A device and method for detecting off-line frequency of pantograph arc

Abstract

The application discloses a kind of detection device and method for catenary electric arc offline frequency, including signal acquisition unit, data processing unit, model construction unit and result output unit.The application has the following advantages and effects: the detection device for catenary electric arc offline frequency proposed in the application can realize the early warning function of catenary system offline, and can send offline frequency information to user mobile terminal in real time;The detection device for catenary electric arc offline frequency proposed in the application can compare the offline frequency of catenary system electric arc with different speeds and roughness, so as to obtain the optimal state, and further reduce the adverse effects of offline arc;By analyzing the loop current signal of catenary system to detect the offline frequency of electric arc, the application has the advantages of high detection precision and simple detection method.

Description

Technical Field

[0001] This invention relates to the field of pantograph-catenary arc detection, and particularly to a device and method for detecting the offline frequency of pantograph-catenary arcs. Background Technology

[0002] High-speed railway trains obtain traction current through the sliding electrical contact between the pantograph and the overhead contact line (i.e., the pantograph-catenary system). During operation, the pantograph and catenary must maintain good contact to ensure the stability of this high-speed current collection. However, in actual operation, various factors such as vehicle vibration, uneven contact wires, and the passage of the neutral zone can cause the pantograph and catenary to momentarily separate, generating an offline arc. This offline arc causes instantaneous high temperatures between the sliding friction pairs, resulting in ablation of the pantograph's contact plate and the contact wire material, accelerating the aging of the contact plate and wires, reducing the service life of the friction pairs, and affecting the quality of current collection.

[0003] Currently, the mainstream methods for detecting offline frequency of pantograph-catenary arcs both domestically and internationally fall into three main categories: optical measurement, resistance measurement, and pantograph current measurement. First, optical measurement mainly includes spark measurement and ultraviolet (UV) measurement. Spark measurement does not require contact with the high-voltage portion of the pantograph; it uses a row of light receivers in a lens system to achieve offline detection. However, due to the very small and weak arc of the offline line, this detection technique cannot accurately detect offline issues in DC systems or when the locomotive is coasting. UV measurement is suitable for high-speed locomotives, achieving offline detection by measuring the UV light emitted when the pantograph is offline. However, it requires continuous monitoring of the locomotive's operating range. Second, resistance measurement utilizes the principle of increased resistance between the pantograph and the contact wire, mainly including the resistance increase method and the resistance-capacitance method. The resistance-increasing method uses the sequence "isolation capacitor—pantograph—contact wire—contact wire-to-ground distributed capacitance—rail." When the pantograph and contact wire are in contact, the current forms a loop. If a "departure" occurs, the impedance between the pantograph and contact wire increases, reducing the current flowing into the detection circuit. This allows current to flow into the circuit, thus detecting the departure signal. The resistance-capacitor method connects two capacitors and a resistor between the pantograph and the ground. The charging and discharging current flowing through the pantograph's capacitors changes the contact wire voltage when the pantograph and contact wire are in good condition; however, the current changes when there is a departure. However, due to the harsh environment and numerous interference sources surrounding train operation, the parameters or waveforms measured by the resistance method are easily affected by environmental noise, resulting in poor detection accuracy. A third method is the pantograph current measurement method. Because the pantograph-contact wire departure arc is an AC arc, when the current passes through the zero point, and the voltage between the plates is less than a certain value, the current in the arc gap gradually decreases to zero, and the arc enters a zero-rest state. The number of zero-rest states is used to further determine the number of departures.

[0004] While current pantograph current measurement methods can easily detect pantograph outages, they rarely consider the impact of pantograph contactor surface roughness on the outage rate. This is because pantograph-catenary line outage arcs cause temperature rise changes on the contact surfaces of the friction pair, leading to ablation of the pantograph contactor and contact wire, thus deteriorating the contact surface. Therefore, this invention provides a method and apparatus for detecting pantograph-catenary line outage frequency under different surface roughness conditions. Summary of the Invention

[0005] The purpose of this invention is to provide a device and method for detecting the offline frequency of pantograph-catenary arc, so as to solve the problems mentioned in the background art.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0007] To achieve the above objectives, the present invention provides a device for detecting the offline frequency of pantograph-catenary arcs, comprising a signal acquisition unit, a data processing unit, a model building unit, and a result output unit; wherein,

[0008] The signal acquisition unit is used to collect offline arc current signals during pantograph-catenary operation and synchronously transmit the signal data to the data processing unit.

