Carbon dioxide laser safety protection method and system based on directional coupling

Abstract

The application discloses a kind of carbon dioxide laser safety protection method and system based on directional coupling, it is related to the field of device using stimulated emission, this method includes: by directional coupler, the forward signal of the power output of carbon dioxide laser and the reverse signal of load reflection are collected, forward voltage and reverse voltage are determined;According to forward voltage and reverse voltage, the reflection coefficient is calculated, and the voltage standing wave ratio is calculated according to the reflection coefficient;Audio data in the running process of carbon dioxide laser is collected by audio sensor, and audio feature is extracted;When the voltage standing wave ratio exceeds normal operation threshold, the first fault determination result is determined based on the voltage standing wave ratio;Audio feature is matched with preset fault feature library, and the second fault determination result is determined;Based on the first fault determination result and the second fault determination result, fault processing instruction is generated and the fault processing instruction is executed.The application can improve equipment maintenance efficiency.

Description

Technical Field

[0001] This application relates to the field of devices utilizing stimulated emission, and more particularly to a safety protection method and system for carbon dioxide lasers based on directional coupling. Background Technology

[0002] Radio frequency (RF) excited carbon dioxide lasers are commonly used in modern industrial processing. To ensure efficient energy transmission from the RF power supply to the laser resonator, the system must maintain good impedance matching. Any mismatch leading to reflected power will not only reduce processing efficiency but may also cause irreversible damage to core components such as the RF power supply. Therefore, effective monitoring of reflected power is the technical foundation for ensuring the stable operation of the laser.

[0003] In related technologies, a hardware protection circuit based on Voltage Standing Wave Ratio (VSWR) is employed. This scheme uses a directional coupler inserted between the RF power amplifier and the matching network to sample the forward transmitted power and the reflected power from the load in real time. The circuit calculates the real-time VSWR based on these two power values ​​and compares it with a pre-set, fixed hardware protection threshold. During laser operation, if the calculated VSWR exceeds this critical threshold, the protection circuit immediately triggers an action, rapidly cutting off the RF power supply output, thereby achieving hardware-level protection of the power supply.

[0004] However, different fault phenomena, such as the gas glow discharge in the resonant cavity turning into an arc discharge, or the optical lens deforming due to thermal effects caused by contamination, can lead to a sharp increase in the standing wave ratio and trigger a shutdown. Because maintenance personnel often struggle to determine the specific direction of troubleshooting, the equipment is shut down for extended periods, impacting production efficiency. Summary of the Invention

[0005] This application provides a safety protection method and system for carbon dioxide lasers based on directional coupling, which can improve equipment maintenance efficiency.

[0006] In a first aspect, this application provides a safety protection method for a carbon dioxide laser based on directional coupling, applied to a safety protection system. The method includes: acquiring the forward signal from the power supply output of the carbon dioxide laser and the reverse signal reflected from the load using a directional coupler to determine the forward and reverse voltages; calculating the reflection coefficient based on the forward and reverse voltages, and calculating the voltage standing wave ratio (VSWR) based on the reflection coefficient; acquiring audio data during the operation of the carbon dioxide laser using an audio sensor to extract audio features; determining a first fault determination result based on the VSWR when the VSWR exceeds the normal operating threshold; matching the audio features with a preset fault feature library to determine a second fault determination result; and generating and executing a fault handling instruction based on the first and second fault determination results.

[0007] In the above embodiments, the safety protection system monitors forward and reverse signals in real time through a directional coupler and combines the acoustic features collected by the audio sensor to form a dual fault determination based on voltage standing wave ratio and audio features. By combining voltage features and acoustic features, the accuracy and reliability of fault identification are improved, the probability of false positives and false negatives is effectively reduced, and the efficiency of equipment maintenance is improved.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of collecting audio data during the operation of a carbon dioxide laser using audio sensors and extracting audio features specifically includes: multiple audio sensors set based on the location of the discharge region of the carbon dioxide laser to collect sound signals from different locations; time synchronization and spatial positioning of the sound signals collected by the multiple audio sensors to obtain the source location of the sound signals; determining the spatial distribution characteristics of the sound signals based on the source location, and extracting the frequency and energy characteristics of each sound signal to generate audio features.

[0009] In the above embodiments, the safety protection system is equipped with multiple audio sensors. Through time synchronization and spatial positioning technology, it accurately determines the source location of the sound signal and constructs a complete acoustic feature map by combining frequency characteristics and energy characteristics. This improves the accuracy of fault location and enables the system to quickly identify the specific location and type of fault.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining a first fault determination result based on the voltage standing wave ratio (VSWR) when the VSWR exceeds the normal operating threshold specifically includes: calculating the rate of change of the VSWR and determining whether the fault type is sudden or gradual based on the rate of change; analyzing the fluctuation curve of the VSWR and determining whether the characteristic pattern is periodic fluctuation or random fluctuation based on the fluctuation curve; combining the rate of change and the fluctuation curve to determine the development stage and severity of the fault; and generating a first fault determination result that includes the fault type, development stage, and severity.

[0011] In the above embodiments, the safety protection system determines the fault type by the rate of change of voltage standing wave ratio, identifies characteristic patterns by fluctuation curves, and determines the development stage and severity of the fault by combining these characteristics. It can not only determine the urgency of the fault, but also predict the development trend of the fault.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, after generating a first fault determination result including fault type, development stage, and severity, the method further includes: using the fault type and development stage in the first fault determination result as filtering conditions to filter corresponding audio feature templates from a preset fault feature library; calculating the similarity between the audio features and the audio feature templates to obtain multiple sets of matching probabilities; performing threshold filtering on the multiple sets of matching probabilities to determine a fault type set, and combining the acoustic feature correlation of each fault in the fault type set to generate a second fault determination result.

[0013] In the above embodiments, the safety protection system establishes a feature template screening mechanism based on fault type and development stage. By calculating similarity and matching probability, it achieves accurate matching of fault features, can accurately identify fault types, and improves the accuracy of fault diagnosis.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of generating and executing a fault handling instruction based on a first fault determination result and a second fault determination result, the method further includes: determining the temporal relationship and consistency degree of the first fault determination result and the second fault determination result; determining the fault triggering order of multiple faults according to the temporal relationship, and determining the fault whose fault triggering order is located at a preset leading position as the dominant fault; determining the fault association strength of multiple faults based on the consistency degree, and determining the faults whose fault association strength is higher than a preset association threshold as common source faults; and establishing a fault development path diagram including the dominant fault and the common source faults.

[0015] In the above embodiments, the safety protection system establishes a correlation analysis mechanism between dominant faults and faults of the same origin by analyzing the temporal relationship and consistency of fault determination results. This mechanism can identify the propagation path of faults and improve the efficiency and accuracy of fault handling.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the step of generating and executing fault handling instructions based on the first fault determination result and the second fault determination result specifically includes: retrieving a basic handling scheme matching the dominant fault from a preset fault handling scheme library; supplementing the basic handling scheme with corresponding preventive handling measures according to the fault characteristics of the same source fault to generate a complete handling scheme; converting the complete handling scheme into a sequence of control instructions executable by the device, and setting an execution priority for each control instruction in the control instruction sequence; and executing each control instruction in the control instruction sequence in the order of execution priority.

[0017] In the above embodiments, the security protection system not only improves the efficiency of fault handling by retrieving matching basic processing schemes, supplementing preventive processing measures, and setting execution priorities, but also effectively prevents the occurrence of related faults and reduces equipment maintenance costs.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after executing each control instruction in the control instruction sequence in the order of execution priority, the method further includes: calculating the change in voltage standing wave ratio during the execution of the control instruction sequence; when the sum of the changes in three consecutive sampling periods exceeds a preset change threshold, reducing the time interval between adjacent control instructions according to the order of execution priority.

