CN117140591A - Adsorption force detection system and method for wafer transfer mechanical arm in ultra-clean environment - Google Patents

Abstract

The invention belongs to the technical field of wafer transfer mechanical arms, and discloses an adsorption force detection method of a wafer transfer mechanical arm in an ultra-clean environment, which comprises the following steps: acquiring historical training data set of the adsorption force of the transfer mechanical arm, training a machine learning model, and acquiring an adsorption force error value through predicted adsorption force data and real-time adsorption force data; based on the adsorption error value, acquiring an adsorption transfer coefficient by combining the wafer motion coefficient; after the adsorption transfer coefficient is compared with a preset safe adsorption threshold value, a safety level and a safety instruction are generated; compared with the prior art, the invention can combine the data parameters of the wafer and the ultra-clean environment, and timely and accurately identify the condition that the suction force of the suction cup is too large or too small, thereby accurately adjusting the operation parameters of the transfer mechanical arm, avoiding the deformation or dropping of the wafer caused by the too large or too small suction force in the suction cup and improving the stability of the wafer suction transfer.

Description

Adsorption force detection system and method for wafer transfer mechanical arm in ultra-clean environment

Technical Field

The invention relates to the technical field of wafer transfer mechanical arms, in particular to an adsorption force detection system and method of a wafer transfer mechanical arm in an ultra-clean environment.

Background

The wafer production and manufacturing process needs to be finished through a plurality of process steps, each process step is completed by a corresponding process machine, when a batch of wafers are finished in the process steps of one process machine, a transfer mechanical arm is needed to transfer the wafers from one process machine to the other process machine, and the transfer mechanical arm is used for forming a negative pressure environment in the suction cup by installing the suction cup on a transfer clamp and connecting the suction cup to vacuum equipment at a factory end, so that the wafer is adsorbed and transferred, and the adsorption force of the suction cup on the wafer directly determines whether the wafer can be safely and stably transferred.

The Chinese patent application publication No. CN116330319A discloses a state monitoring system and a state monitoring method for a carrying manipulator, which further realize the fine control of different suction nozzle assemblies in a suction cup type multi-finger manipulator by introducing external environment factors, thereby being beneficial to preventing failure or damage of carrying semi-product chips caused by insufficient suction force of the suction nozzle assemblies under the influence of the external environment factors.

The traditional wafer transferring mechanical arm detects the adsorption force in the sucker through deploying pressure detection equipment at a machine end, judges whether adsorption force data reach a preset value, so as to monitor the adsorption force of the sucker, but the detection method does not consider the influence of different wafer and ultra-clean environment on the adsorption force, when the data of the wafer and the ultra-clean environment change, the adsorption force of the sucker on the wafer also changes, and then the phenomenon that the adsorption force in the sucker is too large or too small can occur, so that the wafer is easy to deform or fall during the adsorption transfer.

In view of the above, the present invention provides a system and a method for detecting the adsorption force of a wafer transfer robot in an ultra-clean environment to solve the above-mentioned problems.

Disclosure of Invention

In order to overcome the defects in the prior art, the invention provides the following technical scheme for achieving the purposes: the method for detecting the adsorption force of the wafer transfer mechanical arm in the ultra-clean environment comprises the following steps:

collecting a historical training data set of the transfer mechanical arm, wherein the historical training data set comprises vacuum adsorption data and adsorption force data, and the vacuum adsorption data comprises an operating temperature, an operating power, an air extraction rate, a sucking disc adsorption area and a sucking disc surface roughness;

training a machine learning model for predicting the adsorbance data based on the historical training data set;

collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into a machine learning model after training to predict adsorption force data, collecting the real-time adsorption force data, and comparing the predicted adsorption force data with the real-time adsorption force data to obtain an adsorption force error value;

acquiring comprehensive parameters, namely acquiring a wafer motion coefficient according to the comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, and calculating an adsorption transfer coefficient according to the adsorption error value and the wafer motion coefficient;

Comparing and analyzing the adsorption transfer coefficient with a preset safe adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a dropping level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the dropping level;

controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; and sequentially increasing or decreasing the wafer transfer acceleration, increasing or decreasing the wafer transfer inclination and increasing or decreasing the temperature difference according to the hidden danger eliminating instruction, so that the safety level of the transfer mechanical arm is a normal level.

Further, the method for acquiring the pumping rate comprises the following steps:

reading initial pressure at an outlet of a vacuum pump by using a thermal resistance vacuum gauge, starting the vacuum pump to continuously run, reading pressure values at the outlet of the vacuum pump at equal intervals until H pressure values are continuously read and are no longer changed, marking the pressure values which are no longer changed as final pressure, recording air extraction time when the final pressure appears for the first time, wherein the air extraction time is the section time from when the vacuum pump starts to run to when the final pressure appears, and calculating the pressure difference;

The expression of the pressure difference is:；

in the method, in the process of the invention,for pressure difference>For final stress +.>Is the initial pressure;

and obtaining the air extraction rate through the ratio of the pressure difference to the air extraction time, wherein the expression of the air extraction rate is as follows:

；

in the method, in the process of the invention,for the air extraction rate +.>Pumping time for final pressure to occur;

the method for acquiring the suction area of the sucker comprises the following steps:

scanning the surface of the sucker by using a laser scanner to obtain a scanning image of the surface of the sucker, dividing the scanning image into a hollow area and a solid area, wherein the hollow area is an area corresponding to the area of a channel for gas flow in the contact surface of the sucker and a wafer, marking the circle center of the hollow area, taking the circle center as a starting point, measuring the nearest distance from the circle center to the solid area, marking the nearest distance as a contact radius, and calculating the adsorption area of the sucker through a formula;

the expression of the sucking area of the sucker is:；

in the method, in the process of the invention,is the sucking area of the sucking disc->Is of circumference rate>Is the contact radius;

the acquisition method of the surface roughness of the sucker comprises the following steps:

using a laser scanner to irradiate a laser beam on the surface of the sucker for scanning, dividing a scanned image into a sucker area and a non-sucker area, wherein the sucker area is an area corresponding to the sucker surface in the scanned image, and marks are marked in the sucker area Personal point location and measure->The respective heights of the individual spots; sequentially measuring the height difference data between two adjacent points of the sucker area along the same direction, and determining +.>The height difference data are summed up and +.>The average value of the height difference data is recorded as the surface roughness of the sucker;

the expression of the surface roughness of the sucker is:；

in the method, in the process of the invention,for the surface roughness of the sucker->Is->And height difference data.

