Fault diagnosis method, system, detection device, computer equipment and medium for semiconductor light guide tube

Abstract

The application provides a fault diagnosis method, system, detection device, computer equipment and medium for a light pipe for a semiconductor, the application jointly analyzes first detection data and second detection data; generates corresponding fault diagnosis results and maintenance strategies according to the analysis results, so that engineers can distinguish the emergency states of different faults, not only can the hidden risks such as light pipe pollution, mechanical loosening and cooling efficiency reduction be found in advance, the probability of unplanned shutdown caused by light pipe fracture and overheating failure is reduced, excessive maintenance and blind regular maintenance are reduced, predictive maintenance and fine management of the light pipe are realized, and the reliability and availability of the detection device in long-term operation in the reaction cavity of the semiconductor equipment are improved.

Description

Technical Field

[0001] This application relates to the field of semiconductor testing technology, and in particular to a fault diagnosis method, system, testing device, computer equipment, and medium for semiconductor optical guides. Background Technology

[0002] In integrated circuit manufacturing, high-density plasma chemical vapor deposition (HDPCVD) is widely used in key processes such as shallow trench filling and interlayer dielectric deposition due to its high anisotropic etching capability and excellent film quality. These processes are typically performed under high-temperature, high-density plasma and complex atmospheric conditions. The actual thermal state of the wafer is highly complex, and the precise measurement and control of wafer temperature directly affects film stress, linewidth morphology, and device consistency and yield.

[0003] In traditional technologies, to achieve non-contact measurement of wafer temperature, a scheme using a light guide combined with a radiation temperature measurement device is typically employed. This involves inserting a light guide, made of materials such as sapphire, directly into the HDPCVD reaction chamber. Infrared radiation or reflected signals from the wafer surface are collected through the light guide's end face and transmitted to an external optical temperature measurement unit for processing. While these light guide temperature probes have seen some application in engineering practice, they have revealed several limitations and risks under harsh operating conditions such as HDPCVD.

[0004] Specifically, as a component that directly extends into the cavity, the optical guide is susceptible to mechanical collisions or thermal shocks caused by crystal boat transport deviations and interference from internal components. Lacking active protection and health status monitoring methods, it is typically only passively shut down after breakage or severe damage. Furthermore, the deposition of process byproducts on the optical window or the end face of the optical guide can cause optical path contamination, leading to attenuation of the temperature measurement link and drift in temperature readings. Traditional technologies often rely on manual experience or fixed-cycle maintenance, making it difficult to timely and quantitatively assess the degree of contamination and cooling status. The host computer typically only obtains a single temperature value, lacking the ability to perceive and predict the reliability of the temperature measurement data and the health status of the probe itself. Summary of the Invention

[0005] The purpose of this application is to provide a fault diagnosis method, system, detection device, computer equipment and medium for semiconductor optical guides, so as to overcome the shortcomings of traditional technologies in lacking the ability to perceive and predict the reliability of temperature measurement data and the health status of the probe itself.

[0006] In a first aspect, this application provides a fault diagnosis method for a semiconductor optical guide tube, applicable to a detection device, wherein the detection device is partially disposed within a reaction chamber of a semiconductor device, and a cooling tube is disposed within the reaction chamber; the method includes:

[0007] Acquire first detection data and second detection data; wherein, the first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset position in the reaction chamber; the preset position includes the inlet position, outlet position of the cooling pipe, and the body position of the optical guide.

[0008] The first detection data and the second detection data are analyzed respectively to obtain first fault information and second fault information; wherein, the first fault information includes a pollution index, and the second fault information includes at least one of the following: the rate of change of reflected light intensity, the rate of change of vibration, and the cooling efficiency parameter;

[0009] Based on the first fault information and the second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated so that engineers can maintain the optical guide tube according to the maintenance strategy.

[0010] In one embodiment, parsing the first detection data to obtain first fault information includes:

[0011] Image feature data is extracted from the image information, and the image feature data includes the average gray value, the standard deviation of gray distribution, and the image gradient intensity.

[0012] The rate of change of the image feature data is calculated based on the reference value to obtain multiple feature change rates; wherein, the reference value is determined based on a standard image of the protective mirror.

[0013] The pollution index is obtained by weighted summation of the change rates of each of the aforementioned characteristics.

[0014] In one embodiment, when the second detection data includes reflectance spectral data of the wafer surface, the second detection data is parsed to obtain second fault information, including:

[0015] For the reflectance spectral data collected at different times, reflectance spectral data within a preset wavelength range are selected and integrated to obtain the reflected light intensity index corresponding to each collection time.

[0016] The reflected light intensity indicators are arranged in chronological order according to the acquisition time to form a time sequence of reflected light intensity.

[0017] Based on the time series of reflected light intensity, a fitting algorithm is used to obtain the slope of the reflected light intensity changing with time, and this slope is taken as the rate of change of reflected light intensity.

[0018] In one embodiment, when the second detection data includes vibration data of the light guide tube, the second detection data is parsed to obtain second fault information, including:

[0019] Frequency domain analysis is performed on the vibration data collected within a preset time window to obtain the vibration spectrum corresponding to the vibration data at each monitoring time point;

[0020] At least one target frequency is selected from the vibration spectrum, and the spectral amplitude at the target frequency is extracted to obtain a spectral amplitude sequence that varies with time; the target frequency is related to the loosening of the optical guide tube.

[0021] Based on the spectral amplitude sequence, the rate of change of the spectral amplitude is calculated, and the rate of change is used as the vibration rate of the optical guide tube.

[0022] In one embodiment, the cooling performance parameters include effective temperature difference, heat dissipation, and equivalent thermal resistance;

[0023] If the second detection data includes temperature data at a preset location within the reaction chamber, the second detection data is parsed to obtain second fault information, including:

[0024] The effective temperature difference is calculated based on the temperature data at the outlet and inlet positions of the cooling pipe.

[0025] The heat dissipation is obtained by multiplying the medium flow rate of the cooling pipe, the effective temperature difference, and the specific heat capacity of the medium.

[0026] Calculate the temperature difference between the body position of the light guide tube and the inlet position of the cooling pipe, and calculate the ratio of the temperature difference to the heat dissipation to obtain the equivalent thermal resistance.

[0027] In one embodiment, the second fault information further includes the rate of change of the cooling efficiency parameter;

[0028] The process of parsing the second detection data to obtain the second fault information also includes:

[0029] The cooling performance parameters are subjected to moving average and trend fitting processing to obtain their rate of change.

[0030] In one embodiment, the second fault information includes the rate of change of reflected light intensity, the rate of change of vibration, and cooling efficiency parameters, wherein the cooling efficiency parameters include equivalent thermal resistance and its rate of change.

[0031] The step of generating a fault diagnosis result based on the first fault information and the second fault information includes:

[0032] If the increase in the pollution index exceeds a first preset threshold, and if the absolute value of the rate of change of reflected light intensity exceeds a preset attenuation threshold within a preset time, and the rate of change of vibration within a preset time is within a first preset range, and the rate of change of equivalent thermal resistance within a preset time is within a second preset range, then the fault diagnosis result is marked as optical pollution as the main pollution source.

[0033] Based on the fault diagnosis results, a maintenance strategy is generated, including:

[0034] Based on the increase in the contamination index, the remaining number of process cycles required to reach the cleaning threshold is predicted, and after the remaining number of process cycles is reached, a maintenance command is output to perform dust removal operation on the protective mirror during the process interval.

[0035] In one embodiment, predicting the number of remaining process cycles required to reach the cleaning threshold based on the increase in the contamination index includes:

[0036] Obtain historical pollution index and corresponding process cycle number, and construct a historical data sequence to characterize the growth of pollution index with process cycle number;

[0037] Based on the historical data sequence, a preset trend fitting algorithm is used to establish a functional relationship between the pollution index and the number of process cycles.

[0038] Substitute the target contamination index corresponding to the cleaning threshold determined in advance through experiments into the functional relationship to obtain the predicted total number of process cycles when the cleaning threshold is reached.

[0039] Calculate the difference between the predicted total number of process cycles and the number of process cycles currently completed, and use this difference as the remaining number of process cycles required to reach the cleaning threshold.

[0040] In one embodiment, the second fault information includes the vibration rate of change, and the step of generating a fault diagnosis result based on the first fault information and the second fault information further includes:

[0041] If the change rate of vibration exceeds the second preset threshold within a preset time, the fault diagnosis result is marked as a loose mechanical connection fault of the optical guide tube.

[0042] The maintenance strategy generated based on the fault diagnosis results also includes:

[0043] After the current process cycle is completed, a tightening command is output so that the engineer can reinforce the optical guide.

[0044] In one embodiment, the first fault information further includes the standard deviation of the grayscale distribution corresponding to the image information; the second fault information includes the rate of change of reflected light intensity.

[0045] The step of generating a fault diagnosis result based on the first fault information and the second fault information includes:

[0046] If the increase in the pollution index exceeds a first preset threshold, and if the change rate of the reflected light intensity exceeds a preset attenuation range within a preset time, and the characteristic change rate of the grayscale distribution standard deviation exceeds a third preset threshold, then the fault diagnosis result is marked as a pollution source of non-uniform deposition that is prone to causing deviations in temperature data at preset locations; wherein, the characteristic change rate of the grayscale distribution standard deviation is obtained based on a reference value determined by the standard image of the protective mirror.

[0047] Based on the fault diagnosis results, a maintenance strategy is generated, including:

[0048] Based on the rising slope of the equivalent thermal resistance, the required maintenance time window for the cooling system is predicted. Once the maintenance time window is reached, the processing is stopped, and a cleaning plan is output to perform dust removal on the protective mirror.