[0019] In the above embodiments, the safety protection system improves the adaptability and efficiency of fault handling by monitoring changes in voltage standing wave ratio and adjusting the execution strategy of control commands.

[0020] In a second aspect, embodiments of this application provide a security protection system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the security protection system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a security protection system, cause the security protection system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a security protection system, cause the security protection system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the security protection system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. By adopting a dual fault determination mechanism based on voltage standing wave ratio and audio characteristics, as well as a comprehensive monitoring method that collects and analyzes forward signals, reverse signals and acoustic characteristics in real time, the system can monitor the equipment status from both electrical and acoustic characteristics, enabling early detection and accurate judgment of faults. This effectively solves the problem of misjudgment and missed judgment that can easily occur when relying on a single parameter for fault judgment in existing technologies, thereby achieving high precision and high reliability in fault detection.

[0026] 2. By adopting a fault type judgment mechanism based on the rate of change of voltage standing wave ratio and a fault development assessment method combined with the characteristic pattern analysis of the fluctuation curve, the system can accurately identify the type, development stage and severity of the fault, establish a comprehensive fault profile, effectively solve the problem of difficulty in accurately judging the nature and development trend of the fault in the existing technology, and thus realize the precision and predictability of fault diagnosis.

[0027] 3. By adopting a fault correlation analysis mechanism based on time sequence and consistency, as well as a method for identifying dominant faults and common-source faults, the system can accurately grasp the causal relationship and degree of influence between faults, establish a complete fault development path diagram, effectively solve the problem of difficulty in identifying the correlation between multiple faults in existing technologies, and thus achieve systematic and targeted fault handling. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a safety protection method for a carbon dioxide laser based on directional coupling in an embodiment of this application.

[0029] Figure 2 This is another flowchart illustrating the safety protection method for carbon dioxide lasers based on directional coupling in the embodiments of this application;

[0030] Figure 3 This is a schematic diagram of the physical device structure of a security protection system in the embodiments of this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] For ease of understanding, the implementation scenarios and terms of the embodiments of this application are introduced below.

[0034] This application primarily addresses the safety monitoring and fault diagnosis of industrial-grade high-power radio frequency-excited carbon dioxide lasers, such as in precision machining applications like metal sheet cutting, welding, or non-metallic material engraving. In these scenarios, the laser operates at high power for extended periods, placing extremely high demands on the stability of its core components—the radio frequency power supply, resonant cavity, matching network, and optical lenses. This solution integrates a safety protection system between the radio frequency power supply and the matching network to achieve real-time monitoring.

[0035] To clearly explain this solution, we first define the core terms: A directional coupler is a microwave / RF device that can proportionally couple a small portion of power from the main transmission path without interference and can distinguish the direction of signal transmission, thus separating the "forward signal" from the power supply to the load and the "reverse signal" reflected back from the load to the power supply. Voltage Standing Wave Ratio (VSWR) is a dimensionless parameter characterizing the impedance matching degree in an RF system. Its value is equal to the ratio of the maximum to the minimum voltage on the transmission line; it is 1 for ideal matching, and the larger the value, the more severe the mismatch. It is directly related to the magnitude of the reflected power. Audio characteristics refer to a set of data that quantitatively describes the sound characteristics extracted by analyzing the acoustic signals generated during laser operation. For example, when a glow discharge transforms into an arc discharge, it produces a unique "buzzing" sound, with its spectrum mainly concentrated in the high-frequency region, while a gas leak may produce a low-pitched "hissing" sound. The spectral distribution, energy intensity, and time-domain envelope of these sounds together constitute the acoustic fingerprint for distinguishing different faults. The preset fault feature library is a pre-established database that stores standardized data templates of voltage standing wave ratio change patterns (such as change rate and fluctuation pattern) and audio characteristics (such as main frequency, harmonic components, and sound pressure level) of lasers under various known fault modes (such as electrode arcing, lens contamination, and abnormal gas ratio). These templates serve as the benchmark for subsequent fault matching and diagnosis.

[0036] To facilitate understanding, the method provided in this implementation will be described in detail below, using the above scenario as an example. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a safety protection method for a carbon dioxide laser based on directional coupling in an embodiment of this application.

[0037] S101. Acquire the forward signal from the power supply output of the carbon dioxide laser and the reverse signal reflected from the load using a directional coupler to determine the forward voltage and reverse voltage.

[0038] Among them, a directional coupler is a device used to measure the transmission characteristics of radio frequency signals, which can separate the forward and reverse signals in the transmission line; the forward signal represents the signal transmitted from the output of the radio frequency power supply to the load direction, and is used to characterize the output power; the reverse signal refers to the signal reflected back from the load to the power supply direction, and is used to characterize the impedance matching degree; the forward voltage and reverse voltage represent the effective voltage values ​​of these two signals, respectively, and are used to quantify the signal strength of transmission and reflection.

[0039] After the carbon dioxide laser is started up, the safety protection system needs to monitor its power output status and load reflection in real time. Specifically, the safety protection system first installs a directional coupler between the RF power output and the load, and then collects the forward and reverse signals from the transmission line through the coupling port of the directional coupler. For the forward signal, the safety protection system samples and converts the signal using a high-precision sampling circuit to obtain the forward voltage value representing the output power; simultaneously, the reverse signal undergoes the same sampling process to obtain the reverse voltage value representing the reflected power. The safety protection system updates the collected voltage values ​​in real time to maintain continuous monitoring of the laser's operating status.

[0040] In some embodiments, the acquisition and processing of forward and reverse signals can be achieved in various ways: Optionally, the safety protection system can adopt a dual-channel synchronous sampling scheme, sampling both signals simultaneously using a high-speed ADC, then performing digital signal processing to obtain the effective voltage value, and eliminating sampling noise through moving average filtering; Optionally, the safety protection system can adopt a peak detection scheme, converting the radio frequency signal to DC level using a radio frequency detector, then sampling the effective voltage value using a low-speed ADC, and eliminating the influence of temperature drift through a temperature compensation circuit. It is understood that other signal acquisition and processing methods can also be used to obtain the voltage value, and this is not limited here.

[0041] In practical applications, the coupling degree of a directional coupler may drift with changes in temperature and frequency, affecting the accuracy of voltage measurements. To address this, the safety protection system employs a self-calibration compensation mechanism: first, upon system startup, the directional coupler is calibrated using a standard signal source to establish a model relating coupling degree to temperature and frequency; during operation, the coupling degree compensation coefficient is dynamically calculated based on real-time temperature and operating frequency measurements; finally, the compensation coefficient is applied to the voltage measurement results to ensure the accuracy of the measured values. For example, when the ambient temperature rises from 25℃ to 45℃, the system can automatically compensate for approximately 0.5dB of coupling degree drift.

[0042] S102. Calculate the reflection coefficient based on the forward and reverse voltages, and calculate the voltage standing wave ratio based on the reflection coefficient.

[0043] Among them, the reflection coefficient is the ratio of the load reflected voltage to the incident voltage, which is used to characterize the degree of impedance matching; the voltage standing wave ratio (VSWR) is the ratio of the maximum voltage to the minimum voltage on the transmission line, which is a function of the reflection coefficient and is used to quantify the degree of impedance mismatch; the normal operation threshold is the upper limit of the VSWR when the equipment is operating normally, which is used to determine whether the equipment is in an abnormal operating state.

[0044] After acquiring the forward and reverse voltages, the safety protection system needs to calculate key parameters characterizing the impedance matching state. First, the system calculates the complex reflection coefficient Γ from the ratio of the reverse voltage to the forward voltage. The amplitude of this coefficient reflects the reflection intensity, and the phase reflects the impedance characteristics. Then, based on the amplitude of the reflection coefficient, the system calculates the voltage standing wave ratio (VSWR) using the formula (1+|Γ|) / (1-|Γ|). The system updates the calculation results in real time, continuously monitoring the changing trend of the impedance matching state.