Further, the training method of the machine learning model for predicting the adsorption force data comprises the following steps:

converting the collected historical training data set into a corresponding group of feature vectors;

taking each group of feature vectors as input of the machine learning model, taking adsorption force data corresponding to each group of vacuum adsorption data as output, taking adsorption force data actually corresponding to each group of vacuum adsorption data as a prediction target, and taking a minimum loss function value of the machine learning model as a training target; and stopping training when the loss function value of the machine learning model is smaller than or equal to a preset target loss value.

Further, the expression of the adsorption error value is:

in the method, in the process of the invention,for the adsorption error value, ++>For predicted adsorption force data, +. >Is real-time adsorption force data;

when (when)When 0, the real-time adsorption force data is larger than the predicted adsorption force data;

when (when)0, the real-time adsorption force data is equal to the predicted adsorption force data;

when (when)0, real-time adsorption forceThe data is less than predicted adsorbability data.

Further, the method for acquiring the temperature difference value comprises the following steps of;

the temperature of the wafer is measured by an infrared thermometer and is recorded asThe ambient temperature is obtained by a temperature sensor and is recorded asComparing the ambient temperature with the wafer temperature difference to obtain a temperature difference + ->

The expression of the temperature difference is:；

in the method, in the process of the invention,for the temperature difference->

The expression of the wafer motion coefficient is:；

in the method, in the process of the invention,is the wafer motion coefficient +.>Wafer transfer acceleration, which is acquired by an acceleration sensor deployed on a transfer robot arm, < >>For the wafer transfer inclination, the wafer transfer inclination is acquired by an inclination sensor disposed on the transfer robot arm, +.>、/>、/>、/>、/>Greater than 0.

Further, the adsorption transport coefficient comprises an adsorption transport drop coefficient and an adsorption transport deformation coefficient;

when (when)0, the absorption, the transportation and the dropping coefficients are corresponding;

the expression of the absorption transport drop coefficient is:；

in the method, in the process of the invention, The falling coefficient is the absorption, transportation and falling coefficient;

when (when)0, corresponding to the adsorption and transport deformation coefficient;

the expression of the adsorption transport deformation coefficient is:；

in the method, in the process of the invention,the adsorption transport deformation coefficient is;

when (when)At 0, noAnd generating an adsorption transport coefficient.

Further, the safe adsorption threshold value comprises a first safe adsorption threshold value and a second safe adsorption threshold value;

the generation method of the normal level, the deformation level and the dropping level comprises the following steps:

when (when)When 0, generating a normal level; when->Generating a deformation level when 0 is reached; when->Generating a dropping level when 0 is the same;

the deformation level comprises a first deformation level and a second deformation level, and the falling level comprises a first falling level and a second falling level;

the generation method of the primary deformation level and the secondary deformation level comprises the following steps:

presetting a first safe adsorption threshold，/>0；

When 0 isGenerating a first-level deformation level; when->Generating a second level of deformation;

the method for generating the first-stage dropping level and the second-stage dropping level comprises the following steps:

presetting a second safe adsorption threshold，/>0；

When 0 isGenerating a first-level dropping level; when->When a secondary drop level is generated.

Further, the method for generating the dangerous shutdown instruction and the hidden danger eliminating instruction comprises the following steps:

when the transfer mechanical arm is in a secondary deformation level or a secondary falling level, a dangerous shutdown instruction is generated;

When the transferring mechanical arm is at the first-stage deformation level or the first-stage dropping level, a hidden danger eliminating instruction is generated.

Further, the regulation and control method for enabling the security level of the transfer mechanical arm to be a normal level comprises the following steps:

when the transfer mechanical arm is at a first-level deformation level, the regulation and control method comprises the steps of reducing a temperature difference value, increasing the transfer acceleration of the wafer and increasing the transfer inclination of the wafer;

the steps of reducing the temperature difference, increasing the wafer transfer acceleration and increasing the wafer transfer inclination comprise:

when the temperature difference, the wafer transfer acceleration and the wafer transfer inclination do not reach the minimum temperature difference, the maximum wafer transfer acceleration and the maximum wafer transfer inclination;

firstly, increasing the wafer transfer acceleration until the transfer mechanical arm is at a normal level, and stopping increasing the wafer transfer acceleration;

when the wafer transfer acceleration is increased to the maximum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage deformation level, increasing the wafer transfer inclination, and stopping increasing the wafer transfer inclination until the transfer mechanical arm is at a normal level;

when the wafer transferring inclination is increased to the maximum wafer transferring inclination, and the transferring mechanical arm is at a first-stage deformation level, finally reducing the temperature difference until the transferring mechanical arm is at a normal level, and stopping reducing the temperature difference;

When the transfer mechanical arm is at a first-stage dropping level, the regulation and control method comprises the steps of increasing a temperature difference value, reducing the transfer acceleration of the wafer and reducing the transfer inclination of the wafer;

the steps of increasing the temperature difference, decreasing the wafer transfer acceleration, and decreasing the wafer transfer inclination include:

when the temperature difference value, the wafer transfer acceleration and the wafer transfer inclination do not reach the maximum temperature difference value, the minimum wafer transfer acceleration and the minimum wafer transfer inclination, firstly reducing the wafer transfer acceleration until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer acceleration;

when the wafer transfer acceleration is reduced to the minimum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage dropping level, reducing the wafer transfer inclination until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer inclination;

when the wafer transferring inclination is reduced to the minimum wafer transferring inclination, and the transferring mechanical arm is at a first-stage falling level, the temperature difference value is increased finally until the transferring mechanical arm is at a normal level, and the temperature difference value stops increasing.

Further, the operation of controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction is realized through a relay switch electrically connected with the transfer mechanical arm;

When the wafer transfer acceleration is increased or reduced, the rotation speed of the driving motor of the mechanical arm is increased or reduced;

when the wafer transferring gradient is increased or reduced, the driving motor of the mechanical arm increases the tilting movement stroke or reduces the tilting movement stroke;

when the temperature difference is reduced or the temperature difference is increased, the environmental temperature controller is cooled or heated.