[0049] In one embodiment, predicting the required maintenance time window for the cooling system based on the rising slope of the equivalent thermal resistance includes:

[0050] The equivalent thermal resistance of the cooling pipe during the process operation is continuously monitored as a function of process time and / or process cycle number, and the monitoring results are obtained.

[0051] Based on the monitoring results, the unit growth rate of the equivalent thermal resistance relative to the process time and / or the number of process cycles is calculated as the rising slope of the equivalent thermal resistance.

[0052] Based on the current equivalent thermal resistance, the rising slope, and the critical thermal resistance, calculate the remaining time and / or the number of remaining process cycles required to grow from the current state to the critical thermal resistance, and determine the remaining time and / or the number of remaining process cycles as the maintenance time window for the cooling tube; wherein, the critical thermal resistance is determined based on the safe operation requirements of the cooling tube.

[0053] In one embodiment, the cooling performance parameter further includes the body temperature of the light guide; the step of generating a fault diagnosis result based on the first fault information and the second fault information further includes:

[0054] If the temperature rise of the main body of the light guide exceeds the fourth preset threshold, the deviation correction amount caused by the current main body temperature to the intensity of reflected light is further calculated, and the fault diagnosis result is marked as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the light guide.

[0055] The calculation of the deviation correction amount caused by the current temperature to the intensity of reflected light includes:

[0056] A contaminant thermal resistance model is established, and based on the contamination index and the temperature rise of the light guide body, the deviation correction amount caused by the contaminant layer to the temperature measurement accuracy of the light guide is quantitatively calculated.

[0057] The maintenance strategy generated based on the fault diagnosis results also includes:

[0058] The measured wafer surface temperature is compensated in real time based on the deviation correction amount to obtain the calibrated wafer surface temperature value.

[0059] Mark and output the calibrated wafer surface temperature values.

[0060] Secondly, this application provides a fault diagnosis system for a semiconductor optical guide tube, the system being disposed within a detection device, the detection device being partially disposed within a reaction chamber of a semiconductor device, and a cooling pipe being disposed within the reaction chamber; the system includes:

[0061] The acquisition module is used to acquire first detection data and second detection data; wherein, the first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset position in the reaction chamber; the preset position includes the inlet position, the outlet position of the cooling pipe, and the body position of the optical guide.

[0062] The processing module is used to parse the first detection data and the second detection data respectively to obtain first fault information and second fault information; wherein, the first fault information includes a pollution index, and the second fault information includes at least one of the following: the rate of change of reflected light intensity, the rate of change of vibration, and the cooling efficiency parameter; based on the first fault information and the second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated so that engineers can maintain the light guide tube according to the maintenance strategy.

[0063] Thirdly, this application also provides a detection device, partially disposed within the reaction chamber of a semiconductor device, wherein a cooling pipe is disposed within the reaction chamber; the device includes:

[0064] A light guide, partially disposed within the reaction chamber, is used to conduct reflected light from the wafer surface;

[0065] A detector, positioned outside the reaction cavity, is used to generate reflectance spectral data based on the reflected light.

[0066] A vibration detection module is installed at the root of the optical guide tube to acquire vibration data of the optical guide tube.

[0067] A temperature detection module is set at a preset position within the reaction chamber to acquire temperature data at the preset position; the preset position includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide tube.

[0068] An imaging module, located outside the reaction chamber, is used to acquire image information of the protective mirror of the observation window at the front end of the light guide tube;

[0069] The fault diagnosis system for semiconductor optical guides as described in the second aspect is used to acquire and parse first detection data and second detection data to obtain first fault information and second fault information, and generate fault diagnosis results and maintenance strategies based on the first fault information and the second fault information, so that engineers can maintain the optical guide according to the maintenance strategies.

[0070] In one embodiment, a sleeve is provided at the front end of the light guide tube; wherein the sleeve adopts a mechanical quick-change connection mechanism so that the sleeve will detach from the front end of the light guide tube when subjected to an axial or lateral impact force exceeding a design threshold.

[0071] In one embodiment, the connection point between the optical guide and the sleeve is provided with a non-contact identification interface, which is used to acquire the identification information of the sleeve when the front end of the optical guide is connected to the sleeve and transmit it to the fault diagnosis system of the semiconductor optical guide, so that the fault diagnosis system of the semiconductor optical guide can create or update the file information corresponding to the sleeve based on the identification information; wherein, the file information includes the number of times it has been used and vibration data.

[0072] In one embodiment, the optical guide tube is integrated with an in-situ detection unit for detecting the in-situ status of the sleeve and transmitting a trigger signal to the fault diagnosis system of the semiconductor optical guide tube when the sleeve falls off, so as to trigger the alarm mode of the fault diagnosis system of the semiconductor optical guide tube.

[0073] Fourthly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method steps of the first aspect.

[0074] Fifthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method steps of the first aspect.

[0075] The aforementioned fault diagnosis method, system, detection device, computer equipment, and medium for semiconductor optical guides have at least the following advantages:

[0076] This application acquires first and second detection data. The first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide tube. The second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide tube, and temperature data at a preset location within the reaction chamber. The preset location includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide tube. By jointly analyzing the first and second detection data, this application converts the aforementioned multi-source fault characteristics into first and second fault information, and generates corresponding fault diagnosis results and maintenance strategies accordingly. This enables engineers to distinguish the emergency states of different faults and avoid misjudgments caused by a single temperature or alarm signal. This not only allows for the early detection of hidden risks such as optical guide tube contamination, mechanical loosening, and decreased cooling efficiency, reducing the probability of unplanned downtime due to optical guide tube breakage and overheating failure, but also reduces excessive maintenance and blind periodic maintenance, enabling predictive maintenance and refined management of the optical guide tube, and improving the reliability and availability of the detection device during long-term operation within the reaction chamber of semiconductor equipment. Attached Figure Description

[0077] Figure 1 This is a structural block diagram of the detection device in one embodiment;

[0078] Figure 2 This is a flowchart illustrating a fault diagnosis method for a semiconductor optical guide in one embodiment;

[0079] Figure 3 This is a flowchart illustrating the steps for obtaining the first fault information in one embodiment;

[0080] Figure 4 This is a structural block diagram of a fault diagnosis system for a semiconductor optical guide tube in one embodiment;

[0081] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0082] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, unless otherwise specified, the following embodiments and features in the embodiments can be combined with each other.

[0083] Some exemplary embodiments of this application have been described for illustrative purposes. It should be understood that this application may be implemented in other ways not specifically shown in the accompanying drawings.

[0084] Please see Figure 1 In one exemplary embodiment, this application provides a detection device, which is partially disposed within the reaction chamber of a semiconductor device. The reaction chamber includes a base for placing a wafer, the base being made of high-purity graphite and supported by a rotating shaft driven by a motor, allowing continuous rotation during wafer processing. Further, the reaction chamber includes a cooling system for cooling the optical guide tube to prevent damage to the probe from high-temperature plasma. The cooling system includes a cooling pipe, a cooling structure, and a supply unit. The cooling pipe surrounds the optical guide tube body and contains the flowing cooling medium; the cooling structure is a metal shell or clamping structure enclosing the optical guide tube; the supply unit stores the cooling medium and drives its circulation within the cooling pipe.

[0085] Specifically, the detection device includes a light guide, a detector, a vibration detection module, a temperature detection module, a vision module, and a fault diagnosis system for the light guide for semiconductors.

[0086] A light guide, partially disposed within the reaction chamber, is used to conduct reflected light from the wafer surface. It should be understood that the aforementioned detection device also includes a light source for outputting an optical signal, and the light guide further guides the optical signal to the wafer surface and receives reflected light from the wafer surface, transmitting it to the detector. Typically, the aforementioned light source is a broadband infrared light source. Exemplarily, in this embodiment, the light guide is a sapphire light guide.

[0087] The detector, located outside the reaction chamber, is used to generate reflection spectrum data based on the reflected light.

[0088] A vibration detection module is located at the root of the optical guide tube and is used to acquire vibration data of the optical guide tube. For example, in this embodiment, the vibration detection module employs a high-sensitivity MEMS accelerometer for real-time, high-frequency monitoring of the mechanical impact and abnormal vibration conditions of the optical guide tube. The vibration data includes the peak acceleration, spectrum, and duration of the impact event.

[0089] A temperature detection module is positioned at a preset location within the reaction chamber to acquire temperature data at that location. The preset location includes the inlet and outlet positions of the cooling pipe, as well as the body position of the optical guide tube. For example, in this embodiment, the temperature detection module employs a distributed temperature sensor array, with each temperature sensor being a PT100 or a thermocouple. By distributively embedding these high-precision temperature sensors at the preset locations, a three-dimensional temperature field model can be constructed to accurately monitor cooling efficiency and body temperature distribution.

[0090] An imaging module, located outside the reaction chamber, is used to acquire image information of the protective mirror surface of the observation window at the front end of the light guide tube. For example, the detection device also includes a beam splitter, through which a small portion (approximately 2-5%) of the infrared light path transmitted through the light guide tube is guided to a high-temperature resistant miniature CMOS or CCD image sensor, thereby acquiring a high-definition digital image of the protective mirror surface of the observation window at the front end of the light guide tube in real time, and directly monitoring the deposition status of contaminants.

