A low-power collection and transmission system and method for rail health monitoring data

Abstract

The present invention discloses a low-power collection and transmission system and method for rail health monitoring data, which relates to the field of railway measurement technology. The method includes: when the sensor terminal is in a deep sleep state, according to the vibration excitation detected by the low-power acceleration sensor, the sensor terminal is awakened when the first-level threshold is exceeded; the vibration data of the rail is collected by the high-sampling rate acceleration sensor, and the data characteristics of the vibration data are extracted at the same time, and the vibration data is stored in the on-chip FLASH of the BLE module; when the vibration stops, the data collection process of the high-sampling rate acceleration sensor is ended; when the data characteristics exceed the second-level threshold, the collected vibration data is transmitted to the cloud server through the trackside gateway for health monitoring. Through the technical solution of the present invention, a data collection and transmission process based on two-level thresholds is realized, the transmission of invalid data is reduced, energy and bandwidth are saved, and efficient management of sensor data collection and transmission is realized.

Description

Technical Field

[0001] The present invention relates to the field of railway measurement technology, and in particular to a low-power consumption collection and transmission system for rail health monitoring data and a low-power consumption collection and transmission method for rail health monitoring data. Background Art

[0002] The railway department currently still relies mainly on experienced maintenance personnel to conduct regular inspections of turnout rails by visual inspection at the skylight points. However, manual inspections have problems such as low efficiency, high cost, and high experience requirements.

[0003] In addition, there are some customized equipment or methods that can improve the efficiency of inspection. For example, ultrasonic testing, magnetic flux leakage testing, eddy current testing, and machine vision testing.

[0004] The disadvantages of existing customized monitoring equipment are that it will occupy the track during the operation of the equipment, and it is impossible to implement high-frequency inspections due to cumbersome procedures and high deployment and maintenance costs. Therefore, defects may not be discovered in time, leading to accidents. Summary of the invention

[0005] In view of the above problems, the present invention provides a low-power collection and transmission system and method for rail health monitoring data. By embedding a sensor terminal based on low-power Bluetooth BLE technology inside the turnout rail, two acceleration sensors with complementary ranges and functions are used to respectively complete the wake-up and status perception functions of the sensor terminal, so as to reduce the power consumption of the terminal as much as possible while detecting the rail status. Furthermore, through different data sending methods of broadcasting and point-to-point connection, a data collection and transmission process based on a two-level threshold is realized, the transmission of invalid data is reduced, energy and bandwidth are saved, and efficient management of sensor data collection and transmission is realized, so as to realize long-term online monitoring of the turnout rail.

[0006] To achieve the above-mentioned object, the present invention provides a low-power collection and transmission system for rail health monitoring data, comprising: a sensor terminal, a trackside gateway and a cloud server;

[0007] The sensor terminal is embedded and installed inside the switch rail, and the sensor terminal includes a sensor, a low-power Bluetooth BLE module and a multi-layer ceramic antenna;

[0008] The sensor includes a high sampling rate acceleration sensor and a low power consumption acceleration sensor, wherein the high sampling rate acceleration sensor is used to sense the vibration state of the rail, and the low power consumption acceleration sensor is used to provide the sensor terminal with a motion activation function with an ultra-low sleep current and a high-precision inclination calibration function;

[0009] The on-chip FLASH of the BLE module is used to store the data collected by the high sampling rate acceleration sensor and the low power consumption acceleration sensor;

[0010] The multilayer ceramic antenna is used to communicate with the trackside gateway, and the trackside gateway is communicatively connected to the cloud server.

[0011] In the above technical solution, preferably, the sensor terminal is embedded in the jumper hole of the point rail through the sensor base, the sensor terminal and the sensor base are threadedly connected, and the thread is coated with anti-fall glue, the end of the sensor terminal is pressed against the waist surface of the point rail, and the sensor base is installed in the jumper hole by interference fit.

[0012] In the above technical solution, preferably, the sensor, BLE module and multilayer ceramic antenna of the sensor terminal are arranged on a PCB, silicone gaskets are arranged on both sides of the PCB, the PCB is arranged in a hexagonal shape, and the PCB is installed at the front end of the sensor base of the bolt-shaped structure;

[0013] The sensor terminal further comprises a battery compartment in which a battery is installed for supplying power to other components of the sensor terminal.