The system comprises a first data acquisition module, a second data acquisition module, a safety evaluation module, a safety instruction module and an instruction execution module, wherein the modules are connected in a wired and/or wireless network mode;

the first data acquisition module is used for acquiring a historical training data set of the transfer mechanical arm, wherein the historical training data set comprises vacuum adsorption data and adsorption force data, and the vacuum adsorption data comprises operation temperature, operation power, air extraction rate, suction disc adsorption area and suction disc surface roughness;

the second data acquisition module acquires comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, a wafer motion coefficient is acquired, and an adsorption transfer coefficient is calculated according to the adsorption error value and the wafer motion coefficient;

The safety evaluation module is used for training a machine learning model for predicting the adsorption force data based on the historical training data set, collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into the trained machine learning model for predicting the adsorption force data, collecting the real-time adsorption force data, and comparing the predicted adsorption force data with the real-time adsorption force data to obtain an adsorption force error value;

the safety instruction module is used for comparing and analyzing the adsorption transfer coefficient with a preset safety adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a falling level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the falling level;

the instruction execution module is used for controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; and sequentially increasing or decreasing the wafer transfer acceleration, increasing or decreasing the wafer transfer inclination and increasing or decreasing the temperature difference according to the hidden danger eliminating instruction, so that the safety level of the transfer mechanical arm is a normal level.

A computer device, comprising: a processor and a memory;

wherein the memory stores a computer program for the processor to call;

And the processor executes the adsorption force detection method of the wafer transfer mechanical arm in the ultra-clean environment by calling the computer program stored in the memory.

A computer readable storage medium having stored thereon a computer program that is erasable;

when the computer program runs on the computer equipment, the computer equipment is enabled to execute the adsorption force detection method of the wafer transfer mechanical arm in the ultra-clean environment.

The adsorption force detection system and the method of the wafer transfer mechanical arm in the ultra-clean environment have the technical effects and advantages that:

according to the invention, the machine learning model is adopted to predict the adsorption force data of the transfer mechanical arm, the predicted adsorption force data is compared with the real-time adsorption force data to obtain the adsorption force error value, the adsorption transfer coefficient is obtained by combining the wafer motion coefficient on the basis of the adsorption force error value, the adsorption transfer coefficient is compared with the preset safe adsorption threshold value for analysis, the safety level of the transfer mechanical arm is obtained, and then the corresponding safety instruction is generated.

Drawings

FIG. 1 is a schematic diagram of an adsorption force detection method of a wafer transfer robot in an ultra-clean environment in embodiment 1 of the present invention;

FIG. 2 is a schematic diagram of an adsorption force detection system of a wafer transfer robot in an ultra-clean environment according to embodiment 2 of the present invention;

fig. 3 is a schematic structural diagram of an electronic device according to embodiment 3 of the present invention;

fig. 4 is a schematic structural diagram of a computer readable storage medium according to embodiment 4 of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Example 1

Referring to fig. 1, in the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to the embodiment, the method includes:

under an experimental environment, a historical training data set of the transfer mechanical arm is collected, wherein the historical training data set is collected under the condition that the suction force in a sucker is in a normal size, the suction force in the sucker is in the normal size, which means that the sucker can stably and safely adsorb a wafer without causing the phenomenon that the wafer drops due to too small suction force or deforms due to too large suction force, the historical training data set comprises vacuum adsorption data and suction force data, and the vacuum adsorption data comprises operation temperature, operation power, air suction rate, suction area of the sucker and surface roughness of the sucker;

The operating temperature refers to the internal operating temperature of the vacuum pump during operation, and when the internal operating temperature of the vacuum pump is higher, the vacuum pump is easy to generate high-temperature fault phenomenon at the moment, so that the operating stability of the vacuum pump is influenced, and the adsorption force of the sucker is unstable; the operating temperature is obtained by a temperature sensor disposed within the vacuum pump;

the operating power refers to real-time operating power when the vacuum pump operates, and when the operating power is larger or smaller, the operating stability of the vacuum pump is affected, so that the suction force of the suction cup is larger or smaller, and the suction force of the suction cup is unstable;

the factors influencing the operating power comprise operating voltage and operating current, and the operating voltage and the operating current are detected and acquired by a voltage sensor and a current sensor on the transfer mechanical arm;

the expression of the operating power is:；

in the method, in the process of the invention,for operating power, +.>For the operating voltage +.>For operating current;

the pumping speed refers to the speed of the gas flow pumped from the sucker by the vacuum pump in unit time, and when the pumping speed is higher, more gas in the sucker can be pumped out, so that the adsorption force of the sucker is higher;

the method for acquiring the air extraction rate comprises the following steps:

reading initial pressure at an outlet of a vacuum pump by using a thermal resistance vacuum gauge; starting a vacuum pump to continuously run, reading pressure values at an outlet of the vacuum pump at equal intervals until H pressure values are continuously read and are no longer changed, marking the pressure values which are no longer changed as final pressure, recording air extraction time when the final pressure appears for the first time, wherein the air extraction time is the section time from when the vacuum pump starts to run to when the final pressure appears, and calculating the pressure difference;

The expression of the pressure difference is:；

in the method, in the process of the invention,for pressure difference>For final stress +.>Is the initial pressure;

obtaining the air extraction rate through the ratio of the pressure difference to the air extraction time;

the expression of the pumping rate is:；

in the method, in the process of the invention,for the air extraction rate +.>The air extraction time length for the first occurrence of final pressure;

the sucking area of the sucker refers to the area of a channel for gas flow in the contact surface of the sucker and the wafer, and the larger the sucking area of the sucker is, the larger the sucking force of the sucker can provide for the wafer, the larger the sucking force of the sucker is;

the method for acquiring the suction area of the sucker comprises the following steps:

scanning the surface of the sucker by using a laser scanner to obtain a scanning image of the surface of the sucker, dividing the scanning image into a hollow area and a solid area, wherein the hollow area is an area corresponding to the area of a channel for gas flow in the contact surface of the sucker and a wafer, marking the circle center of the hollow area, taking the circle center as a starting point, measuring the nearest distance from the circle center to the solid area, marking the nearest distance as a contact radius, and calculating the adsorption area of the sucker through a formula;

the expression of the sucking area of the sucker is:；

in the method, in the process of the invention,is the sucking area of the sucking disc->Is of circumference rate>Is the contact radius;

the surface roughness of the sucker refers to the surface smoothness of the contact surface of the sucker and the wafer, and the smaller the surface roughness of the sucker is, the larger the sucking force of the sucker on the wafer is;

