pressure trimming method, which is applied to a gas type grinding head. The method includes:

[0007] Determine a plurality of characteristic parameters according to the pressure set value and the actual pressure value of each region; calculate the pressure set value of each region that makes the function output value of the prediction function the smallest according to the prediction function. The prediction function is established according to each characteristic parameter and the weight of each characteristic parameter, and the function output value is used to characterize the difference between the pressure set value and the actual pressure value. The weight of each characteristic parameter is obtained by solving based on the existing data set; control the gas type grinding head according to the pressure set value of each region.

[0008] The beneficial effect is that: the present invention calculates the pressure set value of each region that makes the function output value of the prediction function the smallest. Since the function output value of the prediction function in the embodiment of the present invention is used to characterize the difference between the pressure set value and the actual pressure value, the smaller the difference, the closer the actual pressure value of the gas type grinding head is to the set pressure value when using the set pressure value to control the gas type grinding head, which means the better the pressure control. Therefore, controlling the gas type grinding head to grind the wafer according to the pressure set value calculated in the embodiment of the present invention can achieve stable control of the pressure of the gas type grinding head, improve the stability of the pressure, enhance the precision of the grinding process, and make the ground wafer more in line with the standards.

[0009] Combined with the first aspect, in the first implementation manner of the first aspect, the existing data set includes the pressure set values and actual pressure values of each region. The weights of each characteristic parameter are determined through the following steps: Determine the label quantity vector according to the expected values of the pressure set values, the expected values of the actual pressure values, and the variances of the actual pressure values of each region in the existing data set; Calculate the characteristic quantities corresponding to each characteristic parameter according to the pressure set values and actual pressure values of each region in the existing data set; Input the characteristic quantities into the prediction function, and establish a two-norm matrix based on the difference between the label quantity vector and the prediction function; Take the derivative of the two-norm matrix to obtain the optimal solution and get the weights of each characteristic parameter.

[0010] The beneficial effect is that: By obtaining the optimal solution of the weights of each characteristic parameter through the two-norm matrix, substituting the optimal weights into the prediction function, the optimal output value can be obtained, and thus the most accurate pressure set value can be obtained.

[0011] Combined with the first aspect or the first implementation manner of the first aspect, in the second implementation manner of the first aspect, the grinding pressure trimming method further includes: Obtain the actual pressure values of each region of the pneumatic grinding head when controlling the pneumatic grinding head according to the pressure value set values of each region; Store the pressure set values and actual pressure values of each region into the existing data set; Update the weights of each characteristic parameter according to the updated existing data set.

[0012] The beneficial effect is that: Continuously store new data into the data set, obtain the updated weights of each characteristic parameter according to the updated data, and continuous updating makes the obtained weights more accurate.

[0013] Combined with the first aspect, in the third implementation manner of the first aspect, the prediction function is:

[0014] F(X) = X T W,

[0015] where,

[0016] where, X is the characteristic matrix, represents the m-th characteristic parameter of the n-th region, W is the weight vector, and wm represents the weight of the m-th characteristic parameter.

[0017] Combined with the third implementation manner of the first aspect, in the fourth implementation manner of the first aspect, the two-norm matrix is:

[0018]

[0019] where, Y represents the label quantity vector; F(x) represents the predicted output value;

[0020] Y = [y1 y2 y3…yi] T , yi = log[(Set - Ui)2 *σ];

[0021] Where Y represents the label quantity vector; yi represents the label quantity of the i-th data, Ui represents the expected value of the pressure setting value in the i-th data, Set represents the expected value of the actual pressure, and σ represents the variance of the actual pressure value.

