Metro signal power supply detection system and method based on internet of things technology

Abstract

The present application belongs to the field of power supply detection, and particularly relates to a subway signal power supply detection system and method based on Internet of Things technology, aiming to solve the problems of test data distortion and inability to remotely share in the strong electromagnetic interference environment of the subway. The present application comprises: acquiring multiple analog signals of the measured signal power supply module; obtaining original digital sampling sequences through analog-digital conversion; performing time-frequency multi-resolution decomposition, determining the interference type according to the comparison between the sub-band energy and the preset threshold, and performing retention, zeroing or amplitude shrinkage processing on the coefficients; reconstructing the processed coefficients to obtain the de-noised signal, and then constructing a dynamic limiting threshold according to the local fluctuation characteristics and the change trend to perform collaborative limiting filtering; obtaining the steady-state operation parameters and the starting time based on the filtered signal; and finally sending the test parameters to the cloud platform through the wireless network for storage and display. The present application can effectively suppress the composite electromagnetic interference and realize remote sharing of high-precision test data.

Description

Technical Field

[0001] This invention belongs to the field of power supply detection, and specifically relates to a subway signal power supply detection system and method based on Internet of Things (IoT) technology. Background Technology

[0002] Subway signal power modules provide operating power to the signaling system, and their performance and reliability directly affect train operation safety. When testing signal power modules in a subway operating environment, it is necessary to collect process parameters such as input voltage, input current, output voltage, output current, and module temperature to evaluate their functionality and stability.

[0003] The electromagnetic environment at the subway site exhibits significant complex interference characteristics. Harmonics generated by the traction inverter are mainly distributed in the 2kHz to 20kHz frequency band, forming high-frequency pulse group interference; transient processes of inductive loads introduced by relay contact operations induce periodic oscillations of the 50Hz to 100Hz fundamental frequency and its harmonics; and spatial radiation coupling occurs from track circuits and communication equipment. These interferences are conducted to the signal acquisition link via electromagnetic coupling, causing glitches, clipping, or baseline jitter in the analog signal waveform, reducing the accuracy of analog-to-digital conversion.

[0004] Conventional digital filtering methods have limitations in handling this type of complex interference. Mean filtering includes abnormal amplitudes in the arithmetic mean, causing pulse interference to spread to adjacent data intervals, resulting in waveform distortion. It also dulls signal step edges, losing transient characteristic information. Amplitude limiting filtering relies on a threshold difference between adjacent sampling points. For continuous high-frequency oscillations or periodic interference, it may be mistakenly identified as valid signals and retained because the difference is within the threshold, resulting in poor output waveform smoothness. Furthermore, a fixed threshold is difficult to adapt to differences in signal change rates; a threshold that is too small will incorrectly reject normal fast-changing signals, while a threshold that is too large will miss small-amplitude interference.

[0005] The aforementioned limitations create a technical contradiction between test accuracy and interference suppression, making it difficult to meet the requirements of high-precision testing of subway signal power modules for data authenticity and integrity. Summary of the Invention

[0006] To address the aforementioned problems in the prior art, namely the distortion of test data and the inability to remotely share it under strong electromagnetic interference in subway environments, this invention provides a subway signal power supply detection system and method based on Internet of Things (IoT) technology.

[0007] The first aspect of this invention proposes a method for detecting subway signal power supply based on Internet of Things (IoT) technology, comprising:

[0008] Acquire multiple analog signals during the testing process of the power supply module under test; The multiple analog signals are converted from analog to digital to obtain the original digital sampling sequence; The original digital sampling sequence is decomposed into time-frequency multi-resolution components to obtain multiple sub-frequency band components. The interference type is determined by comparing the energy of the sub-frequency band components with a preset first energy threshold and a second energy threshold. Based on the determination result, the coefficients of the sub-frequency band components are retained, zeroed, or amplitude-shrinked. The first energy threshold is less than the second energy threshold. The processed coefficients are reconstructed to obtain a denoised signal sequence. A dynamic amplitude limiting threshold is constructed based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence. The denoised signal sequence is then subjected to collaborative amplitude limiting filtering to obtain a filtered signal sequence. The interference types include at least pulse interference and periodic oscillation interference in the subway environment. Based on the filtered signal sequence, the steady-state operating parameters and startup time of the signal power supply module are obtained as test parameters; The test parameters are sent to the cloud platform via a wireless network for storage and display.

[0009] Furthermore, the multi-channel analog signal includes at least an input voltage signal, an input current signal, an output voltage signal, an output current signal, and a temperature signal.

[0010] Furthermore, the multiple analog signals are subjected to analog-to-digital conversion to obtain the original digital sampling sequence, including: The analog signal is sampled at a first sampling frequency, and a first number of signal samples are continuously collected. After removing the sample with the largest amplitude and the sample with the smallest amplitude from the first number of signal samples, the average value of the remaining samples is calculated as the first-level filtering sequence. The first-level filtered sequence is sampled at a second sampling frequency, and a second number of samples are continuously collected. After removing the maximum and minimum values, the average value is calculated to obtain the original digital sampling sequence. The first sampling frequency is higher than the second sampling frequency. The sampling performed at the first sampling frequency and the sampling performed at the second sampling frequency are executed in a nested loop manner, and the interrupt priority of the sampling performed at the first sampling frequency is configured to be higher than the interrupt priority of the sampling performed at the second sampling frequency.

[0011] Furthermore, the original digital sampling sequence is subjected to time-frequency multi-resolution decomposition, including: The original digital sampling sequence is decomposed into multiple wavelet packets using wavelet packet basis functions to generate multiple sub-band components corresponding to different frequency ranges. Each sub-band component is composed of a set of wavelet packet coefficients. Based on the spectral distribution of traction inverter harmonic interference and relay operation oscillation interference in the subway environment, the high-frequency sub-bands among the multiple sub-band components are associated with pulse interference frequency bands, and the low-frequency sub-bands are associated with periodic oscillation frequency bands.

[0012] Furthermore, based on a comparison between the energy of the sub-frequency band component and a preset first energy threshold and second energy threshold, the interference type is determined, and a corresponding amplitude adjustment strategy is applied to process the coefficients of the sub-frequency band component, including: Calculate the energy of the sub-frequency band component. If the energy is greater than the second energy threshold, it is determined to be dominated by pulse interference. Set the abnormal coefficients in the sub-frequency band whose amplitude is greater than the preset noise reduction threshold to zero, and retain the other coefficients. If the energy is not less than the first energy threshold and not greater than the second energy threshold, it is determined to be mixed interference. The absolute value of the coefficient is subtracted from the shrinkage threshold and the original sign is retained. If the result is less than zero, it is set to zero. The shrinkage threshold is determined based on the noise standard deviation estimate of the coefficient in the sub-band. If the energy is less than the first energy threshold, the original coefficients are retained; Wherein, the first energy threshold is less than the second energy threshold, and the noise standard deviation estimate is estimated using the median absolute deviation of the sub-band coefficients.

[0013] Furthermore, based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence, a dynamic limiting threshold is constructed, and collaborative limiting filtering is performed on the denoised signal sequence, including: Calculate the variance of the denoised signal sequence within the sliding window, and use it as a local fluctuation characteristic of the short-term fluctuation of the signal; Calculate the first-order difference of the denoised signal sequence and use exponential smoothing to obtain a trend prediction quantity for determining the direction of signal change; The dynamic amplitude limiting threshold at the current moment is obtained by weighted summing of the square root of the local fluctuation feature quantity, the absolute value of the trend prediction quantity, and the absolute value of the difference between the local fluctuation feature quantities at adjacent times. The weighting coefficients in the weighted summation take the first set of preset values ​​when the signal is in the transient stage and the second set of preset values ​​when the signal is in the steady state stage. If the absolute value of the difference between the current signal sample value and the previous filtered output value does not exceed the dynamic amplitude limiting threshold, then the current sample value is used as the filtered output; otherwise, the threshold step is adjusted by one step towards the current sample value based on the previous filtered output value to obtain the current filtered output.