The acquisition method of the surface roughness of the sucker comprises the following steps:

using a laser scanner to scan a laser beamIrradiating the surface of the sucker to scan, dividing a scanned image into a sucker area and a non-sucker area, wherein the sucker area is an area corresponding to the sucker surface in the scanned image, and marks are marked in the sucker areaPersonal point location and measure->The respective heights of the individual spots; sequentially measuring the height difference data between two adjacent points of the sucker area along the same direction, and measuring the height difference data>The height difference data are summed up and +.>The average value of the height difference data is recorded as the surface roughness of the sucker; the same direction is clockwise or anticlockwise, and the clockwise or anticlockwise sequential measurement can ensure +.>The data of the individual points are mutually independent, so that +.>The confusion phenomenon occurs during the calculation of the individual height difference data marks, so that the accuracy of average value calculation is improved;

the expression of the surface roughness of the sucker is:；

in the method, in the process of the invention,for the surface roughness of the sucker->Is->-height difference data;

the adsorption force data refer to the adsorption force for adsorbing the wafer in the sucker, and the adsorption force data are obtained through a pressure sensor arranged in the sucker.

Training a machine learning model for predicting the adsorption force data based on the historical training data set;

The training method of the machine learning model for predicting the adsorption force data comprises the following steps:

converting a historical training data set acquired in an experimental environment into a corresponding group of feature vectors;

taking each group of feature vectors as input of the machine learning model, taking adsorption force data corresponding to each group of vacuum adsorption data as output, taking adsorption force data actually corresponding to each group of vacuum adsorption data as a prediction target, and taking a minimum loss function value of the machine learning model as a training target; stopping training when the loss function value of the machine learning model is smaller than or equal to a preset target loss value;

the machine learning model loss function value is a mean square error;

the mean square error is one of the usual loss functions by summing the loss functions=/>Training a model for the purpose of minimization, so that the machine learning model fits data better, and the performance and accuracy of the model are improved;

in the loss functionThe loss function value of the machine learning model is given, and x is the feature vector group number; m is the number of feature vector groups; />Adsorption force data corresponding to the x-th group of feature vectors,>the adsorption force data actually corresponding to the x-th group of feature vectors;

Other model parameters of the machine learning model, target loss values, optimization algorithms, verification set proportion of training set test sets, loss function optimization and the like are all realized through actual engineering, and the model parameters are obtained after experimental tuning is continuously carried out.

Collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into a machine learning model after training, collecting the real-time adsorption data, and comparing the predicted adsorption data with the real-time adsorption data to obtain an adsorption error value;

the expression of the adsorption error value is:；

in the method, in the process of the invention,for the adsorption error value, ++>For predicted adsorption force data, +.>Is real-time adsorption force data;

when (when)0, it is indicated that the real-time data of the adsorption force is larger than the predicted data of the adsorption force, when +.>When the wafer is larger, the adsorption force applied to the wafer during adsorption and transportation is also larger;

when (when)0, indicating that the real-time adsorption force data is equal to the predicted adsorption force data, and the adsorption force received during the adsorption and transportation of the wafer is proper;

when (when)0, it is indicated that the real-time data of the adsorption force is smaller than the predicted data of the adsorption force, when +.>The smaller the time, the smaller the adsorption force applied to the wafer during the adsorption and transportation;

Acquiring comprehensive parameters, namely acquiring a wafer motion coefficient according to the comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, and calculating an adsorption transfer coefficient according to the adsorption error value and the wafer motion coefficient;

the wafer motion coefficient judges the basis of the change amplitude of the wafer position state when the wafer is adsorbed and transported, when the wafer motion coefficient is larger, the probability of dropping or deforming the wafer when the wafer is adsorbed and transported is smaller, and the stability of the wafer adsorption and transportation is higher;

the wafer transferring acceleration refers to the speed of change of the movement speed in unit time during wafer adsorption transferring, when the wafer transferring acceleration is larger, the inertia force born by the wafer is larger, the wafer movement coefficient is smaller, and the wafer adsorption transferring stability is poorer; the wafer transferring acceleration is obtained by an acceleration sensor arranged on a transferring mechanical arm;

the wafer transferring inclination is the inclination amplitude of the wafer during wafer adsorption transfer, when the wafer transferring inclination is larger, the wafer motion coefficient is smaller, the vertical downward gravity component of the wafer is larger, the probability that the wafer falls down is larger, and the wafer adsorption transfer stability is poorer; the wafer transferring inclination is obtained by an inclination sensor arranged on a transferring mechanical arm;

The temperature difference is the difference between the ambient temperature and the wafer temperature in the ultra-clean environment, when the temperature difference is increased, molecules in the wafer increase heat energy, the larger the distance between the molecules is, the wafer expansion is caused, the contact surface between the wafer and the sucker becomes compact due to the expansion, the interaction force between the molecules is increased, the wafer motion coefficient is increased, and the stability of wafer adsorption and transportation is enhanced;

the temperature difference obtaining method comprises the following steps of;

the temperature of the wafer is measured by an infrared thermometer and is recorded asThe ambient temperature is obtained by a temperature sensor and is recorded asComparing the ambient temperature with the wafer temperature difference value to obtain a temperature difference value;

the expression of the temperature difference is:；

in the method, in the process of the invention,is the temperature difference;

the expression of the wafer motion coefficient is:；

in the method, in the process of the invention,is the wafer motion coefficient +.>Acceleration of wafer transport,/->For wafer transport inclination +.>、/>、/>、/>、/>All are larger than 0, wherein the weight coefficient is acquired by a person skilled in the art, and corresponding weight coefficients are set for each group of comprehensive parameters; substituting the set weight coefficient and the acquired comprehensive parameters into a formula, forming a ternary once equation set by any three formulas, screening the calculated weight coefficient and taking an average value to obtain +. >、/>、/>The average value;

in addition, it should be noted that the size of the weight coefficient is a specific numerical value obtained by quantizing each data, so that the subsequent comparison is convenient, and the size of the weight coefficient depends on the number of the comprehensive parameters and the corresponding weight coefficient is preliminarily set for each group of comprehensive parameters by a person skilled in the art;

the adsorption transfer coefficient is the probability of judging whether the wafer falls or deforms from the sucker when the sucker on the transfer mechanical arm adsorbs the wafer for transfer, and the greater the adsorption transfer coefficient is, the stronger the stability of the wafer adsorption transfer is, and the smaller the probability of the wafer falling or deforming from the sucker is;

the adsorption transfer coefficient comprises an adsorption transfer dropping coefficient and an adsorption transfer deformation coefficient;

when (when)0, indicating that the real-time adsorption force data does not reach the predicted adsorption force data, and corresponding to the adsorption, transportation and dropping coefficients;

the expression of the absorption transport drop coefficient is:；

in the method, in the process of the invention,the falling coefficient is the absorption, transportation and falling coefficient;

when (when)0, indicating that the real-time adsorption force data exceeds the predicted adsorption force data, and corresponding to the adsorption transport deformation coefficient at the moment;

the expression of the adsorption transport deformation coefficient is:；

in the method, in the process of the invention,the adsorption transport deformation coefficient is;

