Pantograph arc detection system and interference suppression method thereof

Abstract

The application discloses a pantograph arc detection system and an interference suppression method thereof, wherein the detection system comprises a sealed metal shell, a high-voltage isolation bin and a signal acquisition bin which are independent of each other are arranged in the metal shell, a first isolation type power conversion module, a high-voltage driving module and a photomultiplier are fixedly arranged in the high-voltage isolation bin and are electrically connected in sequence, a power module and a second isolation type power conversion module are fixedly arranged in the signal acquisition bin and are electrically connected, a main control unit and a signal acquisition module which are electrically connected with the second isolation type power conversion module, the first isolation type power conversion module is electrically connected with the power module, the signal acquisition module comprises a silicon carbide ultraviolet detector, a signal conditioning module and an ADC conversion module which are electrically connected in sequence, and the ADC conversion module and the photomultiplier are electrically connected with the main control unit. The application is not only suitable for quantitative analysis of arc intensity, but also improves monitoring precision.

Description

Technical Field

[0001] This invention relates to the field of pantograph monitoring technology in rail transit, and in particular to a pantograph arc detection system and its interference suppression method. Background Technology

[0002] Subway trains primarily draw electrical energy from the train via the contact between the pantograph and the contact wire (or conductive rail). Therefore, the condition of the pantograph system is crucial for the normal operation of the train. Due to the high-speed movement of the train, the relative movement between the pantograph's carbon contactor and the contact wire (or conductive rail), coupled with wear or physical damage to the carbon contactor itself (such as cracks), or external forces such as train vibrations, can cause a discharge phenomenon between the carbon contactor and the contact wire. This phenomenon is also known as arcing. To monitor the train's operating status, arcing detection is necessary.

[0003] The existing pantograph arc detection methods mainly include the following three methods: I. Vision-based image detection: This method uses high-speed cameras or ultraviolet cameras to continuously monitor the area between the overhead contact line and the pantograph, capturing sparks, flashes, and arcing phenomena. The advantage of this approach is that it provides a relatively intuitive view of the arcing situation, facilitating manual analysis. The disadvantage is that it is affected by lighting conditions; in complex environments such as strong light, nighttime, or fog, the accuracy of arc detection decreases, and it can lead to many false alarms. II. Infrared Thermal Imaging Detection: Arcing is usually accompanied by high temperatures, so infrared thermal imaging equipment can be used to capture abnormal changes in the contact wire temperature to detect arcing. The advantage of this method is that it is unaffected by light and can accurately locate temperature anomalies. However, the disadvantage is a high false alarm rate; other high-temperature equipment around the contact wire (such as substations and cables) can also trigger false alarms. III. Current Waveform Analysis Based on Electrical Signals: Arcing generates electrical disturbances, causing abnormalities in the current or voltage waveforms of the contact network. Arcing can be detected by collecting and analyzing the waveform characteristics of the current and voltage. The advantages of this detection method are high accuracy, immunity to light and external interference, and real-time monitoring. The disadvantages are the need for complex signal processing, high equipment installation requirements, and significant modification difficulties.

[0004] The aforementioned solutions struggle to balance cost and accuracy. Chinese Patent CN 213649583 U discloses a solar-blind pantograph-catenary arc detection device. This device uses an ultraviolet photomultiplier tube (UVP) with a wavelength of 185–260 nm to monitor arcing. The UVP is powered by a 400VDC drive circuit, and arcing time and intensity are monitored through pulse calculation. The arc detection results are then transmitted to a host computer via a network. Using a solar-blind pantograph-catenary arc detection device avoids interference from visible light and various types of lighting, resulting in lower manufacturing costs and improved arc detection accuracy. However, the UVP used in this solution only outputs an electrical signal related to UV light intensity, which can determine the presence and relative strength of an arc, but cannot provide quantitative information. While the UVP has a "solar-blind" characteristic, it can still be affected by interference under extreme light conditions, requiring additional filters to ensure accuracy, resulting in poor reliability. Therefore, further optimization of the pantograph arc detection system is needed. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a pantograph arc detection system and its interference suppression method, which is not only suitable for quantitative analysis of arc intensity but also improves monitoring accuracy.

[0006] Internal interference in pantograph arc detection systems is a complex problem. Ordinary physical isolation inevitably leaves some residual coupling, which cannot meet the stringent safety requirements of rail transit. Moreover, in interference environments without proper physical isolation, the weak signal of the silicon carbide ultraviolet detector is easily submerged by high-voltage noise, resulting in a low signal-to-noise ratio. Software algorithms are unable to accurately extract effective signals from this noise, leading to low arc detection accuracy.