[0014] Furthermore, based on the filtered signal sequence, the start-up time of the signal power supply module is obtained, including: The filtered input voltage signal sequence is continuously monitored. When the input voltage is detected to jump from a preset ratio below the rated value to a preset ratio above the rated value, the start timer is triggered to start counting. The filtered output voltage signal sequence is continuously monitored. When the output voltage first enters the preset stable range near the rated value, the turn-on delay judgment is initiated. If the output voltage remains stable within the preset stable range during the preset delay period, the start-up timing is stopped, and the accumulated time of the start-up timing is taken as the start-up time.

[0015] Furthermore, the step of sending the test parameters to the cloud platform via a wireless network for storage and display includes: The test parameters are encapsulated according to a preset communication protocol and uploaded via a wireless communication link at a set polling interval. The uploaded test parameters are stored in the database in the form of data frames, and are displayed in real time and downloaded as historical data.

[0016] In a second aspect, the present invention proposes a subway signal power supply detection system based on Internet of Things (IoT) technology, the system comprising: The test box integrates an AC / DC power module, a programmable controller, and an IoT module. The AC / DC power module is used to provide test power to the power module of the signal under test. The programmable controller is used to collect test process parameters. The IoT module is used to interact with the cloud platform. An electronic load cell is connected to the output terminal of the power supply module for the signal under test. The sensor is used to detect the input voltage, input current, output voltage, output current, and module temperature of the power supply module under test. The cloud platform establishes a communication connection with the programmable controller through the IoT module, and is used to receive, store and display test data; An operating terminal for accessing the cloud platform; The programmable controller is configured to execute a subway signal power supply detection method based on Internet of Things (IoT) technology.

[0017] Furthermore, the AC / DC power module is encapsulated in a power supply box, and the power supply box is electrically isolated from the test box; The programmable controller and the Internet of Things module are mounted on the control board; the test box is equipped with a cooling fan, a voltage and current display meter, an output switch and a test output interface. The voltage and current display meter is used to display the test voltage and test current, and the output switch is used to control the on / off state of the test output interface.

[0018] The beneficial effects of this invention are: This invention employs a two-layer filtering strategy combining time-frequency multi-resolution decomposition and collaborative amplitude limiting filtering. First, wavelet packet decomposition separates the signal into different sub-frequency bands. Based on the spectral characteristics of the 2kHz-20kHz pulse interference and 50Hz-100Hz periodic oscillation interference unique to the subway environment, hard threshold denoising, soft threshold contraction, or coefficient preservation are adaptively performed. Then, a dynamic amplitude limiting threshold is constructed based on the square root of local fluctuation characteristics, the absolute value of trend prediction, and the absolute value of the difference between local fluctuation characteristics at adjacent times, and collaborative amplitude limiting filtering is applied to the reconstructed signal. This processing improves the signal-to-noise ratio from 28dB to 42dB using conventional methods, and reduces the start-up time measurement error from 7ms to 0.3ms, effectively solving the problem of test data distortion under the strong electromagnetic environment of the subway.

[0019] This invention employs a double-layer nested extreme value removal averaging filter driven by a PLC timer interrupt. In a test channel used for acquiring steady-state operating parameters and identifying low-frequency periodic oscillation interference, the inner layer performs 100 effective samples at a sampling rate of 10kHz and the outer layer at a sampling rate of 1kHz, reducing the random noise variance to 1 / 10. In a test channel used for identifying 2kHz to 20kHz pulse interference, the inner layer sampling rate is configured to be no less than 40kHz to meet the wavelet packet decomposition processing requirements of this pulse interference frequency band. Simultaneously, the cloud platform remotely acquires test parameters at a polling frequency of 1Hz, far exceeding the acquisition frequency of once per minute of conventional test systems, enabling precise capture of power-on transient processes and steady-state fluctuation details.

[0020] This invention automatically collects parameters such as input voltage, input current, output voltage, output current, module temperature, and startup time using a programmable logic controller (PLC), eliminating the need for manual reading and recording. Test data is uploaded in real-time to a cloud platform via an IoT module and stored in a database as data frames. Operators can remotely view real-time curves or download historical data (in Excel format) via a terminal, enabling data sharing among maintenance personnel, users, and third parties, ensuring a fair and transparent testing process.

[0021] The test chamber of this invention integrates an AC / DC power module, a programmable controller, and an Internet of Things (IoT) module. The power module is encapsulated in a plastic power box and electrically isolated from the aluminum alloy casing. A cooling fan ensures long-term operational stability. The system is suitable for environments such as subway maintenance workshops, signal rooms, trackside areas, and inside electric trains. It can perform functional tests, stability tests, and type tests such as high temperature, low temperature, humidity, overload, and vibration tests on signal power modules. Attached Figure Description

[0022] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1This is a flowchart illustrating a subway signal power supply detection method based on Internet of Things (IoT) technology according to the present invention. Figure 2 This is a schematic diagram illustrating the signal interference filtering effect of a subway signal power supply detection method based on Internet of Things technology according to the present invention. Figure 3 This is an overall architecture diagram of a subway signal power supply detection system based on Internet of Things technology according to the present invention; Figure 4 This is a schematic diagram of the front panel layout of a subway signal power supply detection system based on Internet of Things technology according to the present invention; Figure 5 This is a schematic diagram of the rear panel of a subway signal power supply detection system based on Internet of Things technology according to the present invention; Figure 6 This is an internal structural layout diagram of a subway signal power supply detection system based on Internet of Things technology according to the present invention; Figure 7 This is a schematic diagram of the AC / DC power supply module layout in a subway signal power supply detection system based on Internet of Things technology according to the present invention.

[0023] In the diagram, 100 is the test box; 110 is the AC / DC power module; 120 is the programmable logic controller (PLC); 130 is the Internet of Things (IoT) module; 140 is the power supply box; 150 is the control board; 160 is the cooling fan; 170 is the voltage and current display meter; 180 is the power input interface; 200 is the electronic load instrument; 210 is the output switch; 220 is the test output interface; 300 is the sensor; 400 is the cloud platform; 500 is the operation terminal; and 600 is the power supply module for the signal under test. Detailed Implementation

[0024] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] The first embodiment of the present invention provides a method for detecting the power supply of subway signals based on Internet of Things (IoT) technology, comprising: Step S10: Acquire multiple analog signals during the testing process of the power supply module under test; Step S20: Perform analog-to-digital conversion on the multiple analog signals to obtain the original digital sampling sequence; Step S30: Perform time-frequency multi-resolution decomposition on the original digital sampling sequence to obtain multiple sub-frequency band components. Based on the comparison between the energy of the sub-frequency band components and a preset first energy threshold and a second energy threshold, determine the interference type. Then, perform retention, zeroing, or amplitude shrinkage processing on the coefficients of the sub-frequency band components according to the determination result. The first energy threshold is less than the second energy threshold. Step S40: The processed coefficients are reconstructed to obtain a denoised signal sequence. A dynamic amplitude limiting threshold is constructed based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence. Collaborative amplitude limiting filtering is then applied to the denoised signal sequence to obtain a filtered signal sequence. The interference types include at least pulse interference and periodic oscillation interference in the subway environment. Step S50: Based on the filtered signal sequence, obtain the steady-state operating parameters and start-up time of the signal power supply module as test parameters; Step S60: The test parameters are sent to the cloud platform via a wireless network for storage and display.

[0027] To more clearly illustrate the present invention's method for detecting subway signal power supply based on Internet of Things (IoT) technology, the following description is provided in conjunction with... Figure 1 The steps in the embodiments of the present invention are described in detail below: Step S10: Acquire multiple analog signals during the testing process of the power supply module under test; The multi-channel analog signals include at least an input voltage signal, an input current signal, an output voltage signal, an output current signal, and a temperature signal.