[0091] A fault diagnosis system for semiconductor optical guides is used to acquire and analyze first and second detection data to obtain first and second fault information. Based on the first and second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated, enabling engineers to maintain the optical guide according to the maintenance strategy. Specifically, the fault diagnosis system acquires first and second detection data; wherein, the first detection data includes image information of the protective mirror surface of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset location within the reaction chamber; the preset location includes the inlet position, outlet position of the cooling pipe, and the body position of the optical guide; the first and second detection data are analyzed respectively to obtain first and second fault information; wherein, the first fault information includes a contamination index, and the second fault information includes at least one of the following: rate of change of reflected light intensity, rate of change of vibration, and cooling efficiency parameter; based on the first and second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated, enabling engineers to maintain the optical guide according to the maintenance strategy.

[0092] Furthermore, a sleeve is provided at the front end of the optical guide tube; wherein the sleeve employs a mechanical quick-connect mechanism, allowing the sleeve to detach from the front end of the optical guide tube when subjected to an axial or lateral impact force exceeding a design threshold. Exemplarily, the sleeve and the connecting mechanism constitute a sleeve assembly, which is made of a high-temperature resistant alloy such as Inconel, and the connecting mechanism can be either a snap-fit ​​type or a threaded type with a locking ring. Further, the connecting mechanism contains a mechanically weak point with a preset fracture strength, such as a shear pin of a specific size. Using the above solution, when the sleeve is subjected to an axial or lateral impact force exceeding the design threshold, the weak point will break preferentially, safely separating the sleeve assembly from the probe body, thereby preventing force transmission to the expensive optical guide tube.

[0093] Optionally, the optical guide tube integrates an in-situ detection unit for detecting the in-situ status of the sleeve and transmitting a trigger signal to the fault diagnosis system of the semiconductor optical guide tube when the sleeve detaches, thereby triggering the alarm mode of the fault diagnosis system. For example, the in-situ detection unit can be a microswitch or an optical distance sensor. Using the above solution, this application immediately triggers an alarm mode after the sleeve is impacted and detached, reminding engineers to handle the situation promptly and avoiding further damage to the expensive optical guide tube.

[0094] Optionally, the connection point between the optical guide and the sleeve is equipped with a non-contact identification interface. This interface is used to acquire the sleeve's identification information when it is connected to the front end of the optical guide and transmit it to the fault diagnosis system for semiconductor optical guides. This allows the fault diagnosis system to create or update the corresponding file information for the sleeve based on the identification information. The file information includes the number of uses and vibration data. For example, in this embodiment, the non-contact identification interface uses an RFID reader. Each sleeve assembly is assigned a unique identification information during manufacturing. After the sleeve is installed at the front end of the optical guide, the RFID reader automatically acquires this identification information and transmits it to the fault diagnosis system for semiconductor optical guides. If the sleeve is being installed for the first time, the fault diagnosis system automatically identifies and registers the identification information, creates a file information corresponding to this identification information, and begins to accumulate and record the number of uses and vibration data during each use, thus achieving digital lifecycle management of consumables. Furthermore, after the sleeve has reached a preset number of uses, it is discarded to ensure safe use.

[0095] The aforementioned detection device acquires reflectance spectrum data through a light guide and detector, vibration data through a vibration detection module, temperature data at a preset location through a temperature detection module, and image information of the protective mirror of the observation window at the front end of the light guide through a vision module. A fault diagnosis system for semiconductor light guides then performs joint analysis on the above data, converting the multi-source fault characteristics into first and second fault information, and generating corresponding fault diagnosis results and maintenance strategies. This allows engineers to distinguish the emergency states of different faults, avoiding misjudgments caused by a single temperature or alarm signal. This not only allows for the early detection of hidden risks such as light guide contamination, mechanical loosening, and decreased cooling efficiency, reducing the probability of unplanned downtime due to light guide breakage and overheating failure, but also reduces excessive maintenance and blind periodic maintenance, enabling predictive maintenance and refined management of the light guide, and improving the reliability and availability of the detection device during long-term operation within the reaction chamber of semiconductor equipment.

[0096] Please see Figure 2 In one exemplary embodiment, this application provides a fault diagnosis method for a semiconductor optical guide, applicable to a detection device, wherein the detection device is partially disposed within the reaction chamber of a semiconductor device, and a cooling pipe is disposed within the reaction chamber; the fault diagnosis method for the semiconductor optical guide specifically includes the following steps:

[0097] Step 202: Obtain first detection data and second detection data. The first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide tube, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide tube, and temperature data at a preset position within the reaction chamber; the preset position includes the inlet position and outlet position of the cooling pipe, and the body position of the optical guide tube.

[0098] Specifically, considering that byproducts generated during the process easily deposit on the protective lens (viewport) at the end face of the optical guide, gradually forming a contamination film, which significantly attenuates the intensity of the infrared signal passing through the protective lens, resulting in temperature measurement results that are significantly lower than the actual temperature value of the wafer, and that this type of contamination has a high probability of occurring under long-term operating conditions, this embodiment uses the image information of the protective lens surface of the optical guide front end viewing window as the first detection data to intuitively and quantitatively characterize the contamination state of the protective lens surface.

[0099] Furthermore, at least one of the following is used as a second detection data to assist in the judgment: the reflection spectrum data of the wafer surface, the vibration data of the optical guide tube, and the temperature data of the preset position in the reaction chamber. This allows for a multi-dimensional comprehensive evaluation of the changes in the transmission characteristics of the optical link, the dynamic changes in the mechanical stress and cooling state of the optical guide tube, thereby obtaining a more comprehensive health status and providing a reliable data foundation for the subsequent fault diagnosis and maintenance strategy generation of the optical guide tube.

[0100] Step 204: Analyze the first detection data and the second detection data respectively to obtain the first fault information and the second fault information; wherein, the first fault information includes the pollution index, and the second fault information includes at least one of the following: the rate of change of reflected light intensity, the rate of change of vibration, and the cooling efficiency parameter.

[0101] Specifically, the contamination index characterizes the degree of contamination of the protective mirror at the front end of the optical guide tube by process byproduct deposition. By processing the image information of the protective mirror, the originally subjective state of being dirty or clean is quantified into a continuous numerical index. Generally speaking, the higher the contamination index, the thicker or wider the film deposited on the protective mirror, the lower the optical channel transmittance, and the more severe the attenuation of infrared reflected signals, thus indirectly reflecting the optical health status of the temperature measurement link.

[0102] The rate of change of reflected light intensity is used to characterize the dynamic change trend of the reflection spectrum on the wafer surface during the process, especially the rate of increase or decrease of the reflected signal over time or process cycle number. This rate of change of reflected light intensity can reflect the change in the total transmittance of the optical link and the impact of changes in the process surface condition on the reflected signal, and is used to characterize whether there are problems such as abnormal attenuation, increased contamination, or abnormal process conditions in the optical channel.

[0103] The vibration variation rate is used to characterize the trend of mechanical stability changes in the optical guide and sleeve assembly during operation. Generally speaking, if the vibration variation rate continues to increase, it indicates that the optical guide support structure, mounting flange, or surrounding components may be loose, resonating, or experiencing increased impact. Therefore, this vibration variation rate is used to characterize the mechanical health status of the optical guide and the potential risk of loosening.

[0104] Cooling performance parameters are used to characterize whether the cooling system is operating normally and efficiently, and whether there is a gradual decline in cooling capacity or a trend of localized overheating, providing quantitative criteria for assessing the thermal safety and lifespan risks of optical guide tubes. Specific cooling performance parameters include: effective temperature difference, heat dissipation, and equivalent thermal resistance. Effective temperature difference characterizes the temperature difference between the inlet and outlet of the cooling medium, or between the cooling pipe and the optical guide tube body, reflecting whether the driving force of the cooling circuit for heat removal is sufficient. Heat dissipation is the amount of heat removed per unit time based on information such as the flow rate and temperature rise of the cooling medium, characterizing the actual heat dissipation capacity of the current cooling system. Equivalent thermal resistance indicates the overall ease of heat dissipation from the heated part of the optical guide tube to the cooling medium; an increase in equivalent thermal resistance indicates a deterioration in the heat dissipation path or a decrease in cooling efficiency.

[0105] Step 206: Based on the first fault information and the second fault information, generate a fault diagnosis result and a maintenance strategy based on the fault diagnosis result, so that the engineer can maintain the optical guide according to the maintenance strategy.

[0106] Specifically, the fault diagnosis results are used to characterize the current health status and potential risk types of the optical guide tube and its surrounding cooling and mounting structures. For example, the fault diagnosis results can provide the source and severity of the fault, and then automatically generate a maintenance strategy based on the fault diagnosis results according to a preset processing strategy.

[0107] The maintenance strategy is a targeted operation and maintenance recommendation generated based on fault diagnosis results, used to guide engineers to take corresponding maintenance actions. By establishing a mapping relationship between diagnostic conclusions and maintenance actions, engineers can make decisions based on quantitative indicators without relying on experience, thereby enabling graded intervention of optical guides, avoiding over-maintenance or maintenance delays, and improving the reliability and availability of equipment operation.

[0108] The aforementioned fault diagnosis method for semiconductor optical guides acquires first detection data and second detection data. The first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset location within the reaction chamber. The preset location includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide. By jointly analyzing the first and second detection data, this application converts the aforementioned multi-source fault characteristics into first and second fault information, and generates corresponding fault diagnosis results and maintenance strategies accordingly. This enables engineers to distinguish the emergency states of different faults and avoid misjudgments caused by a single temperature or alarm signal. This not only allows for the early detection of hidden risks such as optical guide contamination, mechanical loosening, and decreased cooling efficiency, reducing the probability of unplanned downtime due to optical guide breakage and overheating failure, but also reduces excessive maintenance and blind periodic maintenance, achieving predictive maintenance and refined management of the optical guide, and improving the reliability and availability of the detection device during long-term operation within the reaction chamber of semiconductor equipment.