[0007] The technical solution to achieve the objective of this invention is: A pantograph arc detection system includes a sealed metal casing. Inside the metal casing are an independent high-voltage isolation chamber and a signal acquisition chamber. The high-voltage isolation chamber contains a first isolated power conversion module, a high-voltage drive module, and a photomultiplier tube, all electrically connected in sequence. The signal acquisition chamber contains a power module and a second isolated power conversion module, both electrically connected, as well as a main control unit and a signal acquisition module, both electrically connected to the second isolated power conversion module. The first isolated power conversion module is electrically connected to the power module. The signal acquisition module includes a silicon carbide ultraviolet detector, a signal conditioning module, and an ADC conversion module, all electrically connected in sequence. The ADC conversion module and the photomultiplier tube are both electrically connected to the main control unit.

[0008] Furthermore, the inner surface of the metal casing is subjected to black anodizing treatment.

[0009] Furthermore, the inner cavity of the metal shell is provided with an integrally formed partition, which divides the inner cavity of the metal shell into a high-pressure isolation chamber and a signal acquisition chamber. The wall thickness of the partition is not less than the wall thickness of the metal shell, and the thickness of the metal shell is 4mm.

[0010] Furthermore, the partition plate is provided with a wire-passing hole, and the wire-passing hole is filtered by a feedthrough capacitor.

[0011] Furthermore, the ADC conversion module is connected to the main control unit via an optocoupler-isolated SPI signal line.

[0012] Furthermore, the SPI signal line is a shielded cable, and the shielding layer of the shielded cable is grounded at both ends.

[0013] An interference suppression method for a pantograph arc detection system as described above includes the following steps: Step S1: Signal Acquisition: The main control unit synchronously acquires the pulse counting signal of the photomultiplier tube and the ADC sampling data of the silicon carbide ultraviolet detector to establish the time domain correspondence of the dual-channel data; Step S2: Interference Feature Extraction: During the system initialization phase, the self-generated oscillation characteristics of the high-voltage drive module of the photomultiplier tube are experimentally calibrated. Step S3: Time Domain Compliance Determination: The pantograph arc detection system is in normal working condition. At this time, the high voltage drive module and the external ultraviolet light source are both turned on. A time domain compliance determination window Δt is constructed. When the photomultiplier tube detects the pulse edge, the system automatically opens the determination window, identifies and eliminates residual interference signals, and retains the real arc event.

[0014] Furthermore, in step S2, the calibration method for the self-generated oscillation characteristics includes the following steps: Step S21: Turn off the high-voltage drive module of the photomultiplier tube and only collect the signal from the silicon carbide ultraviolet detector as the background noise reference; Step S22: Turn on the high-voltage drive module, but shield the external ultraviolet light source, so that the photomultiplier tube is in standby mode, and collect the signal of the silicon carbide ultraviolet detector channel. At this time, since there is no external ultraviolet light source, the signal collected by the silicon carbide ultraviolet detector comes entirely from the interference generated by the electromagnetic radiation coupling of the high-voltage drive module, which serves as an interference signal. Step S23: Perform FFT transformation on the acquired interference signal to extract the spectral and time-domain waveform features of the interference. The spectral features include the typical waveform width W of the interference pulse. noise and rise time TR noise The time-domain waveform features include a complete time-domain waveform template sequence, denoted as sequence Y = {y1, y2, …, y}. m}; Step S24: Based on the interference pulse waveform width W extracted in step S23 noise and rise time TR noise Determine the first preset threshold δ w =α×W noise Second preset threshold δ r =β×TR noise Wherein, α and β are allowable relative deviation coefficients, which are determined by stability testing of the high-voltage drive module within its normal operating temperature range.

