Semiconductor equipment-oriented guide rail cleanliness monitoring system and method

Abstract

The invention relates to the technical field of semiconductor equipment monitoring, in particular to a guide rail cleanliness monitoring system and method for semiconductor equipment. The system comprises a signal acquisition module, a signal processing module, a pollution identification module and an evaluation decision module. A signal processing module performs signal analysis on the collected electrostatic induction signals and acoustic emission signals, and performs event detection in combination with motion state information to obtain a characteristic event list and a pollution index vector; the pollution identification module performs space-time correlation analysis on the feature event list to obtain an internal wear source, and identifies an external settlement source in combination with a pollution index vector; applying an excitation signal to the external settlement source, and identifying a silent pollution layer according to a response result; the evaluation decision module generates a comprehensive evaluation result of the cleanliness state of the guide rail based on the internal wear source, the external settlement source and the silent pollution layer; according to the invention, the online monitoring of the cleanliness of the guide rail and the accurate identification of the pollution source are realized, and the operation reliability of semiconductor equipment is improved.

Description

Technical Field

[0001] This application relates to the field of semiconductor equipment monitoring technology, specifically to a system and method for monitoring the cleanliness of guide rails in semiconductor equipment. Background Technology

[0002] Semiconductor manufacturing is a typical industry that relies heavily on cleanliness. As process technology continues to advance to the nanometer scale, the sensitivity of wafer surfaces to particulate contaminants has increased to the submicron and even nanometer levels. As a core transmission component of precision semiconductor equipment such as lithography machines, wafer inspection equipment, and etching machines, the surface cleanliness of guide rails directly determines the positioning accuracy of the equipment and the yield of chip products, making it one of the key links in ensuring the stability of the semiconductor manufacturing process.

[0003] Currently, various technologies for monitoring the cleanliness of guide rails have been developed in the industry. However, all of these technologies have inherent defects and cannot meet the practical application requirements of semiconductor equipment guide rails for high cleanliness, high precision, and online real-time monitoring. Among them, optical particle counting uses laser sensors to scan and monitor the guide rail surface, but it has significant limitations in semiconductor equipment scenarios: the guide rail surface is usually covered with an extremely thin lubricating oil film. After micron-sized metal debris is encased in the oil film, light undergoes complex refraction and scattering phenomena at the oil-gas-solid interface, causing the optical sensor to be unable to effectively penetrate the oil film to identify the encased conductive particles, creating a blind spot and easily leading to missed detections. Vibration monitoring uses accelerometers installed on the guide rail or slider to indirectly determine the wear state by monitoring changes in vibration signals during operation. However, its sensitivity is low; it can only capture effective signals when wear develops to a certain extent and produces significant mechanical impact, making it difficult to identify early, minute wear. Furthermore, the vibration signal is easily interfered with by the external environment, making it impossible to accurately locate the source of contamination. Offline sampling analysis involves periodically stopping the machine to wipe the guide rail surface and taking samples, which are then sent to the laboratory for microscopic analysis or chemical composition detection. Although it has high detection accuracy, it cannot achieve online real-time monitoring of cleanliness. The sampling process may also damage the original contamination state of the guide rail surface, and the detection cycle is long, making it difficult to respond to sudden contamination events in a timely manner. Especially for loose chemical contamination layers, the sampling operation itself may cause contaminants to fall off prematurely, which may increase the risk of wafer contamination.

[0004] In addition, the electrostatic induction method uses the static charge generated by particle friction for detection, and has been initially applied in fields such as pipeline pneumatic conveying and bearing wear monitoring. However, the existing method can only achieve qualitative alarm, and cannot solve the problem of signal instability caused by changes in charge amount with oil film thickness and particle distance. It is also difficult to effectively distinguish the source of contamination, and thus cannot meet the actual needs of accurate monitoring of the cleanliness of semiconductor equipment guide rails.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the overall background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this application is to provide a guide rail cleanliness monitoring system and method for semiconductor equipment, which to some extent solves the problems mentioned in the background art, realizes online monitoring of guide rail cleanliness and accurate identification of contamination sources, thereby improving the reliability of semiconductor equipment operation.

[0007] To achieve the above objectives, this application provides the following technical solution:

[0008] In the first aspect, this application provides a rail cleanliness monitoring system for semiconductor equipment, including a signal acquisition module, a signal processing module, a contamination identification module, and an assessment and decision module;

[0009] The signal acquisition module is used to acquire the electrostatic induction signal and acoustic emission signal generated by the guide rail during operation, and to synchronously acquire the motion status information of the slider through the synchronous acquisition card;

[0010] The signal processing module is used to perform signal analysis on the electrostatic induction signal and the acoustic emission signal, and to perform event detection in combination with the motion state information to generate a list of feature events and a pollution index vector.

[0011] The pollution identification module performs spatiotemporal correlation analysis on the list of feature events to obtain internal wear sources, and combines the pollution index vector to identify external sedimentation sources;

[0012] The pollution identification module also applies an excitation signal to the external sedimentation source and identifies the silent pollution layer based on the response result of the excitation signal;

[0013] The assessment and decision module is used to generate a comprehensive assessment result of the cleanliness status of the guide rail based on the internal wear sources, external sedimentation sources, and silent contamination layer.

[0014] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the feature event list includes an electrostatic event list and an acoustic emission event list; the signal processing module is configured with an electrostatic event list generation strategy; the electrostatic event list generation strategy specifically includes:

[0015] Differential operation is performed on the two voltage signals of each miniature differential electrode pair to obtain the differential voltage signal;

[0016] When the amplitude of the differential voltage signal is greater than a preset adaptive threshold and the differential voltage signal exhibits a zero-crossing waveform, it is determined as a valid granular event, and the corresponding waveform parameters are recorded; the waveform parameters include the zero-crossing time, waveform amplitude, and waveform width;

[0017] Based on waveform parameters and combined with the motion state information of the slider, the characteristic parameters of the corresponding effective particle events are calculated, including the vertical distance of the charged particle from the micro differential electrode pair and the equivalent diameter of the charged particle.