[0109] Please see Figure 3 Optionally, the first detection data is parsed to obtain the first fault information, including:

[0110] Step 302: Extract image feature data from the image information. The image feature data includes the average gray value, the standard deviation of the gray value distribution, and the image gradient intensity.

[0111] Step 304: Calculate the rate of change of image feature data based on the reference value to obtain multiple feature change rates; wherein, the reference value is determined based on the standard image of the protective mirror.

[0112] Step 306: Weighted summation of the change rates of each characteristic to obtain the pollution index.

[0113] Specifically, the image module uses a micro-image CMOS sensor, which acquires an end-face grayscale image I. Image I is preprocessed to improve its signal-to-noise ratio. The preprocessing steps include Gaussian filtering for noise reduction and contrast enhancement.

[0114] Furthermore, the image feature data includes the average gray value, the standard deviation of the gray value distribution, and the image gradient intensity. Contamination deposits lead to a decrease in the average gray value, while clean lenses have high light transmittance and a high overall image gray value. Particulate or non-uniformly deposited contaminants coarsen the image texture, increasing the dispersion of the gray value distribution. Contaminants blur the optical properties of the mirror, causing a decrease in sharp edge features and a reduction in the overall gradient value. The image gradient intensity can be obtained by calculating the sum of the Sobel or Canny edge intensities of the image.

[0115] Furthermore, multiple standard images are acquired in the initial clean state of the probe or after each maintenance. The baseline values ​​of the aforementioned image feature data are calculated, and the rate of change of the image feature data is calculated based on these baseline values. This yields the characteristic rate of change for the average gray value, the standard deviation of the gray value distribution, and the image gradient intensity, expressed as follows:

[0116]

[0117] in, The current average gray value, This serves as the baseline value for the average grayscale value. The standard deviation of the current grayscale distribution, This is the baseline value for the standard deviation of the grayscale distribution; The standard deviation of the current grayscale distribution, This is the baseline value for the standard deviation of the grayscale distribution.

[0118] Using a weighted model or a trained lightweight machine learning model, such as a support vector machine regression model, the rates of change of the above multiple features are combined into a pollution index (CI) ranging from 0 to 100%, expressed as follows: Where ω1, ω2, and ω3 are the weights of the rate of change of each characteristic.

[0119] The above scheme analyzes the protective mirror image information, first extracting multi-dimensional image feature data such as average gray value, gray distribution standard deviation, and image gradient intensity. Then, the rate of change of each feature relative to the benchmark value is weighted and fused to construct a contamination index, achieving a quantitative characterization of the contamination status of the protective mirror. Compared with methods relying solely on a single gray threshold or manual visual judgment, this scheme comprehensively considers changes in multiple dimensions such as overall brightness, brightness uniformity, and edge and texture details, more sensitively reflecting complex situations such as thin film deposition, localized stains, and in-plane non-uniform contamination, significantly improving the robustness and resolution of contamination detection. The single numerical output of the contamination index also facilitates the setting of graded thresholds in the host computer, enabling continuous monitoring and trend analysis of the contamination level of the optical channel. This provides a reliable quantitative basis for subsequent fault diagnosis and preventive maintenance, reducing the risk of over-cleaning or delayed maintenance.

[0120] Optionally, if the second detection data includes reflectance spectral data of the wafer surface, the second detection data is analyzed to obtain second fault information, including:

[0121] For the reflectance spectral data collected at different times, reflectance spectral data within a preset wavelength range are selected and integrated to obtain the reflectance intensity index corresponding to each collection time. The reflectance intensity indexes are arranged in chronological order according to the collection time to form a reflectance intensity time series. Based on the reflectance intensity time series, a fitting algorithm is used to obtain the slope of the reflectance intensity change with time, and this slope is taken as the rate of change of reflectance intensity.

[0122] Specifically, the reflectance spectral data is acquired during the process at a preset sampling period, and includes reflected light intensity values ​​across multiple wavelength channels. For the reflectance spectral data acquired at the i-th acquisition moment, the corresponding light intensity data within a preset wavelength range is selected and integrated to obtain the reflected light intensity index. The integration calculation can be a summation of discrete wavelength channels or a numerical integration calculation. The preset wavelength range refers to one or more pre-selected wavelength intervals within the entire measurement band of the reflectance spectrum, used to extract spectral information from the reflectance spectral data that can stably characterize changes in the reflectance intensity of the wafer surface. This wavelength range can be determined based on the spectral response range of the temperature measuring equipment, the emission spectrum distribution of the target light source, and the transmittance characteristics of the reaction cavity window and the light guide material.

[0123] Furthermore, based on the time series of reflected light intensity, a fitting algorithm is used to fit the relationship between reflected light intensity and time, and the slope in the fitting result is obtained. This slope is used as the rate of change of reflected light intensity to characterize the increasing or decreasing trend of reflected light intensity over time. Specifically, if the rate of change of reflected light intensity is negative and its absolute value is greater than a certain value, the reflected light spectral signal is considered to have a continuous decreasing trend.

[0124] By employing the above scheme, integrating the reflectance spectral data within a preset wavelength range, multi-channel reflectance information can be fused into a single reflectance intensity index. This effectively suppresses the influence of noise fluctuations, local spectral anomalies, or transient interference in individual wavelength channels on the judgment results, thereby improving the stability and robustness of the reflectance signal characterization. Furthermore, by constructing a time-series sequence of reflectance intensity and using a fitting algorithm to extract its slope over time as the rate of change of reflectance intensity, the long-term trend of the reflectance signal can be quantified. This allows the present application to identify reflectance changes caused by optical link attenuation, increased window contamination, or abnormal process surface conditions before significant drift in temperature readings, achieving early warning of risks.

[0125] Optionally, if the second detection data includes vibration data of the light guide tube, the second detection data is analyzed to obtain second fault information, including:

[0126] Frequency domain analysis is performed on the vibration data collected within a preset time window to obtain the vibration spectrum corresponding to each monitoring time point. At least one target frequency is selected in the vibration spectrum, and the spectral amplitude at the target frequency is extracted to obtain a spectral amplitude sequence that changes with time. The target frequency is related to the loosening of the optical guide tube. Based on the spectral amplitude sequence, the rate of change of the spectral amplitude is calculated, and the rate of change is used as the vibration change rate of the optical guide tube.

[0127] Specifically, the vibration data is a time-series vibration sequence continuously collected at a preset sampling frequency within a preset time window. At each monitoring time point, frequency domain analysis is performed on the time-domain vibration sequence within the preset time window to obtain the vibration spectrum at that monitoring time point. The frequency domain analysis can employ Fast Fourier Transform or equivalent spectrum estimation methods.

[0128] The target frequency refers to the frequency associated with the loosening of the optical guide tube. For example, the target frequency can be the natural frequency of the optical guide tube in a tightened state, or the resonant frequency that appears after the fasteners loosen. The spectral amplitude at the target frequency is extracted to obtain the amplitude corresponding to each monitoring time point, and a spectral amplitude sequence is formed in chronological order. The rate of change of the amplitude over time is calculated based on the spectral amplitude sequence. This rate of change can be obtained by calculating the difference between adjacent monitoring time points and / or by obtaining the slope of the sequence through a fitting algorithm. Wherein, a vibration change rate exceeding a certain value indicates that the optical guide tube has a loosening trend or a risk of decreased structural stability.

[0129] By employing the above scheme, frequency domain analysis of vibration data and selection of target frequencies related to loosening within the vibration spectrum can separate features such as enhanced structural resonance caused by loosening from complex broadband vibration noise. This avoids misjudgments caused by relying solely on time-domain amplitude, improving the targeting and anti-interference capabilities of loosening detection. Furthermore, by extracting the spectral amplitude at the target frequency and constructing an amplitude sequence that changes over time, and then calculating the rate of change of the amplitude sequence, the question of whether vibration is increasing can be transformed from qualitative observation into a quantitative indicator of vibration change rate, enabling trend identification and early warning of loosening risk. This allows for timely maintenance alerts before significant mechanical displacement or collision damage occurs in the optical guide tube, reducing unplanned downtime and probe damage probability, and providing an objective basis for hierarchical decision-making in maintenance strategies. This is beneficial for achieving predictive maintenance-oriented optical guide tube health management.

[0130] Optionally, the cooling performance parameters include effective temperature difference, heat dissipation, and equivalent thermal resistance; when the second detection data includes temperature data at a preset location within the reaction chamber, the second detection data is analyzed to obtain second fault information, including:

[0131] Based on the temperature data at the outlet and inlet of the cooling pipe, calculate the effective temperature difference; calculate the product of the medium flow rate, effective temperature difference, and specific heat capacity of the medium to obtain the heat dissipation; calculate the temperature difference between the body of the light guide and the inlet of the cooling pipe, and calculate the ratio of temperature difference to heat dissipation to obtain the equivalent thermal resistance.

[0132] Specifically, the temperature data Tout at the outlet of the cooling pipe and the temperature data Tin at the inlet are obtained, and the difference between the two is calculated to obtain the effective temperature difference, which is expressed as: △T=Tout-Tin.

[0133] Furthermore, the expression for calculating heat dissipation is: ;in, denoted as , where is the flow rate of the cooling pipe; c is the specific heat capacity of the medium.