[0014] In the above technical solution, preferably, the sampling frequency of the high sampling rate acceleration sensor is 6400 Hz, the range of the collected acceleration is 200g, the static current of the low power consumption acceleration sensor is 180nA, and the range of the collected acceleration is 2g.

[0015] In the above technical solution, preferably, the high sampling rate acceleration sensor is configured as a 32-level first-in first-out FIFO buffer.

[0016] The present invention also proposes a low-power collection and transmission method for rail health monitoring data, which is applied to a low-power collection and transmission system for rail health monitoring data disclosed in any one of the above technical solutions, comprising:

[0017] When the sensor terminal is in a deep sleep state, waking up the sensor terminal when the vibration excitation detected by the low-power acceleration sensor exceeds a first level threshold;

[0018] The vibration data of the rail is collected by a high sampling rate acceleration sensor, data features of the vibration data are extracted, and the vibration data is stored in the on-chip FLASH of the BLE module;

[0019] When the vibration stops, the data acquisition process of the high sampling rate acceleration sensor ends;

[0020] When the data feature exceeds the second level threshold, the collected vibration data is transmitted to the cloud server through the trackside gateway for health monitoring.

[0021] In the above technical solution, preferably, the vibration data of the rail is collected by a high sampling rate acceleration sensor, and data features of the vibration data are extracted at the same time, and the specific process includes:

[0022] After the sensor terminal is awakened and activated, the on-chip FLASH of the BLE module erases the FLASH storage area sufficient to store the vibration data;

[0023] The high sampling rate acceleration sensor acquires the vibration data of the rail and performs high frequency sampling at a preset sampling frequency;

[0024] The sampled data is cached in a FIFO register, and data features of the vibration data are extracted in a stream mode before the cached data fills up the FIFO register. The specific process includes:

[0025] Calculate the root mean square value RMS of the vibration data using the following formula:

[0026] ;

[0027] in, a y is the Y-axis acceleration data at the i-th moment, and n is the total number of samples;

[0028] When the FIFO register is filled with the sampled data, an interrupt signal is sent out to transfer the cached sampled data to the on-chip FLASH.

[0029] In the above technical solution, preferably, the data collection process of the high sampling rate acceleration sensor is ended when the vibration stops, and the specific process includes:

[0030] Comparing the time domain amplitude characteristics of the vibration data with a preset time domain amplitude threshold, and determining that the train has departed and the vibration has stopped when the time domain amplitude is lower than the preset time domain amplitude threshold;

[0031] When the vibration stops, the high sampling rate acceleration sensor is controlled to stop the data collection process.

[0032] In the above technical solution, preferably, after determining that the vibration stops, the data characteristics of the vibration data are broadcasted outwardly through the BLE module at a fixed time;

[0033] Receiving data characteristics of vibration data broadcasted and transmitted by each of the sensor terminals in an online connection state according to the connection state between the trackside gateway and the sensor terminal;

[0034] If the sensor terminal is not connected online to the trackside gateway, the sensor terminal is initialized and enters a deep sleep state.

[0035] In the above technical solution, preferably, when the data feature exceeds the second level threshold, the collected vibration data is transmitted to the cloud server through the trackside gateway for health monitoring, and the specific process includes:

[0036] Determine whether the root mean square value RMS of the vibration data exceeds a preset second level threshold;

[0037] If the RMS exceeds the second level threshold, the vibration data stored in the on-chip FLASH is transmitted to the cloud server;

[0038] If the RMS does not exceed the second level threshold, discarding the vibration data collected in the current wake-up state;

[0039] The sensor terminal is initialized and enters a deep sleep state.