[0018] As a preferred embodiment of the rail cleanliness monitoring system for semiconductor equipment described in this application, the electrostatic event list generation strategy further includes:

[0019] Based on the waveform width and the slider's movement speed, the vertical distance between the charged particle and the micro differential electrode pair is calculated, and based on the waveform amplitude and the vertical distance between the charged particle and the micro differential electrode pair, the equivalent diameter of the charged particle is calculated.

[0020] The positions of the micro differential electrode pairs corresponding to the effective particle events are obtained as the corresponding detection positions;

[0021] For each detection location, acquire all valid particle events within a fixed time window, and extract the equivalent diameter of the charged particle corresponding to each valid particle event.

[0022] As a preferred embodiment of the rail cleanliness monitoring system for semiconductor equipment described in this application, the electrostatic event list generation strategy further includes:

[0023] The pollution index corresponding to the effective particle event is obtained by summing the cubes of the equivalent diameters of the charged particles corresponding to all effective particle events.

[0024] For each valid particle event, the position of the corresponding slider is obtained according to the corresponding zero-crossing time, and the position of the corresponding slider is taken as the occurrence position of the corresponding valid particle event;

[0025] Each effective particle event is defined as a single electrostatic event, consisting of its waveform parameters, characteristic parameters, detection location, corresponding pollution index, and occurrence location. This electrostatic event is then added to the electrostatic event list.

[0026] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the signal processing module is configured with an acoustic emission event list generation strategy; the acoustic emission event list generation strategy specifically includes:

[0027] The vibration signal output by each acoustic emission sensor is bandpass filtered, and a detection threshold is set.

[0028] If the amplitude of the bandpass filtered vibration signal is greater than or equal to the detection threshold for a preset first duration, it is determined to be a valid wear event, and the corresponding event parameters are calculated.

[0029] The event parameters include the event start time and the event energy; the event start time is the moment when the amplitude of the vibration signal corresponding to the effective wear event is greater than or equal to the detection threshold.

[0030] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the acoustic emission event list generation strategy further includes:

[0031] Obtain the start and end times of valid wear events, i.e., the start and end times of the corresponding vibration signal amplitude being greater than or equal to the detection threshold.

[0032] Using the start time as the integration starting point and the end time as the integration ending point, the square of the amplitude of the corresponding vibration signal is integrated to obtain the corresponding event energy.

[0033] The arrival time of the same valid wear event on each acoustic emission sensor is extracted, and the propagation speed of the vibration signal in the guide rail is obtained.

[0034] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the acoustic emission event list generation strategy further includes:

[0035] Calculate the arrival time difference between any two adjacent acoustic emission sensors, and calculate the location of the wear source based on the propagation speed, arrival time difference, and the installation location of the corresponding acoustic emission sensors;

[0036] The wear intensity of the effective wear events is classified according to the magnitude of the event energy, including: mild wear, moderate wear, and severe wear;

[0037] The event parameters, location of the wear source, and level of wear intensity of the effective wear event are taken as an acoustic emission event, and the acoustic emission event is added to the acoustic emission event list.

[0038] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the contamination identification module is configured with a spatiotemporal correlation analysis strategy to perform spatiotemporal correlation analysis on a list of characteristic events to obtain internal wear sources; the spatiotemporal correlation analysis strategy specifically includes:

[0039] For each acoustic emission event in the acoustic emission event list, a spatial window is set with the location of its corresponding wear source as the center, and a time window is set with the corresponding event start time as the center.

[0040] Traverse the list of electrostatic events and filter out electrostatic events whose location is within the spatial window and whose zero-crossing time is within the time window, as candidate electrostatic events corresponding to the acoustic emission event.

[0041] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the spatiotemporal correlation analysis strategy further includes:

[0042] Establish a mapping relationship between the wear intensity level and the equivalent diameter of charged particles, and compare the wear intensity level of acoustic emission events, the equivalent diameter of charged particles in candidate electrostatic events, and the mapping relationship.

[0043] If the mapping relationship is satisfied, the match is considered successful; otherwise, the match is considered unsuccessful.

[0044] If a matching candidate electrostatic event exists, the location of the wear source corresponding to the acoustic emission event is marked as an internal wear source.

[0045] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the contamination identification module is configured with an external sedimentation source identification strategy for identifying external sedimentation sources; the external sedimentation source identification strategy specifically includes:

[0046] Set a sliding window, and calculate the background mean and background standard deviation of the pollution index for all micro differential electrode pairs in the sliding window based on the pollution index vector.

[0047] The background threshold is calculated based on the background mean and background standard deviation, and the list of electrostatic events is traversed.

[0048] If the pollution index corresponding to any electrostatic event is greater than the corresponding background threshold within a preset second duration, then the detection location corresponding to the electrostatic event is taken as a pollution anomaly point.

[0049] Within the second spatial window centered on the pollution anomaly point, if there is no internal wear source, the corresponding pollution anomaly point is regarded as an external sedimentation source.

[0050] Otherwise, the pollution anomaly point is determined to be an extension of the corresponding internal wear source.

[0051] As a preferred embodiment of the rail cleanliness monitoring system for semiconductor equipment described in this application, the contamination identification module is configured with an active excitation verification strategy for applying an excitation signal to an external settling source and identifying a silent contamination layer based on the response result; the active excitation verification strategy specifically includes:

[0052] The pollution index of each external sedimentation source is continuously monitored. If the pollution index meets any preset trigger condition, the active excitation verification of the corresponding external sedimentation source is triggered.

[0053] The external settlement source that triggers active excitation verification is taken as the target location, and the active excitation unit is controlled to send an excitation signal to the target location.

[0054] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the active excitation verification strategy further includes:

[0055] After a preset delay time following the excitation signal, the signal acquisition module is activated to collect the electrostatic induction signal, acoustic emission signal, and motion state information of the guide rail.

[0056] The collected electrostatic induction signals and acoustic emission signals are analyzed, and event detection is performed in conjunction with the motion state information.

[0057] If a new acoustic emission event is detected, the characteristic frequency of the new acoustic emission event is extracted; if the characteristic frequency is different from the frequency of metal friction, or if the electrostatic induction signal shows a momentary spike, the target location is determined to be a silent pollution layer.