[0045] Specifically, after obtaining the calibrated and filtered effective values ​​of the forward voltage V_f and the reverse voltage V_r, the system first calculates the magnitude of the voltage reflection coefficient, i.e., |Γ| = V_r / V_f. This ratio intuitively reflects the ratio of the reflected signal amplitude to the incident signal amplitude and is a core parameter for measuring the degree of impedance mismatch. It is worth noting that the voltage value here is obtained through the detection circuit at the coupling port of the directional coupler. It has a fixed coupling degree and detection coefficient relationship with the actual voltage on the main transmission line, but this coefficient is canceled out when calculating the ratio, so the measured voltage can be used directly for calculation. Subsequently, the system calculates the voltage standing wave ratio (VSWR) based on the magnitude of the reflection coefficient |Γ|. The mathematical formula for this calculation is VSWR = (1 + |Γ|) / (1 - |Γ|). This formula shows that when perfectly matched, V_r = 0, therefore |Γ| = 0, and the calculated VSWR = 1, i.e., the ideal state; when total reflection occurs, V_r = V_f, therefore |Γ| = 1, and the calculated VSWR approaches infinity. In practical engineering, to avoid division by zero errors or numerical instability when |Γ| is close to 1, the system sets an upper limit. For example, when |Γ| > 0.99, the VSWR is directly set to a very large preset value (such as 99.9). For example, if the forward voltage measured by the directional coupler is 5.0V and the reverse voltage is 1.2V, the system calculates the magnitude of the reflection coefficient as |Γ| = 1.2 / 5.0 = 0.24, and then calculates the voltage standing wave ratio VSWR as ≈ 1.63 (1 + 0.24) / (1 - 0.24).

[0046] S103. Acquire audio data during the operation of the carbon dioxide laser using an audio sensor, and extract audio features.

[0047] Among them, an audio sensor is a transducer that can convert sound signals into electrical signals and is used to collect the sound of equipment operation; audio features refer to various parameters extracted from the sampled data that can characterize the sound characteristics, including frequency features, energy features, and time domain features.

[0048] The safety protection system requires real-time monitoring of the acoustic characteristics of the laser during operation. Specifically, the system first deploys a high-sensitivity audio sensor array at key locations in the laser discharge region to ensure comprehensive acquisition of different types of fault sounds. The system samples the audio signals using a high-speed data acquisition card, with a sampling frequency of no less than 40kHz to ensure complete preservation of sound details. The system preprocesses and denoises the raw sampled data, then extracts multi-dimensional feature parameters such as spectral characteristics, energy distribution, and modulation characteristics using time-frequency analysis methods to form a complete acoustic feature vector.

[0049] It's important to note that audio feature extraction is a process of transforming the original sound waveform into a set of meaningful numerical vectors. These vectors need to accurately capture the essential differences between various fault sound sources. In this scheme, the system doesn't simply perform a Fourier transform; instead, it calculates a carefully designed set of multi-dimensional acoustic features. First, in the time domain, the system calculates the root mean square energy (RMS Energy) of the signal to reflect the overall loudness of the sound. It also calculates the zero-crossing rate, a feature highly effective in distinguishing between periodic sounds (such as equipment resonance) and non-periodic sounds (such as the hissing sound of a gas leak). Second, in the frequency domain, after obtaining the signal's spectrum through a short-time Fourier transform (STFT), the system extracts the spectral centroid, which indicates the "center of gravity" frequency of the spectral energy. The centroid of high-frequency discharge sounds is significantly higher than that of normal operating noise. Simultaneously, it extracts the spectral roll-off points—frequency points where the spectral energy is below a certain percentage (e.g., 85%)—to characterize the signal's bandwidth and high-frequency components. Furthermore, to capture the timbre perception characteristics of sound, the system also calculates Mel-frequency cepstral coefficients (MFCCs), typically extracting the first 13 coefficients. MFCCs effectively simulate the auditory characteristics of the human ear and have extremely high discriminative power for recognizing complex sound patterns caused by changes in mechanical vibration or discharge modes. Ultimately, these feature parameters from different dimensions (such as RMS energy, ZCR, spectral centroid, and the 13 MFCC coefficients) are combined into a high-dimensional feature vector to completely and quantitatively describe the currently acquired audio event.

[0050] In practical applications, environmental noise and equipment vibration may interfere with the acquisition of audio signals, affecting the accuracy of feature extraction. To address this, the safety protection system employs spatial filtering and adaptive noise reduction techniques: firstly, the spatial arrangement of multiple sensors creates the directionality of the sound field, suppressing interfering sounds from non-target directions; then, an adaptive algorithm is used to estimate the environmental noise characteristics in real time, separating the target sound component from the acquired signal; finally, a feature fusion algorithm is used to improve the robustness of feature extraction.

[0051] S104. When the voltage standing wave ratio exceeds the normal operating threshold, determine the first fault judgment result based on the voltage standing wave ratio.

[0052] Among them, the normal operation threshold refers to the upper limit of the voltage standing wave ratio under the safe operating state of the equipment, which is determined by the equipment technical specifications; the first fault judgment result includes information such as fault type, development stage and severity, which is used to characterize the overall state of the fault.

[0053] When a safety protection system detects an abnormal voltage standing wave ratio (VSWR), it needs to conduct in-depth analysis. Specifically, the system first compares the real-time monitored VSWR with a preset normal operating threshold (usually 1.5). If the threshold is exceeded, a fault analysis process is triggered. The system calculates the instantaneous rate of change and cumulative change of the VSWR to determine whether the fault is sudden or gradual. Simultaneously, the system performs time-series analysis on historical VSWR data, extracting fluctuation characteristics to identify whether periodic changes or random fluctuations exist. Combining these characteristics, the system generates a first fault determination result containing fault information.

[0054] In some embodiments, fault determination can be achieved in multiple ways: Optionally, the safety protection system can employ a rule-based determination method, comprehensively evaluating the fault state by setting multiple determination rules such as change rate threshold and fluctuation period threshold, and calculating the fault level according to the rule weights; Optionally, the safety protection system can employ machine learning methods, mapping the multidimensional features of voltage standing wave ratio to the fault type space by training a support vector machine model to achieve automatic classification and determination. It is understood that other data analysis or artificial intelligence methods can also be used to achieve fault determination, which are not limited here.

[0055] It should be noted that determining the first fault diagnosis result is not a simple listing of isolated indicators, but a process of quantifying and logically combining the dynamic characteristics of the voltage standing wave ratio (VSWR). The system first sets a normal operating threshold, for example, VSWR=1.5. When the VSWR value is detected to exceed this threshold, the fault analysis module is activated. To determine the fault type, the system calculates the first-order difference of the VSWR value over N consecutive sampling periods (e.g., N=10, sampling period is 1ms), i.e., the rate of change v(t) = (VSWR(t) - VSWR(t-1)) / Δt. If the peak value of the rate of change exceeds a high threshold (e.g., v_peak>0.5 / ms) within a short period of time (e.g., within 2ms), the fault type is initially determined to be "sudden," which usually corresponds to instantaneous breakdown between electrodes or arc discharge; conversely, if the peak value of the rate of change is not high, but its moving average value remains positive and steadily increases over a longer period of time (e.g., more than 100ms), it is determined to be "gradual," which may be related to the thermal lensing effect caused by optical lens contamination or slow gas leakage. Next, to analyze the characteristic patterns, the system performs a Fast Fourier Transform (FFT) on the VSWR time series data exceeding the threshold to analyze its power spectral density. If one or more significant narrowband peaks exist in the spectrum, it indicates that the fluctuations are periodic, possibly related to periodic vibrations caused by malfunctions in rotating equipment (such as cooling fans) or periodic instability of the power supply. If the power spectrum exhibits broadband characteristics and has no obvious peaks, it is determined to be random fluctuation, which is more consistent with stochastic processes such as gas discharge instability. Finally, the system integrates these analysis results to determine the development stage and severity. For example, a comprehensive risk index R = w1*|VSWR_max-1.5|+w2*|mean(v(t))| can be defined, where w1 and w2 are weighting coefficients. Based on the range of the risk index R (e.g., R<0.5 is "initial stage / minor", 0.5≤R<1.5 is "development stage / moderate", R≥1.5 is "dangerous stage / severe"), a structured first fault determination result is finally generated, the content of which is: {fault type: sudden, characteristic pattern: random fluctuation, development stage: dangerous, severity: severe, determination basis: {VSWR_max: 3.8, v_peak: 0.9 / ms}}.