[0009] The data processing unit is used to acquire offline current samples based on the offline arc current signal, filter out noise interference by median filtering, add labels to the data samples to obtain the offline arc current dataset, and draw the offline arc current waveform.

[0010] The model building unit is used to construct an input matrix from the data samples in the offline arc current dataset and build an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events.

[0011] The result output unit is used to obtain the number of offline times output by the model building unit, obtain the offline frequency, and push it to the user's mobile terminal.

[0012] A further configuration is as follows: the signal acquisition unit includes a current transformer, a data acquisition card, and a DC power supply; wherein,

[0013] The aforementioned current transformer is used to convert the loop current signal into a voltage signal during the operation of the pantograph-catenary system;

[0014] The data acquisition card is used to perform analog-to-digital conversion on the voltage signal.

[0015] The DC power supply is used to provide power to the current transformer and the data acquisition card.

[0016] A further configuration is as follows: the model construction unit includes an input matrix construction module, a batch reading module, and a state recognition module; wherein,

[0017] The input matrix construction module constructs the input matrix sequentially based on the collected offline arc current data samples;

[0018] The batch reading module is used to read the data samples constructed as an input matrix into the state recognition model in batches;

[0019] The aforementioned state recognition module is used to build an offline state recognition model.

[0020] A further setting is that the matrix constructed by the input matrix construction module has a size of 1,000,000 × 1, and is composed of data samples of the offline arc current collected.

[0021] A further configuration is as follows: the result output unit includes a detection unit and an information push unit; wherein,

[0022] The detection unit is used to obtain the number of offline events output by the model building unit, determine the frequency of offline events generated by the pantograph-catenary system, and send the offline frequency information to the information push unit.

[0023] The information push unit is used to receive offline frequency information sent by the detection unit and push fault information to the user's mobile terminal.

[0024] A further configuration is as follows: the information push unit includes an information receiving module and an information push module;

[0025] in,

[0026] The information receiving module is used to receive arc offline frequency information.

[0027] The information push module is used to push the latest received arc offline frequency information to the user.

[0028] To achieve the above objectives, the present invention also provides a method for detecting the offline frequency of pantograph-catenary arc, comprising the following steps:

[0029] Step S1: Collect the offline arc current signal during pantograph-catenary operation through the signal acquisition unit, and synchronously transmit the signal data to the data processing unit.

[0030] Step S2: The data processing unit obtains offline current samples based on the offline arc current signal, filters out noise interference by using median filtering, adds labels to the data samples, obtains the offline arc current dataset, and plots the offline arc current waveform.

[0031] Step S3: The model building unit constructs an input matrix based on the data samples in the offline arc current dataset and builds an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events.

[0032] Step S4: The result output unit obtains the number of offline times output by the model building unit, gets the offline frequency, and pushes it to the user's mobile device.

[0033] Further settings are as follows: Step S2 specifically involves:

[0034] Step S21: Based on the collected offline arc current data samples, an input matrix is ​​constructed sequentially to obtain the input matrix construction module;

[0035] Step S22: Read the data samples that form the input matrix in batches using the batch reading module;

[0036] Step S23: Build an offline state recognition model.

[0037] Further settings are as follows: Step S23 specifically involves:

[0038] The number of offline events can be obtained by detecting the number of zero-rest periods; when the absolute value of 13 consecutive data points is less than 2A, it can be determined that an electric arc occurred at that moment.

[0039] The further setting is as follows: The specific steps for filtering out noise interference using the median filtering method in step S2 are as follows:

[0040] Step S221: First, define a window of length L with an odd number of digits, where L = 2N + 1 and N is a positive integer;

[0041] Step S222: Suppose that at a certain moment, the signal samples within the window are x(iN), ..., x(i), ..., x(i+N), where x(i) is the signal sample value located at the center of the window;

[0042] Step S223: After arranging the L signal sample values ​​in ascending order, the median value, the sample value at position i, is defined as the output value of the median filter y(i) = Med[x(iN),…,x(i),…x(i+N)]; replace the original value of x(i) in the middle of the window with the sorted median y(i).

[0043] The beneficial effects of this invention are as follows:

[0044] 1. The method for detecting the offline frequency of pantograph-catenary arc proposed in this invention can test the variation law of the offline frequency of pantograph-catenary arc under different surface roughness.

