Infrared heating lamp explosion-proof laser-induced experimental device and experimental method

Abstract

The invention discloses an infrared heating lamp explosion-proof laser induction experiment device and method, and belongs to the technical field of equipment detection, and the method comprises the steps: installing at least three to-be-detected bulbs in the experiment device, carrying out the laser induction of each to-be-detected bulb, and obtaining a first data set; screening characteristic electrical abrupt change signals and thermal response abrupt change signals when the bulb is broken, and performing synchronous judgment; extracting multi-modal data fragments before and after a feature synchronization event occurs, and establishing a feature synchronization event database; and performing feature training and learning based on the feature synchronization event database, and performing risk assessment on the synchronization event of each to-be-tested bulb. In the implementation process of the technical scheme, laser induction and data acquisition are performed on the bulb to be detected, and synchronous judgment is performed according to the time coupling relationship between the electrical signal and the thermal response abrupt change signal, so that the risk level of breakage of the bulb to be detected is determined.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of device detection, in particular to an infrared heating lamp explosion-proof laser induction experiment device and experiment method. BACKGROUND

[0002] Halogen infrared heating lamps are widely used in semiconductor EPI, CVD, RTP and other processes, and have the advantages of fast heating and stable and reliable heating, so they are widely used.

[0003] When the halogen infrared heating lamp reaches the end of its life, the filament will melt. At the moment of filament melting, arc gas will be generated in the lamp. These gases will rapidly expand under high current, which can easily cause the bulb to explode and rupture. Since these halogen infrared heating lamps are used in semiconductor equipment, the explosion will damage semiconductor equipment accessories and chip materials. Therefore, in the prior art, it is necessary to perform an induction test on the halogen infrared heating lamp, such as laser induction, to simulate the arc and gas expansion process at the moment of filament melting, in order to identify high-risk rupture samples in advance.

[0004] However, the existing laser induction experiment method relies on a single signal to determine whether the bulb has ruptured, lacks collaborative analysis of multiple modal signals such as electrical, thermal response and acoustic emission, resulting in a high misjudgment rate. In addition, it is difficult to accurately capture weak sign signals before rupture during the experiment, which cannot realize early risk warning, resulting in difficulty in accurately grading and dynamically predicting the rupture risk.

[0005] Therefore, it is necessary to provide an infrared heating lamp explosion-proof laser induction experiment device and experiment method to solve the above problems.

[0006] It should be noted that the above information disclosed in this background section is only used to understand the background technology of the concept of the present application, and therefore, it can contain information that does not constitute prior art. SUMMARY

[0007] Based on the above problems existing in the prior art, the present application solves the problem of providing an infrared heating lamp explosion-proof laser induction experiment device and experiment method. By performing laser induction experiments on the bulbs to be tested, the multi-modal parameters under the rupture condition are obtained, which facilitates subsequent monitoring.

[0008] The technical solution adopted by the present application to solve its technical problems is: an infrared heating lamp explosion-proof laser induction experiment method, comprising: At least three bulbs to be tested are installed in the experiment device, and laser induction is performed on each bulb to be tested, and a first data set is obtained by a collection device; Based on the first data set, the characteristic electrical mutation signal and the thermal response mutation signal occurring when the bulb breaks are screened out, and the synchronous judgment is carried out according to the time coupling relationship of the characteristic electrical signal and the thermal response mutation signal; Based on the synchronous judgment result, the multi-modal data segments before and after the occurrence of the characteristic synchronous event are extracted, and a characteristic synchronous event database is established, the electrical, thermal response and acoustic emission data within the time segment corresponding to each synchronous event are packaged and stored, and the event type and occurrence time are marked; Based on the characteristic synchronous event database, the characteristic training and learning are carried out, and the risk of each bulb to be tested is evaluated, and the risk level of the bulb to be tested to break under the condition of laser induction is judged.

[0009] In the implementation process of the technical scheme of the present application, the laser induction is carried out on the bulb to be tested, and the data is collected, the synchronous judgment is carried out according to the time coupling relationship of the electrical signal and the thermal response mutation signal, so as to determine the risk level of the bulb to be tested to break.