[0028] In step S10 of this embodiment, multiple analog signals are acquired during the testing process of the power supply module under test. Specifically, the multiple analog signals include an input voltage signal, an input current signal, an output voltage signal, an output current signal, and a temperature signal. The input voltage and input current signals originate from the input side of the power supply module under test, the output voltage and output current signals originate from the output side of the power supply module under test, and the temperature signal originates from a temperature sensor attached to the housing of the power supply module under test. These five analog signals respectively reflect the power supply status, load-carrying output capability, and thermal stability of the power supply module under test during operation. Each physical quantity is converted into a standard analog electrical signal by the corresponding sensor. The standard analog electrical signal is a 4-20mA current-type analog signal or a 1-5V voltage-type analog signal. Shielded cables are used for signal transmission, utilizing the shielding layer's suppression effect on spatial electromagnetic fields to reduce electromagnetic interference intrusion into the transmission link. The acquisition process is continuous real-time sampling, meaning that the five analog signals are sampled uninterruptedly during the test to ensure data integrity and temporal continuity throughout the entire test process.

[0029] Step S20: Perform analog-to-digital conversion on the multiple analog signals to obtain the original digital sampling sequence; In this embodiment, the multiple analog signals are subjected to analog-to-digital conversion to obtain the original digital sampling sequence, including: Step S21: Sample the analog signal at a first sampling frequency, continuously collect a first number of signal samples, remove the sample with the largest amplitude and the sample with the smallest amplitude from the first number of signal samples, and calculate the average value of the remaining samples as the first-level filtering sequence. Step S22: Sample the first-level filtered sequence at the second sampling frequency, continuously collect a second number of sample points, remove the maximum and minimum values ​​and calculate the average value to obtain the original digital sampling sequence, wherein the first sampling frequency is higher than the second sampling frequency; The sampling performed at the first sampling frequency and the sampling performed at the second sampling frequency are executed in a nested loop manner, and the interrupt priority of the sampling performed at the first sampling frequency is configured to be higher than the interrupt priority of the sampling performed at the second sampling frequency.

[0030] In step S20 of this embodiment, the multiple analog signals are converted from analog to digital to obtain the original digital sampling sequence. The programmable logic controller (PLC) receives the standard analog electrical signal from the sensor through its analog input channel and performs a double-layer nested extreme value removal and averaging filter process.

[0031] In the inner loop, the analog signal is sampled at a first sampling frequency. This first sampling frequency is set according to the interference frequency band to be analyzed in the channel under test. For low-frequency test channels used for acquiring steady-state operating parameters and identifying periodic oscillation interference, the first sampling frequency can be driven by the PLC's internal timers T0 to T9 with a 0.1 millisecond period, i.e., the first sampling frequency is 10 kHz. For sampling channels used to identify pulse interference frequency bands, the first sampling frequency is configured to be no less than twice the highest frequency of the pulse interference frequency band to be analyzed. When the upper limit of the pulse interference frequency band to be analyzed is 20 kHz, the first sampling frequency is configured to be no less than 40 kHz. Each time an inner timer interrupt is triggered, the PLC acquires one signal sample point, continuously acquiring a first number of signal samples. In this embodiment, the first number is 10, and the acquired 10 samples are sequentially stored in buffer registers D100 to D109. After 10 sample points are collected, the maximum and minimum values ​​are selected using a comparison instruction. The maximum value is stored in register D111, and the minimum value is stored in register D112. The sum of the remaining 8 sample points is stored in register D110. Then, the maximum and minimum values ​​are subtracted from the sum, and the difference is shifted right by 3 bits (i.e., divided by 8) to obtain a filter value from the inner filter sequence, which is stored in register D120. This process is repeated to form the first-level filter sequence.

[0032] In the outer loop, the first-stage filtered sequence is sampled at a second sampling frequency. This second sampling frequency is driven by PLC internal timers T10 to T19 with a 1-millisecond cycle, meaning the second sampling frequency is 1 kHz, and the first sampling frequency is higher than the second sampling frequency. Each time an outer timer interrupt is triggered, the PLC reads an inner-stage filtered value from register D120 and continuously collects a second number of samples. In this embodiment, the second number is 10, and the 10 collected samples are sequentially stored in the D200 to D209 buffer register group. The same extreme value removal and averaging logic is performed on these 10 samples, that is, after removing the maximum and minimum values, the average value of the remaining 8 samples is calculated to obtain a final filtered value in the original digital sampling sequence.

[0033] The inner and outer loops are executed in a nested loop manner. The interrupt priorities of inner timers T0 to T9 are configured with the highest priority (priority 0), while the interrupt priorities of outer timers T10 to T19 are configured with a lower priority (priority 1). When an inner interrupt is triggered during the execution of an outer interrupt, the inner interrupt immediately preempts the outer interrupt service routine, executes the inner sampling, and then returns to the outer loop to continue execution, thus ensuring that the inner sampling period corresponding to the first sampling frequency is not blocked by the outer program. Conversely, the outer interrupt cannot interrupt the inner interrupt during its execution, preventing interference from the outer loop with the inner's time-critical operations. Through this double-nested extreme value removal averaging filter, the digital sampling sequence obtained after analog-to-digital conversion of the original analog signal has undergone preliminary noise suppression, providing a foundation for subsequent intelligent digital filtering.

[0034] In the inner loop, timers T0 to T9 have the same priority; in the outer loop, timers T10 to T19 also have the same priority. To resolve conflicts arising from simultaneous triggering of timers with the same priority, the PLC program uses the SCR sequential control instruction to schedule multiple timers within the same loop sequentially. Specifically, the LSCR instruction pushes the step bit reference onto the SCR stack, and the stack value determines whether to execute the corresponding program segment. Taking the inner loop in this embodiment as an example, step 1 starts timer T0 for 0.1ms, executes the acquisition program, increments the step bit stack by one after acquisition, and jumps to step 2 to start timer T1, and so on, until all 10 acquisitions from T0 to T9 are completed. The SCR sequential control instruction ensures that timers in the same layer are triggered sequentially in a predetermined order, avoiding resource contention and acquisition timing errors caused by the same priority.

[0035] Step S30: Perform time-frequency multi-resolution decomposition on the original digital sampling sequence to obtain multiple sub-frequency band components. Based on the comparison between the energy of the sub-frequency band components and a preset first energy threshold and a second energy threshold, determine the interference type. Then, perform retention, zeroing, or amplitude shrinkage processing on the coefficients of the sub-frequency band components according to the determination result. The first energy threshold is less than the second energy threshold. In this embodiment, time-frequency multi-resolution decomposition is performed on the original digital sampling sequence, including: Step S31: Select wavelet packet basis functions to perform multi-level wavelet packet decomposition on the original digital sampling sequence to generate multiple sub-band components corresponding to different frequency ranges. Each sub-band component is composed of a set of wavelet packet coefficients. Step S32: Based on the spectral distribution of traction inverter harmonic interference and relay operation oscillation interference in the subway environment, the high-frequency sub-bands among the multiple sub-band components are matched with the pulse interference frequency band, and the low-frequency sub-bands are matched with the periodic oscillation frequency band.

[0036] Based on a comparison of the energy of the sub-frequency band component with a preset first energy threshold and a second energy threshold, the interference type is determined, and a corresponding amplitude adjustment strategy is applied to process the coefficients of the sub-frequency band component, including: Step S33: Calculate the energy of the sub-frequency band component. If the energy is greater than the second energy threshold, it is determined to be dominated by pulse interference. Set the abnormal coefficients with amplitude greater than the preset noise reduction threshold in the sub-frequency band to zero, and retain the other coefficients. Step S34: If the energy is not less than the first energy threshold and not greater than the second energy threshold, it is determined to be mixed interference. The absolute value of the coefficient is subtracted from the shrinkage threshold and the original sign is retained. If the result is less than zero, it is set to zero. The shrinkage threshold is determined based on the noise standard deviation estimate of the coefficient in the sub-band. Step S35: If the energy is less than the first energy threshold, then retain the original coefficients; Wherein, the first energy threshold is less than the second energy threshold, and the noise standard deviation estimate is estimated using the median absolute deviation of the sub-band coefficients.