[0015] Furthermore, the time-domain compliance determination includes the following steps: Step S31: Initial screening using national standard thresholds: Within the judgment window Δt, analyze the optical signal of the silicon carbide ultraviolet detector channel. If the peak signal intensity is less than the third preset threshold V... th Or the waveform width is less than the fourth preset threshold T w If the signal peak intensity is greater than or equal to the third preset threshold V, it is determined to be a non-arc event and is directly excluded; th And the waveform width is greater than or equal to the fourth preset threshold T w Then proceed to the next step: analyze the temporal characteristics of the optical signal in the silicon carbide ultraviolet detector channel; the third preset threshold V th and the fourth preset threshold T w The recommended value is V. th =25uw / cm 2 T w =5ms; Step S32: Fine screening of interference feature fingerprints. For optical signals that have passed the initial screening of national standard thresholds, a fine judgment step is performed. I. Refined comparison of waveform width; Calculate the full width at half maximum (FWHM) and rise time (T) of the optical signal waveform from 10% to 90%. rise The calculation results are compared with the typical waveform width W of the interference signal spectral characteristics extracted in step S23. noise Rising edge feature TR noise Perform a comparison; like: , and , If the signal is initially identified as a pseudo-signal generated by interference coupling, it will be eliminated. Otherwise, it indicates that there is a significant difference between the waveform characteristics of the measured signal and the interference fingerprint, and further determination will be made by waveform morphology correlation matching. II. Waveform morphology correlation matching; The optical signal waveform of the silicon carbide ultraviolet detector channel is cross-correlated with the time-domain waveform characteristics stored in step S2, and the correlation coefficient ρ is calculated: , Where, x i The measured optical signal waveform sequence is y. i This is an interference feature fingerprint template sequence. If ρ > preset correlation coefficient threshold ρ th , ρ th If ρ = 0.85, it is determined to be a pseudo-signal generated by interference coupling, and pseudo-signals are eliminated; if ρ ≤ ρ th If so, it will be retained as a real arcing event; Step S33: Output and processing of the judgment result: 1. Real arc preservation: Output arc alarm signal and record arc parameters, including arc peak intensity, arc duration, arc occurrence time and train current position, and store the above data in local database; 2. False signal rejection: The event is silently discarded without triggering an alarm, but the frequency and intensity of the interference event are recorded for dynamic monitoring of changes in the internal electromagnetic environment of the system. When the frequency of interference events increases abnormally, the system automatically issues a hardware health status warning, prompting maintenance personnel to check whether the high-voltage drive module and the shielding structure of the metal shell are abnormal.

[0016] By adopting the above technical solution, the present invention has the following beneficial effects: (1) In order to achieve high dynamic range detection, the present invention integrates a photomultiplier tube and a silicon carbide ultraviolet detector. The photomultiplier tube has extremely high sensitivity and is used to capture weak ultraviolet signals. The silicon carbide ultraviolet detector has good linear response and is suitable for quantitative analysis of arc intensity. It is insensitive to visible light and infrared light (i.e., "solar blindness") and does not need to rely on complex external filters to eliminate sunlight interference like the ultraviolet photomultiplier tube, thus reducing the false alarm rate from a physical perspective. At the same time, the strong electromagnetic radiation generated by the high voltage drive module of the photomultiplier tube during operation induces high-frequency interference noise at the weak signal acquisition end of the silicon carbide ultraviolet detector through spatial coupling. By encapsulating the two in two independent chambers in the same insulating metal shell, the chamber and the outside form a complete electromagnetic shield. The conductivity of the metal shell constitutes the Faraday cage effect, which confines the high voltage radiation source inside the chamber. This not only makes it more resistant to shock and vibration, but also avoids the false triggering judgment caused by the above problems, thus improving the monitoring accuracy.

[0017] (2) The inner cavity of the metal shell of the present invention is specially treated to enhance the surface insulation performance on the one hand, and reduce the electromagnetic wave reflection of the inner wall of the cavity on the other hand, thereby reducing residual radiation coupling.

[0018] (3) The metal shell and partition of the present invention achieve excellent shielding effect through sufficient wall thickness.

[0019] (4) The present invention suppresses the propagation of high-frequency interference along the wire between the high-voltage isolation chamber and the signal acquisition chamber by setting a feedthrough capacitor.

[0020] (5) The present invention isolates the SPI signal line by optical coupling to block the transmission of digital ground interference to analog ground.

[0021] (6) The SPI signal line of the present invention uses a shielded cable with the shield layer grounded, which further reduces spatial coupling interference.

[0022] (7) The present invention converts the power supply required for the high voltage drive module and the low noise power supply required for the signal acquisition module through two isolated DC-DC modules. The two power supplies are completely independent, the ground wires are separated, and they are connected at a single point only at the grounding point, which can effectively solve the problems of ground loop interference, high voltage surge and power supply noise.

[0023] (8) This invention cleverly utilizes the spatiotemporal consistency of dual-channel signals to solve the "artifact" problem that a single sensor cannot solve. Through time-domain waveform recognition, it cleverly avoids the problem that traditional filters have difficulty filtering out the frequency band overlap interference of the high-voltage power supply itself without damaging the signal. Since the high sensitivity of the photomultiplier tube is reserved for triggering, the system will not miss weak arcs due to setting a high threshold. At the same time, the secondary confirmation of the silicon carbide ultraviolet detector prevents false alarms of the photomultiplier tube, greatly improving the accuracy.