[0022] In combination with the fourth embodiment of the first aspect, in the fifth embodiment of the first aspect, the step of deriving the second normal form matrix, obtaining the optimal solution, and obtaining the weight of each characteristic parameter includes:

[0023]

[0024]

[0025]

[0026] In combination with the first aspect or the fifth embodiment of the first aspect, in the sixth embodiment of the first aspect, the grinding pressure adjustment method provided by the present invention also includes: obtaining the credibility of the weight of each characteristic parameter; if the credibility of the weight of each characteristic parameter meets the preset conditions, executing the step of calculating the pressure setting value of each area according to the prediction function to minimize the function output value.

[0027] The beneficial effect is: verifying the credibility of the weights of each feature parameter, and continuing to the next step if the preset conditions are met. After verifying the credibility of the weights, if the weights have large deviations, they will not be brought into the formula.

[0028] The second aspect of the present invention provides a grinding pressure adjustment device, which is applied to a gas-type grinding head, and includes: a determination module, which is used to determine multiple characteristic parameters based on the pressure setting value and actual pressure value of each area; a calculation module, which is used to calculate the pressure setting value of each area that minimizes the function output value based on a prediction function; and a control module, which is used to control the gas-type grinding head according to the pressure setting value of each area.

[0029] The beneficial effect is: the grinding pressure adjustment device provided by the present invention calculates the pressure setting value of each area that makes the function output value the smallest through a prediction function, and then controls the gas grinding head to grind the wafer according to the calculated pressure setting value, thereby stably controlling the pressure of the grinding head, improving the pressure stability, and improving the accuracy of the grinding process, so that the ground wafers are more in line with the standards.

[0030] The third aspect of the present invention provides a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor, thereby performing the grinding pressure dressing method of the first aspect of the present invention and any one of its optional embodiments.

[0031] In a fourth aspect of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores computer instructions for causing a computer to execute the polishing pressure trimming method according to any one of the first aspect and its embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention.

[0033] Figure 1 It shows a structural diagram of a polishing device provided by an embodiment of the present invention;

[0034] Figure 2 It shows a multi-region display schematic diagram of a polishing head provided by an embodiment of the present invention;

[0035] Figure 3 It shows a flowchart of a polishing pressure trimming method provided by an embodiment of the present invention;

[0036] Figure 4 It shows a cross-sectional structural diagram of a polishing head provided by an embodiment of the present invention;

[0037] Figure 5 It shows a schematic diagram of a polishing pressure trimming device provided by an embodiment of the present invention;

[0038] Figure 6 It shows a schematic diagram of the hardware structure of a computer device provided by an embodiment of the present invention;

[0039] Figure 7 It shows a schematic diagram of the structure of a computer-readable storage medium provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention.

[0041] In the embodiments of the present invention, as Figure 1 shown, the polishing table is composed of a polishing head 101 and a polishing pad 102. During the wafer polishing process, the wafer is adsorbed under the polishing head 101, and the polishing head 101 drives the wafer to perform polishing on the polishing pad 102 according to a set trajectory. Under the combined action of the chemical liquid and physical mechanics, the surface of the wafer is polished, so that the thickness and topography of the wafer meet the requirements.Figure 1 The dotted double-arrow line in the middle represents the motion trajectory of the grinding head 101 . The grinding head 101 can apply pressure to multiple areas, thereby carrying the wafer on the polishing plate 102 for grinding. Figure 1 The dotted line with a single arrow in the middle is the motion trajectory of the polishing disk 102, and the polishing disk 102 rotates along its axis.

[0042] In the embodiment of the present invention, Figure 2 As shown, each area of the grinding head 101 is arranged in a concentric ring structure, and the number of areas varies from one grinding head 101 to another. During the polishing process, the multiple areas of the grinding head 101 use a pneumatic system to set different pressures so that the multiple areas of the wafer are polished with different pressures, thereby achieving the desired morphology.

[0043] The embodiment of the present invention provides a grinding pressure adjustment method, which is applied to a gas-type grinding head 101. Figure 3 As shown, the following steps are included:

[0044] Step S001: determining a plurality of characteristic parameters according to the pressure setting value and the actual pressure value of each area.