When (when)0, explaining the adsorption force data predicted by the real-time adsorption force data, wherein the adsorption force received during the adsorption and transportation of the wafer is proper, and the adsorption and transportation coefficient is not generated;

comparing and analyzing the adsorption transfer coefficient with a preset safe adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a dropping level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the dropping level;

controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; according to hidden danger eliminating instructions, the acceleration of transferring the wafer, the inclination of transferring the wafer and the temperature difference are sequentially increased or decreased, so that the safety level of the transferring mechanical arm is a normal level, and the stability of the wafer during the adsorption and transferring process is ensured;

the safe adsorption threshold value comprises a first safe adsorption threshold value and a second safe adsorption threshold value;

the generation method of the normal level, the deformation level and the dropping level comprises the following steps:

when (when)When 0, generating a normal level;

when (when)Generating a deformation level when 0 is reached;

when (when)Generating a dropping level when 0 is the same;

the deformation level comprises a first deformation level and a second deformation level, and the falling level comprises a first falling level and a second falling level;

The generation method of the primary deformation level and the secondary deformation level comprises the following steps:

presetting a first safe adsorption threshold，/>0；

When 0 isWhen the wafer is deformed, a first-level deformation level is generated, and the probability of deformation of the wafer is small;

when (when)Generating a second-stage deformation level, which indicates that the probability of deformation of the wafer is high;

the method for generating the first-stage dropping level and the second-stage dropping level comprises the following steps:

presetting a second safe adsorption threshold，/>0；

When 0 isWhen the wafer is dropped, a first-level dropping level is generated, which indicates that the probability of dropping the wafer is small;

when (when)When the wafer is dropped, a second-level dropping level is generated, which indicates that the probability of dropping the wafer is high;

the preset safe adsorption threshold value is preset according to the historical data reference of the transfer mechanical arm in a large number of ultra-clean environments, a numerical value capable of distinguishing the adsorption and transfer safety of the wafer is obtained, when the adsorption and transfer coefficient exceeds the safe adsorption threshold value, the fact that the difference between real-time adsorption force data and the predicted adsorption force data is large is indicated, the safety of the wafer in the adsorption and transfer process is low, and the adsorption stability of the sucker is poor;

the generation method of the dangerous shutdown instruction and the hidden danger elimination instruction comprises the following steps:

when the transferring mechanical arm is in a secondary deformation level or a secondary falling level, a dangerous shutdown instruction is generated at the control end of the transferring mechanical arm;

When the transferring mechanical arm is in a first-stage deformation level or a first-stage dropping level, the control end of the transferring mechanical arm generates hidden danger eliminating instructions, and the hidden danger eliminating instructions corresponding to the first-stage dropping level comprise increasing the temperature difference value, reducing the transferring acceleration of the wafer and reducing the transferring inclination of the wafer; the hidden danger eliminating instruction of the first-level deformation level comprises the steps of reducing a temperature difference value, increasing the wafer transfer acceleration and increasing the wafer transfer inclination;

presetting a maximum temperature difference value and a minimum temperature difference value corresponding to the temperature difference value, a maximum wafer transfer acceleration and a minimum wafer transfer acceleration corresponding to the wafer transfer acceleration, and a maximum wafer transfer inclination and a minimum wafer transfer inclination corresponding to the wafer transfer inclination; the temperature adjustable range for wafer adsorption transfer in an ultra-clean environment is smaller, so that the priority of increasing or reducing the temperature difference is lowest, and the priority of increasing or reducing the wafer transfer acceleration is highest because the wafer transfer acceleration directly influences the time for completing the wafer transfer process, and on the whole, the priority of increasing or reducing the wafer transfer inclination is centered;

the regulation and control method for enabling the safety level of the transfer mechanical arm to be a normal level comprises the following steps:

When the transfer mechanical arm is at a first-level deformation level, the regulation and control method comprises the steps of reducing a temperature difference value, increasing the transfer acceleration of the wafer and increasing the transfer inclination of the wafer; when the transfer mechanical arm is at a first-stage dropping level, the regulation and control method comprises the steps of increasing a temperature difference value, reducing the transfer acceleration of the wafer and reducing the transfer inclination of the wafer;

the steps of reducing the temperature difference, increasing the wafer transfer acceleration and increasing the wafer transfer inclination comprise:

when the temperature difference value, the wafer transfer acceleration and the wafer transfer gradient do not reach the minimum temperature difference value, the maximum wafer transfer acceleration and the maximum wafer transfer gradient, the wafer transfer acceleration is increased firstly until the transfer mechanical arm is at a normal level, and the wafer transfer acceleration stops increasing;

when the wafer transfer acceleration is increased to the maximum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage deformation level, increasing the wafer transfer inclination, and stopping increasing the wafer transfer inclination until the transfer mechanical arm is at a normal level;

when the wafer transferring inclination is increased to the maximum wafer transferring inclination, and the transferring mechanical arm is at a first-stage deformation level, finally reducing the temperature difference until the transferring mechanical arm is at a normal level, and stopping reducing the temperature difference;