[0024] (9) This invention reduces the interference coupling strength of the photomultiplier tube to the silicon carbide ultraviolet detector by about 80% through a variety of hardware means such as cavity-type metal shell, black anodizing, feedthrough capacitor filtering, optocoupler isolation, and independent power supply. This changes the signal of the silicon carbide ultraviolet detector from "completely submerged by noise" to "detectable but with residual interference". At the same time, for the residual interference that cannot be eliminated by physical isolation, namely the self-generated oscillation characteristics of the high-voltage drive module, its characteristic fingerprint is calibrated by experiments during the system initialization stage. During operation, the waveforms that match the interference characteristic fingerprint in the silicon carbide ultraviolet detector signal are removed by time domain conformity judgment, while the real arc signal is retained. The two work together to form a complete anti-interference processing chain of physical suppression and feature recognition, realizing high-precision arc detection. Attached Figure Description

[0025] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein: Figure 1 This is a front view of the detection system of the present invention; Figure 2 This is a schematic diagram of the internal structure of the detection system of the present invention; Figure 3This is a schematic diagram showing the installation positions of the photomultiplier tube and the silicon carbide detector in the detection system of the present invention. Figure 4 This is a circuit diagram of the signal acquisition module of the detection system of the present invention; Figure 5 This is the time-domain compliance determination logic diagram of the present invention; Figure 6 This is a waveform diagram of the data before interference suppression in this invention; Figure 7 This is a waveform diagram of the data after interference suppression according to the present invention; Figure 8 This is a flowchart illustrating the physical isolation and algorithmic collaborative processing of this invention.

[0026] The labels in the attached diagram are: 1. Metal casing; 2. Partition; 3. High-voltage isolation chamber; 4. Signal acquisition chamber; 5. Photomultiplier tube; 6. Silicon carbide ultraviolet detector. Detailed Implementation

[0027] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0028] The pantograph arc detection system in this embodiment is as follows: Figures 1 to 3 As shown, the device includes a sealed metal casing 1. An integrally formed partition 2 divides the inner cavity of the metal casing 1 into two independent high-voltage isolation chambers 3 and signal acquisition chambers 4. The high-voltage isolation chamber 3 contains a first isolated power conversion module, a high-voltage drive module, and a photomultiplier tube 5, all electrically connected in sequence. The signal acquisition chamber 4 contains a power module and a second isolated power conversion module, both electrically connected, as well as a main control unit and a signal acquisition module, both electrically connected to the second isolated power conversion module. The first isolated power conversion module is electrically connected to the power module. The signal acquisition module includes a silicon carbide ultraviolet detector 6, a signal conditioning module, and an ADC conversion module, all electrically connected in sequence. The ADC conversion module and the photomultiplier tube are both electrically connected to the main control unit. The photomultiplier tube 5 has extremely high sensitivity for capturing weak ultraviolet signals; the silicon carbide ultraviolet detector 6 has good linear response and is suitable for quantitative analysis of arc intensity. Together, they enable monitoring of pantograph arcing.

[0029] Because photomultiplier tubes operate under high-voltage drives of hundreds of volts, their pulse discharges or the operation of the first isolated power conversion module used for high-frequency DC-DC boost conversion generate strong internal electromagnetic radiation; while the silicon carbide detector outputs a weak current signal in the nanoampere range. Due to the confined space of the entire system, the high-voltage electromagnetic field generated by the photomultiplier tube can easily couple into the signal conditioning module of the silicon carbide detector, forming false positive pulse signals. This internal interference originates from the inherent physical structure of the system, rather than external environmental interference, making it difficult to eliminate it without damaging the arcing characteristics using conventional filtering methods. This embodiment addresses the aforementioned internal interference by optimizing the system hardware structure and employing multi-dimensional circuit isolation and time-domain logic correlation analysis.

[0030] Specifically, in terms of hardware structure, the metal shell is made of a single piece of aluminum alloy through machining. The shell of the high-voltage isolation chamber 3 forms a complete electromagnetic shield. The conductivity of the metal shell constitutes the Faraday cage effect, confining the high-voltage radiation source inside the cavity. The signal acquisition chamber 4 and the high-voltage isolation chamber 3 are physically isolated by the wall thickness of the partition 2, where the thickness of both the partition and the metal shell is 4mm. The inner surface of the entire metal shell cavity is treated with black anodizing, which enhances the surface insulation performance and reduces electromagnetic wave reflection from the inner wall of the cavity, thereby reducing residual radiation coupling. At the same time, the partition 2 is provided with wire holes, which are filtered by feedthrough capacitors to suppress high-frequency interference propagating along the wires.