[0040] Compared with the prior art, the beneficial effects of the present invention are as follows: by embedding a sensor terminal based on low-power Bluetooth BLE technology inside the turnout rail, two acceleration sensors with complementary ranges and functions are used to respectively complete the wake-up and status perception functions of the sensor terminal, thereby detecting the rail status while reducing the power consumption of the terminal as much as possible. Furthermore, through different data sending methods of broadcasting and point-to-point connection, a data collection and transmission process based on a two-level threshold is realized, the transmission of invalid data is reduced, energy and bandwidth are saved, and efficient management of sensor data collection and transmission is realized, so that long-term online monitoring of the turnout rail can be realized. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic diagram of the architecture of a low-power collection and transmission system for rail health monitoring data disclosed in an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of a PCB layout of a sensor terminal disclosed in an embodiment of the present invention;

[0043] Figure 3 A schematic flow chart of a method for low-power collection and transmission of rail health monitoring data disclosed in an embodiment of the present invention;

[0044] Figure 4 A schematic diagram of data buffering and data processing periods of a FIFO buffer of a sensor terminal disclosed in an embodiment of the present invention;

[0045] Figure 5 A schematic diagram of the content structure of a broadcast packet disclosed in an embodiment of the present invention;

[0046] Figure 6 A schematic diagram of temperature and humidity data of a trackside gateway disclosed in an embodiment of the present invention;

[0047] Figure 7 A schematic diagram of broadcast data of a sensor terminal disclosed in an embodiment of the present invention;

[0048] Figure 8 A schematic diagram of vibration data transmitted back by a sensor terminal disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0051] like Figure 1 As shown, a low-power collection and transmission system for rail health monitoring data provided according to the present invention includes: a sensor terminal, a trackside gateway and a cloud server;

[0052] The sensor terminal is embedded and installed inside the switch rail. The sensor terminal includes a sensor, a low-power Bluetooth BLE module and a multi-layer ceramic antenna.

[0053] The sensors include high sampling rate accelerometers and low power consumption accelerometers. The high sampling rate accelerometers are used to sense the vibration state of the rails, and the low power consumption accelerometers are used to provide the sensor terminal with ultra-low sleep current motion activation function and high-precision inclination calibration function.

[0054] The on-chip FLASH of the BLE module is used to store the data collected by the high sampling rate accelerometer and the low power consumption accelerometer;

[0055] The multilayer ceramic antenna is used to communicate with the trackside gateway, and the trackside gateway is used to communicate with the cloud server.

[0056] In this implementation, by embedding a sensor terminal based on low-power Bluetooth BLE technology inside the turnout rail, two acceleration sensors with complementary ranges and functions are used to respectively complete the wake-up and status perception functions of the sensor terminal, thereby detecting the rail status while reducing the power consumption of the terminal as much as possible. Furthermore, through different data sending methods of broadcasting and point-to-point connection, a data collection and transmission process based on a two-level threshold is realized, the transmission of invalid data is reduced, energy and bandwidth are saved, and efficient management of sensor data collection and transmission is achieved, so that long-term online monitoring of the turnout rail can be realized.

[0057] Specifically, the present invention proposes a new beacon system based on low-power Bluetooth (BLE) technology with a terminal embedded inside the turnout rail, which is used for real-time monitoring of the turnout rail status. The sensor terminal of the system uses a deep event-driven method to collect key data such as the temperature and vibration of the point rail, and transmits the data to the outside through broadcast or point-to-point connection. At the same time, it includes an efficient data link between the trackside gateway and the cloud server to ensure the real-time and stability of data transmission while being compatible with the new beacon.

[0058] Among them, the sensor terminal is equipped with a large-range and high-sampling rate acceleration sensor ADXL372, which is used to sense the vibration state of the fork. At the same time, in order to reduce the sleep power consumption of the terminal, a small-range and low-power ADXL367 is separately equipped to provide the node with ultra-low sleep current motion activation function and high-precision inclination calibration function. This complementary dual sensor configuration can solve the contradiction between range and resolution, separate the wake-up and perception functions, and give full play to the advantages of different devices in different functions. The factors affecting power consumption of all sensor terminals have been optimized and designed as the basis for low power consumption of the terminal. For example, the selection of the three main chips, the use of SPI instead of IIC for the communication protocol, and the optimization of data throughput to 44Kbyte / s.

[0059] The on-chip FLASH of the BLE module chip is used to implement the functions of hibernation storage and extended storage space for the terminal. Since the terminal will be in hibernation mode most of the time, all power domains of the BLE module except the wake-up circuit are cut off to achieve the lowest hibernation current, but the node will lose all variables after hibernation. The on-chip FLASH of the BLE module 4M is used to implement hibernation storage and provide enough space for the storage of original data. All peripherals communicate with the host through the high-speed and low-power SPI protocol.