[0134] Establish and update the equivalent thermal resistance model of the probe in real time: Rth=(Tbody-Tin) / Q, where the slow increase of the equivalent thermal resistance Rth is a direct indicator of the degradation of cooling efficiency.

[0135] Optionally, the second fault information also includes the rate of change of cooling performance parameters; parsing the second detection data to obtain the second fault information further includes:

[0136] The cooling performance parameters were processed by moving average and trend fitting to obtain their rate of change.

[0137] Specifically, the cooling performance parameters are processed by moving average to suppress parameter fluctuations caused by plasma pulsation, transient disturbances, and measurement noise; the length of the moving window can be preset according to the sampling period and process cycle.

[0138] The cooling performance parameters after moving average are subjected to trend fitting to obtain the slope of the cooling performance parameters changing with time, and the slope is output as the rate of change of the cooling performance parameters. When the rate of change of the cooling performance parameters exceeds a certain value, it is considered that the cooling capacity has a continuous deterioration trend or the risk of local overheating is increased.

[0139] The above scheme calculates the effective temperature difference using temperature data from the cooling pipe inlet and outlet, and multiplies the cooling medium flow rate, effective temperature difference, and specific heat capacity of the medium to obtain the heat dissipation. Furthermore, it constructs an equivalent thermal resistance based on the ratio of the temperature difference between the optical guide tube body and the cooling pipe inlet to the heat dissipation, thus achieving a multi-dimensional quantitative characterization of the cooling system. Specifically, the effective temperature difference reflects the overall strength of heat absorption by the cooling circuit, the heat dissipation characterizes the amount of heat actually removed per unit time, and the equivalent thermal resistance comprehensively reflects the heat conduction or heat exchange efficiency between the optical guide tube and the cooling medium. This provides a more differentiated diagnostic basis for different failure modes, such as decreased flow rate, water circuit blockage, scaling leading to deteriorated heat exchange, and localized overheating.

[0140] Furthermore, applying a moving average to the aforementioned cooling performance parameters can effectively suppress short-term fluctuations caused by plasma pulsation, process cycle changes, and temperature measurement noise. By combining trend fitting to extract the rate of change, the cooling capacity can be converted into a quantifiable indicator. This allows the present application to identify risks and issue early warnings in the early stages of a gradual decline in cooling performance, thereby providing a clear time window for maintenance strategies, reducing the probability of overheating damage to the optical guide tube and unplanned downtime, and improving the reliability and maintainability of equipment operation.

[0141] Optionally, the second fault information includes the rate of change of reflected light intensity, the rate of change of vibration, and cooling efficiency parameters, wherein the cooling efficiency parameters include equivalent thermal resistance and its rate of change; based on the first and second fault information, a fault diagnosis result is generated, including:

[0142] If the increase in the pollution index exceeds the first preset threshold, and if the absolute value of the change rate of reflected light intensity exceeds the preset attenuation threshold within a preset time, and the change rate of vibration within a preset time is within the first preset range, and the change rate of equivalent thermal resistance within a preset time is within the second preset range, then the fault diagnosis result will be marked as optical pollution as the main pollution source.

[0143] Specifically, during wafer fabrication, if the increase in the contamination index exceeds a first preset threshold, optical contamination is considered to exist, requiring further identification of the contamination source and impact assessment. If the absolute value of the rate of change of reflected light intensity exceeds a preset attenuation threshold, the reflected spectral signal is considered to exhibit synchronous linear attenuation, and at this point, the vibration is stable and the equivalent thermal resistance is stable. The fault diagnosis result is then marked as optical contamination as the primary contamination source. It should be noted that if the reflected light signal exhibits attenuation, the rate of change of reflected light intensity is negative, and the smaller the negative value, the faster the attenuation rate.

[0144] Optionally, a maintenance strategy is generated based on the fault diagnosis results, including:

[0145] Based on the increase in the contamination index, predict the number of remaining process cycles required to reach the cleaning threshold, and after reaching the remaining number of process cycles, output a maintenance command to perform dust removal operation on the protective mirror during the process interval.

[0146] Optionally, based on the increase in the contamination index, the remaining number of process cycles required to reach the cleaning threshold is predicted, including:

[0147] Historical contamination indexes and corresponding process cycle numbers are obtained to construct a historical data sequence characterizing the increase of the contamination index with the number of process cycles. Based on the historical data sequence, a pre-defined trend fitting algorithm is used to establish a functional relationship between the contamination index and the number of process cycles. The target contamination index corresponding to the cleaning threshold determined in advance through experiments is substituted into the functional relationship to solve for the predicted total number of process cycles when the cleaning threshold is reached. The difference between the predicted total number of process cycles and the number of process cycles currently completed is calculated, and this difference is used as the remaining number of process cycles required to reach the cleaning threshold.

[0148] Specifically, when optical contamination is the primary source of pollution, and this contamination is considered to have a certain impact on the production process, dust removal operations can be performed on the protective mirror during process breaks to ensure the long-term stability of the temperature measurement link. For example, the dust removal operation can be automatically performed by an air blowing system. Upon receiving a dust removal command, the air blowing system controls the air source valve to open and supplies air to a preset nozzle to blow clean the protective mirror and / or the end of the optical guide tube. Optionally, the dust removal operation can also be manually performed by an engineer according to a maintenance strategy. For example, the engineer can blow clean or clean the protective mirror and / or the end of the optical guide tube using an external air gun or maintenance interface; this application does not limit this approach.

[0149] Furthermore, different levels can be set based on the specific value of the increase in the pollution index. For example, the value corresponding to the normal level is the first preset threshold. When the first preset threshold is reached, the maintenance strategy is to perform dust removal operation during the process interval. If the value reaches the threshold corresponding to the medium level, dust removal operation will be performed immediately after the current cycle ends. If the value reaches the threshold corresponding to the high level, an alarm mode should be triggered immediately to remind engineers to follow up and handle the situation in a timely manner.

[0150] Optionally, in the presence of optical contamination, this application further detects the body temperature of the light guide tube. In this case, based on the first fault information and the second fault information, a fault diagnosis result is generated, which further includes:

[0151] If the temperature rise of the light guide tube exceeds the fourth preset threshold, the deviation correction amount caused by the current body position temperature to the reflected light intensity is further calculated, and the fault diagnosis result is marked as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the light guide tube.

[0152] Optionally, the deviation correction amount caused by the current temperature to the intensity of reflected light is calculated, including:

[0153] A contaminant thermal resistance model was established, and the deviation correction amount caused by the contaminant layer to the temperature measurement accuracy of the light guide was quantitatively calculated based on the contamination index and the temperature rise of the light guide body.

[0154] Specifically, based on the pollution index and historical sediment composition data, the pollutant characteristic mapping table is queried to obtain the equivalent thermal conductivity keff and optical attenuation coefficient μ of the current pollutant layer.

[0155] A one-dimensional composite layer thermal model of the front end of the light guide is constructed, which includes a base lens layer and a contaminant layer. Using the equivalent thermal conductivity keff, the temperature rise of the light guide body, and the heat flow boundary conditions estimated from the process conditions in the reaction chamber, the current equivalent thickness deff of the contaminant layer is backfitted.

[0156] Based on the above composite layer thermal model, the temperature deviation ΔTbias at the measurement point caused by the additional thermal resistance Rp (=deff / keff) of the contaminant layer is calculated.

[0157] The combined temperature deviation ΔTbias and optical attenuation coefficient μ are used to synthesize the overall deviation correction amount ΔIcorrectio for the original reflectance spectral data using the following formula: ΔIcorrectio = γ1×ΔTbias + γ2×μ×Iraw; where Iraw is the measured original reflected light intensity, and γ1 and γ2 are weighting coefficients determined through pre-calibration.

[0158] In this case, the maintenance strategy generated based on the fault diagnosis results also includes:

[0159] The measured wafer surface temperature is compensated in real time based on the deviation correction amount to obtain the calibrated wafer surface temperature value; the calibrated wafer surface temperature value is then marked and output.

[0160] Specifically, the calibrated measurement values ​​are provided to engineers for reference. For example, engineers can replace the original measurement values ​​with the calibrated values ​​and mark the replaced values ​​as calibrated to improve the accuracy of temperature measurements.

[0161] Optionally, if the second fault information includes the vibration change rate, generating a maintenance strategy based on the fault diagnosis results further includes:

[0162] If the rate of change of vibration exceeds the second preset threshold within a preset time, the fault diagnosis result will be marked as a loose mechanical connection fault of the optical guide tube.

[0163] In this case, the maintenance strategy generated based on the fault diagnosis results also includes:

[0164] After the current process cycle is completed, a tightening command is output so that engineers can reinforce the light guide.

[0165] Specifically, vibration data of the optical guide tube is continuously collected during wafer fabrication, and the vibration change rate is output according to a preset sampling period. This application uses a continuous preset time period as a monitoring window, performs statistical analysis on the vibration change rate within this monitoring window, and calculates the increment or slope of the vibration change rate within the preset time period. When the vibration change rate is detected to exceed a second preset threshold within the preset time period, it is determined that the vibration state of the optical guide tube has a significant upward trend, and a fault diagnosis result is marked.

[0166] After obtaining the above fault diagnosis results, this application further generates a maintenance strategy that matches the fault type. Without affecting the stability of the current process, the maintenance execution time is determined to be at the end node of the current process cycle. When the current process cycle ends, a tightening command is output to the host computer or human-machine interface to prompt the engineer to strengthen the mechanical connection parts of the optical guide tube to avoid measurement inaccuracies or collision risks caused by increased loosening.