[0031] In the circuit design, optimizations were made from multiple angles, including analog front-end isolation, power supply isolation, and signal line shielding. First, the weak current signal output from the silicon carbide ultraviolet detector is converted into a voltage signal by the preamplifier circuit of the signal conditioning module, and then sent to the ADC conversion module for analog-to-digital conversion. The ADC conversion module communicates with the main control unit via an SPI interface. The SPI signal line is optocoupled to block the conduction of digital ground interference to analog ground. The system's total power supply is DC 12V, which is converted by the first and second isolated power conversion modules (two isolated DC-DC modules) to provide power for the high-voltage drive module (including boosting to several hundred volts) and low-noise power (±5V) for the signal acquisition module. The two power supplies are completely independent, with separate ground wires, connected only at a single point on the metal casing. The circuit schematic of the signal acquisition module is shown below. Figure 4 As shown. The SPI signal line leading from the signal acquisition cavity to the main control unit uses a shielded cable, with both ends of the shielding layer grounded to further reduce spatial coupling interference.

[0032] The aforementioned hardware isolation measures can reduce the interference coupling strength between the photomultiplier tube and the silicon carbide ultraviolet detector by more than 80%. However, due to the compact installation space on the train roof and the complex electromagnetic environment, a small amount of residual interference still exists, and the system's false alarm rate remains above 10%. Therefore, the detection system in this embodiment further introduces an algorithm based on signal feature correlation to eliminate residual interference, reducing the system's false alarm rate to below 0.1% and increasing the arc detection rate to over 95%.

[0033] Specifically, an interference suppression method for a pantograph arc detection system as described above includes the following steps: Step S1: Signal Acquisition: The main control unit synchronously acquires the pulse counting signal of the photomultiplier tube and the ADC sampling data of the silicon carbide ultraviolet detector to establish the time domain correspondence of the dual-channel data; Step S2: Interference Feature Extraction: During the system initialization phase, the self-generated oscillation characteristics of the high-voltage drive module of the photomultiplier tube are experimentally calibrated. The specific calibration method is as follows: Step S21: Turn off the high-voltage drive module of the photomultiplier tube and only collect the signal from the silicon carbide ultraviolet detector as the background noise reference; Step S22: Turn on the high-voltage drive module, but shield the external ultraviolet light source, so that the photomultiplier tube is in standby mode and the signal of the silicon carbide ultraviolet detector channel is collected. At this time, since there is no external ultraviolet light source, the signal collected by the silicon carbide ultraviolet detector comes entirely from the interference generated by the electromagnetic radiation coupling of the high-voltage drive module, which serves as an interference signal. Step S23: Perform FFT transformation on the acquired interference signal to extract the spectral and time-domain waveform features of the interference. The spectral features include the typical waveform width W of the interference pulse. noise and rise time TR noise The time-domain waveform features include a complete time-domain waveform template sequence, denoted as sequence Y = {y1, y2, …, y m}; Step S24: Based on the interference pulse waveform width W extracted in step S23 noise and rise time TR noise Determine the first preset threshold δ w =α×W noise Second preset threshold δ r =β×TR noiseWherein, α and β are the allowable relative deviation coefficients, which are determined by stability testing of the high-voltage drive module within its normal operating temperature range (-20℃ to +70℃). Experiments show that the maximum fluctuation of the self-generated oscillation characteristics (pulse waveform width and rise time) of the high-voltage drive module within this temperature range does not exceed ±15% of the calibrated value. To ensure a judgment margin and avoid misjudgments due to temperature fluctuations, 20% is taken as the judgment threshold, i.e., in this embodiment, α=β=0.2, meaning a relative deviation of 20% is allowed. For different models of high-voltage drive modules or different operating environments, W can be recalibrated during the system initialization phase. noise and TR noise And automatically calculate δ based on this. w and δ r To achieve adaptive decision-making; Step S3: Time Domain Compliance Determination. This step builds upon the aforementioned physical isolation measures. At this point, the pantograph arc detection system is operating normally, with both the high-voltage drive module and the external ultraviolet light source activated. Through physical isolation, the interference coupling strength between the photomultiplier tube and the silicon carbide ultraviolet detector is reduced by approximately 80%, transforming the output signal of the silicon carbide ultraviolet detector from a state of "completely submerged in noise" to a state of "detectable but still with residual interference." This transformation provides the prerequisite for subsequent processing. Without physical isolation preprocessing, the output signal of the silicon carbide ultraviolet detector would be completely covered by noise, making it impossible to extract effective information, thus hindering the effective operation of the algorithm.