When the wafer transfer acceleration and the wafer transfer gradient are respectively increased to the maximum wafer transfer acceleration, the maximum wafer transfer gradient and the minimum temperature difference, the transfer mechanical arm is at a normal level, and is adjusted on the basis that the transfer mechanical arm is at a first-stage deformation level, the transfer mechanical arm is at an adjustable state at the moment, the condition that the transfer mechanical arm is at the normal level after adjustment is satisfied, if the transfer mechanical arm is at a second-stage deformation level, the transfer mechanical arm needs to be stopped, and the transfer mechanical arm is at an unadjustable state at the moment;

by way of example, the way of increasing the acceleration of transferring the wafer is to send the information of increasing the acceleration to the driving motor of the mechanical arm through the control end of the transferring mechanical arm, so that the driving motor of the mechanical arm increases the rotating speed and increases the acceleration of transferring the wafer; the method for increasing the transfer inclination of the wafer comprises the steps that information of inclination increase is sent to a driving motor of the mechanical arm through a control end of the transfer mechanical arm, so that the driving motor of the mechanical arm increases the inclination movement stroke and increases the transfer inclination of the wafer; the mode of reducing the temperature difference sends information of reducing the ambient temperature to the ambient temperature controller through the control end of the transfer mechanical arm, so that the ambient temperature controller increases the refrigerating capacity and reduces the temperature difference;

The steps of increasing the temperature difference, decreasing the wafer transfer acceleration, and decreasing the wafer transfer inclination include:

when the temperature difference value, the wafer transfer acceleration and the wafer transfer inclination do not reach the maximum temperature difference value, the minimum wafer transfer acceleration and the minimum wafer transfer inclination, firstly reducing the wafer transfer acceleration until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer acceleration;

when the wafer transfer acceleration is reduced to the minimum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage dropping level, reducing the wafer transfer inclination until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer inclination;

when the wafer transferring inclination is reduced to the minimum wafer transferring inclination, and the transferring mechanical arm is at a first-stage dropping level, finally increasing the temperature difference value until the transferring mechanical arm is at a normal level, and stopping increasing the temperature difference value;

when the wafer transfer acceleration and the wafer transfer gradient are respectively reduced to the minimum wafer transfer acceleration, the reduced wafer transfer gradient and the maximum temperature difference, the transfer mechanical arm is at a normal level, and is adjusted on the basis that the transfer mechanical arm is at a first-stage dropping level, the transfer mechanical arm is at an adjustable state at the moment, the condition that the transfer mechanical arm is at the normal level after adjustment is satisfied, if the transfer mechanical arm is at a second-stage dropping level, the transfer mechanical arm needs to be stopped, and the transfer mechanical arm is at an unadjustable state at the moment;

By way of example, the way of reducing the acceleration of transferring the wafer is to send the information of reducing the acceleration to the driving motor of the mechanical arm through the control end of the transferring mechanical arm, so that the driving motor of the mechanical arm reduces the rotating speed and reduces the acceleration of transferring the wafer; the mode of reducing the transfer inclination of the wafer sends information of the inclination reduction to a driving motor of the mechanical arm through a control end of the transfer mechanical arm, so that the driving motor of the mechanical arm reduces the inclination movement stroke and reduces the transfer inclination of the wafer; the mode of increasing the temperature difference sends information of increasing the ambient temperature to the ambient temperature controller through the control end of the transfer mechanical arm, so that the ambient temperature controller reduces the refrigerating capacity and increases the temperature difference;

the method comprises the steps of realizing the stop operation of a transfer mechanical arm, enabling the safety level of the transfer mechanical arm to be a normal level, and enabling relevant execution hardware, a drive motor and an environmental temperature controller to be relay switches, wherein the relevant execution hardware is existing equipment;

the operation of controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction is realized through a relay switch electrically connected with the transfer mechanical arm, and the relay switch can be used for switching on or switching off a power supply of the transfer mechanical arm;

when the wafer transfer acceleration is increased or reduced, the rotation speed of the driving motor of the mechanical arm is increased or reduced;

When the wafer transferring gradient is increased or reduced, the driving motor of the mechanical arm increases the tilting movement stroke or reduces the tilting movement stroke;

when the temperature difference is reduced or the temperature difference is increased, the environmental temperature controller is cooled or heated.

Example 2

Referring to fig. 2, the embodiment is not described in detail in embodiment 1, and provides an adsorption force detection system of a wafer transfer mechanical arm in an ultra-clean environment, which is used for implementing an adsorption force detection method of the wafer transfer mechanical arm in the ultra-clean environment, and the system comprises a first data acquisition module, a second data acquisition module, a security evaluation module, a security instruction module and an instruction execution module, wherein the modules are connected by a wired and/or wireless network manner;

the first data acquisition module is used for acquiring a historical training data set of the transfer mechanical arm, wherein the historical training data set comprises vacuum adsorption data and adsorption force data, and the vacuum adsorption data comprises operation temperature, operation power, air extraction rate, suction disc adsorption area and suction disc surface roughness;

the second data acquisition module acquires comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, a wafer motion coefficient is acquired, and an adsorption transfer coefficient is calculated according to the adsorption error value and the wafer motion coefficient;

The safety evaluation module is used for training a machine learning model for predicting the adsorption force data based on the historical training data set, collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into the trained machine learning model for predicting the adsorption force data, collecting the real-time adsorption force data, and comparing the predicted adsorption force data with the real-time adsorption force data to obtain an adsorption force error value;

the safety instruction module is used for comparing and analyzing the adsorption transfer coefficient with a preset safety adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a falling level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the falling level;

the instruction execution module is used for controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; and sequentially increasing or decreasing the wafer transfer acceleration, increasing or decreasing the wafer transfer inclination and increasing or decreasing the temperature difference according to the hidden danger eliminating instruction, so that the safety level of the transfer mechanical arm is a normal level.

Example 3

Referring to fig. 3, the disclosure provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor implements the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to any one of the methods provided by the above when executing the computer program.

Since the electronic device described in this embodiment is an electronic device used for implementing the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment in embodiment 1 of the present application, based on the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment described in this embodiment, a person skilled in the art can understand the specific implementation of the electronic device and various modifications thereof, so how to implement the method in this embodiment of the present application for this electronic device will not be described in detail herein. As long as the person skilled in the art implements the electronic device adopted by the method for detecting the adsorption force of the wafer transfer mechanical arm in the ultra-clean environment in the embodiment of the application, the method belongs to the scope of protection required by the application.

Example 4

Referring to fig. 4, the disclosure provides a computer readable storage medium, which includes a memory, a processor, and a computer program stored on the memory and capable of running on the processor, wherein the processor implements the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to any one of the methods provided by the above when executing the computer program.