[0167] Optionally, if the first fault information also includes the standard deviation of the grayscale distribution corresponding to the image information, and the second fault information includes the rate of change of reflected light intensity, a fault diagnosis result is generated based on the first fault information and the second fault information, including:

[0168] If the increase in the pollution index exceeds the first preset threshold, and if the change rate of reflected light intensity exceeds the preset attenuation range within a preset time, and the characteristic change rate of the grayscale distribution standard deviation exceeds the third preset threshold, then the fault diagnosis result will be marked as a pollution source with non-uniform deposition, which is likely to cause deviation in the temperature data at the preset location; wherein, the characteristic change rate of the grayscale distribution standard deviation is obtained from the benchmark value determined based on the standard image of the protective mirror.

[0169] Then, a maintenance strategy is generated based on the fault diagnosis results, including:

[0170] Based on the rising slope of the equivalent thermal resistance, the required maintenance time window for the cooling system is predicted. Once the maintenance time window is reached, the processing is stopped, and a cleaning plan is output to remove dust from the protective mirror.

[0171] Optionally, the required maintenance time window for the cooling system can be predicted based on the rising slope of the equivalent thermal resistance, including:

[0172] The equivalent thermal resistance of the cooling tube during process operation is continuously monitored to obtain monitoring results. Based on the monitoring results, the unit growth rate of the equivalent thermal resistance relative to the process time and / or the number of process cycles is calculated as the rising slope of the equivalent thermal resistance. Based on the current equivalent thermal resistance value, the rising slope, and the critical thermal resistance value, the remaining time and / or the remaining number of process cycles required to grow from the current state to the critical thermal resistance value are calculated, and the remaining time and / or the remaining number of process cycles are determined as the maintenance time window of the cooling tube. Among them, the critical thermal resistance value is determined based on the safe operation requirements of the cooling tube.

[0173] Using the above scheme, this application models and predicts the cooling performance degradation process based on the rising slope of the equivalent thermal resistance to determine the optimal maintenance intervention time; and when the predicted maintenance time arrives, it will automatically arrange process interruption and simultaneously output a mirror cleaning scheme aimed at correcting the heat conduction interface to curb the increase in thermal resistance and temperature measurement deviation caused by non-uniform deposition.

[0174] Furthermore, the maintenance strategy of this embodiment can also incorporate the above-mentioned scheme of adjusting the measured value of the wafer surface temperature according to the deviation correction amount.

[0175] Optionally, the above-mentioned fault diagnosis method for semiconductor optical guides further includes:

[0176] The first fault information and the second fault information are weighted and summed to obtain a health index for engineers' reference; the weights of the first fault information and the second fault information are dynamically adjusted based on historical fault diagnosis results.

[0177] Optionally, the above-mentioned fault diagnosis method for semiconductor optical guides further includes:

[0178] Generate and output a status report for engineers' reference. The status report includes health index, fault diagnosis results, and recommended maintenance strategies. The health index provides an overall status overview, while the fault diagnosis results and maintenance strategies provide specific fault location and operation guidance.

[0179] Furthermore, in addition to the diagnostic conclusions mentioned above, the fault diagnosis results can also provide fault levels or risk levels, such as "normal," "warning," and "severe," to distinguish different degrees of severity in operating conditions. By comprehensively judging fault information such as contamination index, rate of change of reflected light intensity, rate of change of vibration, and cooling efficiency parameters, the fault diagnosis results can reflect whether the optical guide is currently in a safe operating range, whether there is a trend of deterioration, and potential failure modes.

[0180] For example, the above maintenance strategies include: generating a planned maintenance recommendation to "perform protective mirror cleaning or cooling circuit inspection during the next planned shutdown window" when the diagnostic result is mild contamination or a slight decrease in cooling efficiency; generating a structural maintenance recommendation to "prioritize inspection of the light guide support structure and fasteners" when the diagnostic result is a significant increase in vibration variation rate or a risk of loosening; and generating an emergency maintenance recommendation to "immediately stop the relevant process, prioritize replacement of the light guide or troubleshooting the cooling system" when the diagnostic result is severe contamination, cooling failure, or a high-risk condition.

[0181] It should be noted that the order of maintenance strategies given in the above embodiments is only an example. In practical applications, different fault priorities can be set according to the processing environment, and different combinations of maintenance strategies and / or the order of processing can be set according to the priorities.

[0182] Furthermore, the values ​​of the first preset threshold, the second preset threshold, the third preset threshold, and the fourth preset threshold, as well as the values ​​of the first preset range and the second preset range, can be set as needed, and no specific value is limited in this application.

[0183] The aforementioned fault diagnosis method for semiconductor optical guides acquires first detection data and second detection data. The first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset location within the reaction chamber. The preset location includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide. By jointly analyzing the first and second detection data, this application converts the aforementioned multi-source fault characteristics into first and second fault information, and generates corresponding fault diagnosis results and maintenance strategies accordingly. This enables engineers to distinguish the emergency states of different faults and avoid misjudgments caused by a single temperature or alarm signal. This not only allows for the early detection of hidden risks such as optical guide contamination, mechanical loosening, and decreased cooling efficiency, reducing the probability of unplanned downtime due to optical guide breakage and overheating failure, but also reduces excessive maintenance and blind periodic maintenance, achieving predictive maintenance and refined management of the optical guide, and improving the reliability and availability of the detection device during long-term operation within the reaction chamber of semiconductor equipment.

[0184] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0185] Based on the same inventive concept, this application also provides a fault diagnosis system for semiconductor optical guides. This system is applicable to the fault diagnosis method for semiconductor optical guides described above. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations in one or more system embodiments provided below can be found in the limitations of the method above, and will not be repeated here.

[0186] Please see Figure 4 In one embodiment, this application provides a fault diagnosis system for a semiconductor optical guide tube, wherein the system is disposed within a detection device, the detection device is partially disposed within a reaction chamber of a semiconductor device, and a cooling tube is disposed within the reaction chamber; the system includes: an acquisition module and a processing module.

[0187] The acquisition module is used to acquire first detection data and second detection data; wherein, the first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide tube, and the second detection data includes at least one of the reflection spectrum data of the wafer surface, vibration data of the optical guide tube, and temperature data at a preset position in the reaction chamber; the preset position includes the inlet position of the cooling pipe, the outlet position, and the body position of the optical guide tube.

[0188] The processing module is used to parse the first detection data and the second detection data respectively to obtain the first fault information and the second fault information; wherein the first fault information includes a pollution index, and the second fault information includes at least one of the reflected light intensity change rate, vibration change rate, and cooling efficiency parameter; based on the first fault information and the second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated, so as to maintain the light guide tube according to the maintenance strategy.

[0189] Optionally, the processing module parses the first detection data to obtain the first fault information, including: extracting image feature data from the image information, the image feature data including the average gray value, the standard deviation of gray distribution and the image gradient intensity; calculating the rate of change of the image feature data according to the benchmark value to obtain multiple feature change rates; wherein, the benchmark value is determined based on the standard image of the protective mirror; and weighting and summing the change rates of each feature to obtain the contamination index.

[0190] Optionally, if the second detection data includes reflectance spectral data of the wafer surface, the processing module parses the second detection data to obtain second fault information, including: selecting reflectance spectral data within a preset wavelength range and performing integral calculations for reflectance spectral data collected at different times to obtain the reflectance intensity index corresponding to each collection time; arranging each reflectance intensity index according to the time sequence of the collection time to form a reflectance intensity time sequence; and using a fitting algorithm to obtain the slope of the reflectance intensity change with time based on the reflectance intensity time sequence, and using the slope as the reflectance intensity change rate.

[0191] Optionally, when the second detection data includes vibration data of the optical guide tube, the processing module parses the second detection data to obtain second fault information, including: performing frequency domain analysis on the vibration data collected within a preset time window to obtain the vibration spectrum corresponding to each monitoring time point; selecting at least one target frequency in the vibration spectrum, extracting the spectral amplitude at the target frequency, and obtaining a spectral amplitude sequence that changes over time; wherein, the target frequency is related to the loosening of the optical guide tube; and calculating the rate of change of the spectral amplitude based on the spectral amplitude sequence, and using the rate of change as the vibration change rate of the optical guide tube.

[0192] Optionally, if the second detection data includes temperature data at a preset location within the reaction chamber, the processing module parses the second detection data to obtain second fault information, including: calculating the effective temperature difference based on the temperature data at the outlet and inlet locations of the cooling pipe; calculating the product of the medium flow rate, effective temperature difference, and specific heat capacity of the medium in the cooling pipe to obtain the heat dissipation; calculating the temperature difference between the body location of the optical guide tube and the inlet location of the cooling pipe, and calculating the ratio of the temperature difference to the heat dissipation to obtain the equivalent thermal resistance.

[0193] Optionally, the second fault information also includes the rate of change of cooling performance parameters; the processing module parses the second detection data to obtain the second fault information, and further includes: performing moving average and trend fitting processing on the cooling performance parameters to obtain their rate of change.