[0045] In an optional embodiment, for example, a grinding head 101 having five regions is used as an example. Figure 4 As shown, the pressures in different areas interact with each other, so the characteristic parameters that affect the pressure include: the set pressure of this area, the set pressure of other areas, the swing of the pneumatic valve and other factors.

[0046] In an optional embodiment, the grinding head 101 is affected by the interaction of pressures in multiple regions, and thus a multi-input multi-output system is adopted, where multi-input refers to inputting multiple characteristic quantities, and multi-output refers to outputting pressure setting values for multiple regions.

[0047] In an optional embodiment, for example, assuming that the polishing head 101 has n regions and m characteristic quantities are set, wherein the characteristic quantities include the set value of each pressure, the difference between the actual value and the set value of the pressure in each region, etc., then the input X = [x1x2x3…xm] T .

[0048] In an optional embodiment, each row of input values represents an m-dimensional feature vector, and a feature vector has m eigenvalues.

[0049] In an optional embodiment, all characteristic quantities are used as input values. For example, if the set value is 5, then 5 will be used as an input value; if the actual value during processing is 5.3, then the difference between the actual value and the set value, 0.3, will also be used as an input value.

[0050] Step S002: Calculate the pressure setting value of each area that minimizes the function output value based on the prediction function. The prediction function is established based on each characteristic parameter and the weight of each characteristic parameter. The function output value is used to represent the difference between the pressure setting value and the actual pressure value. The weight of each characteristic parameter is obtained based on the existing data set.

[0051] In an optional embodiment, the function output value is used to represent the difference between the pressure setting value and the actual pressure value. The smaller the difference is, the closer the actual pressure value of the gas grinding head 101 is to the set pressure value when the set pressure value is used to control the gas grinding head 101, and the better the pressure control is. Therefore, in this embodiment of the present invention, the pressure setting value of each area that minimizes the function output value can be calculated based on the prediction function.

[0052] In an optional embodiment, the present invention needs to train a set of appropriate weights so that the function output value is as close to the label value as possible.

[0053] Step S003: Control the gas polishing head 101 according to the pressure setting value of each area.

[0054] In actual applications, the pressure value required for work can be set as the actual pressure value, and then the set pressure value with the smallest difference from the pressure value required for work is calculated, so that the gas grinding head 101 can be controlled according to the set pressure value, thereby achieving precise control of the gas grinding head 101.

[0055] The embodiment of the present invention calculates the pressure setting value of each area that minimizes the function output value through a prediction function. Since the function output value of the prediction function in the embodiment of the present invention is used to represent the difference between the pressure setting value and the actual pressure value, the smaller the difference is, the closer the actual pressure value of the gas grinding head 101 is to the set pressure value when the set pressure value is used to control the gas grinding head 101, and the better the pressure control is. Therefore, controlling the gas grinding head 101 to grind the wafer according to the pressure setting value calculated in the embodiment of the present invention can achieve stable control of the pressure of the gas grinding head 101, improve the pressure stability, improve the accuracy of the grinding process, and make the ground wafers more in line with the standards.

[0056] In an optional embodiment, the existing data set includes the pressure setting value and actual pressure value of each area, and the weight of each characteristic parameter is determined by the following steps:

[0057] First, a label quantity vector is determined according to the expected value of the pressure setting value of each region in the existing data set, the mean value of the actual pressure value, and the variance of the actual pressure value.

[0058] In an optional embodiment, the existing data set includes multiple pieces of data, and each piece of data includes the set pressure value and the actual pressure value of each region. That is, for the same region, there are multiple set pressure values and multiple actual pressure values corresponding to it. The tag quantity vector includes multiple tag quantities, and the number of tag quantities is the same as the number of regions of the pneumatic polishing head 101. One tag quantity in the tag quantity vector is calculated based on the expected value of the pressure set value, the expected value of the actual pressure value, and the variance of the actual pressure value corresponding to one region.