[0056] S105. Match the audio features with the preset fault feature library to determine the second fault determination result.

[0057] Among them, the preset fault feature library refers to a database that stores various known fault audio feature templates; the matching process refers to calculating the similarity between the feature to be tested and the template feature; the similarity threshold is used to judge the degree of feature matching; the second fault judgment result contains a set of possible fault types and their corresponding confidence levels, which are used to characterize the fault judgment result based on acoustic features.

[0058] Safety protection systems require fault identification based on acoustic features. Specifically, the system first loads standard feature templates for various faults from a pre-set fault feature library. These templates contain multi-dimensional information such as frequency features, energy features, and time-domain features. The system employs a multi-level matching strategy: first, it performs coarse matching to filter out potential fault types; then, it performs fine matching on candidate faults, calculating the similarity scores for each feature dimension. The system integrates the matching results from each feature dimension, generates matching probabilities through weighted fusion, and sets a confidence threshold to filter out high-probability fault types, forming the second fault determination result.

[0059] The establishment of the pre-defined fault feature library is a systematic data acquisition and modeling process, not simply data entry. This process first requires reproducing various typical faults of the carbon dioxide laser in a controlled experimental environment, either through artificial guidance or by utilizing historical data. For example, this can be done by intentionally introducing contaminants to simulate lens contamination or adjusting the gas ratio to simulate gas anomalies. Under each fault mode, the safety protection system simultaneously records complete time-series data of the voltage standing wave ratio (VSWR) and multi-channel audio signals. Subsequently, the system applies the aforementioned analysis methods (such as calculating the rate of change and fluctuation patterns) to the recorded VSWR data to form a "VSWR feature template" for that fault. Simultaneously, the system extracts the aforementioned multi-dimensional feature vectors from the audio data and performs statistical averaging and variance analysis on the fault feature vectors from multiple experiments to form an "audio feature template" for that fault. Each template contains the mean vector and covariance matrix of the features, used to describe the typical values ​​and fluctuation ranges of the features under that fault mode. For example, the audio template for the "electrode arcing" fault might exhibit an extremely high spectral centroid and a wide spectral roll-off point, while its VSWR template would exhibit an extremely high instantaneous rate of change. The feature templates of all known faults are stored in a structured manner, along with detailed fault labels (including fault name, physical cause, severity level, etc.), which together constitute this rich preset fault feature library.

[0060] S106. Based on the first fault determination result and the second fault determination result, generate a fault handling instruction and execute the fault handling instruction.

[0061] The safety protection system needs to comprehensively analyze fault determination results from two dimensions and take corresponding measures. Specifically, the safety protection system first performs a consistency analysis on the first and second fault determination results to determine their degree of matching. When the two results are highly consistent, the system directly determines the fault type; when discrepancies exist, the system conducts in-depth analysis, identifying the dominant and related faults through the correlation and temporal sequence of fault characteristics. Based on the fault type and severity, the system retrieves matching basic processing solutions from a pre-set processing solution library, optimizes and adjusts them in conjunction with the current operating status, and generates a sequence of processing instructions containing specific control parameters. The system executes fault handling measures step by step according to the priority and dependency of the instruction sequence, monitors the processing effect in real time, and dynamically adjusts the processing strategy when necessary.

[0062] In some embodiments, fault handling instructions can be generated and executed in multiple ways: Optionally, the safety protection system can employ a rule-based processing scheme generation method, first establishing a mapping relationship between fault types and basic processing schemes, then adjusting processing parameters according to fault severity and equipment status, and finally generating an execution sequence containing specific control instructions; Optionally, the safety protection system can employ a reinforcement learning method, establishing a decision model of fault state-processing action-reward feedback to learn the optimal processing strategy and achieve adaptive optimization of the processing scheme. It is understood that other decision optimization or intelligent control methods can also be used to implement fault handling, which are not limited here.

[0063] In practical applications, some fault handling measures may fail due to equipment response delays or sudden changes in status. To address this, the safety protection system employs a dynamic feedback control mechanism: First, after executing each processing instruction, the system evaluates the processing effect by monitoring the changing trends of voltage standing wave ratio (VSWR) and audio characteristics. Then, based on the feedback results, the system dynamically adjusts the execution strategy of subsequent instructions, including modifying control parameters, adjusting execution timing, or switching to alternative solutions. Finally, successful processing experiences are recorded in a solution library for optimizing future processing strategies. For example, when a processing measure fails to produce an effect within the expected timeframe, the system automatically switches to an alternative solution and appropriately extends the monitoring time to ensure the fault is effectively handled.

[0064] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is another flowchart illustrating the safety protection method for a carbon dioxide laser based on directional coupling in this application embodiment.

[0065] S201. Acquire the forward signal from the power supply output of the carbon dioxide laser and the reverse signal reflected from the load using a directional coupler to determine the forward voltage and reverse voltage.

[0066] Referring to step S101, the safety protection system will monitor the laser power output status in real time.

[0067] S202. Calculate the reflection coefficient based on the forward and reverse voltages, and calculate the voltage standing wave ratio based on the reflection coefficient.

[0068] Referring to step S102, the safety protection system will calculate voltage parameters to assess the matching status.

[0069] S203. Acquire audio data during the operation of the carbon dioxide laser using an audio sensor, and extract audio features.

[0070] Referring to step S103, the safety protection system will analyze the acoustic characteristics of the laser operation.

[0071] In some embodiments, the safety protection system performs multi-point acoustic monitoring and analysis of the discharge area. Specifically, the safety protection system uses multiple audio sensors set up based on the location of the discharge area of ​​the carbon dioxide laser to collect sound signals from different locations; it performs time synchronization and spatial positioning on the sound signals collected by the multiple audio sensors to obtain the source location of the sound signals; it determines the spatial distribution characteristics of the sound signals based on the source location, and extracts the frequency and energy characteristics of each sound signal to generate audio features.

[0072] Among them, spatial distribution characteristics represent the propagation and attenuation patterns of sound signals in space; frequency characteristics refer to the spectral composition and dominant frequency distribution of sound signals; and energy characteristics are used to characterize the intensity and energy density distribution of sound signals.

[0073] After the carbon dioxide laser is started, the safety protection system needs to monitor the acoustic characteristics of the discharge area in real time. Specifically, the system first installs a high-sensitivity audio sensor array around the discharge tube in a grid layout, with the sensor spacing not exceeding 10cm to ensure complete coverage of the discharge area. The system uses a unified clock source to trigger synchronous sampling of all sensors, with the sampling frequency set above 100kHz to capture high-frequency fault sounds. By calculating the time difference and intensity ratio of multiple sensor signals, the system uses a triangulation algorithm to determine the sound source location. Based on the sound source location, the system establishes a sound field intensity distribution map and performs a Fast Fourier Transform on the signals collected by each sensor to extract spectral features in the 0-50kHz range. Simultaneously, the system calculates the energy distribution and temporal envelope features of each frequency band, ultimately generating a feature vector containing spatial, frequency, and energy dimensions.