[0010] Further, the first data set includes the following data: The electrical parameter change data of the bulb to be tested before and after the start of laser induction; The time node of the filament fusion and the corresponding laser irradiation cumulative time; The surface temperature distribution data of the filament at the moment of fusion; The image or video data of the bulb at the moment of fusion or breakage; The acoustic signal or micro-pressure change data generated during laser induction; And the deformation and crack propagation process data of the bulb shell under the action of thermal stress.

[0011] Further, the synchronous judgment process includes the following steps: Time axis unification and sampling synchronization are carried out, and all original data from different sources are time-stamped corrected according to a unified time reference; The mutation points of various data in the first data set are extracted respectively, and the time proximity analysis of these mutation points is carried out, and the synchronous events caused by the filament fusion and the corresponding data are screened out to generate a characteristic synchronous event sequence; Based on the characteristic synchronous event sequence, the time domain and frequency domain characteristic parameters of each modality signal before and after the mutation moment are further extracted, including the current drop slope, the infrared radiation peak rise time, the acoustic emission signal main frequency energy distribution, the multi-modal feature vector is constructed and normalized.

[0012] Further, the mutation point shows the step change of signal amplitude or the sharp rise of derivative, the current sudden drop and voltage fluctuation in the electrical signal are detected by the difference algorithm or wavelet transform method based on sliding window, the instantaneous jump of infrared radiation intensity in the thermal response data is identified, and the burst pulse of high frequency component in the acoustic emission signal is marked, then the time proximity analysis of various mutation points is carried out.

[0013] Further, if the electrical mutation, thermal response mutation and acoustic signal burst occur simultaneously in a millisecond time window, it is determined as a synchronous event caused by the filament burning out, and these data are collected in time sequence to form a characteristic synchronous event sequence.

[0014] Further, the characteristic synchronous event database supports index query by time range and event type.

[0015] Further, when a combination mode of steep current drop slope, short infrared response rise time and high intensity acoustic emission main frequency energy is identified, it is marked as a high-risk precursor of rupture.

[0016] Further, after feature training and feature learning, the risk level of bulb rupture is judged, the similarity between the feature vector of the to-be-tested bulb under the same excitation condition and the high-risk event marked in the database is calculated, and the Mahalanobis distance or cosine similarity is used for quantitative comparison to determine whether it is a high-risk rupture event.

[0017] An infrared heating lamp explosion-proof laser induction experiment system, comprising: A data acquisition module is configured to install at least three to-be-tested bulbs into an experimental device, induce each to-be-tested bulb with laser, and obtain a first data set through a collection device; A synchronous determination module is configured to filter out characteristic electrical mutation signals and thermal response mutation signals occurring when the bulb ruptures based on the first data set, and perform synchronous determination according to the time coupling relationship between the characteristic electrical signals and the thermal response mutation signals; A multi-modal data segment extraction module is configured to extract multi-modal data segments before and after the occurrence of characteristic synchronous events based on the synchronous determination result, and establish a characteristic synchronous event database, pack and store electrical, thermal response and acoustic emission data in the time segment corresponding to each synchronous event, and mark the event type and occurrence time; A risk assessment module is configured to perform feature training and learning based on the characteristic synchronous event database, and perform risk assessment on the synchronous events of each to-be-tested bulb to determine the risk level of rupture under laser induction.

[0018] The infrared heating lamp explosion-proof laser induction experiment device and experiment method provided by the present application have the beneficial effects that the to-be-tested bulb is induced with laser and data is collected, synchronous determination is performed according to the time coupling relationship between the electrical signals and the thermal response mutation signals, and thus the risk level of rupture of the to-be-tested bulb is determined.