[0034] This step primarily addresses residual interference that cannot be completely eliminated by physical isolation measures. This residual interference manifests as self-generated oscillations from the high-voltage drive module, as detailed below: A time-domain conformance judgment window Δt is constructed. When the photomultiplier tube detects a pulse edge, the system automatically opens the judgment window to identify and eliminate residual interference signals while retaining the actual arcing event. The specific judgment method is as follows: Figure 5 As shown, the content is as follows: Step S31: Initial screening using national standard thresholds: Within the judgment window Δt, analyze the optical signal of the silicon carbide ultraviolet detector channel. If the peak signal intensity is less than the third preset threshold V... th Or the waveform width is less than the fourth preset threshold T w If the signal peak intensity is greater than or equal to the third preset threshold V, it is determined to be a non-arc event and is directly excluded; th And the waveform width is greater than or equal to the fourth preset threshold T w Then proceed to the next step: analyze the temporal characteristics of the optical signal in the silicon carbide ultraviolet detector channel; the third preset threshold V th and the fourth preset threshold T w The recommended value is V. th=25uw / cm 2 T w =5ms; Step S32: Fine screening of interference feature fingerprints. For optical signals that have passed the initial screening of national standard thresholds, a fine judgment step is performed. I. Refined comparison of waveform width; Calculate the full width at half maximum (FWHM) and rise time (T) of the optical signal waveform from 10% to 90%. rise The calculation results are compared with the typical waveform width W of the interference signal spectral characteristics extracted in step S23. noise Rising edge feature TR noise Perform a comparison; like: , and , By comparing the width of the measured pulse with the width of the pre-stored interference pulse, and comparing the steepness of the rising edge of the measured pulse with the steepness of the rising edge of the pre-stored interference pulse, if the results are not significantly different, it is initially determined to be a pseudo signal generated by interference coupling, and pseudo signal elimination is performed; otherwise, it indicates that there is a significant difference between the waveform characteristics of the measured optical signal and the interference characteristic fingerprint, and further determination is made by waveform morphology correlation matching. II. Waveform morphology correlation matching; The optical signal waveform of the silicon carbide ultraviolet detector channel is cross-correlated with the time-domain waveform characteristics stored in step S2, and the correlation coefficient ρ is calculated: , Where, x i The measured optical signal waveform sequence is y. i This is an interference feature fingerprint template sequence. If ρ > preset correlation coefficient threshold ρ th , ρ th If ρ = 0.85, it is determined to be a pseudo-signal generated by interference coupling, and pseudo-signals are eliminated; if ρ ≤ ρ th If so, it will be retained as a real arcing event; The national standard threshold is only used as a preliminary screening condition. Its purpose is to quickly pre-filter the continuous signal stream, eliminating event signals that clearly do not conform to the characteristics of arcing, such as environmental noise and non-continuous micro-discharge signals. This reduces the data size entering the fine screening stage and reduces the computational burden on subsequent algorithms. The national standard threshold is not used as the final basis for determining arcing events. Interference signals at or above the national standard threshold, such as pseudo-signals generated by the self-generated oscillation coupling of the high-voltage drive module, may still pass the preliminary screening stage and proceed to the fine screening stage for further judgment. The final identification result of the arcing event is determined by the interference feature fingerprint matching result of the fine screening stage. Through the above two-stage processing mechanism, the target arcing signal is screened and finely identified step by step, improving the accuracy of the final judgment while ensuring detection efficiency.

[0035] Step S33: Output and processing of the judgment result: 1. Real arc preservation: Output arc alarm signal and record arc parameters, including arc peak intensity, arc duration, arc occurrence time and train current position, and store the above data in local database; 2. False signal rejection: The event is silently discarded without triggering an alarm, but the frequency and intensity of the interference event are recorded for dynamic monitoring of changes in the internal electromagnetic environment of the system. When the frequency of interference events increases abnormally, the system automatically issues a hardware health status warning, prompting maintenance personnel to check whether the high-voltage drive module and the shielding structure of the metal shell are abnormal.