[0074] In some embodiments, acoustic feature acquisition and processing can be achieved in multiple ways: Optionally, the safety protection system can employ beamforming technology for sound source localization. First, the sensor array is spatially calibrated and its sensitivity compensated. Then, a directional beam is formed using a delay-sum algorithm. The sound source location is obtained by scanning different spatial directions. Finally, complete features are extracted by combining sound pressure level calculation and spectral analysis. Optionally, the safety protection system can employ blind source separation technology for feature extraction. Alternating independent component analysis (ACI) algorithms are used to separate aliased sound signals. Cross-correlation analysis is used to determine the spatial relationship of the signals, and wavelet transform is combined to achieve accurate extraction of time-frequency features. It is understood that other acoustic signal processing methods can also be used for feature extraction, and this is not limited here.

[0075] It should be noted that, in order to achieve accurate spatial positioning of signals acquired by multiple audio sensors, the system first employs a hardware time synchronization scheme based on Network Time Protocol (NTP) or Precise Time Protocol (PTP) to ensure that the sampling clock error of all sensors is controlled within the microsecond level. This is a prerequisite for effective sound source localization. After obtaining the audio signal sequence s_i(t) (where i is the sensor number) synchronously acquired by each sensor, the system uses a localization algorithm based on Time Difference of Arrival (TDOA) to determine the sound source location. Specifically, when a faulty sound source generates a sound pulse at spatial location P(x, y, z), the time it takes for this pulse to arrive at sensors at different spatial locations P_i(x_i, y_i, z_i) is different. The system calculates the cross-correlation function R_ij(τ) = ∫s_i(t)s_j(t+τ)dt of the same sound event signal received by any two sensors (such as sensors i and j), and the delay time τ_ij corresponding to its peak value is the time difference between the arrival of the signal at these two sensors. Based on τ_ij and the speed of sound c in the mixed gas inside the laser, the distance difference between the sound source and the two sensors can be determined as Δd_ij = c * τ_ij. Mathematically, this defines a hyperboloid with the two sensors as foci, on which the sound source must lie. By selecting at least three non-collinear sensor pairs (e.g., sensors 1-2, 1-3, 1-4), the system can obtain multiple hyperboloid equations. The intersection of these hyperboloids is the unique spatial location P(x, y, z) of the sound source. This solution process can be completed using the least squares method or an iterative algorithm, thus precisely associating the abstract audio data with specific physical locations within the laser resonant cavity (such as near an electrode or on the surface of a lens).

[0076] S204. Calculate the rate of change of the voltage standing wave ratio and determine whether the fault type is sudden or gradual based on the rate of change.

[0077] Sudden faults refer to fault types where the voltage standing wave ratio changes in a very short time; gradual faults refer to fault types where the voltage standing wave ratio changes slowly.

[0078] Safety protection systems need to determine fault types by analyzing the changes in voltage standing wave ratio (VSWR). Specifically, the system first calculates the VSWR difference between adjacent sampling points and divides it by the sampling time interval to obtain the instantaneous rate of change. The system then performs statistical analysis on the rate of change over multiple consecutive sampling periods, calculating the average rate of change and the peak rate of change. When the peak rate of change exceeds a preset abrupt change threshold (e.g., 0.5 / ms), the system classifies it as a sudden fault; when the average rate of change is below this threshold but continues to rise, it is classified as a gradual fault. The system further subdivides the specific fault type by considering the fluctuation characteristics of the rate of change.

[0079] S205. Analyze the fluctuation curve of the voltage standing wave ratio and determine whether the characteristic pattern is periodic fluctuation or random fluctuation based on the fluctuation curve.

[0080] Among them, the fluctuation curve refers to the trend of voltage standing wave ratio changing over time; periodic fluctuations represent a change pattern with regular and repetitive characteristics; random fluctuations refer to irregular change patterns without obvious patterns; and characteristic patterns are used to characterize the overall characteristics of fluctuations.

[0081] Safety protection systems require in-depth analysis of the fluctuation characteristics of voltage standing wave ratio (VSWR). Specifically, the system first performs time-domain analysis on the fluctuation curve, extracting characteristic moments such as peaks, troughs, and zero-crossing points. Then, it performs frequency-domain analysis, calculating the power spectral density using Fast Fourier Transform (FFT) to identify the main frequency components. When characteristic frequencies are present and their energy is concentrated, the system classifies it as periodic fluctuation; when the spectral energy distribution is dispersed and there is no obvious dominant frequency, it is classified as random fluctuation. The system also calculates the autocorrelation function of the fluctuation to further verify the degree of periodicity.

[0082] In practical applications, fluctuation curves may simultaneously contain multiple characteristic patterns, increasing the difficulty of judgment. To address this, the safety protection system employs a hierarchical analysis strategy: first, the fluctuation curve is separated into trend and fluctuation components; then, principal component analysis is performed on the fluctuation components to extract the main characteristic patterns; finally, by setting an energy contribution rate threshold, the dominant fluctuation type is determined. For example, when the energy contribution rate of the periodic component exceeds 70%, the system will determine it as a periodic-dominant fluctuation pattern.

[0083] S206. Combine the rate of change and fluctuation curves to determine the development stage and severity of the fault.

[0084] The development stage includes three phases: initial, development, and danger, which characterize the degree of evolution of the fault; the severity is divided into three levels: minor, moderate, and severe, which quantifies the degree of harm caused by the fault; the comprehensive evaluation index is an evaluation system that transforms the rate of change and fluctuation characteristics into a unified quantitative standard.

[0085] Safety protection systems require multi-dimensional analysis to determine fault states. Specifically, the system first establishes an evaluation matrix based on the rate of change and fluctuation characteristics, using the magnitude of the rate of change and the degree of fluctuation as two main dimensions. The system determines the development stage based on the cumulative value of the rate of change; a continuously rising rate indicates that the fault is worsening. The severity is assessed based on the amplitude and frequency of fluctuations; more intense fluctuations indicate a more severe fault. The system also considers the fault's duration and development trend to comprehensively determine the final evaluation result.

[0086] In practical applications, the evaluation criteria for different types of faults may differ, affecting the accuracy of the evaluation. To address this, the safety protection system employs an adaptive evaluation mechanism: first, it selects the appropriate evaluation template based on the fault type; then, it dynamically adjusts the evaluation parameters based on historical data; and finally, it ensures the reliability of the evaluation results through consistency verification of multiple evaluations. For example, for gradual-change faults, the system appropriately reduces the weight of the rate of change and increases the influence of fluctuation characteristics.

[0087] S207. Generate a first fault determination result that includes fault type, development stage, and severity.

[0088] The first fault determination result is a structured dataset containing a complete feature description of the fault; the determination confidence level is used to indicate the reliability of the determination result; the feature correlation degree indicates the logical relationship between the features; and the determination timestamp is used to record the precise moment of the fault determination.

[0089] The security protection system needs to integrate fault analysis results into standardized judgment results. Specifically, the security protection system first constructs a fault information template, including descriptive fields for three main dimensions: fault type, development stage, and severity. The system then fills the corresponding fields with the results obtained from the preliminary analysis and adds auxiliary information such as timestamps and confidence levels. For the judgment results of each dimension, the system attaches specific judgment criteria and key characteristic parameters. The system also analyzes the logical relationships between the results of each dimension to ensure the internal consistency of the judgment results, ultimately generating a complete fault judgment report.

[0090] In some embodiments, the judgment result can be generated in multiple ways: Optionally, the security protection system can use a template matching method to generate a standardized judgment result by combining a preset fault description template with actual feature parameters, and enrich the fault description through text generation technology; Optionally, the security protection system can use a knowledge graph method to construct an association network of fault features and generate multi-level judgment results through a reasoning mechanism. It is understood that other information organization methods can also be used to generate the judgment result, which are not limited here.