[0058] As a preferred embodiment of the guide rail cleanliness monitoring system for semiconductor equipment described in this application, the signal acquisition module is configured with a signal acquisition strategy; the signal acquisition strategy specifically includes:

[0059] An electrostatic induction array is arranged on the guide rail of the semiconductor device, and electrostatic induction signals are collected based on the electrostatic induction array; the electrostatic induction array is arranged by arranging miniature differential electrode pairs at equal intervals in the extension direction of the guide rail to obtain the electrostatic induction array.

[0060] At least three acoustic emission sensors are arranged at both ends and the middle of the guide rail to form an acoustic emission array, and acoustic emission signals are collected based on the acoustic emission array; the acoustic emission sensors are fixed to the guide rail bracket by magnetic bases, and the acoustic emission sensors are in close contact with the surface of the guide rail;

[0061] A motion encoder is installed on the shaft of the slider drive motor, and the position and speed of the slider are output through the motion encoder.

[0062] Secondly, this application provides a method for monitoring the cleanliness of guide rails in semiconductor equipment, comprising the following steps:

[0063] The system collects electrostatic induction signals and acoustic emission signals generated by the guide rail during operation, and synchronously collects the motion status information of the slider through a synchronous acquisition card.

[0064] Signal analysis is performed on the electrostatic induction signal and acoustic emission signal, and event detection is performed in combination with the motion state information to generate a feature event list and a pollution index vector;

[0065] Spatiotemporal correlation analysis is performed on the list of characteristic events to obtain internal wear sources, and external sedimentation sources are identified by combining the pollution index vector; an excitation signal is applied to the external sedimentation sources, and the silent pollution layer is identified based on the response result of the excitation signal.

[0066] Based on the internal wear sources, external sedimentation sources, and silent contamination layer, a comprehensive evaluation result of the guide rail cleanliness status is generated.

[0067] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0068] By synchronously acquiring electrostatic induction signals, acoustic emission signals, and slider motion state information of the guide rail, and performing signal analysis and event detection on these signals and motion state information, a list of characteristic events and a contamination index vector are obtained. This allows for the extraction of effective contamination and wear characteristics from the raw signals, filtering out noise interference and improving the sensitivity and reliability of cleanliness monitoring. Spatiotemporal correlation analysis is performed based on the characteristic event list to identify internal wear sources, and the contamination index vector is used to identify external settling sources. An excitation signal is applied to the external settling sources, and the silent contamination layer is identified based on the response to the excitation signal. This achieves automatic identification and precise location of contamination types, avoiding misjudgments and omissions, and improving the ability to identify complex and weak signal contamination and the completeness of monitoring. Based on the internal wear sources, external settling sources, and silent contamination layers, a comprehensive evaluation result of the guide rail cleanliness status is generated, achieving a comprehensive quantitative assessment of guide rail cleanliness and providing a reliable basis for preventative maintenance and precise operation and maintenance of semiconductor equipment. Attached Figure Description

[0069] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0070] Figure 1This is an architecture diagram of the rail cleanliness monitoring system for semiconductor equipment provided in this application.

[0071] Figure 2 A flowchart of the method for monitoring the cleanliness of guide rails in semiconductor equipment provided in this application. Detailed Implementation

[0072] The technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof. In the absence of conflict, the embodiments and technical features in the embodiments can be combined with each other.

[0073] Example 1

[0074] like Figure 1 As shown in the figure, this embodiment introduces a rail cleanliness monitoring system for semiconductor equipment, including a signal acquisition module, a signal processing module, a contamination identification module, and an evaluation and decision module;

[0075] The signal acquisition module is used to synchronously acquire the electrostatic induction signals, acoustic emission signals, and slider motion state information generated by the guide rail during operation in real time. The slider motion state information includes the slider's position and speed. The synchronous acquisition of the electrostatic induction signals, acoustic emission signals, and slider motion state information is achieved through a synchronous acquisition card. The synchronous acquisition card synchronously acquires the electrostatic induction signals, acoustic emission signals, and slider motion state information at a fixed sampling rate and adds a unified timestamp to the electrostatic induction signals, acoustic emission signals, and slider motion state information. The fixed sampling rate is greater than or equal to 1MHz.

[0076] The electrostatic induction signal and acoustic emission signal of the guide rail are acquired as follows: an electrostatic induction array is deployed on the guide rail of the semiconductor device, and the electrostatic induction signal is acquired based on the electrostatic induction array; the electrostatic induction array is deployed by arranging miniature differential electrode pairs at equal intervals along the extension direction of the guide rail to obtain the electrostatic induction array; each miniature differential electrode pair of the electrostatic induction array independently outputs two voltage signals. , The signal is an electrostatic induction signal; where i is the number of the micro differential electrode pair, and i = 1, 2, ..., N, where N is the total number of micro differential electrode pairs, and t is a time index representing the acquisition time of the electrostatic induction signal and the acoustic emission signal; an acoustic emission array is arranged on the guide rail of the semiconductor device, and acoustic emission signals are acquired based on the acoustic emission array; the acoustic emission array is arranged as follows: at least 3 acoustic emission sensors are arranged at both ends and the middle part of the guide rail to form an acoustic emission array; the acoustic emission sensors are fixed to the guide rail bracket by magnetic bases, and the acoustic emission sensors are in close contact with the guide rail surface to ensure effective transmission of vibration signals; each acoustic emission sensor in the acoustic emission array outputs a vibration signal A. j (t), i.e., acoustic emission signal; where j is the number of the acoustic emission sensor, j=1,2,...,M, M is the number of acoustic emission sensors, and M≥3;

[0077] In this embodiment, the micro differential electrode pairs are made of corrosion-resistant and interference-resistant precious metal materials, such as platinum-iridium alloy, to avoid contamination caused by electrode wear. Each micro differential electrode pair of the electrostatic induction array includes two semi-ring electrodes, which are symmetrically wrapped around both sides of the guide rail. The semi-ring electrodes do not contact the guide rail surface, leaving a 0.5mm gap to prevent friction damage to the guide rail. The total number of micro differential electrode pairs and the number of acoustic emission sensors are determined according to the length of the guide rail. For example, 100 micro differential electrode pairs are arranged per meter of guide rail. When the length of the guide rail is less than 2 meters, one acoustic emission sensor is arranged at each end and the middle of the guide rail. When the length of the guide rail is greater than or equal to 2 meters, one more acoustic emission sensor is added for every additional meter of guide rail length. The acoustic emission sensor is a piezoelectric acoustic emission sensor, and its encapsulation adopts a sealed structure to prevent corrosive gases and dust in the semiconductor manufacturing environment from damaging the acoustic emission sensor.