[0194] Optionally, the second fault information includes the rate of change of reflected light intensity, the rate of change of vibration, and cooling efficiency parameters, whereby the cooling efficiency parameters include equivalent thermal resistance and its rate of change. The processing module generates a fault diagnosis result based on the first and second fault information, including: if the increase in the contamination index exceeds a first preset threshold, and if the absolute value of the rate of change of reflected light intensity exceeds a preset attenuation threshold within a preset time, and the rate of change of vibration is within a first preset range and the rate of change of equivalent thermal resistance is within a second preset range within a preset time, then the fault diagnosis result is marked as optical contamination being the primary source of contamination. A maintenance strategy is generated based on the fault diagnosis result, including: predicting the remaining number of process cycles required to reach the cleaning threshold based on the increase in the contamination index, and outputting a maintenance command after reaching the remaining number of process cycles to perform dust removal operations on the protective mirror during process intervals. The process of predicting the remaining number of process cycles required to reach the cleaning threshold based on the increase in the contamination index includes: obtaining historical contamination indices and corresponding process cycle numbers to construct a historical data sequence characterizing the increase in the contamination index with the number of process cycles; establishing a functional relationship between the contamination index and the number of process cycles based on the historical data sequence using a preset trend fitting algorithm; substituting the target contamination index corresponding to the cleaning threshold determined in advance through experiments into the functional relationship to solve for the predicted total number of process cycles when the cleaning threshold is reached; and calculating the difference between the predicted total number of process cycles and the number of process cycles currently completed, using this difference as the remaining number of process cycles required to reach the cleaning threshold.

[0195] Optionally, in the presence of optical contamination, this application further detects the body temperature of the optical guide tube. In this case, the processing module generates a fault diagnosis result based on the first and second fault information, further including: if the increase in the body temperature of the optical guide tube exceeds a fourth preset threshold, then further calculating the deviation correction amount caused by the current body temperature on the reflected light intensity, and marking the fault diagnosis result as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the optical guide tube. Calculating the deviation correction amount caused by the current temperature on the reflected light intensity includes: establishing a contaminant thermal resistance model, and quantifying the deviation correction amount caused by the contaminant layer on the temperature measurement accuracy of the optical guide tube based on the contamination index and the increase in the body temperature of the optical guide tube. Then, generating a maintenance strategy based on the fault diagnosis result further includes: performing real-time compensation calculations on the measured wafer surface temperature based on the deviation correction amount to obtain a calibrated wafer surface temperature value; marking and outputting the calibrated wafer surface temperature value.

[0196] Optionally, if the second fault information includes the vibration change rate, the processing module generates a maintenance strategy based on the fault diagnosis result, and further includes: if the change in the vibration change rate exceeds the second preset threshold within a preset time, then the fault diagnosis result is marked as a loose mechanical connection fault of the optical guide tube; then the maintenance strategy generated based on the fault diagnosis result at this time also includes: after the current process cycle ends, outputting a tightening command so that the engineer can perform reinforcement treatment on the optical guide tube.

[0197] Optionally, if the first fault information also includes the grayscale distribution standard deviation corresponding to the image information, and the second fault information includes the rate of change of reflected light intensity, the processing module generates a fault diagnosis result based on the first and second fault information. This includes: if the increase in the contamination index exceeds a first preset threshold, and if the change in the rate of change of reflected light intensity exceeds a preset attenuation range within a preset time, and the characteristic rate of change of the grayscale distribution standard deviation exceeds a third preset threshold, then the fault diagnosis result is marked as a source of contamination that is non-uniformly deposited and prone to causing deviations in temperature data at preset locations; wherein, the characteristic rate of change of the grayscale distribution standard deviation is obtained from a benchmark value determined based on a standard image of the protective mirror. Then, a maintenance strategy is generated based on the fault diagnosis result, including: predicting the required maintenance time window for the cooling system based on the rising slope of the equivalent thermal resistance, and stopping the processing after the maintenance time window is reached, outputting a cleaning plan to perform dust removal operations on the protective mirror. The process of predicting the required maintenance time window for the cooling system based on the rising slope of the equivalent thermal resistance includes: continuously monitoring the change in the equivalent thermal resistance of the cooling tube with process time and / or process cycle number during process operation, and obtaining monitoring results; calculating the unit growth rate of the equivalent thermal resistance relative to process time and / or process cycle number based on the monitoring results, as the rising slope of the equivalent thermal resistance; calculating the remaining time and / or remaining process cycle number required to grow from the current state to the critical thermal resistance value based on the current equivalent thermal resistance value, the rising slope, and the critical thermal resistance value, and determining the remaining time and / or remaining process cycle number as the maintenance time window for the cooling tube; wherein, the critical thermal resistance value is determined based on the safe operation requirements of the cooling tube.

[0198] The aforementioned fault diagnosis system for semiconductor optical guides acquires first detection data and second detection data. The first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset location within the reaction chamber. The preset location includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide. By jointly analyzing the first and second detection data, this application converts the aforementioned multi-source fault characteristics into first and second fault information, and generates corresponding fault diagnosis results and maintenance strategies accordingly. This enables engineers to distinguish the emergency states of different faults and avoid misjudgments caused by a single temperature or alarm signal. This not only allows for the early detection of hidden risks such as optical guide contamination, mechanical loosening, and decreased cooling efficiency, reducing the probability of unplanned downtime due to optical guide breakage and overheating failure, but also reduces excessive maintenance and blind periodic maintenance, achieving predictive maintenance and refined management of the optical guide, and improving the reliability and availability of the detection device during long-term operation within the reaction chamber of semiconductor equipment.

[0199] Each module in the aforementioned fault diagnosis system for semiconductor optical guides can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0200] In one feasible embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the aforementioned fault diagnosis method for semiconductor optical guides. The display unit is used to form a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0201] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer equipment on which the present application is applied. Specific computer equipment may include, for example, [the following]. Figure 5 The diagram shows more or fewer components, or combinations of certain components, or different component arrangements.

[0202] In one feasible embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method steps in the above-described method for diagnosing faults in semiconductor optical guides.

[0203] In one feasible embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method steps in the above-described fault diagnosis method for semiconductor optical guides.

[0204] In one feasible embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the method steps in the above-described fault diagnosis method for semiconductor optical guides.

[0205] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0206] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fault diagnosis method for semiconductor optical guides, characterized in that, The method is applicable to a detection device, wherein the detection device is partially disposed within a reaction chamber of a semiconductor device, and a cooling pipe is disposed within the reaction chamber; the method includes: Acquire first detection data and second detection data; wherein, the first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the following: reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset position in the reaction chamber; the preset position includes the inlet position, outlet position of the cooling pipe, and the body position of the optical guide. The first detection data and the second detection data are analyzed respectively to obtain first fault information and second fault information; wherein, the first fault information includes a pollution index, and the second fault information includes at least one of the following: the rate of change of reflected light intensity, the rate of change of vibration, and the cooling efficiency parameter; based on the first fault information and the second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated so that engineers can maintain the light guide according to the maintenance strategy. When the second fault information includes cooling efficiency parameters, and the cooling efficiency parameters include the body temperature of the light guide tube, then generating a fault diagnosis result based on the first fault information and the second fault information includes: If the temperature rise of the main body of the light guide exceeds the fourth preset threshold, the deviation correction amount caused by the current main body temperature to the intensity of reflected light is further calculated, and the fault diagnosis result is marked as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the light guide. The calculation of the deviation correction amount caused by the current temperature to the intensity of reflected light includes: establishing a pollutant thermal resistance model, and quantifying the deviation correction amount caused by the pollutant layer to the temperature measurement accuracy of the light guide based on the pollution index and the rise value of the temperature at the position of the light guide body. The maintenance strategy is generated based on the fault diagnosis results, including: performing real-time compensation calculation on the measured value of the wafer surface temperature according to the deviation correction amount to obtain the calibrated wafer surface temperature value; and marking and outputting the calibrated wafer surface temperature value.

2. The method according to claim 1, characterized in that, The first fault information is obtained by parsing the first detection data, including: Image feature data is extracted from the image information, and the image feature data includes the average gray value, the standard deviation of gray distribution, and the image gradient intensity. The rate of change of the image feature data is calculated based on the reference value to obtain multiple feature change rates; wherein, the reference value is determined based on a standard image of the protective mirror. The pollution index is obtained by weighted summation of the change rates of each of the aforementioned characteristics.

3. The method according to claim 1, characterized in that, When the second detection data includes reflectance spectral data of the wafer surface, the second detection data is analyzed to obtain second fault information, including: For the reflectance spectral data collected at different times, reflectance spectral data within a preset wavelength range are selected and integrated to obtain the reflected light intensity index corresponding to each collection time. The reflected light intensity indicators are arranged in chronological order according to the acquisition time to form a time sequence of reflected light intensity. Based on the time series of reflected light intensity, a fitting algorithm is used to obtain the slope of the reflected light intensity changing with time, and this slope is taken as the rate of change of reflected light intensity.

4. The method according to claim 1, characterized in that, If the second detection data includes the vibration data of the light guide tube, the second detection data is analyzed to obtain second fault information, including: Frequency domain analysis is performed on the vibration data collected within a preset time window to obtain the vibration spectrum corresponding to the vibration data at each monitoring time point; At least one target frequency is selected from the vibration spectrum, and the spectral amplitude at the target frequency is extracted to obtain a spectral amplitude sequence that changes over time; the target frequency is related to the loosening of the optical guide tube. Based on the spectral amplitude sequence, the rate of change of the spectral amplitude is calculated, and the rate of change is used as the vibration rate of the optical guide tube.

5. The method according to claim 1, characterized in that, The cooling performance parameters include effective temperature difference, heat dissipation, and equivalent thermal resistance; If the second detection data includes temperature data at a preset location within the reaction chamber, the second detection data is parsed to obtain second fault information, including: The effective temperature difference is calculated based on the temperature data at the outlet and inlet positions of the cooling pipe. The heat dissipation is obtained by multiplying the medium flow rate of the cooling pipe, the effective temperature difference, and the specific heat capacity of the medium. Calculate the temperature difference between the body position of the light guide tube and the inlet position of the cooling pipe, and calculate the ratio of the temperature difference to the heat dissipation to obtain the equivalent thermal resistance.