The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (13)

1. The method for detecting the adsorption force of the wafer transfer mechanical arm in the ultra-clean environment is characterized by comprising the following steps of:

collecting a historical training data set of the transfer mechanical arm, wherein the historical training data set comprises vacuum adsorption data and adsorption force data, and the vacuum adsorption data comprises an operating temperature, an operating power, an air extraction rate, a sucking disc adsorption area and a sucking disc surface roughness;

training a machine learning model for predicting the adsorbance data based on the historical training data set;

collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into a machine learning model after training to predict adsorption force data, collecting the real-time adsorption force data, and comparing the predicted adsorption force data with the real-time adsorption force data to obtain an adsorption force error value;

acquiring comprehensive parameters, namely acquiring a wafer motion coefficient according to the comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, and calculating an adsorption transfer coefficient according to the adsorption error value and the wafer motion coefficient;

Comparing and analyzing the adsorption transfer coefficient with a preset safe adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a dropping level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the dropping level;

controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; and sequentially increasing or decreasing the wafer transfer acceleration, increasing or decreasing the wafer transfer inclination and increasing or decreasing the temperature difference according to the hidden danger eliminating instruction, so that the safety level of the transfer mechanical arm is a normal level.

2. The method for detecting the adsorption force of the wafer transfer robot arm in the ultra-clean environment according to claim 1, wherein the method for acquiring the pumping rate comprises the following steps:

reading initial pressure at an outlet of a vacuum pump by using a thermal resistance vacuum gauge, starting the vacuum pump to continuously run, reading pressure values at the outlet of the vacuum pump at equal intervals until H pressure values are continuously read and are no longer changed, marking the pressure values which are no longer changed as final pressure, recording air extraction time when the final pressure appears for the first time, wherein the air extraction time is the section time from when the vacuum pump starts to run to when the final pressure appears, and calculating the pressure difference;

The expression of the pressure difference is:；

in the method, in the process of the invention,for pressure difference>For final stress +.>Is the initial pressure;

and obtaining the air extraction rate through the ratio of the pressure difference to the air extraction time, wherein the expression of the air extraction rate is as follows:

；

in the method, in the process of the invention,for the air extraction rate +.>Pumping time for the first occurrence of final pressure;

the method for acquiring the suction area of the sucker comprises the following steps:

scanning the surface of the sucker by using a laser scanner to obtain a scanning image of the surface of the sucker, dividing the scanning image into a hollow area and a solid area, wherein the hollow area is an area corresponding to the area of a channel for gas flow in the contact surface of the sucker and a wafer, marking the circle center of the hollow area, taking the circle center as a starting point, measuring the nearest distance from the circle center to the solid area, marking the nearest distance as a contact radius, and calculating the adsorption area of the sucker through a formula;

the expression of the sucking area of the sucker is as follows:；

in the method, in the process of the invention,is the sucking area of the sucking disc->Is of circumference rate>Is the contact radius;

the acquisition method of the surface roughness of the sucker comprises the following steps:

using a laser scanner to irradiate a laser beam on the surface of the sucker for scanning, dividing a scanned image into a sucker area and a non-sucker area, wherein the sucker area is an area corresponding to the sucker surface in the scanned image, and marks are marked in the sucker area Personal point location and measure->The respective heights of the individual spots; sequentially measuring the height difference data between two adjacent points of the sucker area along the same direction, and determining +.>The height difference data are summed up and +.>The average value of the height difference data is recorded as the surface roughness of the sucker;

the expression of the surface roughness of the sucker is:；

in the method, in the process of the invention,for the surface roughness of the sucker->Is->Height of eachAnd (5) degree difference data.

3. The method for detecting the adsorption force of the wafer transfer robot arm in the ultra-clean environment according to claim 2, wherein the training method for predicting the machine learning model of the adsorption force data comprises the following steps:

converting the collected historical training data set into a corresponding group of feature vectors;

taking each group of feature vectors as input of the machine learning model, taking adsorption force data corresponding to each group of vacuum adsorption data as output, taking adsorption force data actually corresponding to each group of vacuum adsorption data as a prediction target, and taking a minimum loss function value of the machine learning model as a training target; and stopping training when the loss function value of the machine learning model is smaller than or equal to a preset target loss value.

4. The method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to claim 3, wherein the expression of the adsorption force error value is: ；

In the method, in the process of the invention,for the adsorption error value, ++>For predicted adsorption force data, +.>Is real-time adsorption force data;

when (when)When 0, the real-time adsorption force data is larger than the predicted adsorption force data;

when (when)0, the real-time adsorption force data is equal to the predicted adsorption force data;

when (when)And 0, wherein the real-time adsorption force data is smaller than the predicted adsorption force data.

5. The method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to claim 4, wherein the method for obtaining the temperature difference value comprises the steps of;

the temperature of the wafer is measured by an infrared thermometer and is recorded asThe ambient temperature is obtained by a temperature sensor and is recorded asComparing the ambient temperature with the wafer temperature difference value to obtain a temperature difference value;

the expression of the temperature difference is:；

in the method, in the process of the invention,is the temperature difference;

the expression of the wafer motion coefficient is:；

in the method, in the process of the invention,is the wafer motion coefficient +.>Acceleration of wafer transportThe wafer transferring acceleration is obtained by an acceleration sensor arranged on a transferring mechanical arm, +.>For the wafer transfer inclination, the wafer transfer inclination is acquired by an inclination sensor disposed on the transfer robot arm, +.>、/>、/>、/>、/>Are all greater than 0.

6. The method for detecting the adsorption force of the wafer transfer robot arm in the ultra-clean environment according to claim 5, wherein the adsorption transfer coefficient comprises an adsorption transfer drop coefficient and an adsorption transfer deformation coefficient;

When (when)0, the absorption, the transportation and the dropping coefficients are corresponding;

the expression of the absorption transport drop coefficient is:；

in the method, in the process of the invention,the falling coefficient is the absorption, transportation and falling coefficient;

when (when)0, corresponding to the adsorption and transport deformation coefficient;

the expression of the adsorption transport deformation coefficient is:；

in the method, in the process of the invention,the adsorption transport deformation coefficient is;

when (when)At 0, no adsorption transport coefficient is generated.

7. The method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to claim 6, wherein the safe adsorption threshold value comprises a first safe adsorption threshold value and a second safe adsorption threshold value;

the generation method of the normal level, the deformation level and the dropping level comprises the following steps:

when (when)When 0, generating a normal level; when->Generating a deformation level when 0 is reached; when->Generating a dropping level when 0 is the same;

the deformation level comprises a first deformation level and a second deformation level, and the falling level comprises a first falling level and a second falling level;

the generation method of the primary deformation level and the secondary deformation level comprises the following steps:

presetting a first safe adsorption threshold，/>0；

When 0 isGenerating a first-level deformation level; when->Generating a second level of deformation;

the method for generating the first-stage dropping level and the second-stage dropping level comprises the following steps:

presetting a second safe adsorption threshold ，/>0；

When 0 isGenerating a first-level dropping level; when->When a secondary drop level is generated.