[0078] The motion state information of the slider is acquired by installing a motion encoder on the shaft of the slider drive motor and outputting the position and speed of the slider through the motion encoder; the motion encoder is an incremental photoelectric encoder, which is installed coaxially with the slider drive motor and has a resolution greater than or equal to 1000 lines.

[0079] Specifically, the miniature differential electrode pair is connected to the synchronous acquisition card via a shielded cable to reduce the impact of electromagnetic interference on the electrostatic induction signal; the output terminal of the acoustic emission sensor is connected to the synchronous acquisition card via a differential signal cable, and differential amplification technology is used during the transmission of the acoustic emission signal to suppress common-mode noise; the motion encoder is connected to the synchronous acquisition card via an RS485 interface, and the output signal of the motion encoder is decoded by the synchronous acquisition card after differential transmission, converting the output signal into digital position and motion speed; the synchronous acquisition card is a high-speed multi-channel synchronous acquisition card, whose channels correspond to the 2N electrostatic induction signals output by the electrostatic induction array, the M acoustic emission signals output by the acoustic emission array, and the 2 signals output by the motion encoder.

[0080] The signal processing module is used to perform signal analysis on the electrostatic induction signal and the acoustic emission signal, and to perform event detection in conjunction with the motion state information, generating a feature event list and a pollution index vector; the feature event list includes an electrostatic event list and an acoustic emission event list; the electrostatic event list is generated as follows:

[0081] Because the two semi-ring electrodes of a miniature differential electrode pair are located in the same electromagnetic environment, they are susceptible to common-mode noise, such as electromagnetic interference from equipment and electrostatic interference from the environment, leading to output signal distortion. Therefore, the two voltage signals of each miniature differential electrode pair are... , Differential operations are performed to obtain differential voltage signals, thereby eliminating common-mode interference and simultaneously enhancing the differential-mode signal generated by charged particles; the differential operation specifically involves: calculating... and The difference, i.e. minus ;

[0082] When a charged particle passes by the micro differential electrode pair, the differential voltage signal will exhibit a distinct zero-crossing waveform, meaning that when a charged particle approaches the micro differential electrode pair... and The difference will change in opposite directions. Therefore, based on the differential voltage signal, an adaptive threshold method is used to identify valid particle events: when the amplitude of the differential voltage signal is greater than a preset adaptive threshold and the differential voltage signal shows a zero-crossing waveform, it is determined to be a valid particle event, and the corresponding waveform parameters are recorded; the adaptive threshold is adjusted in real time according to the background noise, and the adaptive threshold is the sum of the mean of the background noise and twice the standard deviation of the background noise; the mean and standard deviation of the background noise are obtained by collecting pure background noise signals and analyzing the pure background noise signals before the guide rail generates obvious particle or wear interference. The waveform parameters are obtained through statistical analysis. These parameters include zero-crossing time, waveform amplitude, and waveform width. Zero-crossing time refers to the moment the differential voltage signal changes from positive to negative or vice versa. Waveform amplitude refers to the absolute peak value of the differential voltage signal, reflecting the charge intensity of the charged particle, and is related to the size of the charged particle and the distance between the charged particle and the micro-differential electrode pair. Waveform width refers to the half-width at half-maximum (WHM) of the differential voltage signal, i.e., the time interval when the amplitude of the differential voltage signal reaches half the waveform amplitude, reflecting the time it takes for the charged particle to pass through the micro-differential electrode pair, and is related to the speed and size of the charged particle.

[0083] Based on the waveform parameters and the motion state information of the slider, the characteristic parameters of the corresponding effective particle events are calculated. These characteristic parameters include: the vertical distance between the charged particle and the micro-differential electrode pair, and the equivalent diameter of the charged particle. Specifically, the vertical distance between the charged particle and the micro-differential electrode pair is calculated based on the waveform width and the slider's motion speed. The vertical distance between the charged particle and the micro-differential electrode pair is directly proportional to both the waveform width and the slider's motion speed, and the proportionality coefficient corresponding to this proportionality is obtained through offline calibration experiments. In this embodiment, standard particles with known particle sizes, such as monodisperse particles, are used. Polystyrene microspheres pass at a fixed speed through a known distance between charged particles and a micro-differential electrode pair, and the waveform width is measured to calculate the proportionality coefficient. Based on the waveform amplitude and the perpendicular distance between the charged particles and the micro-differential electrode pair, the equivalent diameter of the charged particles is calculated. Since the charge of a charged particle is proportional to its volume, and the waveform amplitude is proportional to the charge of the charged particle and inversely proportional to the square of the perpendicular distance between the charged particles and the micro-differential electrode pair, the equivalent diameter of the charged particles is positively correlated with both the waveform amplitude and the perpendicular distance between the charged particles and the micro-differential electrode pair.

[0084] The positions of the micro differential electrode pairs corresponding to the effective particle events are obtained as the corresponding detection positions. For each detection position, all effective particle events within a fixed time window are obtained, and the equivalent diameters of the charged particles corresponding to all effective particle events are extracted. The sum of the cubes of the equivalent diameters of the charged particles corresponding to all effective particle events is calculated to obtain the pollution index corresponding to the effective particle event. The fixed time window is determined based on the expected arrival rate of the effective particle events per unit time to ensure timely response to changes in pollution levels, such as 1 second.

[0085] For each valid particle event, the position of the corresponding slider is obtained according to the corresponding zero-crossing time, and the position of the corresponding slider is taken as the occurrence position of the corresponding valid particle event;

[0086] The waveform parameters, characteristic parameters, detection location, corresponding pollution index, and occurrence location of each effective particle event are taken as an electrostatic event, and the electrostatic event is added to the electrostatic event list.