[0037] In step S30 of this embodiment, the original digital sampling sequence is decomposed into time-frequency multi-resolution components to obtain multiple sub-frequency band components. Specifically, wavelet packet basis functions are selected to perform multi-level wavelet packet decomposition on the original digital sampling sequence to generate multiple sub-frequency band components corresponding to different frequency ranges. Each sub-frequency band component is composed of a set of wavelet packet coefficients. This embodiment uses three-level wavelet packet decomposition to generate eight sub-frequency band components. Based on the spectral distribution of traction inverter harmonic interference and relay operation oscillation interference in the subway environment, the high-frequency sub-bands in the multiple sub-frequency band components are associated with the pulse interference frequency band, which corresponds to 2kHz to 20kHz. When performing wavelet packet decomposition on the pulse interference frequency band, the first sampling frequency used to form the original digital sampling sequence is set according to the highest frequency of the pulse interference frequency band to be analyzed and configured to be no less than twice the highest frequency. When the upper limit of the pulse interference frequency band to be analyzed is 20kHz, the first sampling frequency is configured to be no less than 40kHz. The low-frequency sub-bands are associated with periodic oscillation bands, which correspond to 50Hz to 100Hz and their harmonics.

[0038] Based on a comparison of the energy of the sub-band components with a preset first energy threshold and a second energy threshold, the interference type is determined, and a corresponding amplitude adjustment strategy is applied to process the coefficients of the sub-band components. First, the energy of each sub-band component is calculated; this energy is the sum of the squares of all wavelet packet coefficients within that sub-band divided by the sum of the squares of all sub-band coefficients. A first energy threshold and a second energy threshold are preset, with the first energy threshold being less than the second energy threshold. The first energy threshold is used to identify clean signals, and the second energy threshold is used to identify dominant interference. In this embodiment, the first energy threshold ranges from 0.15 to 0.40, preferably 0.25; the second energy threshold ranges from 0.35 to 0.65, preferably 0.5.

[0039] If the calculated sub-band energy is greater than the second energy threshold, the sub-band is determined to be dominated by impulse interference. For such sub-bands, a hard threshold noise reduction strategy is employed: abnormal coefficients within the sub-band with amplitudes greater than a preset noise reduction threshold are set to zero, while other coefficients are retained. The preset noise reduction threshold is determined based on the estimated noise standard deviation of the coefficients within the sub-band.

[0040] If the calculated sub-band energy is not less than the first energy threshold and not greater than the second energy threshold, it is determined to be mixed interference. For such sub-bands, a soft-threshold adaptive shrinkage strategy is activated: the absolute value of each coefficient is subtracted from the shrinkage threshold, and the original sign is retained; if the subtraction result is less than zero, the coefficient is set to zero. The shrinkage threshold is determined based on the estimated noise standard deviation of the coefficients within the sub-band, and the specific calculation formula is as follows: ,in This is an estimate of the noise standard deviation. The number of coefficients within the sub-band is given. The noise standard deviation estimate is obtained by estimating the absolute deviation of the median of the sub-band coefficients, that is, dividing the median of the absolute values ​​of the coefficients within the sub-band by 0.6745 as a robust estimate of the noise standard deviation.

[0041] If the calculated sub-band energy is less than the first energy threshold, it is determined to be a clean signal, the original coefficients are directly retained, and the filtering process is skipped and the signal is directly output.

[0042] Through the adaptive processing based on the comparison between sub-band energy and preset threshold, this embodiment can adopt corresponding coefficient adjustment strategies for different types of interference in the subway environment, laying the foundation for subsequent signal reconstruction.

[0043] In this embodiment, before determining the interference type based on the comparison between the energy of the sub-frequency band component and a preset first energy threshold and a second energy threshold, the method further includes: calculating the energy percentage of the same sub-frequency band component in multiple consecutive sliding analysis windows, and generating an energy stability marker based on the change in the energy percentage in adjacent sliding analysis windows; when the energy percentage continuously reaches a preset number of confirmations and is greater than the second energy threshold, the corresponding sub-frequency band is determined to be dominated by pulse interference; when the energy percentage falls back from being greater than the second energy threshold, an exit threshold lower than the second energy threshold is used for exit determination, and the exit threshold is obtained by subtracting a preset hysteresis amount from the second energy threshold.

[0044] As an optional implementation, in step S30, to avoid frequent switching of sub-band interference types caused by occasional pulses within a single window, the energy proportion of the same sub-band component can be continuously confirmed within a window. Specifically, for the first... Each sub-band component, in the first Calculate the energy percentage within each sliding analysis window. The energy percentage For the first The ratio of the sum of squares of wavelet packet coefficients of each sub-band component within the sliding analysis window to the sum of squares of wavelet packet coefficients of all sub-band components within the sliding analysis window. Further calculations are made of the change in energy proportion between adjacent sliding analysis windows. When the change in the energy percentage is less than a preset stable change threshold, an energy stability marker is generated.

[0045] When determining the type of interference, if the same sub-band component is in continuous Each sliding analysis window has an energy stability marker, and the continuous If the energy percentage in each sliding analysis window is greater than the second energy threshold, then the sub-band component is determined to be dominated by impulse interference. A value of 2-5 is acceptable. If the sub-frequency band component has been determined to be dominated by pulse interference, when the energy proportion in the subsequent sliding analysis window decreases, the pulse interference dominance determination is not immediately lifted. Instead, the second energy threshold is subtracted from the preset hysteresis value to obtain the exit threshold. When the energy proportion in the subsequent sliding analysis window is less than the exit threshold, the pulse interference dominance determination is lifted. The preset hysteresis value can be 0.03-0.08. Through the above continuous window confirmation and hysteresis exit determination, the misjudgment and frequent switching of interference types caused by occasional pulse spikes in the subway can be reduced, and the stability of subsequent coefficient zeroing, amplitude contraction, or coefficient retention processing can be improved.

[0046] Step S40: The processed coefficients are reconstructed to obtain a denoised signal sequence. A dynamic amplitude limiting threshold is constructed based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence. Collaborative amplitude limiting filtering is then applied to the denoised signal sequence to obtain a filtered signal sequence. The interference types include at least pulse interference and periodic oscillation interference in the subway environment. In this embodiment, a dynamic amplitude limiting threshold is constructed based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence, and collaborative amplitude limiting filtering is performed on the denoised signal sequence, including: Step S41: Calculate the variance of the denoised signal sequence within the sliding window, as a local fluctuation feature of the short-term fluctuation degree of the signal; Step S42: Calculate the first-order difference of the denoised signal sequence and use exponential smoothing to obtain the trend prediction quantity used to determine the direction of signal change; Step S43: The square root of the local fluctuation feature quantity, the absolute value of the trend prediction quantity, and the absolute value of the difference between the local fluctuation feature quantities at adjacent times are weighted and summed to obtain the dynamic amplitude limiting threshold at the current time. The weighting coefficients in the weighted summation take the first set of preset values ​​when the signal is in the transient stage and the second set of preset values ​​when the signal is in the steady state stage. Step S44: If the absolute value of the difference between the current signal sample value and the previous filtered output value does not exceed the dynamic amplitude limiting threshold, then the current sample value is used as the filtered output; otherwise, the threshold step size is adjusted in the direction of the current sample value based on the previous filtered output value to obtain the current filtered output.

[0047] In step S40 of this embodiment, the sub-band coefficients processed in step S30 are reconstructed using wavelet packets to obtain a denoised signal sequence. Based on this, a dynamic amplitude limiting threshold is constructed according to the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence, and the denoised signal sequence is subjected to collaborative amplitude limiting filtering to further suppress residual interference components, resulting in a filtered signal sequence.

[0048] The process of constructing the dynamic amplitude limiting threshold is as follows: Calculate the variance of the denoised signal sequence within the sliding window, and use this variance as a local fluctuation characteristic of the short-term fluctuation degree of the signal. For the current time n, take W consecutive samples, including the current sample, to form a sliding window, and calculate the variance of the sample values ​​within this window, denoted as... Its calculation formula is ,in This represents the local fluctuation characteristic at the current moment. This indicates the length of the sliding window, i.e., the number of sample points. This represents the denoised signal value of the i-th sample point. This represents the arithmetic mean of the sample values ​​within the window. This local fluctuation characteristic reflects the dispersion of the signal near the current time; the greater the fluctuation, the greater the variance.