[0019] In addition to the purposes, features and advantages described above, the present application has other purposes, features and advantages. The present application will be described in further detail below with reference to the drawings. Attached Figure Description

[0020] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic diagram of the experimental method for explosion-proof laser-induced infrared heating lamps according to this application; Figure 2 This is a schematic diagram of the module structure of an explosion-proof laser-induced experimental system for infrared heating lamps according to this application; Figure 3 This is a schematic diagram of the structure of an explosion-proof laser-induced experimental device for infrared heating lamps according to this application; The following are the labeling elements in the figure: 1. First slide; 2. Second slide; 3. Fixing plate; 4. Mounting slot; 11. Ball screw; 12. Mounting platform. Detailed Implementation

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0023] Example 1: As Figure 1 As shown, this application provides a laser-induced explosion-proof experimental method for infrared heating lamps. This method is mainly applied to the induction experiment of infrared heating lamps. Infrared heating lamps are devices used to generate infrared radiation and play a heating role in the semiconductor EPI process. However, there is a risk of cracking due to local overheating. Therefore, it is necessary to simulate extreme energy input through laser induction and monitor the infrared heating lamp to determine whether there are safety hazards such as thermal cracking, laser-induced explosion, and fragmentation. Specifically, the method includes the following steps: Step 101: Install at least three test bulbs into the experimental setup, and perform laser induction on each test bulb, and acquire the first data set through the acquisition device; When laser induction experiment of infrared heating lamp is carried out, a test device is needed, which is used for fixing the to-be-tested bulb and providing controllable and stable laser irradiation environment. After the to-be-tested bulb is installed, laser induction is carried out. Under the action of laser irradiation, the filament in the to-be-tested bulb will be fused due to high temperature, thereby simulating the condition when the bulb reaches the end of life in actual use and accelerating the bulb aging process. In the embodiment, at least three to-be-tested bulbs are needed for testing (preferably 3 to 10), so as to obtain sufficient sample data. In the embodiment, at least three to-be-tested bulbs are needed for testing (preferably 3 to 10), so as to obtain sufficient sample data. In the experiment, the electrical parameters of each to-be-tested bulb when the filament is fused are collected and taken as a first data set. The collection equipment includes but is not limited to a high-speed camera, an electrical parameter collection device, an infrared thermal imager, an acoustic emission sensor and the like. The multi-dimensional data of each to-be-tested bulb in the laser induction process and at the moment of fusion are collected in real time. The collected data collectively constitute the first data set. The first data set includes the following data: The electrical parameter change data of the to-be-tested bulb before and after the start of laser induction, such as the change of voltage, current, power and resistance value; the time node of filament fusion and the corresponding laser irradiation cumulative time; the surface temperature distribution data of the filament at the moment of fusion, which is collected by an infrared thermal imager; the image or video data of the bulb at the moment of fusion or rupture, which is collected by a high-speed camera; the acoustic signal or micro-pressure change data generated in the laser induction process, which is collected by an acoustic emission sensor and a pressure sensor; and the deformation and crack propagation process data of the bulb shell under the action of thermal stress. Through the collection of the above data, consistent laser-induced effects can be applied to multiple infrared heating lamp samples under the same experimental conditions, and data such as electrical characteristic changes, thermal response characteristics, and structural integrity changes of the samples when the laser energy is input at the near-end-of-life state can be obtained, serving as the basis for subsequent multi-dimensional feature analysis and life prediction model construction. Step 102: Based on the first data set, the characteristic electrical and thermal response mutation signals occurring when the bulb breaks are screened, and the time coupling relationship between the characteristic electrical and thermal response mutation signals is determined synchronously. In the first data set, not only electrical parameters but also thermal response, acoustic, and mechanical deformation multi-modal data are included. These multi-modal data usually have correlations in the time dimension, especially at the moment of filament melting, the mutation of electrical parameters is often accompanied by the sudden change of thermal radiation intensity and the acoustic emission signal caused by structural rupture. Due to the difference in sensor sampling frequency and the difference in signal transmission path, there may be a slight time offset between the modal signals. In this case, the precise occurrence time and cause of filament melting or structural rupture cannot be directly determined at the original data level, so synchronous determination based on the time coupling relationship between the characteristic electrical and thermal response mutation signals is required. Specifically, the synchronous determination process includes the following steps: Time axis unification and sampling synchronization are performed, and all original data from different sources are timestamp corrected according to a unified time reference. All original data from different sources are timestamp corrected according to a unified time reference, and an interpolation resampling method is used to align data sequences with different sampling rates to the same time axis, thereby eliminating the time offset caused by the difference in sampling rate and hardware difference, and ensuring that the modal data are aligned in the time dimension. The mutation points of each type of data in the first data set are extracted, and time proximity analysis is performed on these mutation points. The synchronous events caused by filament melting and the corresponding data are screened to generate a characteristic synchronous event sequence. After the time alignment of the various data included in the first data set, the mutation points can be extracted. These mutation points usually exhibit step changes in signal amplitude or sharp rises in derivative. Therefore, the current drop and voltage fluctuation in the electrical signal can be detected by a difference algorithm based on a sliding window or a wavelet transform method, the instantaneous jump in infrared radiation intensity in the thermal response data can be identified, and the burst pulses of high-frequency components in the acoustic emission signal can be marked. Subsequently, time proximity analysis is performed on each type of mutation point. If the electrical mutation, thermal response mutation, and acoustic signal burst occur simultaneously within a millisecond time window, it is determined to be a synchronous event caused by filament melting, and these data are collected in chronological order to form a characteristic synchronous event sequence. Based on the sequence of characteristic synchronization events, further extract the time domain and frequency domain feature parameters of each modal signal before and after the mutation moment, including current drop slope, infrared radiation peak rise time, acoustic emission signal main frequency energy distribution, construct multi-modal feature vector and normalize processing; On the basis of the sequence of characteristic synchronization events, extract the key features of each modal signal in the time window before and after the mutation, such as current drop slope, infrared radiation peak rise time, acoustic emission main frequency band energy ratio, and construct a multi-dimensional feature vector. Then, through the normalization method, the dimension of each feature parameter is unified, and the weight deviation caused by the difference in signal amplitude is eliminated, thereby providing the subsequent standardized input; Step 103: Based on the synchronization determination result, extract the multi-modal data segment before and after the occurrence of the characteristic synchronization event, and establish a characteristic synchronization event database. The electrical, thermal response and acoustic emission data in the time segment corresponding to each synchronization event are packaged and stored, and the event type and occurrence time are labeled; After synchronization determination, based on the multi-modal data segment in the sequence of characteristic synchronization events, a characteristic synchronization event database is established. The current, voltage, infrared thermal image and acoustic emission signal in the time segment corresponding to each filament melting or breaking related synchronization event are intercepted and packaged and stored, and the event type is labeled as "filament melting" or "filament breaking", and the specific time stamp of the event occurrence is labeled, which is convenient for subsequent model training or fault mode comparison, and also supports indexing query according to time range and event type, improving data analysis efficiency; Step 104: Based on the characteristic synchronization event database, perform feature training and learning, and evaluate the risk of each synchronization event of the to-be-tested bulb, and determine the risk level of the bulb under the condition of laser induction.