[0059] Secondly, calculate the characteristic quantities corresponding to each characteristic parameter according to the pressure set value and the actual pressure value of each region in the existing data set.

[0060] Thirdly, input the characteristic quantities into the prediction function, and establish a two-norm matrix according to the difference between the tag quantity vector and the prediction function.

[0061] In the embodiment of the present invention, the prediction function is a multi-input multi-output function, and the output value of the prediction function includes the difference between the pressure set value and the actual pressure value of each region of the pneumatic polishing head 101. It can be seen that in the embodiment of the present invention, both the tag quantity vector and the output value of the prediction function are vectors containing multiple data, and the number of data contained in the tag quantity vector and the prediction function is the same, both equal to the number of regions of the pneumatic polishing head 101. Therefore, a matrix can be established based on the difference between the tag quantity vector and the prediction function.

[0062] Finally, take the derivative of the two-norm matrix, find the optimal solution, and obtain the weights of each characteristic parameter.

[0063] In the embodiment of the present invention, by finding the optimal solution of the weights of each characteristic parameter through the two-norm matrix, substituting the optimal weights into the prediction function, the optimal output value can be obtained, and thus the most accurate pressure set value can be obtained.

[0064] In an optional embodiment, the polishing pressure trimming method provided by the embodiment of the present invention further includes:

[0065] Firstly, obtain the actual pressure value of each region of the pneumatic polishing head 101 when controlling the pneumatic polishing head 101 according to the pressure value set of each region;

[0066] Secondly, store the pressure set value and the actual pressure value of each region into the existing data set;

[0067] Finally, update the weights of each characteristic parameter according to the updated existing data set.

[0068] In the embodiment of the present invention, new data is continuously stored in the data set, and the updated weights of each characteristic parameter are obtained according to the updated data. Continuous updating makes the obtained weights more accurate.

[0069] In an optional embodiment, the following formula is established to predict the difference between the set pressure value and the actual pressure value:

[0070] F(X) = WX + b,

[0071] where b is the intercept, so the above formula is changed to:

[0072]

[0073] W = [w1 w2 w3…wm w(m + 1)] T ,

[0074] where X is the feature matrix, represents the m-th feature parameter of the n-th region, W is the weight vector, and wm represents the weight of the m-th feature parameter.

[0075] Through the above change, the intercept is unified as a one-dimensional feature quantity in the following formula:

[0076] F(X) = X T W,

[0077] Thus, F(X) = X T W can be determined as the prediction function.

[0078] In an optional embodiment, each feature parameter corresponds to a weight.

[0079] In an optional embodiment, X is the feature matrix, and each row in X is an m-dimensional feature vector and a constant term. Exemplarily, the first vector has m eigenvalues: Then the result of multiplying X T and W is

[0080] In an optional embodiment, the two-norm matrix is:

[0081]

[0082] where Y represents the label quantity vector; F(x) represents the predicted output value;

[0083] Y = [y1 y2 y3…yi] T , yi = log[(Set - Ui) 2 *σ];

[0084] where Y represents the label quantity vector; yi represents the label quantity of the i-th data, Ui represents the expected value of the pressure set value in the i-th data, Set represents the actual pressure expected value, and σ represents the variance of the actual pressure value.

[0085] In an embodiment of the present invention, the label quantity of the i-th data refers to the label quantity calculated based on the mean and variance of multiple actual pressure values corresponding to the i-th region.

[0086] In an optional embodiment, when a set of values is taken for the feature quantity, there will be a set of observed values, that is, the actual pressure mean. (Set - Ui) in the yi formula represents the distance between the expected value of the actual pressure value and the expected value of the pressure set value. Therefore, yi can characterize the quality of the actually output pressure value. The smaller yi is, the closer the actual pressure mean is to the pressure set value, and the better the pressure control.

[0087] In an optional embodiment, σ represents the variance of the actual pressure value. The smaller σ is, the more concentrated the data is.