[0049] Calculate the first-order difference of the denoised signal sequence, and use exponential smoothing to obtain a trend prediction value for determining the direction of signal change. Calculate the first-order difference between the current sample and the previous sample, and perform exponential smoothing on this difference value to obtain the trend prediction value. Its calculation formula is ,in This represents the trend forecast at the current moment. Represents the exponential smoothing coefficient. This represents the denoised signal value of the current sample point. This represents the denoised signal value of the previous point. This represents the trend prediction from the previous moment. Exponential smoothing can filter out high-frequency noise in the difference sequence, allowing the trend prediction to accurately reflect the direction and rate of signal change. When the signal rises rapidly... When the value is positive and has a large absolute value, and decreases rapidly... When it is negative and has a large absolute value, it is stable. Approaching zero.

[0050] The dynamic limiting threshold for the current moment is obtained by weighted summing the square root of the local fluctuation characteristic, the absolute value of the trend prediction, and the absolute value of the difference between the local fluctuation characteristic at adjacent times. The formula for calculating the dynamic limiting threshold is as follows: ,in, This indicates the dynamic threshold value at the current moment. This indicates the preset basic threshold. , , Indicates non-negative weighting coefficients. This represents the local fluctuation characteristic at the current moment. This represents the trend forecast at the current moment. It represents the absolute value of the difference in local fluctuation characteristics between adjacent time points.

[0051] The weighting coefficients are adjusted according to the stage of the signal. The initial start time is taken as the moment when the input voltage triggers the start timing. At the current moment satisfy When the signal is in a transient phase, it is considered to be in a transient phase; when the current moment... satisfy When the output voltage has entered the preset stable range and continues to reach the preset delay period, the signal is determined to be in a steady state; when the current time... satisfy At that time, the signal is identified as a transition phase.

[0052] During the transient phase, the basic limiting threshold and weighting coefficients in the dynamic limiting threshold formula are taken from the first set of preset values; during the steady-state phase, the basic limiting threshold and weighting coefficients in the dynamic limiting threshold formula are taken from the second set of preset values; during the transition phase, the basic limiting threshold and weighting coefficients in the dynamic limiting threshold formula are determined by linear interpolation between the first set of preset values ​​and the second set of preset values.

[0053] As an optional implementation, when the processed signal is a voltage signal, the rated voltage value of the channel is used as the reference quantity. The basic limiting threshold for the transient phase can be 1.5%-3% of the rated voltage value, the weighting coefficient corresponding to the local fluctuation characteristic quantity can be 1.2-2.0, the weighting coefficient corresponding to the trend prediction quantity can be 1.0-1.8, and the weighting coefficient corresponding to the difference in local fluctuation characteristic quantities between adjacent time points can be 0.2-0.8. For the steady-state phase, the basic limiting threshold can be 0.3%-1% of the rated voltage value, the weighting coefficient corresponding to the local fluctuation characteristic quantity can be 0.5-1.0, the weighting coefficient corresponding to the trend prediction quantity can be 0.2-0.8, and the weighting coefficient corresponding to the difference in local fluctuation characteristic quantities between adjacent time points can be 0.1-0.4. For current and temperature signals, the rated range or full-scale range of the corresponding channel can be used as the reference quantity, and the basic limiting threshold can be determined according to the same proportional relationship. With the above settings, the dynamic limiting threshold in the transient phase is greater than the dynamic limiting threshold in the steady-state phase to avoid excessive clamping of the real step change during the startup process; the dynamic limiting threshold in the steady-state phase is relatively narrowed to improve the fluctuation suppression effect of the steady-state test parameters.

[0054] In this embodiment, the collaborative amplitude limiting filtering of the denoised signal sequence includes: calculating the sign of the difference between the current signal sample value and the previous filtered output value, and comparing the sign of the difference with the sign of the trend prediction; if the sign of the difference is consistent with the sign of the trend prediction, then the current signal sample value is output according to the dynamic amplitude limiting threshold; if the sign of the difference is inconsistent with the sign of the trend prediction, and the absolute value of the difference between the current signal sample value and the previous filtered output value exceeds the dynamic amplitude limiting threshold, then the current signal sample value is determined to be a reverse glitch sample, and the current filtered output is generated based on the previous filtered output value, the trend prediction, and a reverse correction step size less than the dynamic amplitude limiting threshold; when the sign of the difference of multiple consecutive subsequent samples is consistent with the sign of the trend prediction, the reverse glitch determination is lifted.

[0055] As an optional implementation, in step S40, when performing collaborative amplitude limiting filtering on the denoised signal sequence, the reverse spike samples can also be identified by combining the trend prediction value. Specifically, the denoised signal sample value at the current moment is recorded as... The filtered output value from the previous moment is denoted as The current trend forecast is denoted as Calculate the difference .when symbols and The signs are consistent, or When the absolute value is less than the preset trend lower limit, the current signal sample value is followed and output according to the dynamic amplitude limiting threshold, that is, based on the previous filtered output value. Based on, along The direction adjustment does not exceed the step size of the dynamic limiting threshold to obtain the current filter output. .

[0056] when symbols and The signs are inconsistent, and When the current signal sample value exceeds the dynamic limiting threshold, it is identified as a reverse glitch sample. For reverse glitch samples, the adjustment is not directly based on the direction of the current signal sample value, but rather on the previous filtered output value. Based on trend forecast The direction of the trend prediction is corrected in the opposite direction, with the step size being less than the dynamic limit threshold at the current time, preferably 0.3-0.6 times the dynamic limit threshold at the current time; if the trend prediction amount If the absolute value is less than the preset trend lower limit, then the previous filter output value is maintained. This is the current filter output. If subsequent consecutive... If, in each sample point, the sign of the difference between the current signal sample value and the previous filtered output value is consistent with the sign of the trend prediction, then the reverse glitch determination is lifted. Option 2-4 is acceptable. Through the above trend consistency verification, it is possible to distinguish between the actual start-up step change and the glitch interference that is opposite to the overall trend, thereby reducing false amplitude limiting and false following under strong electromagnetic interference.

[0057] After obtaining the dynamic amplitude limiting threshold, collaborative amplitude limiting filtering is performed on the denoised signal sequence. The sample value of the denoised signal at the current moment is denoised as... The filtered output value at the previous moment is denoted as . judge and Does the absolute value of the difference not exceed the dynamic limiting threshold at the current moment? If the value does not exceed the specified limit, the current sample value is considered a valid signal, and the current sample value is directly used as the filtered output at the current moment. ,in This represents the current filtered output value. If it exceeds the threshold, it is determined that the current sample value contains abnormal fluctuations. In this case, the threshold step size is adjusted from the previous filtered output value towards the current sample value to obtain the current filtered output. ,in The sign function takes a value of +1 when the independent variable is positive, -1 when it is negative, and zero when it is zero. In this embodiment, the threshold step size is equal to the dynamic limiting threshold at the current moment. This adjustment method preserves the true trend of signal change while limiting the impact of abnormal fluctuations on the filtered output.

[0058] Through the above-mentioned collaborative amplitude limiting filtering, based on the time-frequency decomposition and coefficient adaptive processing in step S30, this embodiment further utilizes the local statistical characteristics and trend information of the signal to dynamically adjust the amplitude limiting threshold, effectively suppressing the residual components of pulse interference and periodic oscillation interference in the subway environment, and obtaining a smooth and undistorted filtered signal sequence.