[0024] After establishing the characteristic synchronization event database, model training and feature learning are also needed according to the known data in the database, and pattern recognition and labeling are needed for high-risk events. For example, identify the combination pattern of steep current drop slope, short infrared response rise time and high intensity acoustic emission main frequency energy, and mark it as a high-risk breaking precursor. Or when the current drops slowly, the infrared radiation rises smoothly, and the acoustic emission signal is weak, it is determined to be a low-risk melting mode; After feature training and feature learning, the risk level of the bulb breaking needs to be determined. Specifically, the method calculates the similarity of the feature vector of the to-be-tested bulb under the same excitation condition with the high-risk events labeled in the database, and uses Mahalanobis distance or cosine similarity for quantitative comparison. If the steepness of the current drop slope exceeds the threshold, the infrared peak rise time is shorter than the typical melting precursor, and the acoustic emission high frequency energy ratio is abnormal, it is determined as a high-risk breaking event, and the corresponding risk level warning is output, and the hierarchical alarm mechanism is triggered; When the similarity index exceeds the preset threshold multiple times in a row, and the consistency of multimodal features deviates from the normal range, the risk level will be automatically increased, and an early warning signal will be sent to the control system to start the protection program, cut off the laser excitation source and record the experimental parameters to avoid equipment damage. After the aforementioned process, as the amount of data increases, during subsequent bulb use, it is only necessary to collect its electrical parameters, thermal response, and acoustic emission signals in real time. The trained model can then dynamically match the feature synchronization event pattern to achieve continuous monitoring and risk prediction of the filament status, thereby reducing the impact of bulb breakage on the equipment.