[0088] In an optional embodiment, the steps of deriving the two-norm matrix to obtain the optimal solution and getting the weights of each feature parameter include:

[0089]

[0090]

[0091] In an optional embodiment, the grinding pressure trimming method provided by the embodiment of the present invention further includes:

[0092] First, obtain the credibility of the weights of each feature parameter;

[0093] Second, if the credibility of the weights of each feature parameter meets the preset conditions, execute the step of calculating the pressure set value of each region that makes the function output value the smallest according to the prediction function.

[0094] In an optional embodiment, obtaining the credibility of the weights of each feature parameter verifies the rationality of the currently trained weight W. All values are divided into a training set and a verification set. The specific verification can be carried out on the verification set, and MSELoss can be used for verification. The specific formula is If the weight W is trained more reasonably, MSELoss will be smaller.

[0095] In an embodiment of the present invention, verifying the credibility of the weights of each feature parameter, if it meets the preset conditions, continue to execute the next step. If the credibility of the weight is verified and a weight with a large deviation is detected, it will not be brought into the formula.

[0096] An embodiment of the present invention further provides a grinding pressure trimming device, which is applied to a gas type grinding head 101, as Figure 5 shown, and includes the following modules:

[0097] Determination module 501 is configured to determine multiple characteristic parameters according to the pressure set values and actual pressure values of each region. For detailed content, refer to the description of step S001 in the above embodiment, which will not be elaborated here.

[0098] Calculation module 502 is configured to calculate the pressure set values of each region that make the function output value minimum according to the prediction function. For detailed content, refer to the description of step S002 in the above embodiment, which will not be elaborated here.

[0099] Control module 503 is configured to control the gas abrasive head 101 according to the pressure set values of each region. For detailed content, refer to the description of step S003 in the above embodiment, which will not be elaborated here.

[0100] The abrasive pressure trimming device provided by the embodiment of the present invention calculates the pressure set values of each region that make the function output value minimum through the prediction function, and then controls the gas abrasive head 101 to grind the wafer according to the calculated pressure set values, so as to stably control the pressure of the abrasive head 101, improve the stability of the pressure, enhance the precision of the grinding process, and make the ground wafer more in line with the standard.

[0101] The embodiment of the present invention also provides a computer device, such as Figure 6 is a schematic hardware structure diagram of a computer device proposed according to an exemplary embodiment.

[0102] Such as Figure 6 shown, the device includes one or more processors 601 and a memory 602. The memory 602 includes persistent memory, volatile memory, and a hard disk. Figure 6 Taking one processor 601 as an example. The device may further include: an input device 603 and an output device 604.

[0103] The processor 601, the memory 602, the input device 603, and the output device 604 may be connected through a bus or other means. Figure 6 Taking the connection through the bus as an example.

[0104] The processor 601 may be a Central Processing Unit (CPU). The processor 601 may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., or a combination of the above types of chips. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0105] As a non-transitory computer-readable storage medium, the memory 602 includes persistent memory, volatile memory, and a hard disk, and can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instruction modules corresponding to the service management method in the embodiments of the present application. The processor 601 executes various functional applications and data processing of the server by running the non-transitory software programs, instructions, and modules stored in the memory 602, that is, implements any one of the above grinding pressure dressing methods.

[0106] The memory 602 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data used according to needs, etc. In addition, the memory 602 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory 602 may optionally include a memory remotely set relative to the processor 601, and these remote memories may be connected to the data processing device through a network. Examples of the above networks include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0107] The input device 603 can receive input digital or character information, and generate key signal inputs related to user settings and function controls. The output device 604 may include display devices such as a display screen.

[0108] One or more modules are stored in the memory 602 and, when executed by one or more processors 601, execute as Figure 1 shown in the method.

[0109] The above product can execute the method provided by the embodiments of the present invention, and has corresponding functional modules and beneficial effects for executing the method. For technical details not described in detail in this embodiment, reference may be made specifically to, for example,Figure 1 The relevant description in the illustrated embodiment.