6. The method according to claim 5, characterized in that, The second fault information also includes the rate of change of the cooling efficiency parameters; The process of parsing the second detection data to obtain the second fault information also includes: The cooling performance parameters are subjected to moving average and trend fitting processing to obtain their rate of change.

7. The method according to claim 1, characterized in that, The second fault information includes the rate of change of reflected light intensity, the rate of change of vibration, and cooling efficiency parameters, wherein the cooling efficiency parameters include the equivalent thermal resistance and its rate of change; The step of generating a fault diagnosis result based on the first fault information and the second fault information includes: If the increase in the pollution index exceeds a first preset threshold, and if the absolute value of the rate of change of the reflected light intensity exceeds a preset attenuation threshold within a preset time, and the rate of change of the vibration within a preset time is within a first preset range, and the rate of change of the equivalent thermal resistance within a preset time is within a second preset range, then the fault diagnosis result is marked as optical pollution as the main pollution source. Based on the fault diagnosis results, a maintenance strategy is generated, including: Based on the increase in the contamination index, the remaining number of process cycles required to reach the cleaning threshold is predicted, and after the remaining number of process cycles is reached, a maintenance command is output to perform dust removal operation on the protective mirror during the process interval.

8. The method according to claim 7, characterized in that, The step of predicting the number of remaining process cycles required to reach the cleaning threshold based on the increase in the contamination index includes: Obtain historical pollution index and corresponding process cycle number, and construct a historical data sequence to characterize the growth of pollution index with process cycle number; Based on the historical data sequence, a preset trend fitting algorithm is used to establish a functional relationship between the pollution index and the number of process cycles. Substitute the target contamination index corresponding to the cleaning threshold determined in advance through experiments into the functional relationship to obtain the predicted total number of process cycles when the cleaning threshold is reached. Calculate the difference between the predicted total number of process cycles and the number of process cycles currently completed, and use this difference as the remaining number of process cycles required to reach the cleaning threshold.

9. The method according to claim 1, characterized in that, The second fault information includes the vibration change rate. The step of generating a fault diagnosis result based on the first fault information and the second fault information further includes: If the change rate of vibration exceeds the second preset threshold within a preset time, the fault diagnosis result is marked as a loose mechanical connection fault of the optical guide tube. The maintenance strategy generated based on the fault diagnosis results also includes: After the current process cycle is completed, a tightening command is output so that the engineer can reinforce the optical guide.

10. The method according to claim 1, characterized in that, The first fault information also includes the standard deviation of the grayscale distribution corresponding to the image information; The second fault information includes the rate of change of reflected light intensity; The step of generating a fault diagnosis result based on the first fault information and the second fault information includes: If the increase in the pollution index exceeds a first preset threshold, and if the change rate of the reflected light intensity exceeds a preset attenuation range within a preset time, and the characteristic change rate of the grayscale distribution standard deviation exceeds a third preset threshold, then the fault diagnosis result is marked as a pollution source of non-uniform deposition that is prone to causing deviations in temperature data at preset locations; wherein, the characteristic change rate of the grayscale distribution standard deviation is obtained based on a reference value determined by the standard image of the protective mirror. Based on the fault diagnosis results, a maintenance strategy is generated, including: Based on the rising slope of the equivalent thermal resistance, the required maintenance time window for the cooling system is predicted. Once the maintenance time window is reached, the processing is stopped, and a cleaning plan is output to perform dust removal on the protective mirror.

11. The method according to claim 10, characterized in that, The method of predicting the required maintenance time window for the cooling system based on the rising slope of the equivalent thermal resistance includes: The equivalent thermal resistance of the cooling pipe during the process operation is continuously monitored as a function of process time and / or process cycle number, and the monitoring results are obtained. Based on the monitoring results, the unit growth rate of the equivalent thermal resistance relative to the process time and / or the number of process cycles is calculated as the rising slope of the equivalent thermal resistance. Based on the current equivalent thermal resistance, the rising slope, and the critical thermal resistance, calculate the remaining time and / or the number of remaining process cycles required to grow from the current state to the critical thermal resistance, and determine the remaining time and / or the number of remaining process cycles as the maintenance time window for the cooling tube; wherein, the critical thermal resistance is determined based on the safe operation requirements of the cooling tube.

12. A fault diagnosis system for a semiconductor optical guide tube, characterized in that, The system is disposed within a detection device, the detection device being partially disposed within a reaction chamber of a semiconductor device, and a cooling pipe being disposed within the reaction chamber; the system includes: The acquisition module is used to acquire first detection data and second detection data; wherein, the first detection data includes image information of the protective mirror of the observation window at the front end of the optical guide, and the second detection data includes at least one of the reflectance spectrum data of the wafer surface, vibration data of the optical guide, and temperature data at a preset position in the reaction chamber; the preset position includes the inlet position, the outlet position of the cooling pipe, and the body position of the optical guide. The processing module is used to parse the first detection data and the second detection data respectively to obtain first fault information and second fault information; wherein, the first fault information includes a contamination index, and the second fault information includes at least one of the reflected light intensity change rate, vibration change rate, and cooling efficiency parameter; based on the first fault information and the second fault information, a fault diagnosis result and a maintenance strategy generated based on the fault diagnosis result are generated so that engineers can maintain the light guide according to the maintenance strategy; When the second fault information includes cooling efficiency parameters, and the cooling efficiency parameters include the body temperature of the light guide tube, the processing module generates a fault diagnosis result based on the first fault information and the second fault information, including: If the temperature rise of the main body of the light guide exceeds the fourth preset threshold, the deviation correction amount caused by the current main body temperature to the intensity of reflected light is further calculated, and the fault diagnosis result is marked as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the light guide. The calculation of the deviation correction amount caused by the current temperature to the intensity of reflected light includes: establishing a pollutant thermal resistance model, and quantifying the deviation correction amount caused by the pollutant layer to the temperature measurement accuracy of the light guide based on the pollution index and the rise value of the temperature at the position of the light guide body. The processing module generates a maintenance strategy based on the fault diagnosis results, including: performing real-time compensation calculations on the measured values ​​of the wafer surface temperature according to the deviation correction amount to obtain the calibrated wafer surface temperature value; and marking and outputting the calibrated wafer surface temperature value.

13. A detection device, characterized in that, Partially disposed within the reaction chamber of a semiconductor device, the reaction chamber being provided with a cooling pipe; the device includes: A light guide, partially disposed within the reaction chamber, is used to conduct reflected light from the wafer surface; A detector, positioned outside the reaction cavity, is used to generate reflectance spectral data based on the reflected light. A vibration detection module is installed at the root of the optical guide tube to acquire vibration data of the optical guide tube. A temperature detection module is set at a preset position within the reaction chamber to acquire temperature data at the preset position; the preset position includes the inlet and outlet positions of the cooling pipe and the body position of the optical guide tube. An imaging module, located outside the reaction chamber, is used to acquire image information of the protective mirror of the observation window at the front end of the light guide tube; The fault diagnosis system for semiconductor optical guides as described in claim 12 is configured to acquire and parse first detection data and second detection data to obtain first fault information and second fault information, and generate a fault diagnosis result and a maintenance strategy based on the fault diagnosis result, so that engineers can maintain the optical guide according to the maintenance strategy; when the second fault information includes cooling efficiency parameters, and the cooling efficiency parameters include the body temperature of the optical guide, the processing module generates a fault diagnosis result based on the first fault information and the second fault information, including: If the temperature rise of the main body of the light guide exceeds the fourth preset threshold, the deviation correction amount caused by the current main body temperature to the intensity of reflected light is further calculated, and the fault diagnosis result is marked as the contaminant layer increasing thermal resistance and affecting the heat dissipation of the light guide. The calculation of the deviation correction amount caused by the current temperature to the intensity of reflected light includes: establishing a pollutant thermal resistance model, and quantifying the deviation correction amount caused by the pollutant layer to the temperature measurement accuracy of the light guide based on the pollution index and the rise value of the temperature at the position of the light guide body. The processing module generates a maintenance strategy based on the fault diagnosis results, including: performing real-time compensation calculations on the measured values ​​of the wafer surface temperature according to the deviation correction amount to obtain the calibrated wafer surface temperature value; and marking and outputting the calibrated wafer surface temperature value.

14. The apparatus according to claim 13, characterized in that, A sleeve is provided at the front end of the optical guide tube; wherein the sleeve adopts a mechanical quick-change connection mechanism so that the sleeve will detach from the front end of the optical guide tube when subjected to an axial or lateral impact force exceeding the design threshold.

15. The apparatus according to claim 14, characterized in that, The connection point between the optical guide and the sleeve is provided with a non-contact identification interface, which is used to acquire the identification information of the sleeve when the front end of the optical guide is connected to it and transmit it to the fault diagnosis system of the semiconductor optical guide, so that the fault diagnosis system of the semiconductor optical guide can create or update the file information corresponding to the sleeve based on the identification information; wherein, the file information includes the number of times it has been used and vibration data.

16. The apparatus according to claim 14, characterized in that, The optical guide is integrated with an in-situ detection unit for detecting the in-situ status of the sleeve and transmitting a trigger signal to the fault diagnosis system of the semiconductor optical guide when the sleeve falls off, so as to trigger the alarm mode of the fault diagnosis system of the semiconductor optical guide.

17. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-11.

18. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-11.

Images