[0059] like Figure 2As shown, the green curve represents the signal curve after noise reduction before collaborative amplitude limiting filtering, while the purple curve represents the filtered signal curve after collaborative amplitude limiting filtering. The green curve still contains several high-amplitude pulse interference segments and low-amplitude periodic oscillation interference segments. The high-amplitude pulse interference segments manifest as localized sharp spikes, while the low-amplitude periodic oscillation interference segments exhibit continuous small fluctuations around the actual signal trend. After collaborative amplitude limiting filtering in this embodiment, the dynamic amplitude limiting threshold can adaptively adjust based on local fluctuation characteristics and trend prediction, limiting and correcting the residual components of high-amplitude pulse interference and continuously suppressing the residual components of low-amplitude periodic oscillation interference. This results in a significantly reduced number of spikes and periodic jitters in the purple curve compared to the green curve, while preserving the overall signal trend, thus obtaining a smoother filtered signal curve that more closely approximates the actual output changes of the power supply module under test.

[0060] Step S50: Based on the filtered signal sequence, obtain the steady-state operating parameters and start-up time of the signal power supply module as test parameters; In this embodiment, obtaining the start-up time of the signal power supply module based on the filtered signal sequence includes: Step S51: Continuously monitor the filtered input voltage signal sequence. When the input voltage jumps from a preset ratio below the rated value to a preset ratio above the rated value, trigger the start timing to begin counting. Step S52: Continuously monitor the filtered output voltage signal sequence. When the output voltage first enters the preset stable range near the rated value, start the turn-on delay judgment. If the output voltage remains stable within the preset stable range during the preset delay period, stop the start-up timing and use the accumulated time of the start-up timing as the start-up time.

[0061] In step S50 of this embodiment, based on the filtered signal sequence, the steady-state operating parameters and start-up time of the signal power module are obtained as test parameters. The steady-state operating parameters include output voltage, output current, ripple voltage, module operating temperature, and input voltage and input current. The programmable logic controller (PLC) calculates the values ​​of the above parameters according to the filtered signal sequences of each channel and a preset test standard. Specifically, the output voltage is the average time value of the filtered signal sequence during the steady-state phase, the output current is also the average value during the steady-state phase, the ripple voltage is obtained by extracting the AC component from the filtered output voltage signal and calculating its peak-to-peak value, and the module operating temperature is converted into an actual temperature value from the filtered signal sequence corresponding to the temperature sensor.

[0062] The specific process for obtaining the startup time of the signal power module is as follows.

[0063] The filtered input voltage signal sequence is continuously monitored. A rated voltage value and a preset ratio are set; in this embodiment, the preset ratio is 10% of the rated voltage. When the input voltage is detected to jump from 10% below the rated voltage to 10% above the rated voltage, a rising edge trigger instruction inside the PLC starts a high-precision internal timer, denoted as timer T20. This timer starts accumulating time in 0.1ms increments.

[0064] Simultaneously, the filtered output voltage signal sequence is continuously monitored. A preset stable range near the rated voltage is set; in this embodiment, the preset stable range is 95% to 105% of the rated voltage. When the output voltage first enters this preset stable range, the PLC starts a turn-on delay judgment timer TON1, with a preset delay period of 10ms. During the 10ms delay period, it continuously judges whether the output voltage remains stable within the preset stable range. If the output voltage does not exceed this range within 10ms, the contact of the turn-on delay judgment timer TON1 closes, triggering the stop of the start timer T20. The accumulated time of the start timer T20 is the start time of the signal power supply module, and this time value is stored in the D403 register as a test parameter.

[0065] If the output voltage exceeds the preset stable range within the 10ms delay period, the delay judgment timer is reset, and the delay judgment is restarted after the output voltage returns to the preset stable range. Through the above mechanism, combined with the high signal-to-noise ratio filtered signal sequence output in step S40, this embodiment can stably capture the starting point of the input voltage rise and the stable establishment time of the output voltage, reducing the impact of interference glitches on the start-up time judgment.

[0066] After obtaining the above steady-state operating parameters and startup time, the PLC encapsulates these test parameters according to the Modbus RTU protocol and stores them in the corresponding holding registers, waiting for data transmission in subsequent steps.

[0067] Step S60: The test parameters are sent to the cloud platform via a wireless network for storage and display.

[0068] In this embodiment, step S60 includes: Step S61: Encapsulate the test parameters according to a preset communication protocol and upload them through a wireless communication link at a set polling interval; Step S62: The uploaded test parameters are stored in the database in the form of data frames, and real-time display and historical data download are performed.

[0069] In step S60 of this embodiment, the test parameters are sent to the cloud platform via a wireless network for storage and display.

[0070] The programmable logic controller (PLC) encapsulates the steady-state operating parameters and start-up time test parameters obtained in step S50 according to the Modbus RTU protocol. The PLC is configured as a Modbus RTU slave, with the station address set to 1, the communication baud rate set to 9600, data bits set to 8, parity set to none, and stop bits set to 1. The Internet of Things (IoT) module (DTU) acts as a communication gateway, establishing a wired communication connection with the PLC's RS485 interface via an RVSP shielded cable. The DTU is configured as a Modbus RTU master, polling and reading the test parameters from the PLC's holding registers. The DTU's uplink establishes a wireless communication connection with the IoT cloud platform via its built-in 4G wireless communication module, associating the DTU's device serial number with the cloud platform server to achieve point-to-point communication between the cloud platform and the DTU.

[0071] The IoT cloud platform uses cloud-based polling for data collection, with a polling interval of 1 second. This means the cloud platform sends a data read request to the remote DTU every second. The DTU then uploads the test parameters obtained from the PLC at the current moment to the cloud platform via a 4G wireless link. This configuration ensures that test parameters are continuously uploaded at a second-level frequency, significantly improving the temporal resolution of the testing process.

[0072] After receiving the uploaded test parameters, the cloud platform stores them persistently in the cloud database as data frames. Each data frame corresponds to a timestamp and its associated test parameter values. Simultaneously, the cloud platform displays the received test parameters in real-time on the monitoring interface of the operating terminal. Operators can observe the changes in parameters such as output voltage, output current, and module temperature throughout the entire test process via a webpage or client application. Operators can also download historical data from the database of the cloud platform and export it as an Excel spreadsheet for data analysis and compliance assessment after the test is completed. Through the above data transmission, storage, and display mechanisms, this embodiment achieves remote monitoring and data sharing of the testing process. Repair personnel, users, and third parties can view the test data at any time, ensuring the impartiality of the testing process.

[0073] like Figures 3 to 7 As shown, the second embodiment of the present invention proposes a subway signal power supply detection system based on Internet of Things (IoT) technology, the system comprising: The test box 100 integrates an AC / DC power module 110, a programmable controller 120, and an Internet of Things (IoT) module 130. The AC / DC power module 110 is used to provide test power to the power module 600 under test. The programmable controller 120 is used to collect test process parameters. The IoT module 130 is used to interact with the cloud platform 400. The electronic load cell 200 is connected to the output terminal of the power supply module 600 for the signal under test. Sensor 300 is used to detect the input voltage, input current, output voltage, output current and module temperature of the power supply module 600 under test; The cloud platform 400 establishes a communication connection with the programmable controller 120 through the Internet of Things module 130, and is used to receive, store and display test data; The operation terminal 500 is used to access the cloud platform 400; The programmable controller 120 is configured to execute a subway signal power supply detection method based on Internet of Things (IoT) technology.

[0074] See Figure 3 In this embodiment, the test box 100 integrates an AC / DC power module 110, a programmable controller 120, and an Internet of Things (IoT) module 130. The AC / DC power module 110 steps down, rectifies, and regulates the external AC power input, converting it into the DC test power required by the power supply module 600 under test, and providing test power to the module. The programmable controller 120, as the underlying data acquisition and test control unit, receives test process parameters such as input voltage, input current, output voltage, output current, and module temperature generated during the test. The IoT module 130, as a communication gateway, establishes a downlink communication connection with the programmable controller 120 and an uplink communication connection with the cloud platform 400, enabling data interaction between the programmable controller 120 and the cloud platform 400.

[0075] The electronic load tester 200 is connected to the output terminal of the power supply module 600 under test. The electronic load tester 200 is used to simulate the load conditions of the power supply module 600 under test in actual operation, and can adjust the load current within the range of no-load to full-load to meet the requirements of different load conditions in functional testing, load testing and stability testing.