[0045] 2. This method performs median filtering on the loop current signal, thus making the results of offline arc quantity detection more accurate.

[0046] 3. The offline frequency detection device for pantograph-catenary arc proposed in this invention can compare the offline frequencies of pantograph-catenary system arcs with different speeds and roughnesses, thereby obtaining the optimal state and reducing the adverse effects of offline arcs.

[0047] 4. The pantograph-catenary arc offline frequency detection device proposed in this invention can realize the early warning function of pantograph-catenary system offline and can send offline frequency information to the user's mobile terminal in real time.

[0048] 5. The offline arc state recognition model proposed in this invention can run online in embedded devices. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the structure of the pantograph-catenary arc offline frequency detection device according to the present invention;

[0050] Figure 2 This is a schematic diagram of the median filtering method in an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the composition of the result output unit in an embodiment of the present invention;

[0052] Figure 4 This is a zero-rest feature diagram of the arc current waveform in the embodiment of the present invention;

[0053] Figure 5 This is a flowchart of the information push unit in an embodiment of the present invention. Detailed Implementation

[0054] The present invention will be further described in detail below with reference to the accompanying drawings.

[0055] As attached Figures 1 to 5 As shown;

[0056] This embodiment discloses a device for detecting the offline frequency of pantograph-catenary arcs, including a signal acquisition unit, a data processing unit, a model building unit, and a result output unit; wherein,

[0057] The signal acquisition unit is used to collect offline arc current signals during pantograph-catenary operation and synchronously transmit the signal data to the data processing unit.

[0058] The data processing unit is used to acquire offline current samples based on the offline arc current signal, filter out noise interference by median filtering, add labels to the data samples to obtain the offline arc current dataset, and draw the offline arc current waveform.

[0059] The model building unit is used to construct an input matrix from the data samples in the offline arc current dataset and build an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events.

[0060] The result output unit is used to obtain the number of offline times output by the model building unit, and to obtain the offline frequency and push it to the user's mobile device.

[0061] Specifically, the signal acquisition unit includes a current transformer, a data acquisition card, and a DC power supply; among which,

[0062] Current transformers are used to convert loop current signals into voltage signals during the operation of the pantograph-catenary system. The current transformer is a Hall type, which converts the primary signal into a voltage signal based on the Hall closed-loop zero flux principle.

[0063] A data acquisition card is used to perform analog-to-digital conversion on the voltage signal; specifically, it converts the voltage signal into a corresponding digital quantity through ADC sampling.

[0064] DC power supply, used to provide power to current transformers and data acquisition cards.

[0065] In this embodiment, model building blocks are implemented using Python 3.6 and Tensorflow 2.5.

[0066] Specifically, the model building unit includes an input matrix construction module, a batch reading module, and a state recognition module; among them,

[0067] The input matrix construction module constructs the input matrix sequentially based on the collected offline arc current data samples.

[0068] The batch reading module is used to read the data samples constructed as the input matrix into the state recognition model in batches; in this embodiment, every 200 data samples are input into the arc offline state recognition model as a batch;

[0069] The state recognition module is used to build an offline state recognition model. In this embodiment of the invention, MATLAB is used to build an offline arc state recognition model. When the absolute value of 13 consecutive data is less than 2A, it can be determined that an arc has occurred at that moment.

[0070] Specifically, the matrix constructed by the input matrix construction module has a size of 1,000,000 × 1, and is composed of data samples of the offline arc current collected.

[0071] Specifically, the result output unit includes a detection unit and an information push unit; among which,

[0072] Specifically, the detection unit is used to obtain the number of offline events output by the model building unit, determine the frequency of offline events generated by the pantograph-catenary system, and send the offline frequency information to the information push unit.

[0073] The information push unit is used to receive offline frequency information sent by the detection unit and push fault information to the user's mobile terminal.

[0074] It should be noted that, in this embodiment, an embedded device is used to perform the data processing unit and the model building unit.

[0075] An embedded device is used to acquire the current signal of the circuit, construct it into an input matrix, and input it into a state recognition model. Based on the recognition results of the offline recognition model, it determines whether an offline state exists in the pantograph-catenary system. When an offline state is detected in the line, the offline frequency information is stored and recorded, and the offline frequency information is synchronously sent to an information push unit.