8. The method for detecting the adsorption force of the wafer transfer robot arm in the ultra-clean environment according to claim 7, wherein the method for generating the dangerous shutdown command and the hidden danger elimination command comprises the steps of:

when the transfer mechanical arm is in a secondary deformation level or a secondary falling level, a dangerous shutdown instruction is generated;

when the transferring mechanical arm is at the first-stage deformation level or the first-stage dropping level, a hidden danger eliminating instruction is generated.

9. The method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to claim 8, wherein the adjusting and controlling method for enabling the safety level of the transfer robot to be a normal level is as follows:

when the transfer mechanical arm is at a first-level deformation level, the regulation and control method comprises the steps of reducing a temperature difference value, increasing the transfer acceleration of the wafer and increasing the transfer inclination of the wafer;

the steps of reducing the temperature difference, increasing the wafer transfer acceleration and increasing the wafer transfer inclination comprise:

when the temperature difference, the wafer transfer acceleration and the wafer transfer inclination do not reach the minimum temperature difference, the maximum wafer transfer acceleration and the maximum wafer transfer inclination;

Firstly, increasing the wafer transfer acceleration until the transfer mechanical arm is at a normal level, and stopping increasing the wafer transfer acceleration;

when the wafer transfer acceleration is increased to the maximum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage deformation level, increasing the wafer transfer inclination, and stopping increasing the wafer transfer inclination until the transfer mechanical arm is at a normal level;

when the wafer transferring inclination is increased to the maximum wafer transferring inclination, and the transferring mechanical arm is at a first-stage deformation level, finally reducing the temperature difference until the transferring mechanical arm is at a normal level, and stopping reducing the temperature difference;

when the transfer mechanical arm is at a first-stage dropping level, the regulation and control method comprises the steps of increasing a temperature difference value, reducing the transfer acceleration of the wafer and reducing the transfer inclination of the wafer;

the steps of increasing the temperature difference, decreasing the wafer transfer acceleration, and decreasing the wafer transfer inclination include:

when the temperature difference value, the wafer transfer acceleration and the wafer transfer inclination do not reach the maximum temperature difference value, the minimum wafer transfer acceleration and the minimum wafer transfer inclination, firstly reducing the wafer transfer acceleration until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer acceleration;

When the wafer transfer acceleration is reduced to the minimum wafer transfer acceleration, and the transfer mechanical arm is at a first-stage dropping level, reducing the wafer transfer inclination until the transfer mechanical arm is at a normal level, and stopping reducing the wafer transfer inclination;

when the wafer transferring inclination is reduced to the minimum wafer transferring inclination, and the transferring mechanical arm is at a first-stage falling level, the temperature difference value is increased finally until the transferring mechanical arm is at a normal level, and the temperature difference value stops increasing.

10. The method for detecting the adsorption force of the wafer transfer mechanical arm in the ultra-clean environment according to claim 9, wherein the operation of controlling the transfer mechanical arm to stop running according to the dangerous shutdown command is realized by a relay switch electrically connected with the transfer mechanical arm;

when the wafer transfer acceleration is increased or reduced, the rotation speed of the driving motor of the mechanical arm is increased or reduced;

when the wafer transferring gradient is increased or reduced, the driving motor of the mechanical arm increases the tilting movement stroke or reduces the tilting movement stroke;

when the temperature difference is reduced or the temperature difference is increased, the environmental temperature controller is cooled or heated.

11. The adsorption force detection system of the wafer transfer mechanical arm in the ultra-clean environment is applied to a control end of the transfer mechanical arm and is used for realizing the adsorption force detection method of the wafer transfer mechanical arm in the ultra-clean environment according to any one of claims 1-10, and is characterized by comprising a first data acquisition module, a second data acquisition module, a safety evaluation module, a safety instruction module and an instruction execution module, wherein the modules are connected in a wired and/or wireless network mode;

the first data acquisition module is used for acquiring a historical training data set of the transfer mechanical arm, wherein the historical training data set comprises vacuum adsorption data and adsorption force data, and the vacuum adsorption data comprises operation temperature, operation power, air extraction rate, suction disc adsorption area and suction disc surface roughness;

the second data acquisition module acquires comprehensive parameters, wherein the comprehensive parameters comprise wafer transfer acceleration, wafer transfer inclination and temperature difference, a wafer motion coefficient is acquired, and an adsorption transfer coefficient is calculated according to the adsorption error value and the wafer motion coefficient;

the safety evaluation module is used for training a machine learning model for predicting the adsorption force data based on the historical training data set, collecting real-time vacuum adsorption data, inputting the real-time vacuum adsorption data into the trained machine learning model for predicting the adsorption force data, collecting the real-time adsorption force data, and comparing the predicted adsorption force data with the real-time adsorption force data to obtain an adsorption force error value;

The safety instruction module is used for comparing and analyzing the adsorption transfer coefficient with a preset safety adsorption threshold value to obtain a comparison result, generating a safety level of the transfer mechanical arm according to the comparison result, wherein the safety level comprises a normal level, a deformation level and a falling level, and respectively generating a dangerous shutdown instruction or a hidden danger elimination instruction for the deformation level and the falling level;

the instruction execution module is used for controlling the transfer mechanical arm to stop running according to the dangerous shutdown instruction; and sequentially increasing or decreasing the wafer transfer acceleration, increasing or decreasing the wafer transfer inclination and increasing or decreasing the temperature difference according to the hidden danger eliminating instruction, so that the safety level of the transfer mechanical arm is a normal level.

12. A computer device, comprising: a processor and a memory;

wherein the memory stores a computer program for the processor to call;

the processor executes the method for detecting the adsorption force of the wafer transfer robot in the ultra-clean environment according to any one of claims 1 to 10 by calling the computer program stored in the memory.

13. A computer readable storage medium having stored thereon a computer program that is erasable;

when the computer program is run on a computer device, the computer device is caused to execute the method for detecting the adsorption force of the wafer transfer mechanical arm in the ultra-clean environment according to any one of claims 1 to 10.