[0087] The list of acoustic emission events is generated as follows:

[0088] Vibration signal A output by each acoustic emission sensor j (t) Bandpass filtering is performed; the bandpass filtering method is: using an infinite impulse response filter with a frequency range of 50kHz to 500kHz to filter out equipment operating noise below 50kHz and electromagnetic interference signals above 500kHz, so as to ensure the purity of the vibration signal.

[0089] Set a detection threshold, and based on the detection threshold and the bandpass-filtered vibration signal, identify valid wear events: if the bandpass-filtered A j If the amplitude of (t) is greater than or equal to the detection threshold within a preset first duration, it is determined to be a valid wear event, and the corresponding event parameters are calculated; the first duration is greater than 10 μs; the detection threshold is set by calculating the root mean square of the vibration signal within a preset sliding time window, and setting a second proportional coefficient according to the operating noise characteristics of the semiconductor device; multiplying the root mean square by the second proportional coefficient to obtain the corresponding detection threshold; the sliding time window is set according to the noise change cycle of the semiconductor device, such as 100 ms; the operating noise characteristics include the amplitude level, fluctuation range, spectral distribution, and time-varying pattern of the background noise; for example, the second proportional coefficient takes a smaller value, such as 1.5, when the noise amplitude is low and stable, to improve the detection sensitivity, and takes a larger value, such as 3.0, when the noise amplitude is high or fluctuates violently, to suppress false alarms; the value range of the second proportional coefficient is [1.5, 3.0];

[0090] The event parameters include the event start time and the event energy; the event start time is: A corresponding to the effective wear event. j The amplitude of (t) is greater than or equal to the start time of the detection threshold; the event energy is used to quantify the intensity of the effective wear event, and the greater the event energy, the more severe the wear; the event energy is calculated by obtaining the start time and end time of the effective wear event, i.e., the corresponding A j The amplitude of (t) is greater than or equal to the start and end times of the detection threshold; taking the start time as the integration start point and the end time as the integration end point, for the corresponding A j Integrate the square of the amplitude of (t) to obtain the corresponding event energy;

[0091] For the same effective wear event, the resulting vibration signal will propagate through the guide rail to different acoustic emission sensors; due to the different placement locations of the acoustic emission sensors, A j Since the arrival times of the same effective wear event differ across acoustic emission sensors (t), the peak detection method is used to extract the arrival time of each acoustic emission sensor, i.e., the A corresponding to each acoustic emission sensor. j The time corresponding to the peak value of (t);

[0092] Get A j (t) The propagation speed in the guide rail, and calculate the arrival time difference between any two adjacent acoustic emission sensors, that is, the absolute value of the difference in arrival time of the acoustic emission signals between any two adjacent acoustic emission sensors; based on the propagation speed, arrival time difference and the installation position of the corresponding acoustic emission sensor, the location of the wear source is calculated using the arrival time difference positioning algorithm.

[0093] The wear intensity of the effective wear events is classified according to the magnitude of the event energy, including: if the event energy is less than or equal to a preset first energy threshold, it is classified as mild wear; if the event energy is less than or equal to a preset second energy threshold and greater than the first energy threshold, it is classified as moderate wear; if the event energy is greater than the second energy threshold, it is classified as severe wear. The first energy threshold and the second energy threshold are determined based on the service life and maintenance cycle of the semiconductor device rail. In this embodiment, the first energy threshold is 100mV. 2 ·s, the second energy threshold is 500mV 2 ·s;

[0094] The event parameters, location of the wear source, and level of wear intensity of the effective wear event are taken as an acoustic emission event, and the acoustic emission event is added to the acoustic emission event list.

[0095] The dimension of the pollution index vector is the same as the total number of micro-differential electrode pairs; the pollution index vector is constructed by using the positions of the micro-differential electrode pairs as discrete points along the length of the guide rail. ; where q i is the pollution index corresponding to the position of the i-th micro differential electrode pair.

[0096] The pollution identification module performs spatiotemporal correlation analysis on the feature event list to obtain internal wear sources, and combines the pollution index vector to identify external sedimentation sources; the internal wear sources refer to wear particles generated by friction between the slider and the guide rail, and their generation method is as follows:

[0097] For each acoustic emission event in the acoustic emission event list, a spatial window is set centered on the location of its corresponding wear source, and a time window is set centered on the corresponding event start time. The length of the spatial window is adjusted according to the positioning accuracy of the wear source. The length of the time window is determined based on the time delay from the generation of charged particles from the wear source to their movement to the micro differential electrode pair. The time delay is negatively correlated with the running speed of the slider. In this embodiment, the length of the spatial window is 10mm by default, and the length of the time window is 100ms by default.

[0098] Traverse the list of electrostatic events and filter out electrostatic events whose location is within the spatial window and whose zero-crossing time is within the time window, as candidate electrostatic events corresponding to the acoustic emission event;

[0099] Based on the wear intensity level of the acoustic emission event and the equivalent diameter of the charged particles in each corresponding candidate electrostatic event, the candidate electrostatic events are matched. If a successfully matched candidate electrostatic event exists, the location of the wear source corresponding to the acoustic emission event is marked as an internal wear source. The specific method for matching the candidate electrostatic events is as follows: a mapping relationship between the wear intensity level and the equivalent diameter of the charged particles is established, and the wear intensity level of the acoustic emission event, the equivalent diameter of the charged particles in each corresponding candidate electrostatic event, and the mapping relationship are compared. If the mapping relationship is satisfied, the match is considered successful; otherwise, the match is considered unsuccessful.

[0100] The mapping relationship between the wear intensity level and the equivalent diameter of the charged particles was obtained through offline calibration experiments. In this embodiment, the equivalent diameter of the charged particles corresponding to mild wear is less than 1 μm; the equivalent diameter of the charged particles corresponding to moderate wear is greater than or equal to 1 μm and less than 5 μm; and the equivalent diameter of the charged particles corresponding to severe wear is greater than or equal to 5 μm.