[0060] In the above embodiment, preferably, the sensor terminal is embedded in the electrical jumper hole of the point rail through the sensor base, the sensor terminal and the sensor base are threadedly connected, and the thread is coated with anti-falling glue, and the end of the sensor terminal is pressed against the waist surface of the point rail to ensure the transmission quality of the vibration signal. The sensor base is installed in the jumper hole with an interference fit to achieve the design principle of anti-loosening and anti-rotation.

[0061] Specifically, the structure of the sensor terminal protrudes only 7mm from the rail waist plane, ensuring that the installation in the tip area of ​​the point rail will not affect the closure and close fit between the point rail and the base rail. For example, the gap between the tip of the SC330 turnout point rail and the base rail waist is 11-13mm, and after installing the sensor terminal, there is still a 4-6mm safety space.

[0062] During the implementation process, the structure of the sensor terminal can imitate the bolt design, and the bolt head can be cut open as the sensor base. The PCB circuit board of the sensor terminal can be designed into a hexagon and embedded in the bolt head. The advantages are good waterproofness, sufficient rigidity of the acceleration sensor, and easy installation and disassembly.

[0063] like Figure 2 As shown, in the above embodiment, preferably, the sensor, BLE module and multilayer ceramic antenna of the sensor terminal are arranged on a PCB, and silicone gaskets are arranged on both sides of the PCB to meet the sealing requirements. The PCB is arranged in a hexagonal shape, and the PCB is installed at the front end of the sensor base with a bolt-shaped structure;

[0064] The sensor terminal also includes a battery compartment in which batteries are installed for powering other components of the sensor terminal.

[0065] Specifically, the PCB area of ​​the sensor terminal node is only 2.6 square centimeters, and for easy burning, a burning pin is reserved. The PCB has six through holes for passing the bolts connecting the cover plate and the battery compartment.

[0066] The battery compartment can be adjusted according to the thickness of the point rail. During implementation, the space in the battery compartment can install 8 CR1025 button batteries with a total capacity of 240 mAh.

[0067] In the above implementation, preferably, the sampling frequency of the high sampling rate acceleration sensor is 6400 Hz, the range of the collected acceleration is 200 g, the static current of the low power consumption acceleration sensor is 180 nA, and the range of the collected acceleration is 2 g.

[0068] In the above implementation, preferably, the high sampling rate acceleration sensor is configured as a 32-level first-in first-out FIFO buffer to cache the sampled data.

[0069] like Figure 3As shown, the present invention also proposes a low-power collection and transmission method for rail health monitoring data, which is applied to the low-power collection and transmission system for rail health monitoring data disclosed in any one of the above embodiments, including:

[0070] When the sensor terminal is in a deep sleep state, the sensor terminal is awakened when the vibration excitation detected by the low-power acceleration sensor exceeds the first level threshold;

[0071] The vibration data of the rail is collected through a high sampling rate acceleration sensor, and the data features of the vibration data are extracted and stored in the on-chip FLASH of the BLE module;

[0072] When the vibration stops, the data acquisition process of the high sampling rate acceleration sensor ends;

[0073] When the data feature exceeds the second-level threshold, the collected vibration data will be transmitted to the cloud server through the trackside gateway for health monitoring.

[0074] In this implementation, by embedding a sensor terminal based on low-power Bluetooth BLE technology inside the turnout rail, two acceleration sensors with complementary ranges and functions are used to respectively complete the wake-up and status perception functions of the sensor terminal, thereby detecting the rail status while reducing the power consumption of the terminal as much as possible. Furthermore, through different data sending methods of broadcasting and point-to-point connection, a data collection and transmission process based on a two-level threshold is realized, the transmission of invalid data is reduced, energy and bandwidth are saved, and efficient management of sensor data collection and transmission is achieved, so that long-term online monitoring of the turnout rail can be realized.

[0075] Specifically, once the sensor terminal node is powered on, it immediately enters a deep sleep state after completing the initialization of the low-power acceleration sensor used for wake-up, until the node begins to be awakened by vibration stimulation.