[0036] Existing technologies for suppressing internal electromagnetic interference in equipment mainly employ two types of solutions: one is a purely hardware solution, which reduces interference intensity through physical means such as shielding, filtering, and grounding; the other is a purely software solution, which processes and eliminates interference signals through algorithms such as digital filtering and threshold judgment. However, the above solutions have significant shortcomings when facing internal mutual interference from high-voltage drive modules to weak signal acquisition terminals: the purely hardware solution is limited by installation space and cost, and cannot completely eliminate interference, always leaving residual interference components; the purely software solution, when faced with strong interference, the signal is completely submerged by noise, making it impossible to effectively extract the valid signal.

[0037] To achieve high dynamic range detection, this invention integrates a photomultiplier tube (PMT) and a silicon carbide ultraviolet (UV) detector. The PMT possesses extremely high sensitivity for capturing weak UV signals, while the UV detector exhibits excellent linear response, making it suitable for quantitative analysis of arc intensity. It is insensitive to visible and infrared light (i.e., "solar-blind"), eliminating the need for complex external filters to filter sunlight interference, unlike UV PMTs, thus physically reducing the false alarm rate. Simultaneously, the strong electromagnetic radiation generated by the PMT's high-voltage drive module induces high-frequency interference noise at the weak signal acquisition end of the UV detector through spatial coupling. By encapsulating both PMTs within two independent chambers of the same insulating metal casing, the chambers form complete electromagnetic shielding from the outside. Utilizing the conductivity of the metal casing to create a Faraday cage effect, the high-voltage radiation source is confined within the chambers, making it more resistant to shock and vibration and preventing false triggering caused by the aforementioned problems. Simultaneously, it cleverly utilizes the spatiotemporal consistency of dual-channel signals to solve the "artifact" problem that a single sensor cannot address. Through time-domain waveform recognition, it cleverly avoids the difficulty of traditional filters in filtering out the frequency band overlap interference of the high-voltage power supply itself without damaging the signal. Since the high sensitivity of the photomultiplier tube is reserved for triggering, the system will not miss weak arcing even with a high threshold setting. At the same time, the secondary confirmation by the silicon carbide ultraviolet detector prevents false alarms from the photomultiplier tube, greatly improving accuracy. The system as a whole forms a closed-loop anti-interference system with hardware physical isolation and software logic complementarity, achieving effective suppression of system interference.

[0038] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A pantograph arc detection system, characterized in that: The device includes a sealed metal casing, within which are independent high-voltage isolation chambers and signal acquisition chambers. The high-voltage isolation chamber contains a first isolated power conversion module, a high-voltage drive module, and a photomultiplier tube, all electrically connected in sequence. The signal acquisition chamber contains a power module and a second isolated power conversion module, both electrically connected, as well as a main control unit and a signal acquisition module, both electrically connected to the second isolated power conversion module. The first isolated power conversion module is electrically connected to the power module. The signal acquisition module includes a silicon carbide ultraviolet detector, a signal conditioning module, and an ADC conversion module, all electrically connected in sequence. The ADC conversion module and the photomultiplier tube are both electrically connected to the main control unit. An integrally formed partition separates the inner cavity of the metal casing into the high-voltage isolation chamber and the signal acquisition chamber. The partition wall thickness is not less than the wall thickness of the metal casing, which is not less than 4 mm. The main control unit is adapted to establish a time-domain correspondence between the pulse counting signal of the photomultiplier tube and the ADC sampling data of the silicon carbide ultraviolet detector by collecting the pulse counting signal of the photomultiplier tube and the sampling data of the ADC of the silicon carbide ultraviolet detector. By experimentally calibrating the self-generated oscillation characteristics of the high-voltage drive module of the photomultiplier tube, when the pantograph arc detection system is in normal working condition, a time-domain conformity judgment window Δt is constructed so that when the photomultiplier tube detects a pulse edge, the judgment window is automatically opened to identify and eliminate residual interference signals and retain the real arc event.

2. The pantograph arc detection system according to claim 1, characterized in that: The inner surface of the metal casing is treated with black anodizing.

3. The pantograph arc detection system according to claim 2, characterized in that: The partition is provided with a wire-passing hole, and the wire-passing hole is filtered by a feedthrough capacitor.

4. The pantograph arc detection system according to claim 1, characterized in that: The ADC conversion module is connected to the main control unit via an optocoupler-isolated SPI signal line.

5. The pantograph arc detection system according to claim 4, characterized in that: The SPI signal line uses a shielded cable, and the shielding layer of the shielded cable is grounded at both ends.