[0101] The external settling source refers to the contamination formed by dust and impurities from the external environment settling onto the guide rail surface. Its identification method is as follows:

[0102] Set up a sliding window, such as 10 micro-differential electrode pairs, and calculate the background mean and background standard deviation of the pollution index of all micro-differential electrode pairs in the sliding window based on the pollution index vector; the background mean is the arithmetic mean of the pollution index of all micro-differential electrode pairs in the sliding window, reflecting the average pollution level of the guide rail area within the sliding window; the background standard deviation is the standard deviation of the pollution index of all micro-differential electrode pairs in the sliding window, reflecting the degree of deviation of the pollution index of each micro-differential electrode pair from the background mean.

[0103] The background threshold is calculated based on the background mean and background standard deviation, and the list of electrostatic events is traversed. If the pollution index corresponding to any electrostatic event is greater than the corresponding background threshold within a preset second duration, the detection location corresponding to the electrostatic event is taken as a pollution anomaly point. In this embodiment, the background threshold is the sum of the background mean and 3 times the background standard deviation. The second duration is set according to the sampling period of the electrostatic induction signal, for example, it is set to the total duration corresponding to 5 consecutive sampling periods.

[0104] Within the second spatial window centered on the pollution anomaly point, if there is no internal wear source, the corresponding pollution anomaly point is regarded as an external sedimentation source; otherwise, the pollution anomaly point is determined to be an extension of the corresponding internal wear source.

[0105] The contamination identification module also actively excites and verifies the external settling source by applying an excitation signal to it and identifying the silent contamination layer based on the response to the excitation signal. The silent contamination layer refers to a loose contamination layer attached to the guide rail surface, which is normally stable and has no obvious electrostatic or acoustic emission signals. It is easily detached under equipment vibration or slider movement disturbances, causing serious contamination to the wafer. The method for identifying the silent contamination layer is as follows:

[0106] The pollution index of each external sedimentation source is continuously monitored. If the pollution index meets any preset trigger condition, active excitation verification of the corresponding external sedimentation source is triggered. The trigger conditions include: the pollution index continuously rises, and the current pollution index exceeds a preset percentage of the historical peak value; the preset percentage is set according to actual monitoring needs; for example, the current pollution index exceeds 50% of the peak pollution index in the past hour; the absolute value of the pollution index of any external sedimentation source is greater than a preset excessive threshold for K consecutive sampling periods; where K is a positive integer; the excessive threshold is set according to actual monitoring needs; for example, the absolute value of the pollution index is greater than 5 for 10 consecutive sampling periods. The specific method of active excitation verification is as follows:

[0107] The external settlement source that triggers active excitation verification is taken as the target location, and the active excitation unit is controlled to send an excitation signal to the target location; the active excitation unit is a pulsed laser or an ultrasonic transducer; the excitation signal is laser excitation or ultrasonic excitation.

[0108] After a preset delay time following the excitation signal, the signal acquisition module is activated to collect the electrostatic induction signal, acoustic emission signal, and motion state information of the slider from the guide rail. The acquired electrostatic induction signal and acoustic emission signal are analyzed, and event detection is performed in conjunction with the motion state information. If a new acoustic emission event is detected, the characteristic frequency of the new acoustic emission event is extracted. If the characteristic frequency differs from the frequency of metal friction, or if the electrostatic induction signal exhibits a momentary spike, the target location is determined to be a silent contamination layer. In this embodiment, the delay time is set based on the duration of the excitation signal itself, such as 10 μs; the frequency of metal friction is 50~500 kHz.

[0109] The assessment and decision-making module is used to generate a comprehensive assessment result of the guide rail cleanliness status based on the internal wear sources, external settling sources, and silent contamination layer. The generation of this comprehensive assessment result includes: visualization of contamination distribution and generation of maintenance recommendations. The specific steps for visualizing the contamination distribution are as follows:

[0110] The risk of detachment of the silent contamination layer is assessed based on the new acoustic emission events and electrostatic induction signals: if the corresponding event energy is greater than or equal to a preset third energy threshold, or the peak amplitude of the electrostatic induction signal is greater than or equal to a preset first amplitude threshold, it is determined to be of high risk of detachment; if the corresponding event energy is greater than or equal to a preset fourth energy threshold and less than the third energy threshold, or the peak amplitude of the electrostatic induction signal is greater than or equal to a preset second amplitude threshold and less than the first amplitude threshold, it is determined to be of medium risk of detachment; if the corresponding event energy is less than the fourth energy threshold, or the peak amplitude of the electrostatic induction signal is less than the second amplitude threshold, it is determined to be of low risk of detachment; the third and fourth energy thresholds are determined based on the service life and maintenance cycle of the semiconductor equipment guide rail; the first and second amplitude thresholds are determined based on the peak induced voltage generated by standard particles, such as monodisperse polystyrene microspheres, at a known distance obtained from offline calibration experiments;

[0111] A two-dimensional visualization chart is constructed with the guide rail length as the horizontal axis and the pollution index as the vertical axis, and three types of pollution source markers are superimposed: internal wear sources are represented by red dots, where the size of the red dots is positively correlated with the wear intensity level, that is, the smallest red dot indicates light wear, and the largest red dot indicates heavy wear; external sedimentation sources are represented by blue squares, where the size of the blue squares is positively correlated with the pollution index; and silent pollution layers are represented by yellow triangles, and the flashing frequency of the yellow triangles is positively correlated with the risk of detachment.

[0112] The maintenance recommendations are generated in the following way:

[0113] For internal wear sources, maintenance recommendations are generated based on the wear intensity level: If the internal wear source is minor, it is recommended to check the wear condition of the corresponding guide rail or slider every 24 hours, and check whether there are scratches on the guide rail surface and whether the slider lubrication is sufficient; if the slider lubrication is insufficient, add appropriate lubricant in time; if the internal wear source is moderate, it is recommended to check the wear condition of the corresponding guide rail or slider every 12 hours and measure the wear amount of the guide rail; if the wear amount of the guide rail is less than or equal to 0.1mm, it can continue to be used; otherwise, the slider position needs to be adjusted; if the internal wear source is severe, it is recommended to stop the machine immediately to check the wear condition of the corresponding guide rail or slider, and replace the severely worn guide rail or slider to avoid aggravated wear and the generation of a large amount of particulate contamination;

[0114] For external sedimentation sources, maintenance recommendations are generated based on the pollution index: if the pollution index is less than or equal to the preset high index threshold, it is recommended to perform local cleaning on the corresponding location within 24 hours; in this embodiment, the high index threshold is the sum of twice the background mean and three times the standard deviation; if the pollution index is greater than the high index threshold, it is recommended to perform local cleaning on the corresponding location within 12 hours and check the sealing of the semiconductor equipment to prevent external dust from settling again.