[0091] In practical applications, multi-dimensional judgment results may suffer from information redundancy or inconsistency. To address this, the security protection system employs an information optimization mechanism: first, it prioritizes the information across all dimensions, highlighting key features; then, it eliminates logical contradictions through rule-based checks; and finally, it extracts the core content using an information compression algorithm, ensuring the simplicity and accuracy of the judgment results. For example, when certain feature parameters are found to contribute little to the judgment result, the system automatically categorizes them as auxiliary information.

[0092] In some embodiments, the safety protection system performs multi-dimensional matching based on fault characteristics. Specifically, the safety protection system uses the fault type and development stage in the first fault determination result as a filtering condition to select corresponding audio feature templates from a preset fault feature library; calculates the similarity between the audio features and the audio feature templates to obtain multiple sets of matching probabilities; performs threshold filtering on the multiple sets of matching probabilities to determine the fault type set; and combines the acoustic feature correlation of each fault in the fault type set to generate a second fault determination result.

[0093] Among them, the audio feature template refers to the standard pattern of various typical fault acoustic features that are stored in advance; the acoustic feature correlation indicates the degree of correlation between different fault sound features; and the fault type set refers to the combination of possible fault types obtained after screening.

[0094] After obtaining the initial fault determination result, the safety protection system needs to perform precise acoustic feature matching. Specifically, the system first retrieves relevant audio feature templates from the fault feature database based on the fault type and development stage in the initial determination result. For each candidate template, the system calculates its Euclidean distance and cosine similarity with the measured audio features in three dimensions: spectrum, energy, and spatial distribution. The system converts the similarity values ​​into standardized matching probabilities using a Softmax function and sets a probability threshold of 0.75 for initial screening. For fault types that pass the threshold screening, the system further analyzes the acoustic feature correlation coefficient matrix between them to identify correlated fault combinations. Finally, based on the matching probability and correlation analysis results, the system generates a second fault determination result that includes the fault type, confidence level, and correlation strength.

[0095] In practical applications, acoustic characteristics may differ under different operating conditions, affecting the accuracy of feature matching. To address this, the safety protection system employs a dynamic template update mechanism: first, feature templates are categorized by operating condition, establishing a template library with multiple sets of condition-related parameters; then, through online learning methods, feature templates for each operating condition are continuously accumulated and updated; finally, an adaptive threshold strategy is used to dynamically adjust matching parameters based on the operating condition characteristics. For example, when the laser power is adjusted from 2kW to 3kW, the system can automatically switch to the template library for the corresponding power level and increase the matching threshold by 10% to accommodate feature deviations.

[0096] S208. Match the audio features with the preset fault feature library to determine the second fault determination result.

[0097] Referring to step S105, the safety protection system will identify the fault type based on audio characteristics.

[0098] S209. Determine the timing relationship and consistency between the first fault determination result and the second fault determination result.

[0099] Among them, the temporal relationship represents the order and overlap of the two judgment results in the time dimension; the consistency degree is used to quantify the degree of matching between the two judgment results in terms of fault type, characteristics, etc.

[0100] The security protection system needs to perform correlation analysis on the judgment results from two dimensions. Specifically, the system first compares the timestamps of the two judgment results to determine their temporal relationship, including whether they occurred simultaneously, their order, and the time interval. Then, the system performs consistency calculations on various feature dimensions of the judgment results, including the matching degree of fault types and differences in severity. The system also analyzes whether the changing trends of the two results are synergistic, assessing their correlation in the fault development process. Finally, the system integrates the temporal and consistency characteristics to generate a complete correlation analysis result.

[0101] In some embodiments, association analysis can be implemented in multiple ways: Optionally, the security protection system can employ time series analysis methods to evaluate time-series correlation by calculating cross-correlation functions and combining the cosine similarity of feature vectors to calculate the degree of consistency; alternatively, the security protection system can employ probabilistic graphical models to establish conditional dependencies between judgment results and evaluate the correlation strength through inference algorithms. It is understood that other data analysis methods can also be used to implement association analysis, and this is not limited here.

[0102] In practical applications, the timing of two decision results may deviate due to acquisition delays or processing latency. To address this, the security protection system employs a time synchronization mechanism: first, clock calibration ensures the time consistency of the two decision modules; then, a dynamic time window is set to compensate for acquisition and processing delays; finally, methods such as moving averages are used to smooth timing characteristics and improve the stability of correlation analysis. For example, when a timing difference within 5ms is detected, the system treats it as a simultaneous event.

[0103] S210. Determine the fault triggering order of multiple faults according to the timing relationship, and determine the fault that is located at the top of the preset fault triggering order as the dominant fault.

[0104] Among them, the fault triggering order refers to the order in which multiple faults occur; the preset front position represents the key sorting threshold in the triggering order; the dominant fault refers to the core fault that plays a leading role in the fault chain; and the time sequence weight is used to quantify the importance of different time positions.

[0105] Safety protection systems require time-series analysis to determine the causal relationships of faults. Specifically, the system first sorts all fault determination results by timestamp to establish a complete fault trigger sequence. The system calculates the temporal position weight for each fault in the sequence, with the weight decreasing as the sequence number increases. When a fault's trigger time is at the beginning of the sequence (e.g., within the first 20%), and multiple related faults subsequently occur, the system identifies it as the dominant fault. The system also analyzes the time interval distribution between faults and identifies fault nodes with initiating effects through cluster analysis.

[0106] In some embodiments, the identification of dominant faults can be achieved in several ways: Optionally, the security protection system can employ causal network analysis to construct a fault propagation graph, calculate the out-degree and influence range of nodes, and comprehensively assess the dominance of the fault; Optionally, the security protection system can employ time-series pattern mining to identify frequently occurring fault sequence patterns and determine key triggering nodes through statistical analysis. It is understood that other data mining methods can also be used to identify dominant faults, and this is not limited here.

[0107] In practical applications, some faults may exhibit spurious timing relationships due to differences in sensor response characteristics. To address this, the safety protection system employs a response time compensation mechanism: first, it establishes detection delay models for different types of faults; then, it compensates for the delay in the original timing data; and finally, it ensures the reliability of timing judgments through confidence interval analysis. For example, when the response delay of the audio sensor is 5ms slower than that of the voltage detection, the system will compensate accordingly for audio-related fault times.

[0108] S211. Determine the fault association strength of multiple faults based on the degree of consistency, and determine the faults with a fault association strength higher than the preset association threshold as faults of the same origin.

[0109] Among them, fault correlation strength refers to the degree of mutual influence between different faults; preset correlation threshold represents the standard value for judging faults of the same origin; faults of the same origin refer to a group of related faults caused by the same root cause; feature similarity is used to quantify the degree of matching of fault features.

[0110] Safety protection systems need to identify fault groups with common root causes through correlation analysis. Specifically, the system first calculates the similarity matrix between fault feature vectors, including the matching degree of voltage and acoustic features. The system sets multiple levels of correlation features, such as feature similarity, temporal correlation, and spatial distribution, and assigns different weights to them. When the overall correlation strength exceeds a preset threshold (e.g., 0.8), the system classifies the related faults into a common-source fault group. The system also analyzes the characteristic evolution patterns of the common-source fault group and establishes a fault propagation model.