6. An interference suppression method for a pantograph arc detection system as described in any one of claims 1 to 5, characterized in that: Includes the following steps: Step S1: Signal acquisition: The main control unit synchronously acquires the pulse counting signal of the photomultiplier tube and the ADC sampling data of the silicon carbide ultraviolet detector to establish the time domain correspondence of the dual-channel data; Step S2: Interference Feature Extraction: During the system initialization phase, the self-generated oscillation characteristics of the high-voltage drive module of the photomultiplier tube are experimentally calibrated. Step S3: Time Domain Compliance Determination: The pantograph arc detection system is in normal working condition. At this time, the high voltage drive module and the external ultraviolet light source are both turned on. A time domain compliance determination window Δt is constructed. When the photomultiplier tube detects the pulse edge, the system automatically opens the determination window, identifies and eliminates residual interference signals, and retains the real arc event.

7. The interference suppression method according to claim 6, characterized in that, In step S2, the calibration method for the self-generated oscillation characteristics includes the following steps: Step S21: Turn off the high-voltage drive module of the photomultiplier tube and only collect the signal from the silicon carbide ultraviolet detector as the background noise reference; Step S22: Turn on the high-voltage drive module, but shield the external ultraviolet light source, so that the photomultiplier tube is in standby mode, and collect the signal of the silicon carbide ultraviolet detector channel. At this time, since there is no external ultraviolet light source, the signal collected by the silicon carbide ultraviolet detector comes entirely from the interference generated by the electromagnetic radiation coupling of the high-voltage drive module, which serves as an interference signal. Step S23: Perform FFT transformation on the acquired interference signal to extract the spectral features and time-domain waveform features of the interference signal. The spectral features include the typical waveform width W of the interference pulse. noise and rise time TR noise The time-domain waveform features include a complete time-domain waveform template sequence, denoted as sequence Y = {y1, y2, …, y}. m }; Step S24: Based on the interference pulse waveform width W extracted in step S23 noise and rise time TR noise Determine the first preset threshold δ w =α×W noise Second preset threshold δ r =β×TR noise Wherein, α and β are allowable relative deviation coefficients, which are determined by stability testing of the high-voltage drive module within its normal operating temperature range.

8. The interference suppression method according to claim 7, characterized in that, The time-domain compliance determination includes the following steps: Step S31: Initial screening using national standard thresholds: Within the judgment window Δt, analyze the optical signal of the silicon carbide ultraviolet detector channel. If the peak signal intensity is less than the third preset threshold V... th Or the waveform width is less than the fourth preset threshold T w If the signal peak intensity is greater than or equal to the third preset threshold V, it is determined to be a non-arc event and is directly excluded; th And the waveform width is greater than or equal to the fourth preset threshold T w Then proceed to the next step: analyze the temporal characteristics of the optical signal in the silicon carbide ultraviolet detector channel; the third preset threshold V th and the fourth preset threshold T w The recommended value is V. th =25uw / cm 2 T w =5ms; Step S32: Fine screening of interference feature fingerprints. For optical signals that have passed the initial screening of national standard thresholds, a fine judgment step is performed. I. Refined comparison of waveform width; Calculate the full width at half maximum (FWHM) and rise time (T) of the optical signal waveform from 10% to 90%. rise The calculation results are compared with the typical waveform width W of the interference signal spectral characteristics extracted in step S23. noise Rising edge feature TR noise Perform a comparison; like: and If the signal is initially identified as a pseudo-signal generated by interference coupling, it will be eliminated. Otherwise, it indicates that there is a significant difference between the waveform characteristics of the measured signal and the interference fingerprint, and further determination will be made by waveform morphology correlation matching. II. Waveform morphology correlation matching; The optical signal waveform of the silicon carbide ultraviolet detector channel is cross-correlated with the time-domain waveform characteristics stored in step S2, and the correlation coefficient ρ is calculated: Where, x i The measured optical signal waveform sequence is y. i For interfering feature fingerprint template sequences; if ρ > preset correlation coefficient threshold ρ th , ρ th If ρ = 0.85, it is determined to be a pseudo-signal generated by interference coupling, and pseudo-signals are eliminated; if ρ ≤ ρ th If so, it will be retained as a real arcing event; Step S33: Output and processing of the judgment result:

1. Real arc preservation: Output arc alarm signal and record arc parameters, including arc peak intensity, arc duration, arc occurrence time and train current position, and store the above data in local database; 2. False signal rejection: The event is silently discarded without triggering an alarm, but the frequency and intensity of the interference event are recorded for dynamic monitoring of changes in the internal electromagnetic environment of the system. When the frequency of interference events increases abnormally, the system automatically issues a hardware health status warning, prompting maintenance personnel to check whether the high-voltage drive module and the shielding structure of the metal shell are abnormal.

Images