[0110] An embodiment of the present invention also provides a computer-readable storage medium, such as Figure 7 As shown, computer-executable instructions 701 are stored in the computer-readable storage medium, and the computer-executable instructions 701 can execute the grinding pressure dressing method in any of the above method embodiments.

[0111] The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0112] Obviously, the above embodiments are merely examples for clear illustration and not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to enumerate all the implementation manners here. And the obvious changes or modifications derived therefrom are still within the protection scope of the present invention.

Claims

1. A grinding pressure dressing method, characterized in that: Applied to a gas-type grinding head, the gas-type grinding head includes multiple areas, and the method includes: Determine multiple characteristic parameters based on the pressure setting value and actual pressure value of each area; the multiple characteristic parameters include: the set pressure of the area, the set pressure of other areas, and the swing of the pneumatic valve; Calculating the pressure setting value for each region that minimizes the function output value based on a prediction function, wherein the prediction function is established based on each characteristic parameter and the weight of each characteristic parameter. The function output value is used to represent the difference between the pressure setting value and the actual pressure value. The weight of each characteristic parameter is obtained based on an existing data set. controlling the gas-type grinding head according to the pressure setting value of each zone; The existing data set includes the pressure setting value and actual pressure value of each area. The weight of each characteristic parameter is determined by the following steps: Determine a label quantity vector according to the expected value of the pressure setting value, the expected value of the actual pressure value, and the variance of the actual pressure value of each area in the existing data set; Calculate the characteristic value corresponding to each characteristic parameter according to the pressure setting value and the actual pressure value of each area in the existing data set; Inputting the feature quantity into the prediction function, and establishing a two-normal form matrix according to the difference between the label quantity vector and the prediction function; Derivative the second normal form matrix to obtain the optimal solution and obtain the weights of the characteristic parameters; The method further comprises: Obtaining actual pressure values of each region of the gas-type grinding head when the gas-type grinding head is controlled according to the pressure value set value of each region; Storing the pressure setting value and actual pressure value of each area in the existing data set; Updating the weights of the feature parameters according to the updated existing data set; The method further comprises: Obtaining the credibility of the weights of the characteristic parameters; If the credibility of the weights of the characteristic parameters meets the preset conditions, the step of calculating the pressure setting value of each area that minimizes the function output value according to the prediction function is performed.

2. The grinding pressure dressing method according to claim 1, characterized in that: The prediction function is: F(X)=X T W, in, ， W=[w1 w2 w3 … wm w(m+1)] T Where X is the feature matrix, represents the mth feature parameter of the nth region, W is the weight vector, and wm represents the weight of the mth feature parameter.

3. The grinding pressure dressing method according to claim 2, characterized in that: The second normal form matrix is: J= ， Among them, Y represents the label quantity vector; F(x) represents the predicted output value; Y=[y1 y2 y3 … yi] T , yi = log[ * ]; Among them, Y represents the label quantity vector; yi represents the label quantity of the i-th data, Ui represents the expected value of the pressure setting value in the i-th data, represents the expected value of the actual pressure, and σ represents the variance of the actual pressure value.

4. The grinding pressure dressing method according to claim 3, wherein: The step of deriving the second normal form matrix to obtain the optimal solution and obtaining the weights of the characteristic parameters includes: = = = = 。 5. A grinding pressure dressing device, characterized in that: Used to perform the grinding pressure dressing method according to any one of claims 1 to 4, applied to a gas-type grinding head, comprising the following modules: A determination module, configured to determine a plurality of characteristic parameters according to a pressure setting value and an actual pressure value of each area; A calculation module, configured to calculate, based on the prediction function, a pressure setting value for each region that minimizes the function output value; The control module is used to control the gas-type grinding head according to the pressure setting value of each area.

6. A computer device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to perform the grinding pressure dressing method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the grinding pressure dressing method according to any one of claims 1 to 4.

Images