[0076] While performing data collection and processing tasks, the node will extract the time domain amplitude characteristics of the vibration signal to determine whether the train has left the switch area. If the train has not completely left, the node will repeat the data collection and processing. If the train has passed, the node will stop collecting data and enter the data packaging and sending state.

[0077] Furthermore, the introduction of the first-level threshold is due to the fact that the length of time when a train passes through the switch area is much shorter than the length of time when no train passes through. That is to say, in most cases, no train passes through the switch area. If the data is continuously collected, most of the acceleration data is just meaningless noise data, and the data collected at this time is redundant. Therefore, by setting a trigger mode, the terminal starts to collect data only after being awakened by the vibration excitation of the train, avoiding unnecessary sampling in irrelevant situations (such as when no train passes), thereby saving system resources. The introduction of the second-level threshold is to take into account some wake-up events that are not caused by the excitation of the train, such as the vibration of the point rail caused by the switch switching, and is set to reduce the amount of wireless data transmission of the terminal. A considerable part of the data of the train passing is invalid (low-speed, light-loaded train passing, switch switching, adjacent track passing, etc.). This part of the data will consume a lot of power for the terminal to send, and it is completely redundant and cumbersome for the back-end database. Setting the second-level threshold can filter out this part of useless data, thereby reducing the power consumption of the terminal and improving the purity of the data.

[0078] The first-level threshold determines whether to collect data, and the second-level threshold determines whether to transmit a large amount of raw data back. The introduction of the second-level threshold on the acceleration data feature allows the system to discard invalid data, thereby reducing the amount of data transmitted back by the sensor terminal and avoiding the transmission of invalid or low-impact events (such as light-loaded trains or activities on adjacent tracks), thereby reducing transmission bandwidth and energy consumption. In addition, the introduction of the second-level threshold may even allow the system to transmit all raw data only when the structure is damaged, and only transmit a feature value extracted locally at the terminal when the structure is normal.

[0079] In this system, the first-level threshold can be set to any three-axis amplitude of the acceleration sample exceeding 62.5 mg. The second-level threshold is implemented by manual decision-making. Specifically, for the second-level threshold, the sensor terminal calculates the root mean square value of the acceleration data locally as the feature of the data and transmits it to the outside in the form of broadcast, while the decision on whether to transmit the original data back is left to the cloud maintenance personnel.

[0080] In the above implementation, preferably, the vibration data of the rail is collected by a high sampling rate acceleration sensor, and data features of the vibration data are extracted at the same time. The specific process includes:

[0081] After the sensor terminal is awakened and activated, the on-chip FLASH of the BLE module erases the FLASH storage area sufficient to store vibration data, for example, the FLASH storage area sufficient to store 30 seconds of raw vibration data;

[0082] The high sampling rate acceleration sensor acquires the vibration data of the rail and performs high-frequency sampling at a preset sampling frequency;

[0083] The sampled data is cached in the FIFO register, and the data features of the vibration data are extracted in a stream mode before the cached data fills up the FIFO register. The specific process includes:

[0084] Calculate the root mean square value RMS of the vibration data using the following formula:

[0085] ;

[0086] in, a y is the Y-axis acceleration data at the i-th moment, and n is the total number of samples;

[0087] When the FIFO register is filled with sampled data, an interrupt signal is issued to transfer the cached sampled data to the on-chip FLASH.

[0088] In this embodiment, the time taken by the sensor terminal to erase the FLASH storage area is approximately between 400 milliseconds and 1000 milliseconds depending on the sampling frequency.

[0089] Specifically, the high sampling rate acceleration sensor uses the strategy of reading first, storing second, and sending second when sampling at high frequency (6400Hz). When reading high-frequency sampled data, the FIFO buffer is first used to temporarily store the sampled data. When the FIFO is full, the host is informed through an interrupt signal, and then the host quickly reads the sampled data. While reading and after reading, the sampled data continues to accumulate in the FIFO buffer. The advantage of reading first, storing second, and sending second is that the sampled data can be stored for a long time when the on-chip resources are insufficient. The advantage of using the FIFO buffer and data interrupt is that the closed loop of the reading operation can be guaranteed while reducing the host's compliance, so that the host can perform other operations. The combination of the two can achieve the purpose of accurate, closed-loop reading and storage of a large number of acceleration samples. Unlike wireless sensor nodes that send in real time, due to the subsequent second-level threshold, the original sampled data should be retained locally. However, due to the limitation of memory space, directly storing it in the running memory will lead to serious memory shortage. This requires the main control chip to write a large amount of sampled data into the external memory. At this time, in order not to lose data, the FIFO buffer and data interrupt must be used.