[0076] Taking a detection process of an embedded device as an example, the embedded device first receives the voltage signal processed by the data acquisition card. Then, it constructs an input matrix from the data samples in the offline arc current dataset and inputs it into the offline recognition model for judgment. If the absolute value of 13 consecutive offline current data is less than 2A, it can be determined that an arc has occurred at that moment, and the arc count is incremented by 1. After sequentially retrieving all the data, the frequency of the offline arc can be obtained. The offline arc frequency information is stored and recorded, and synchronously sent to the information push unit.

[0077] Specifically, the information push unit includes an information receiving module and an information push module; among which,

[0078] The information receiving module is used to receive offline arc frequency information.

[0079] The information push module is used to push the latest received arc offline frequency information to the user.

[0080] like Figure 5 As shown, in this embodiment, an information push unit is constructed using HTML, CSS, and JavaScript. The information push unit includes a fault information receiving module and a fault information pushing module; wherein,

[0081] An offline information receiving module is used to receive offline frequency information sent by an embedded device when the pantograph-catenary system goes offline. In this embodiment of the invention, the offline information receiving module is located on the user's mobile terminal. After obtaining user authorization, it first uses the WebSocket protocol to handshake with the embedded device of the online detection unit to establish a persistent connection and transmit information. The offline frequency information receiving module requests a read operation from the embedded device every 1.5 seconds and transmits the read offline frequency information to the offline information push module.

[0082] The offline information push module is used to push the latest received offline frequency information to the user. In this embodiment of the invention, the offline frequency information push module first determines whether the offline information is the latest information. If it is, the offline frequency information is pushed to the user in a timely manner through the app. Otherwise, no information push is performed.

[0083] This embodiment also discloses a method for detecting the offline frequency of pantograph-catenary arc, comprising the following steps:

[0084] Step S1: Collect the offline arc current signal during pantograph-catenary operation through the signal acquisition unit, and synchronously transmit the signal data to the data processing unit.

[0085] Step S2: The data processing unit obtains offline current samples based on the offline arc current signal, filters out noise interference by using median filtering, adds labels to the data samples to obtain the offline arc current dataset, and plots the offline arc current waveform.

[0086] Step S3: The model building unit constructs an input matrix based on the data samples in the offline arc current dataset and builds an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events.

[0087] Step S4: The result output unit obtains the number of offline times output by the model building unit, gets the offline frequency, and pushes it to the user's mobile device.

[0088] Specifically, step S2 is as follows:

[0089] Step S21: Based on the collected offline arc current data samples, an input matrix is ​​constructed sequentially to obtain the input matrix construction module;

[0090] Step S22: Read the data samples that form the input matrix in batches using the batch reading module;

[0091] Step S23: Build an offline state recognition model.

[0092] Specifically, step S23 is as follows:

[0093] The number of offline events can be obtained by detecting the number of zero-rest periods; when the absolute value of 13 consecutive data points is less than 2A, it can be determined that an electric arc occurred at that moment.

[0094] The specific steps for filtering out noise interference using the median filtering method in step S2 are as follows:

[0095] Step S221: First, define a window of length L with an odd number of digits, where L = 2N + 1 and N is a positive integer;

[0096] Step S222: Suppose that at a certain moment, the signal samples within the window are x(iN), ..., x(i), ..., x(i+N), where x(i) is the signal sample value located at the center of the window;

[0097] Step S223: After arranging the L signal sample values ​​in ascending order, the median value, the sample value at position i, is defined as the output value of the median filter y(i) = Med[x(iN),…,x(i),…x(i+N)]; replace the original value of x(i) in the middle of the window with the sorted median y(i).

[0098] This invention processes data samples using median filtering, removing noise with minimal impact on the original image and achieving high accuracy. It has low hardware requirements and fast current data acquisition and analysis. Furthermore, this invention deploys a series offline state recognition model in an embedded device, designing an offline arc detection device that detects the offline frequency of the pantograph-catenary arc.

[0099] This specific embodiment is merely an explanation of the present invention and is not intended to limit the invention. After reading this specification, those skilled in the art can make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they are within the scope of the claims of the present invention.