[0115] For silent contamination layers, maintenance recommendations are based on the risk of detachment: If the silent contamination layer has a high risk of detachment, it is recommended to immediately stop the machine and clean the corresponding location using vacuum adsorption and wiping with a lint-free cloth to prevent the silent contamination layer from detaching. After cleaning, the guide rail surface should be inspected to ensure there is no residual contamination. If the silent contamination layer has a medium risk of detachment, it is recommended to clean the corresponding location within 6 hours. If the silent contamination layer has a low risk of detachment, it is recommended to clean the corresponding location within 24 hours.

[0116] If there are two or more internal wear sources, or three or more external sedimentation sources or silent contamination layers in the same area, it is recommended to carry out a comprehensive overhaul of the corresponding area, check the installation accuracy of the guide rail, the operating status of the slider, and investigate the causes of abnormalities in the semiconductor equipment.

[0117] Example 2

[0118] This embodiment is the second embodiment of this application; it is based on the same inventive concept as Embodiment 1, and refers to... Figure 2 This embodiment describes a method for monitoring the cleanliness of guide rails in semiconductor equipment, including the following steps:

[0119] The system collects electrostatic induction signals and acoustic emission signals generated by the guide rail during operation, and synchronously collects the motion state information of the slider through a synchronous acquisition card; the motion state information of the slider includes the slider's position and speed.

[0120] Signal analysis is performed on the electrostatic induction signal and the acoustic emission signal, and event detection is performed in conjunction with the motion state information to generate a feature event list and a pollution index vector; the feature event list includes an electrostatic event list and an acoustic emission event list;

[0121] Spatiotemporal correlation analysis is performed on the list of characteristic events to obtain internal wear sources, and external sedimentation sources are identified by combining the pollution index vector; an excitation signal is applied to the external sedimentation sources, and the silent pollution layer is identified based on the response result of the excitation signal.

[0122] Based on the internal wear sources, external settling sources, and silent contamination layer, a comprehensive assessment result of the guide rail cleanliness status is generated; the comprehensive assessment result of the guide rail cleanliness status includes: visualization of contamination distribution and generation of maintenance recommendations.

[0123] The specific functions of each step described above are explained in the relevant content of the guide rail cleanliness monitoring system for semiconductor equipment described in Example 1, and will not be repeated here.

[0124] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0125] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of protection of this application, and these forms are all within the protection scope of this application.

Claims

1. A guide rail cleanliness monitoring system for semiconductor equipment, characterized in that, It includes a signal acquisition module, a signal processing module, a pollution identification module, and an assessment and decision-making module; The signal acquisition module is used to acquire the electrostatic induction signal and acoustic emission signal generated by the guide rail during operation, and to synchronously acquire the motion status information of the slider through the synchronous acquisition card; The signal processing module is used to perform signal analysis on the electrostatic induction signal and the acoustic emission signal, and to perform event detection in combination with the motion state information to generate a list of feature events and a pollution index vector. The pollution identification module performs spatiotemporal correlation analysis on the list of feature events to obtain internal wear sources, and combines the pollution index vector to identify external sedimentation sources; The pollution identification module also applies an excitation signal to the external sedimentation source and identifies the silent pollution layer based on the response result of the excitation signal; The assessment and decision module is used to generate a comprehensive assessment result of the cleanliness status of the guide rail based on the internal wear sources, external sedimentation sources, and silent contamination layer.

2. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 1, characterized in that, The feature event list includes an electrostatic event list and an acoustic emission event list; The signal processing module is configured with an electrostatic event list generation strategy; The electrostatic event list generation strategy specifically includes: Differential operation is performed on the two voltage signals of each miniature differential electrode pair to obtain the differential voltage signal; When the amplitude of the differential voltage signal is greater than a preset adaptive threshold and the differential voltage signal exhibits a zero-crossing waveform, it is determined as a valid granular event, and the corresponding waveform parameters are recorded; the waveform parameters include the zero-crossing time, waveform amplitude, and waveform width; Based on waveform parameters and combined with the motion state information of the slider, the characteristic parameters of the corresponding effective particle events are calculated, including the vertical distance of the charged particle from the micro differential electrode pair and the equivalent diameter of the charged particle.

3. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 2, characterized in that, The electrostatic event list generation strategy also includes: Based on the waveform width and the slider's movement speed, the vertical distance between the charged particle and the micro differential electrode pair is calculated, and based on the waveform amplitude and the vertical distance between the charged particle and the micro differential electrode pair, the equivalent diameter of the charged particle is calculated. The positions of the micro differential electrode pairs corresponding to the effective particle events are obtained as the corresponding detection positions; For each detection location, acquire all valid particle events within a fixed time window, and extract the equivalent diameter of the charged particle corresponding to each valid particle event.

4. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 3, characterized in that, The electrostatic event list generation strategy also includes: The pollution index corresponding to the effective particle event is obtained by summing the cubes of the equivalent diameters of the charged particles corresponding to all effective particle events. For each valid particle event, the position of the corresponding slider is obtained according to the corresponding zero-crossing time, and the position of the corresponding slider is taken as the occurrence position of the corresponding valid particle event; Each effective particle event is defined as a single electrostatic event, consisting of its waveform parameters, characteristic parameters, detection location, corresponding pollution index, and occurrence location. This electrostatic event is then added to the electrostatic event list.

5. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 2, characterized in that, The signal processing module is configured with an acoustic emission event list generation strategy; the acoustic emission event list generation strategy specifically includes: The vibration signal output by each acoustic emission sensor is bandpass filtered, and a detection threshold is set. If the amplitude of the bandpass filtered vibration signal is greater than or equal to the detection threshold for a preset first duration, it is determined to be a valid wear event, and the corresponding event parameters are calculated. The event parameters include the event start time and the event energy; the event start time is the moment when the amplitude of the vibration signal corresponding to the effective wear event is greater than or equal to the detection threshold.

6. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 5, characterized in that, The acoustic emission event list generation strategy also includes: Obtain the start and end times of valid wear events, i.e., the start and end times of the corresponding vibration signal amplitude being greater than or equal to the detection threshold. Using the start time as the integration starting point and the end time as the integration ending point, the square of the amplitude of the corresponding vibration signal is integrated to obtain the corresponding event energy. The arrival time of the same valid wear event on each acoustic emission sensor is extracted, and the propagation speed of the vibration signal in the guide rail is obtained.

7. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 6, characterized in that, The acoustic emission event list generation strategy also includes: Calculate the arrival time difference between any two adjacent acoustic emission sensors, and calculate the location of the wear source based on the propagation speed, arrival time difference, and the installation location of the corresponding acoustic emission sensors; The wear intensity of the effective wear events is classified according to the magnitude of the event energy, including: mild wear, moderate wear, and severe wear; The event parameters, location of the wear source, and level of wear intensity of the effective wear event are taken as an acoustic emission event, and the acoustic emission event is added to the acoustic emission event list.

8. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 1, characterized in that, The pollution identification module is configured with a spatiotemporal correlation analysis strategy to perform spatiotemporal correlation analysis on the list of characteristic events to obtain internal wear sources; the spatiotemporal correlation analysis strategy specifically includes: For each acoustic emission event in the acoustic emission event list, a spatial window is set with the location of its corresponding wear source as the center, and a time window is set with the corresponding event start time as the center. Traverse the list of electrostatic events and filter out electrostatic events whose location is within the spatial window and whose zero-crossing time is within the time window, as candidate electrostatic events corresponding to the acoustic emission event.

9. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 8, characterized in that, The spatiotemporal correlation analysis strategy also includes: Establish a mapping relationship between the wear intensity level and the equivalent diameter of charged particles, and compare the wear intensity level of acoustic emission events, the equivalent diameter of charged particles in candidate electrostatic events, and the mapping relationship. If the mapping relationship is satisfied, the match is considered successful; otherwise, the match is considered unsuccessful. If a matching candidate electrostatic event exists, the location of the wear source corresponding to the acoustic emission event is marked as an internal wear source.

10. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 2, characterized in that, The pollution identification module is configured with an external sedimentation source identification strategy for identifying external sedimentation sources; the external sedimentation source identification strategy specifically includes: Set a sliding window, and calculate the background mean and background standard deviation of the pollution index for all micro differential electrode pairs in the sliding window based on the pollution index vector. The background threshold is calculated based on the background mean and background standard deviation, and the list of electrostatic events is traversed. If the pollution index corresponding to any electrostatic event is greater than the corresponding background threshold within a preset second duration, then the detection location corresponding to the electrostatic event is taken as a pollution anomaly point. Within the second spatial window centered on the pollution anomaly point, if there is no internal wear source, the corresponding pollution anomaly point is regarded as an external sedimentation source. Otherwise, the pollution anomaly point is determined to be an extension of the corresponding internal wear source.

11. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 1, characterized in that, The pollution identification module is configured with an active excitation verification strategy, used to apply excitation signals to external sedimentation sources and identify silent pollution layers based on the response results; the active excitation verification strategy specifically includes: The pollution index of each external sedimentation source is continuously monitored. If the pollution index meets any preset trigger condition, the active excitation verification of the corresponding external sedimentation source is triggered. The external settlement source that triggers active excitation verification is taken as the target location, and the active excitation unit is controlled to send an excitation signal to the target location.

12. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 11, characterized in that, The active incentive verification strategy also includes: After a preset delay time following the excitation signal, the signal acquisition module is activated to collect the electrostatic induction signal, acoustic emission signal, and motion state information of the guide rail. The collected electrostatic induction signals and acoustic emission signals are analyzed, and event detection is performed in conjunction with the motion state information. If a new acoustic emission event is detected, the characteristic frequency of the new acoustic emission event is extracted; if the characteristic frequency is different from the frequency of metal friction, or if the electrostatic induction signal shows a momentary spike, the target location is determined to be a silent pollution layer.

13. The guide rail cleanliness monitoring system for semiconductor equipment as described in claim 1, characterized in that, The signal acquisition module is configured with a signal acquisition strategy; The signal acquisition strategy specifically includes: An electrostatic induction array is arranged on the guide rail of the semiconductor device, and electrostatic induction signals are collected based on the electrostatic induction array; the electrostatic induction array is arranged by arranging miniature differential electrode pairs at equal intervals in the extension direction of the guide rail to obtain the electrostatic induction array. At least three acoustic emission sensors are arranged at both ends and the middle of the guide rail to form an acoustic emission array, and acoustic emission signals are collected based on the acoustic emission array; the acoustic emission sensors are fixed to the guide rail bracket by magnetic bases, and the acoustic emission sensors are in close contact with the surface of the guide rail; A motion encoder is installed on the shaft of the slider drive motor, and the position and speed of the slider are output through the motion encoder.

14. A method for monitoring the cleanliness of guide rails in semiconductor equipment, implemented based on the guide rail cleanliness monitoring system for semiconductor equipment as described in any one of claims 1-13, characterized in that, Includes the following steps: The system collects electrostatic induction signals and acoustic emission signals generated by the guide rail during operation, and synchronously collects the motion status information of the slider through a synchronous acquisition card. Signal analysis is performed on the electrostatic induction signal and acoustic emission signal, and event detection is performed in combination with the motion state information to generate a feature event list and a pollution index vector; Spatiotemporal correlation analysis is performed on the list of characteristic events to obtain internal wear sources, and external sedimentation sources are identified by combining the pollution index vector; an excitation signal is applied to the external sedimentation sources, and the silent pollution layer is identified based on the response result of the excitation signal. Based on the internal wear sources, external sedimentation sources, and silent contamination layer, a comprehensive evaluation result of the guide rail cleanliness status is generated.

Images