[0090] For example, the second-level threshold is used to decide whether to transmit, which is RMS current > RMS Threshold , if the current RMS value RMS current Exceeding the preset threshold RMS Threshold , then data is transmitted, otherwise it is discarded.

[0091] Through the above two-level threshold mechanism, the power consumption optimization formula is: P = P sampling + P storage + P transmission , which reduces P sampling and P transmission consumption.

[0092] During a wake-up cycle, the motion activation interrupt pin of the node controlled by the low-power accelerometer ADXL367, the data interrupt pin controlled by the high-sampling-rate accelerometer ADXL372, and the status indication pin controlled by the DA14531 are monitored. The results are as follows: Figure 4 As shown. The high sampling rate acceleration sensor ADXL372 is configured as a 32-level first-in first-out FIFO buffer, and the stream mode processing of data is completed by using the time when the sampled data collected by the sensor fills the FIFO buffer. In order to prevent the loss of original data, the time for feature extraction processing of sampled data should be less than the time when the FIFO is filled. That is, when the data output rate of the sensor is 1600HZ, the data processing time cannot exceed 20ms ( Figure 4 Partially enlarged image). During the sampling process, in addition to completing feature extraction, the original data is also stored in the FLASH buffer for future reference. Figure 4 It is shown in Figure 6 that it takes 6.9ms to calculate the RMS and save the raw data every two times the FIFO buffer is filled.

[0093] Specifically, the "read first, store then, send" method of the FIFO buffer specifically includes:

[0094] (1) Read first: The FIFO buffer is used to cache high-frequency sampled data. The first-in-first-out mechanism of the FIFO buffer enables the processor to read the data in a timely manner and notify the processor to process it through an interrupt signal when the FIFO buffer is full.

[0095] (2) Restoration: The read high-frequency sampling data will be temporarily stored in an external storage device (such as Flash) instead of being processed directly in the memory.

[0096] (3) Retransmission: After the sampled data is stored, the system screens the sampled data according to the second-level threshold. Only useful data that meets the second-level threshold will be transmitted through the wireless module, reducing the transmission of invalid data and saving energy and bandwidth.

[0097] In the above implementation, preferably, the data collection process of the high sampling rate acceleration sensor is ended when the vibration stops, and the specific process includes:

[0098] Compare the time domain amplitude characteristics of the vibration data with a preset time domain amplitude threshold, and determine that the train has left and the vibration has stopped when it is lower than the preset time domain amplitude threshold;

[0099] When the vibration stops, the high sampling rate acceleration sensor is controlled to stop the data acquisition process.

[0100] In the above implementation, preferably, after it is determined that the vibration stops, the data characteristics of the vibration data are broadcasted outwardly through the BLE module at a fixed time;

[0101] According to the connection status between the trackside gateway and the sensor terminal, the data characteristics of the vibration data broadcasted and transmitted by each sensor terminal in the online connection state are received;

[0102] If the sensor terminal is not connected online to the trackside gateway, the sensor terminal is initialized and enters a deep sleep state.

[0103] Specifically, the vibration characteristics of the turnout extracted from the vibration and the status information of the sensor terminal node itself (including battery power and temperature data) are packaged into the broadcast package of the BLE module. The sensor terminal node starts to broadcast regularly, broadcasting the status information of the turnout and the sensor terminal node. At the same time, the broadcast of the sensor terminal node can be connected. After the connection, the sensor terminal node reads the original vibration data from the on-chip FLASH and uploads it.