[0111] To objectively quantify the consistency between the first and second fault determination results, and thus determine the correlation strength of the faults, the system employs a weighted scoring model based on multi-dimensional feature fusion. First, the system extracts comparable features from the two determination results, primarily including fault type and severity. For the consistency of fault types, the system queries a predefined fault type correlation matrix M, where the value of M(i, j) represents the correlation probability (range 0-1) between fault i determined by VSWR and fault j determined by audio in terms of physical cause. For example, the correlation probability between arc discharge (VSWR determination) and high-frequency discharge sound (audio determination) is close to 1. The consistency score is Score_type = M(type1, type2). For severity, the system maps its levels (e.g., mild, moderate, severe) to numerical values ​​(e.g., 1, 2, 3), and then calculates the absolute value of the difference after normalization as a measure of inconsistency. The consistency score can be defined as Score_severity = 1 - |level1 - level2| / (max_level - 1). Finally, the Correlation Strength is calculated by weighted summation of these individual scores: Correlation Strength = w_type * Score_type + w_severity * Score_severity, where w_type and w_severity are weighting coefficients (e.g., w_type = 0.6, w_severity = 0.4), reflecting the importance of different features in the correlation determination. When the calculated Correlation Strength is higher than a preset correlation threshold (e.g., 0.8), the system determines that the two anomalies detected by different sensor systems point to the same physical root cause, i.e., they are common source faults. For example, if the VSWR determines it as a sudden / severe fault, and the audio determines it as a high-frequency discharge sound / severe, their type correlation is 0.95, and their severity is completely consistent. The final correlation strength is 0.6 * 0.95 + 0.4 * 1.0 = 0.97, which is much higher than the threshold of 0.8, thus confirming them as common source faults.

[0112] S212. Establish a fault development path diagram that includes the dominant fault and the faults of the same origin.

[0113] Among them, the fault development path diagram refers to the directed graph structure that describes the fault evolution process; the node relationship represents the way faults propagate and affect each other; the path weight is used to quantify the probability of fault propagation; and the time sequence label is used to record the time process of fault development.

[0114] A comprehensive fault development model is required for a security protection system. Specifically, the system first establishes a tree structure with the dominant fault as the root node, organizing faults of the same origin into different levels according to their timing and correlation strength. The system labels each propagation path with attributes such as propagation delay and impact level, and calculates the path's reliability. For cases with multiple propagation branches, the system analyzes the development characteristics of each branch and identifies critical propagation paths. The system also verifies the rationality of the paths based on historical data, optimizing and updating the structure as necessary.

[0115] It's important to note that the fault development path graph is internally stored and processed as a directed weighted graph. In this graph, each node represents an identified independent fault event (whether a dominant fault or a related fault). The node itself stores detailed information about the fault, such as fault type, severity, timestamp of occurrence, and all details of the first and second fault determinations. Directed edges in the graph represent causal or correlational relationships between faults, with the edge pointing from cause to effect, for example, from a dominant fault to a related fault it triggers. Each edge is accompanied by multiple weights to quantify this correlation. These weights include: 1) Time Lag: the time difference between the endpoint node (fault) and the origin node (fault) of the edge; 2) Correlation Strength: the calculated value (between 0 and 1) that combines VSWR and audio feature consistency; 3) Transition Probability: based on historical data statistics, the probability that the originating fault will trigger the endpoint fault. Through this structured representation, the system can not only clearly depict the chain of fault propagation, but also perform complex queries and reasoning. For example, it can find key fault nodes by analyzing the topology of the graph, or predict the possibility of further development of the fault chain by calculating the weight product on the path.

[0116] In some embodiments, the path graph can be constructed in multiple ways: Optionally, the security protection system can employ graph theory analysis methods, using the minimum spanning tree algorithm to construct the fault propagation network and using graph traversal algorithms to analyze the characteristics of the propagation path; alternatively, the security protection system can employ Bayesian network methods to establish conditional probability relationships between faulty nodes and analyze fault propagation patterns through probabilistic reasoning. It is understood that other network analysis methods can also be used to construct the path graph, and this is not limited here. In practical applications, fault development paths may dynamically change due to changes in equipment status. To address this, the security protection system employs a dynamic path update mechanism: first, it verifies the effectiveness of existing paths through real-time monitoring; then, it updates the path structure based on newly added fault data; and finally, it visually displays the development trend through path visualization technology. For example, when a deviation is detected between the actual development of a propagation path and the expected one, the system will promptly adjust the path weights and structure.

[0117] S213. Based on the first fault determination result and the second fault determination result, generate a fault handling instruction and execute the fault handling instruction.

[0118] Referring to step S106, the safety protection system will execute an intelligent fault handling process.

[0119] In some embodiments, the safety protection system formulates a hierarchical processing strategy based on the fault type. That is, the safety protection system retrieves a basic processing scheme that matches the dominant fault from a preset fault processing scheme library; supplements the basic processing scheme with corresponding preventive processing measures according to the fault characteristics of the same source fault to generate a complete processing scheme; converts the complete processing scheme into a sequence of control instructions that can be executed by the device, and sets the execution priority for each control instruction in the sequence of control instructions; and executes each control instruction in the sequence of control instructions in the order of execution priority.

[0120] Among them, the basic handling plan refers to the standard handling process for a single fault type; the preventive handling measures refer to the additional control measures taken to prevent the spread of related faults; and the complete handling plan refers to the overall solution that includes the main fault handling and preventive measures.

[0121] After identifying the dominant fault and related faults, the safety protection system needs to develop a systematic handling plan. Specifically, the system first retrieves the most relevant basic handling plan from its fault handling plan database based on the type and development stage of the dominant fault. For the retrieved basic plan, the system analyzes the characteristic parameters and development trends of the related faults, supplementing necessary preventative measures, such as adjusting control parameters and adding monitoring points. The system then decomposes the complete handling plan into specific control commands, including power adjustment commands, temperature control commands, and air pressure adjustment commands. For each command, the system comprehensively scores it based on its importance to fault control, execution timeliness, and operational dependence, assigning an execution priority level of 1-5. Finally, the system strictly follows the priority order, executing commands one by one through the equipment control interface and monitoring the execution effect in real time.

[0122] In practical applications, the execution of certain control commands may cause drastic fluctuations in equipment parameters, affecting system stability. To address this, the safety protection system employs a gradual execution strategy: first, it assesses the risk level of control commands to identify critical commands that could lead to drastic fluctuations; then, it breaks these commands down into multiple smaller steps, achieving a smooth transition by setting buffer times and transition parameters; finally, it dynamically adjusts execution parameters through real-time feedback control. For example, when it is necessary to reduce RF power by 50%, the system divides the adjustment process into five steps, adjusting by 10% each step, and ensuring that the system parameters stabilize within the allowable range after each adjustment before proceeding to the next step.

[0123] In some embodiments, the safety protection system adjusts the fault handling parameters, that is, the safety protection system calculates the change in voltage standing wave ratio during the execution of the control command sequence; when the sum of the changes in three consecutive sampling periods exceeds a preset change threshold, the time interval between adjacent control commands is reduced according to the execution priority.

[0124] In some embodiments, the safety protection system also adjusts the processing timing, namely, calculating the change amplitude of audio features over time, calculating the attenuation period of audio features based on the point when the voltage standing wave ratio drops to within the normal operating threshold, comparing the attenuation rate of audio features corresponding to the same source fault with the point when the audio features corresponding to the dominant fault attenuate to a preset reference value, and supplementing the associated timing in the fault development path diagram; and generating timing feature data of the complete processing scheme based on the attenuation rate change points of audio features and the corresponding audio amplitude, and writing it into the fault processing scheme library.

[0125] Among them, the decay period represents the time required for the intensity of the audio feature to decrease to a specific level; the decay rate describes how quickly the fault feature weakens over time; and the time-series feature data refers to the data set describing each key time node in the fault handling process.

[0126] During the execution of fault handling commands, the safety protection system needs to dynamically optimize its execution strategy and accumulate handling experience. Specifically, the safety protection system first monitors the voltage standing wave ratio (VSWR) with a sampling period of 10ms, calculates the difference between adjacent sampling points, and performs sliding accumulation. When the accumulated value of three consecutive cycles exceeds a preset threshold (e.g., ±0.3), the system shortens the execution interval of adjacent commands by 20%, accelerating the processing pace.