[0025] Example 2: Figure 2 As shown, this application also proposes an infrared heating lamp explosion-proof laser-induced experimental system, which operates the experimental method in Embodiment 1. The system includes: The data acquisition module is used to install at least three test bulbs into the experimental device, perform laser induction on each test bulb, and acquire the first data set through the acquisition device. The synchronization determination module is used to filter out the characteristic electrical change signal and thermal response change signal that occur when the bulb breaks based on the first data set, and to make a synchronization determination based on the time coupling relationship between the characteristic electrical signal and the thermal response change signal. The multimodal data fragment extraction module is used to extract multimodal data fragments before and after the occurrence of characteristic synchronization events based on the synchronization determination results, and to establish a characteristic synchronization event database. It packages and stores the electrical, thermal response and acoustic emission data within the time segment corresponding to each synchronization event, and marks the event type and occurrence time. The risk assessment module is used to train and learn features based on the feature synchronization event database, and to assess the risk of each light bulb under test for synchronization events, determining the risk level of breakage under laser-induced conditions.

[0026] Example 3: Figure 3 As shown, this application also proposes an explosion-proof laser-induced experimental device for infrared heating lamps. This device is used to fix an infrared laser source and can adjust the position of the infrared laser source to achieve directional irradiation of different bulbs. The experimental device includes a first slide 1, on which a ball screw 11 is provided. The ball screw 11 is connected to a drive motor (not shown in the figure), and the ball screw 11 is rotatably connected to a mounting platform 12. Under the drive of the drive motor, the mounting platform 12 can reciprocate along the axial direction of the ball screw 11. The mounting platform 12 is provided with a second slide 2, which has the same structure as the first slide 1. A fixing plate 3 is slidably provided on the second slide 2. Under the cooperative action of the first slide 1 and the second slide 2, the fixing plate 3 can be adjusted in the horizontal and vertical directions. The side surface of the fixed plate 3 is provided with a mounting groove 4 for mounting a laser lamp holder, and an infrared laser source is fixed in the laser lamp holder. When the irradiation position of the laser source needs to be adjusted, the ball screw 11 is driven to rotate by a driving motor, the mounting platform 12 is moved along the first sliding seat 1, and the position of the fixed plate 3 on the second sliding seat 2 is adjusted, so that the laser source is accurately positioned in the horizontal direction and the vertical direction. The structure of the laser lamp holder can refer to the prior art, and will not be described in detail in the embodiment.

[0027] The above only describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for an explosion-proof laser-induced experiment using an infrared heating lamp, characterized in that: The application relates to a method for evaluating the risk of lamp rupture under laser induction. The method comprises the following steps: At least three to-be-tested bulbs are installed into an experimental device, and laser induction is carried out on each to-be-tested bulb respectively, and a first data set is obtained through a collection device; Based on the first data set, characteristic electrical mutation signals and thermal response mutation signals occurring when the bulb ruptures are screened out, and synchronous determination is carried out according to the time coupling relationship of the characteristic electrical signals and the thermal response mutation signals; Based on the synchronous determination result, multi-modal data segments before and after the occurrence of a characteristic synchronous event are extracted, and a characteristic synchronous event database is established, the electrical, thermal response and acoustic emission data in the time segment corresponding to each synchronous event are packed and stored, and the event type and occurrence time are labeled; 2. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 1, characterized in that: Based on the characteristic synchronous event database, characteristic training and learning are carried out, and the risk of rupture of each to-be-tested bulb under laser induction is evaluated to determine the risk level of rupture. The first data set comprises the following data:

3. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 1, characterized in that: Electrical parameter change data of the to-be-tested bulb before and after the start of laser induction; time nodes of filament fusing and corresponding laser irradiation cumulative time; surface temperature distribution data of the filament at the moment of fusing; image or video data of the bulb at the moment of fusing or rupturing; acoustic signals or micro-pressure change data generated during laser induction; and deformation and crack propagation process data of the bulb shell under thermal stress. The synchronous determination process comprises the following steps: Time axis unification and sampling synchronization are carried out, and all original data from different sources are time-stamped corrected according to a unified time reference; Mutation points of various data in the first data set are extracted respectively, time proximity analysis is carried out on the mutation points, synchronous events caused by filament fusing and corresponding data are screened out, and a characteristic synchronous event sequence is generated; 4. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 3, characterized in that: Based on the characteristic synchronous event sequence, time domain and frequency domain characteristic parameters of various modal signals before and after the mutation moment are further extracted, including current drop slope, infrared radiation peak rise time, acoustic emission signal main frequency energy distribution, a multi-modal feature vector is constructed and normalized.

5. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 4, characterized in that: The mutation point is a step change of signal amplitude or a sharp rise of derivative, current sudden drop and voltage fluctuation in the electrical signal are detected through a difference algorithm based on a sliding window or a wavelet transform method, instantaneous jump of infrared radiation intensity in the thermal response data is identified, and burst pulses of high-frequency components in the acoustic emission signal are marked, and then time proximity analysis is carried out on various mutation points.

6. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 1, characterized in that: If the electrical mutation, thermal response mutation and acoustic signal burst simultaneously appear in a millisecond-level time window, the synchronous event caused by filament fusing is determined, and the data are collected according to time sequence to form a characteristic synchronous event sequence.

7. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 1, characterized in that: The characteristic synchronous event database supports index query according to time range and event type. When a combination mode with steep current drop slope, short infrared response rise time and high-intensity acoustic emission main frequency energy is identified, it is marked as a high-risk rupture precursor.

8. The explosion-proof laser-induced experiment method of an infrared heating lamp according to claim 1, characterized in that: After the feature training and feature learning are performed, the risk level of the bulb breakage is judged, the similarity of the feature vector of the to-be-tested bulb under the same excitation condition and the high-risk event labeled in the database is calculated, the Mahalanobis distance or cosine similarity is used for quantitative comparison, and it is determined whether it is a high-risk breakage event.

9. An explosion-proof laser-induced experiment system of an infrared heating lamp, used for implementing the explosion-proof laser-induced experiment method of the infrared heating lamp according to any one of claims 1 to 8, characterized in that: The method comprises the steps of: a data acquisition module is configured to install at least three to-be-tested bulbs into an experimental device, perform laser induction on each to-be-tested bulb, and obtain a first data set through a collection device; a synchronous judgment module is configured to filter out characteristic electrical mutation signals and thermal response mutation signals occurring when the bulb breaks based on the first data set, and perform synchronous judgment according to the time coupling relationship of the characteristic electrical signals and the thermal response mutation signals; a multi-modal data segment extraction module is configured to extract multi-modal data segments before and after the characteristic synchronous event based on the synchronous judgment result, establish a characteristic synchronous event database, pack and store the electrical, thermal response and acoustic emission data in the time segment corresponding to each synchronous event, and label the event type and occurrence time; a risk assessment module is configured to perform feature training and learning based on the characteristic synchronous event database, and perform risk assessment on the synchronous event of each to-be-tested bulb to judge the risk level of the bulb breakage under the laser induction condition.

10. An explosion-proof laser-induced experiment apparatus of an infrared heating lamp, characterized in that: An infrared heating lamp explosion-proof laser induction experiment method is used to implement the method according to any one of claims 1 to 8.

Images