[0104] The BLE module's broadcast packet consists of Figure 5 The first two bytes of the broadcast packet are used to mark whether the broadcast is connectable. This time, the broadcast of the sensor terminal is set to connectable. The middle 8 bytes are used to number the node (GT-00X, 7 bytes, plus a byte describing the length, a total of eight bytes), and the last two parts of the manufacturer information are used to contain some characteristics of the original data and the status information of the node.

[0105] In the above implementation, preferably, when the data feature exceeds the second level threshold, the collected vibration data is transmitted to the cloud server through the trackside gateway for health monitoring. The specific process includes:

[0106] Determine whether the root mean square value RMS of the vibration data exceeds a preset second level threshold;

[0107] If the RMS exceeds the second level threshold, the vibration data stored in the on-chip FLASH is transmitted to the cloud server;

[0108] If the RMS does not exceed the second level threshold, the vibration data collected in the current wake-up state is discarded;

[0109] Initialize the sensor terminal and enter deep sleep state.

[0110] In this implementation, the sensor terminal calculates a characteristic value locally, but does not use it directly as a threshold value. Instead, it broadcasts it, handing over the decision-making power to the gateway or cloud platform with much richer computing and storage resources. A simple example is that without returning the original data, we can use the root mean square value RMS broadcast by the terminal to determine whether the wake-up time is caused by the switch.

[0111] There are two keys to achieve the second level threshold. The first is the calculation of the feature. The calculation of the feature requires a stream mode so that the calculation task can be put together while waiting for the acceleration sample to fill the FIFO buffer. The calculation content is special. In this embodiment, the sensor terminal calculates the root mean square value.

[0112] The second is the effective transmission of data. As the second-level threshold (in fact, the second-level threshold is multimodal and also includes the terminal status data, such as temperature, battery voltage, and number of samples), if the trackside gateway or cloud server decides that the original data needs to be sent back, then the sensor terminal must be online instead of sleeping (this is also the reason why the broadcast time is set relatively long). This puts forward requirements for the real-time performance of the entire data link (in addition to distance, this is also the reason for using ceramic antennas). Considering the huge number of sensor terminals corresponding to the same trackside gateway, in order to achieve the purpose of quickly transmitting a small amount of data between each sensor terminal and the same trackside gateway, only the broadcast communication method can meet the requirements.

[0113] In this embodiment, a connection is directly initiated to a designated sensor terminal according to the name of the broadcast packet, and it is preset in the sensor terminal program that once connected, the original data is returned, and if not connected, the broadcast is continued until the timer ends.

[0114] According to the low-power collection and transmission system and method for rail health monitoring data disclosed in the above-mentioned implementation mode, the above-mentioned system and method of the present invention are tested during the implementation process. During the implementation process, the system is deployed on the straight point rail of one set of turnouts, and the two sensor terminals are respectively located at the rail head and rail bottom of the point rail. The trackside gateway is arranged at a distance of about 11m in front of the two sensors through a bracket and placed in a sealed acrylic box for rain protection.

[0115] The monitoring interface server is located in Chongqing and is used to receive data sent back by the gateway. It monitored the switch point rail for one month between July and August.

[0116] like Figure 6The temperature and humidity changes of the gateway within one month are shown. In the early stage of system deployment, the temperature is low and the humidity is high, and it is rainy. In the later stage, the temperature is high and the humidity is low, and it is sunny. Among them, the temperature and humidity cycle between day and night is large, and the temperature difference between day and night exceeds 20℃.

[0117] like Figure 7 As shown in Figure 1, the broadcast data of the sensor terminal is displayed. Figure 8 As shown, part of the original data during the monitoring process is displayed. The characteristics of the wheelset passed by the train are obvious and the data is complete, achieving the predetermined monitoring purpose.

[0118] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A low-power collection and transmission system for rail health monitoring data, characterized in that: include: Sensor terminals, trackside gateways and cloud servers; The sensor terminal is embedded and installed inside the switch rail, and the sensor terminal includes a sensor, a low-power Bluetooth BLE module and a multi-layer ceramic antenna; The sensor includes a high sampling rate acceleration sensor and a low power consumption acceleration sensor, wherein the high sampling rate acceleration sensor is used to sense the vibration state of the rail, and the low power consumption acceleration sensor is used to provide the sensor terminal with a motion activation function with an ultra-low sleep current and a high-precision inclination calibration function; The on-chip FLASH of the BLE module is used to store the data collected by the high sampling rate acceleration sensor and the low power consumption acceleration sensor; The multilayer ceramic antenna is used to communicate with the trackside gateway, and the trackside gateway is communicated with the cloud server; The sensor terminal is embedded in the jumper hole of the point rail through the sensor base. The sensor terminal and the sensor base are threadedly connected, and the thread is coated with anti-fall glue. The end of the sensor terminal is pressed tightly against the waist surface of the point rail, and the sensor base is installed in the jumper hole with interference fit.