[0127] Simultaneously, the system records real-time changes in audio characteristics. When the voltage standing wave ratio (VSWR) first falls back to the normal range (e.g., below 1.5), the system marks this moment and calculates the attenuation process of the audio characteristics from the peak value to that moment. The system focuses on the time point when the dominant fault characteristic attenuates to a reference value (e.g., 30% of the peak value), and uses this as a benchmark to analyze the relative attenuation rate of characteristics of the same fault source, establishing a time-series relationship model for fault elimination. Finally, the system extracts the inflection point time and corresponding amplitude of the attenuation curve to form a standardized time-series characteristic description, which is stored in the solution library as processing experience.

[0128] In some embodiments, the optimization and experience accumulation of the processing can be achieved in multiple ways: Optionally, the safety protection system can employ adaptive control methods for execution optimization, by establishing a system response model to predict parameter change trends, dynamically adjusting the command interval in conjunction with a model predictive control algorithm, and simultaneously using fuzzy control rules to handle special operating conditions, ultimately achieving precise adjustment of the processing; Optionally, the safety protection system can employ deep reinforcement learning methods to accumulate processing experience, by constructing a temporal deep network to learn fault evolution patterns, optimizing processing strategies using an experience replay mechanism, and expanding the applicability of the model through transfer learning methods. It is understood that other intelligent control or machine learning methods can also be used to optimize the processing, and this is not limited here.

[0129] In practical applications, the interaction between different faults can lead to nonlinear changes in attenuation characteristics, affecting the accuracy of time-series feature extraction. To address this, the safety protection system employs a multi-scale analysis strategy: first, wavelet decomposition is performed on the attenuation curve to separate the variation characteristics at different time scales; then, singular value decomposition is used to identify the main change patterns and eliminate the influence of random fluctuations; finally, a feature extraction model based on piecewise linear approximation is established to accurately capture key time nodes in the attenuation process. For example, when the attenuation process of a fault exhibits multiple inflection points, the system can identify the most representative change characteristics and use them as key descriptive points for the time-series features.

[0130] In this embodiment, by employing a dual fault determination mechanism based on voltage standing wave ratio (VSWR) and audio characteristics, combined with a correlation analysis method for dominant and co-originating faults, the safety protection system can simultaneously monitor equipment status from both electrical and acoustic dimensions. This allows for accurate identification of fault types and development trends, effectively solving the problems of misjudgment and missed judgment in existing technologies, as well as the difficulty in analyzing fault correlation and propagation paths. Consequently, it achieves high precision and reliability in fault detection, as well as systematic and targeted fault handling. This fault monitoring and handling method improves the accuracy of fault judgment through complementary feature verification and correlation analysis. Furthermore, by establishing a complete fault development path map and a dynamically optimized handling mechanism, it enables rapid fault location and precise handling, providing strong protection for the safe and stable operation of equipment.

[0131] The security protection system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 3 This is a schematic diagram of the physical device structure of a security protection system in an embodiment of this application.

[0132] It should be noted that, Figure 3 The structure of the security protection system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0133] like Figure 3 As shown, the security protection system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in ROM 302 or a program loaded into RAM 303 from storage section 308, such as executing the methods described in the above embodiments. RAM 303 also stores various programs and data required for system operation. CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. I / O interface 305 is also connected to bus 304.

[0134] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including hard disks, etc.; and communication section 309 including network interface cards such as LAN (Local Area Network) cards, modems, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0135] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0136] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0137] Specifically, the safety protection system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the carbon dioxide laser safety protection method based on directional coupling provided in the above embodiment.

[0138] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the security protection system described in the above embodiments; or it may exist independently and not assembled into the security protection system. The storage medium carries one or more computer programs that, when executed by a processor of the security protection system, cause the security protection system to implement the directional coupling-based carbon dioxide laser security protection method provided in the above embodiments.

[0139] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0140] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

Claims

1. A safety protection method for a carbon dioxide laser based on directional coupling, characterized in that, The method, applied to a security protection system, includes: The forward signal from the power supply output of the carbon dioxide laser and the reverse signal reflected from the load are collected by a directional coupler to determine the forward and reverse voltages. The reflection coefficient is calculated based on the forward voltage and the reverse voltage, and the voltage standing wave ratio is calculated based on the reflection coefficient. Audio data during the operation of the carbon dioxide laser is collected using an audio sensor, and audio features are extracted. When the voltage standing wave ratio (VSWR) exceeds the normal operating threshold, a first fault determination result is determined based on the VSWR. Specifically, this step includes: calculating the rate of change of the VSWR and determining whether the fault type is sudden or gradual based on the rate of change; analyzing the fluctuation curve of the VSWR and determining whether the characteristic pattern is periodic or random based on the fluctuation curve; combining the rate of change and the fluctuation curve to determine the development stage and severity of the fault; and generating a first fault determination result containing the fault type, the development stage, and the severity. The audio features are matched with a preset fault feature library to determine the second fault determination result; Based on the first fault determination result and the second fault determination result, a fault handling instruction is generated and executed.

2. The method according to claim 1, characterized in that, The step of collecting audio data during the operation of the carbon dioxide laser using an audio sensor and extracting audio features specifically includes: Multiple audio sensors, positioned based on the discharge region of the carbon dioxide laser, collect sound signals from different locations. The sound signals collected by the multiple audio sensors are synchronized in time and located spatially to obtain the source location of the sound signals; Based on the source location, the spatial distribution characteristics of the sound signal are determined, and the frequency and energy characteristics of each sound signal are extracted to generate audio features.

3. The method according to claim 1, characterized in that, After the step of generating a first fault determination result including the fault type, the development stage, and the severity, the method further includes: Using the fault type and development stage in the first fault determination result as filtering conditions, corresponding audio feature templates are filtered from the preset fault feature library; Calculate the similarity between the audio features and the audio feature template to obtain multiple sets of matching probabilities; Threshold filtering is performed on the multiple sets of matching probabilities to determine the set of fault types, and the acoustic feature correlation of each fault in the set of fault types is combined to generate a second fault determination result.

4. The method according to claim 1, characterized in that, Before the step of generating a fault handling instruction based on the first fault determination result and the second fault determination result and executing the fault handling instruction, the method further includes: Determine the temporal relationship and consistency between the first fault determination result and the second fault determination result; The fault triggering order of multiple faults is determined according to the timing relationship, and the fault that is located in the preset front position of the fault triggering order is determined as the dominant fault. Based on the degree of consistency, the fault association strength of multiple faults is determined, and faults with a fault association strength higher than a preset association threshold are identified as faults of the same origin. Establish a fault development path diagram that includes the dominant fault and the faults of the same origin.

5. The method according to claim 4, characterized in that, The step of generating a fault handling instruction based on the first fault determination result and the second fault determination result, and then executing the fault handling instruction, specifically includes: Retrieve a basic handling solution that matches the dominant fault from a pre-defined fault handling solution library; Based on the fault characteristics of the common source fault, corresponding preventive measures are added to the basic processing scheme to generate a complete processing scheme; The complete processing scheme is converted into a sequence of control instructions that can be executed by the device, and an execution priority is set for each control instruction in the sequence of control instructions. Each control instruction in the sequence of control instructions is executed in the order of the execution priority.

6. The method according to claim 5, characterized in that, After the step of executing each control instruction in the control instruction sequence according to the execution priority, the method further includes: Calculate the change in voltage standing wave ratio during the execution of the control command sequence; When the sum of the changes in three consecutive sampling periods exceeds a preset change threshold, the time interval between adjacent control commands is reduced according to the execution priority.

7. A safety protection system, characterized in that, The security protection system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the security protection system to perform the method as described in any one of claims 1-6.

8. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the security protection system, it causes the security protection system to perform the method as described in any one of claims 1-6.

9. A computer program product, characterized in that, When the computer program product is run on the security protection system, the security protection system performs the method as described in any one of claims 1-6.

Images