[0076] The sensor 300 is used to detect the input voltage, input current, output voltage, output current, and module temperature of the power supply module 600 under test. In this embodiment, the input voltage and output voltage can be DC voltage sensors, the input current and output current can be DC current sensors, and the module temperature can be a PT100 platinum resistance temperature sensor. The sensor 300 converts the detected physical quantity under test into an analog or digital signal that can be acquired by the programmable controller 120, and connects it to the corresponding acquisition channel of the programmable controller 120 so that the programmable controller 120 can obtain the test process parameters of the power supply module 600 under test.

[0077] The cloud platform 400 establishes a communication connection with the programmable controller 120 via the Internet of Things (IoT) module 130. The cloud platform 400 receives test data uploaded by the programmable controller 120 through the IoT module 130, and stores and displays the received test data. Specifically, the cloud platform 400 can store the test data in the form of data frames or data records in a database, and display the change curves of parameters such as input voltage, input current, output voltage, output current, and module temperature in real time on the monitoring interface, so that operators can view the test status of the power supply module 600 under test.

[0078] The operating terminal 500 is used to access the cloud platform 400. The operating terminal 500 can be a personal computer, tablet computer, or smartphone. Operators can log in to the cloud platform 400 through a web browser or a dedicated client application to remotely monitor the testing process, view the test results, and download test data files for subsequent analysis.

[0079] As a further explanation of the present invention, the AC / DC power module 110 is encapsulated in the power box 140, and the power box 140 is electrically isolated from the test box 100. The programmable controller 120 and the Internet of Things module 130 are mounted on the control board 150; the test box 100 is equipped with a cooling fan 160, a voltage and current display meter 170, an output switch 210 and a test output interface 220. The voltage and current display meter 170 is used to display the test voltage and test current, and the output switch 210 is used to control the on / off state of the test output interface 220.

[0080] See Figure 4 , Figure 5 , Figure 6 and Figure 7 Regarding the internal structure of the test box 100, the AC / DC power module 110 is encapsulated within a separate power supply box 140. The power supply box 140 may be made of insulating material, or insulating gaskets, insulating supports, or insulating partitions may be provided between the power supply box 140 and the test box 100 to form electrical isolation between the power supply section containing the AC / DC power module 110 and the box of the test box 100, thereby reducing the possibility of leakage, short circuit, or accidental contact from the AC input side or DC conversion side to the box.

[0081] The programmable logic controller (PLC) 120 and the Internet of Things (IoT) module 130 are mounted on a control board 150. The control board 150 can be a printed circuit board, a control circuit mounting board, or a mounting board with terminal blocks. The PLC 120 and IoT module 130 are fixed to the control board 150 and connected for power and communication via lines, terminals, or communication interfaces on the control board 150. The PLC 120 receives test process parameters collected by the sensor 300, and the IoT module 130 uploads the test data obtained by the PLC 120 to the cloud platform 400.

[0082] The test chamber 100 is equipped with a cooling fan 160, a voltage and current display meter 170, an output switch 210, a test output interface 220, and a power input interface 180. The cooling fan 160 is located on the side wall, rear wall, or a ventilation point of the test chamber 100 to dissipate heat generated inside the test chamber 100 by the operation of the AC / DC power module 110, the programmable controller 120, and the IoT module 130, thus maintaining the internal temperature of the test chamber 100 within the allowable operating range of the devices. The power input interface 180 is used to introduce external AC power and connect the external AC power to the input side of the AC / DC power module 110.

[0083] A voltage and current display meter 170 is installed on the panel of the test chamber 100 to display the test voltage and test current. Specifically, multiple voltage and current displays 170 can be configured, each corresponding to a different test voltage level or a different test channel. For example, multiple voltage and current displays 170 can correspond to a 110VDC test channel, a 60VDC test channel, a 48VDC test channel, and a 24VDC test channel, respectively, to display the voltage and current values ​​of the corresponding test channels. Thus, the operator can directly observe the operating status of each test channel on the panel of the test chamber 100.

[0084] Output switches 210 are located on the panel of the test box 100 and are used to control the on / off state of the test output interface 220. Specifically, multiple output switches 210 can be configured, each connected to a corresponding test output interface 220, allowing different test channels to be independently switched on or off. For example, the multiple output switches 210 can correspond to 110VDC, 60VDC, 48VDC, and 24VDC test channels, respectively. Before testing or when switching test voltage levels, the operator can control whether the corresponding test output interface 220 outputs test power to the power supply module 600 or the test circuit through the corresponding output switch 210.

[0085] Test output interface 220 is disposed on the panel of test box 100 and is used to output the DC test power generated by AC / DC power module 110. Multiple test output interfaces 220 can be configured, each corresponding to a different DC output voltage level. For example, test output interfaces 220 include a 110VDC test output interface, a 60VDC test output interface, a 48VDC test output interface, and a 24VDC test output interface. Each test output interface 220 may include a positive terminal and a negative terminal for connection to the power module 600 under test or a test circuit via a test cable.

[0086] Regarding the panel structure of the test chamber 100, the voltage and current display meter 170, output switch 210, and test output interface 220 can be arranged sequentially in a vertical direction. Specifically, the voltage and current display meter 170 is located at the top of the panel for easy observation of the test voltage and current by the operator; the output switch 210 is located below the voltage and current display meter 170 for easy control of connecting or disconnecting each test channel; and the test output interface 220 is located below the output switch 210 for easy connection of test cables. The voltage and current display meter 170, output switch 210, and test output interface 220 of each test channel can be arranged horizontally on the panel to form a group of display, control, and output ports of the same voltage level, reducing the possibility of wiring and operational errors. The power input interface 180 can be located on the rear, side, or other convenient location for external power cord access in the test chamber 100 to avoid confusion with the test output interface 220.

[0087] Through the structural design of the test box 100, the AC / DC power module 110, the programmable controller 120, and the Internet of Things module 130 can be integrated into the same test box 100; the AC / DC power module 110 is electrically isolated from the test box 100 through the power supply box 140; the power input interface 180 is used to introduce external AC power; multiple voltage and current display meters 170, multiple output switches 210, and multiple test output interfaces 220 can correspond to different test channels respectively, realizing multi-voltage level test output, independent on / off control of channels, and on-site display of test status.

[0088] During the system setup process, the programmable controller 120, the IoT module 130, and the cloud platform 400 need to be configured. The communication port of the programmable controller 120 can be configured using ISPSoft software. For example, the station address can be set to 1, the communication baud rate to 9600, the data bits to 8, the parity to no parity, the stop bits to 1, and the communication mode to Modbus RTU mode. After configuration, the settings are downloaded to the programmable controller 120.

[0089] The IoT module 130 can be a DTU module, which can be configured using 4G CAT1 setup software. Specifically, the target server address and port for uplink communication of the DTU module can be set, and the global parameters for downlink communication can be set to 9600 baud rate, 8 data bits, 1 stop bit, and no parity, ensuring that the communication parameters of the IoT module 130 and the programmable controller 120 are consistent. Simultaneously, the device serial number of the DTU module is associated and bound to the server of the cloud platform 400 so that the cloud platform 400 can identify the corresponding test box 100.

[0090] The cloud platform 400 can establish a communication gateway and write the device serial number of the IoT module 130 into this gateway. A remote programmable controller (PLC) device is established within the cloud platform 400, written into the cloud configuration, and its communication serial port number and slave address of the PLC 120 are set. The communication protocol between the cloud platform 400 and the remote PLC 120 is configured as Modbus RTU. Furthermore, the register address, data type, and storage method for data exchange are defined in the cloud platform 400 and associated with the established communication gateway. After completing the above configuration, the test box 100, IoT module 130, cloud platform 400, and operation terminal 500 can then be put into operation.