2. The low-power collection and transmission system for rail health monitoring data according to claim 1 is characterized in that: The sensor, BLE module and multilayer ceramic antenna of the sensor terminal are arranged on a PCB, silicone gaskets are arranged on both sides of the PCB, the PCB is arranged in a hexagonal shape, and the PCB is installed at the front end of the sensor base of the bolt-shaped structure; The sensor terminal further comprises a battery compartment in which a battery is installed for supplying power to other components of the sensor terminal.

3. The low-power collection and transmission system for rail health monitoring data according to claim 2 is characterized in that: The sampling frequency of the high sampling rate acceleration sensor is 6400 Hz, and the range of the collected acceleration is 200 g. The static current of the low power consumption acceleration sensor is 180 nA, and the range of the collected acceleration is 2 g.

4. The low-power collection and transmission system for rail health monitoring data according to claim 3 is characterized in that: The high sampling rate acceleration sensor is configured as a 32-level first-in first-out FIFO buffer.

5. A low-power collection and transmission method for rail health monitoring data, characterized in that: A low-power collection and transmission system for rail health monitoring data according to any one of claims 1 to 4, comprising: When the sensor terminal is in a deep sleep state, waking up the sensor terminal when the vibration excitation detected by the low-power acceleration sensor exceeds a first level threshold; The vibration data of the rail is collected by a high sampling rate acceleration sensor, data features of the vibration data are extracted, and the vibration data is stored in the on-chip FLASH of the BLE module; Comparing the time domain amplitude characteristics of the vibration data with a preset time domain amplitude threshold, and determining that the train has departed and the vibration has stopped when the time domain amplitude is lower than the preset time domain amplitude threshold; When the vibration stops, the high sampling rate acceleration sensor is controlled to stop the data acquisition process; When the data feature exceeds the second level threshold, the collected vibration data is transmitted to the cloud server through the trackside gateway for health monitoring. The specific process includes: Determine whether the root mean square value RMS of the vibration data exceeds a preset second level threshold; If the RMS exceeds the second level threshold, the vibration data stored in the on-chip FLASH is transmitted to the cloud server; If the RMS does not exceed the second level threshold, discarding the vibration data collected in the current wake-up state; The sensor terminal is initialized and enters a deep sleep state.

6. The low-power collection and transmission method for rail health monitoring data according to claim 5 is characterized in that: The process of collecting the vibration data of the rail by using a high sampling rate acceleration sensor and extracting the data features of the vibration data includes: After the sensor terminal is awakened and activated, the on-chip FLASH of the BLE module erases the FLASH storage area sufficient to store the vibration data; The high sampling rate acceleration sensor acquires the vibration data of the rail and performs high frequency sampling at a preset sampling frequency; The sampled data is cached in a FIFO register, and data features of the vibration data are extracted in a stream mode before the cached data fills up the FIFO register. The specific process includes: Calculate the root mean square value RMS of the vibration data using the following formula: ； in, a y is the Y-axis acceleration data at the i-th moment, and n is the total number of samples; When the FIFO register is filled with the sampled data, an interrupt signal is sent out to transfer the cached sampled data to the on-chip FLASH.

7. The low-power collection and transmission method for rail health monitoring data according to claim 5, characterized in that: After determining that the vibration stops, the data characteristics of the vibration data are broadcasted outwardly through the BLE module at a fixed time; Receiving data characteristics of vibration data broadcasted and transmitted by each of the sensor terminals in an online connection state according to the connection state between the trackside gateway and the sensor terminal; If the sensor terminal is not connected online to the trackside gateway, the sensor terminal is initialized and enters a deep sleep state.

Images