Claims

1. A device for detecting the offline frequency of pantograph-catenary arc, characterized in that: It includes a signal acquisition unit, a data processing unit, a model building unit, and a result output unit; among which, The signal acquisition unit is used to collect offline arc current signals during pantograph-catenary operation and synchronously transmit the signal data to the data processing unit. The data processing unit is used to acquire offline current samples based on the offline arc current signal, filter out noise interference by median filtering, add labels to the data samples to obtain the offline arc current dataset, and draw the offline arc current waveform. The model building unit is used to construct an input matrix from the data samples in the offline arc current dataset and build an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events. The result output unit is used to obtain the number of offline times output by the model building unit, obtain the offline frequency, and push it to the user's mobile terminal.

2. The detection device for offline frequency of pantograph-catenary arc according to claim 1, characterized in that: The signal acquisition unit includes a current transformer, a data acquisition card, and a DC power supply; wherein, The aforementioned current transformer is used to convert the loop current signal into a voltage signal during the operation of the pantograph-catenary system; The data acquisition card is used to perform analog-to-digital conversion on the voltage signal. The DC power supply is used to provide power to the current transformer and the data acquisition card.

3. The detection device for offline frequency of pantograph-catenary arc according to claim 1, characterized in that: The model building unit includes an input matrix construction module, a batch reading module, and a state recognition module; wherein, The input matrix construction module constructs the input matrix sequentially based on the collected offline arc current data samples; The batch reading module is used to read the data samples constructed as an input matrix into the state recognition model in batches; The aforementioned state recognition module is used to build an offline state recognition model.

4. The detection device for offline frequency of pantograph-catenary arc according to claim 3, characterized in that: The matrix constructed by the input matrix construction module has a size of 1,000,000 × 1 and is composed of data samples of the offline arc current collected.

5. The detection device for offline frequency of pantograph-catenary arc according to claim 1, characterized in that: The result output unit includes a detection unit and an information push unit; wherein... The detection unit is used to obtain the number of offline events output by the model building unit, determine the frequency of offline events generated by the pantograph-catenary system, and send the offline frequency information to the information push unit. The information push unit is used to receive offline frequency information sent by the detection unit and push fault information to the user's mobile terminal.

6. The detection device for offline frequency of pantograph-catenary arc according to claim 5, characterized in that: The information push unit includes an information receiving module and an information push module; wherein... The information receiving module is used to receive arc offline frequency information; The information push module is used to push the latest received arc offline frequency information to the user.

7. A method for detecting the offline frequency of pantograph-catenary arc, applied to the detection device for the offline frequency of pantograph-catenary arc as described in any one of claims 1-6, characterized in that, It includes the following steps: Step S1: Collect the offline arc current signal during pantograph-catenary operation through the signal acquisition unit, and synchronously transmit the signal data to the data processing unit. Step S2: The data processing unit obtains offline current samples based on the offline arc current signal, filters out noise interference by using median filtering, adds labels to the data samples, obtains the offline arc current dataset, and plots the offline arc current waveform. Step S3: The model building unit constructs an input matrix based on the data samples in the offline arc current dataset and builds an offline state recognition model. Using the offline arc current dataset constructed as the input matrix, zero-time data points are collected to obtain the number of offline events. Step S4: The result output unit obtains the number of offline times output by the model building unit, gets the offline frequency, and pushes it to the user's mobile device.

8. The method for detecting the offline frequency of pantograph-catenary arc according to claim 7, characterized in that, Step S2 is as follows: Step S21: Based on the collected offline arc current data samples, an input matrix is ​​constructed sequentially to obtain the input matrix construction module; Step S22: Read the data samples that form the input matrix in batches using the batch reading module; Step S23: Build an offline state recognition model.

9. A method for detecting the offline frequency of pantograph-catenary arc according to claim 8, characterized in that, Step S23 is as follows: The number of offline events can be obtained by detecting the number of zero-rest periods; when the absolute value of 13 consecutive data points is less than 2A, it can be determined that an electric arc occurred at that moment.

10. A method for detecting the offline frequency of pantograph-catenary arc according to claim 7, characterized in that, The specific steps for filtering out noise interference using the median filtering method in step S2 are as follows: Step S221: First, define a window of length L with an odd number of digits, where L = 2N + 1, and N is a positive integer; Step S222: Suppose that at a certain moment, the signal samples within the window are x(iN), ..., x(i), ..., x(i+N), where x(i) is the signal sample value located at the center of the window; Step S223: After arranging the L signal sample values ​​in ascending order, among which... The sample value at position i is defined as the output value of the median filter, y(i) = Med[x(iN),…,x(i),…x(i+N)]; the value of x(i) in the middle of the original window is replaced with the sorted median y(i).

Images