[0091] This subway signal power supply testing system is particularly suitable for the maintenance and testing of subway signal power supply modules, as well as type testing of new power supply products, including high-temperature testing, low-temperature testing, damp heat testing, high-voltage testing, low-voltage testing, overload testing, and vibration testing. Test data can be displayed and played back in real-time via a cloud platform, and can be downloaded in Excel format. The data acquisition frequency can be set to once per second, facilitating various performance analyses of the tested signal power supply module 600.

[0092] The terms “first”, “second”, etc., are used to distinguish similar objects, not to describe or indicate a specific order or sequence.

[0093] The term "comprising" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent in such process, method, article, or apparatus / device.

[0094] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A method for detecting subway signal power supply based on Internet of Things (IoT) technology, characterized in that, include: Acquire multiple analog signals during the testing process of the power supply module under test; The multiple analog signals are converted from analog to digital to obtain the original digital sampling sequence; The original digital sampling sequence is decomposed into time-frequency multi-resolution components to obtain multiple sub-frequency band components. The interference type is determined by comparing the energy of the sub-frequency band components with a preset first energy threshold and a second energy threshold. Based on the determination result, the coefficients of the sub-frequency band components are retained, zeroed, or amplitude-shrinked. The first energy threshold is less than the second energy threshold. The processed coefficients are reconstructed to obtain a denoised signal sequence. A dynamic amplitude limiting threshold is constructed based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence. The denoised signal sequence is then subjected to collaborative amplitude limiting filtering to obtain a filtered signal sequence. The interference types include at least pulse interference and periodic oscillation interference in the subway environment. Based on the filtered signal sequence, the steady-state operating parameters and startup time of the signal power supply module are obtained as test parameters; The test parameters are sent to the cloud platform via a wireless network for storage and display. The process includes constructing a dynamic amplitude limiting threshold based on the local statistical fluctuation characteristics and signal change trend of the denoised signal sequence, and performing collaborative amplitude limiting filtering on the denoised signal sequence, including: Calculate the variance of the denoised signal sequence within the sliding window, and use it as a local fluctuation characteristic of the short-term fluctuation of the signal; Calculate the first-order difference of the denoised signal sequence and use exponential smoothing to obtain a trend prediction quantity for determining the direction of signal change; The dynamic amplitude limiting threshold at the current moment is obtained by weighted summing of the square root of the local fluctuation feature quantity, the absolute value of the trend prediction quantity, and the absolute value of the difference between the local fluctuation feature quantities at adjacent times. The weighting coefficients in the weighted summation take the first set of preset values ​​when the signal is in the transient stage and the second set of preset values ​​when the signal is in the steady state stage. If the absolute value of the difference between the current signal sample value and the previous filtered output value in the denoised signal sequence does not exceed the dynamic limiting threshold, then the current signal sample value in the denoised signal sequence is taken as the current filtered output. Otherwise, based on the filtered output value of the previous moment, a threshold step size is adjusted towards the signal sample value of the current moment in the denoised signal sequence to obtain the filtered output of the current moment.

2. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, The multi-channel analog signals include at least the input voltage signal, the input current signal, the output voltage signal, the output current signal, and the temperature signal.

3. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, The multiple analog signals are subjected to analog-to-digital conversion to obtain the original digital sampling sequence, including: The multi-channel analog signal is sampled at a first sampling frequency, and a first number of signal samples are continuously collected. After removing the sample with the largest amplitude and the sample with the smallest amplitude from the first number of signal samples, the average value of the remaining samples is calculated as the first-level filtering sequence. The first-level filtered sequence is sampled at a second sampling frequency, and a second number of samples are continuously collected. After removing the maximum and minimum values, the average value is calculated to obtain the original digital sampling sequence. The first sampling frequency is higher than the second sampling frequency. The sampling performed at the first sampling frequency and the sampling performed at the second sampling frequency are executed in a nested loop manner, and the interrupt priority of the sampling performed at the first sampling frequency is configured to be higher than the interrupt priority of the sampling performed at the second sampling frequency.

4. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, The original digital sampling sequence is subjected to time-frequency multi-resolution decomposition, including: The original digital sampling sequence is decomposed into multiple wavelet packets using wavelet packet basis functions to generate multiple sub-band components corresponding to different frequency ranges. Each sub-band component is composed of a set of wavelet packet coefficients. Based on the spectral distribution of traction inverter harmonic interference and relay operation oscillation interference in the subway environment, the high-frequency sub-bands among the multiple sub-band components are associated with pulse interference frequency bands, and the low-frequency sub-bands are associated with periodic oscillation frequency bands.

5. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, Based on a comparison of the energy of the sub-frequency band component with a preset first energy threshold and a second energy threshold, the interference type is determined, and a corresponding amplitude adjustment strategy is applied to process the coefficients of the sub-frequency band component, including: Calculate the energy of the sub-frequency band component. If the energy is greater than the second energy threshold, it is determined to be dominated by pulse interference. Set the abnormal coefficients in the sub-frequency band whose amplitude is greater than the preset noise reduction threshold to zero, and retain the other coefficients. If the energy is not less than the first energy threshold and not greater than the second energy threshold, it is determined to be mixed interference. The absolute value of the coefficient is subtracted from the shrinkage threshold and the original sign is retained. If the result is less than zero, it is set to zero. The shrinkage threshold is determined based on the noise standard deviation estimate of the coefficient in the sub-band. If the energy is less than the first energy threshold, the original coefficients are retained; Wherein, the first energy threshold is less than the second energy threshold, and the noise standard deviation estimate is estimated using the median absolute deviation of the sub-band coefficients.

6. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, Based on the filtered signal sequence, the startup time of the signal power supply module is obtained, including: The filtered signal sequence includes a filtered input voltage signal sequence and a filtered output voltage signal sequence; The filtered input voltage signal sequence is continuously monitored. When the input voltage in the filtered input voltage signal sequence jumps from a preset proportion below the rated value to a higher proportion above the preset value, the timing is triggered to start counting. The filtered output voltage signal sequence is continuously monitored. When the output voltage in the filtered output voltage signal sequence first enters a preset stable range near the rated value, a turn-on delay judgment is initiated. If the output voltage in the filtered output voltage signal sequence remains stable within the preset stable range during the preset delay period, the start-up timing is stopped, and the accumulated time of the start-up timing is taken as the start-up time.

7. The method for detecting subway signal power supply based on Internet of Things technology according to claim 1, characterized in that, The step of sending the test parameters to the cloud platform via a wireless network for storage and display includes: The test parameters are encapsulated according to a preset communication protocol and uploaded via a wireless communication link at a set polling interval. The uploaded test parameters are stored in the database in the form of data frames, and are displayed in real time and downloaded as historical data.

8. A subway signal power supply detection system based on Internet of Things (IoT) technology, characterized in that, The system includes: The test box (100) integrates an AC / DC power module (110), a programmable controller (120), and an Internet of Things (IoT) module (130). The AC / DC power module (110) is used to provide test power to the power module (600) under test. The programmable controller (120) is used to collect test process parameters. The IoT module (130) is used to interact with the cloud platform (400) for data exchange. An electronic load cell (200) is connected to the output terminal of the power supply module (600) for the signal under test; The sensor (300) is used to detect the input voltage, input current, output voltage, output current and module temperature of the power supply module (600) under test; The cloud platform (400) establishes a communication connection with the programmable controller (120) through the Internet of Things module (130) for receiving, storing and displaying test data; An operating terminal (500) is used to access the cloud platform (400). The programmable controller (120) is configured to execute the subway signal power supply detection method based on Internet of Things technology as described in any one of claims 1-7.

9. A subway signal power supply detection system based on Internet of Things technology according to claim 8, characterized in that, The AC / DC power module (110) is encapsulated in a power box (140), and the power box (140) is electrically isolated from the test box (100); The programmable controller (120) and the Internet of Things module (130) are mounted on the control board (150); the test box (100) is equipped with a cooling fan (160), a voltage and current display meter (170), an output switch (210) and a test output interface (220). The voltage and current display meter (170) is used to display the test voltage and test current, and the output switch (210) is used to control the on / off state of the test output interface (220).

Images