Railway construction protection alarm method and system based on wireless intercom relay

Abstract

The application provides a railway construction protection alarm method and system based on wireless intercom relay, belongs to the field of railway driving safety protection and digital trunking communication cross technology, and comprises the following steps: a front-end sensing device is used to monitor the train approaching state of a railway construction section in real time and generate a trigger signal; an automatic trigger controller receives the trigger signal, simultaneously outputs a PTT control signal and an alarm audio signal to a digital intercom relay host through hardwiring, wherein the PTT control signal is used to forcibly preempt and start the transmission channel of the relay host; the relay host modulates the alarm audio signal into a digital radio frequency signal and broadcasts it through an antenna; and a digital handheld intercom in the construction area receives the digital radio frequency signal on a preset channel and outputs an alarm voice. The application realizes full automation from train approaching sensing to full alarm, eliminates manual response delay, is compatible with existing digital intercom equipment, and has the advantages of low deployment cost, fast response speed and strong anti-interference ability.

Description

Technical Field

[0001] This invention relates to the technical field of railway traffic safety protection and digital trunking communication, and in particular to a railway construction protection alarm method and system based on wireless intercom relay. Background Technology

[0002] Railway construction safety is a crucial aspect of ensuring the safety of construction workers and train operations. Currently, railway construction safety mainly relies on two independent systems.

[0003] The first type is the manual voice protection system. This system relies on the station liaison officer to call the on-site protection officer via walkie-talkie (usually on the 400MHz band), and then the protection officer to notify the workers to ditch and avoid the train. The response time of this method depends entirely on the human reaction speed and communication efficiency. From the time the train approaches and is detected by the station liaison officer to the time the on-site personnel receive the instruction, there is usually a delay of several seconds to tens of seconds. On high-speed rail lines, the train speed is as high as 300 kilometers per hour or more, and it travels more than 80 meters per second. A delay of several seconds means a braking distance gap of hundreds of meters, which poses a great safety hazard.

[0004] The second type is an independent audible and visual alarm system. This system is installed separately at the construction site using radar or infrared detectors, and only generates audible and visual alarm signals on-site. However, the alarm information of this system cannot be automatically converted into signals that conform to the railway 400MHz digital intercom protocol and injected into the communication network. The safety officer must be distracted to observe the status of the alarm and then manually press the button to call it, which introduces a double delay of "manual confirmation + manual operation". In addition, there is a lot of electromagnetic interference (contact network, construction machinery) and clutter interference (birds, dust, temporary obstacles) at the railway construction site. Coupled with the random fluctuations in radar ranging and speed measurement, direct use may lead to false triggering or missed triggering, affecting the accuracy of the alarm.

[0005] Essentially, the two systems mentioned above are heterogeneous systems completely isolated at the physical layer. The root cause of their technical defects is not a simple lack of functionality, but rather the inherent limitations of their underlying theoretical architecture. On the one hand, ordinary sensor signals (switching signals) cannot directly drive the channel coding of digital walkie-talkies (such as the DMR / PDT protocol), making it impossible to use existing digital relay networks for broadcasting in construction areas. On the other hand, the digital walkie-talkies commonly used by railway construction workers are usually in group call monitoring mode. How to force the repeater's transmission channel to be preempted through hardware logic when the digital walkie-talkie is in group call or normal call mode, thus ensuring the highest priority of alarm commands, is also a technical problem that urgently needs to be solved.

[0006] Therefore, there is an urgent need for a railway construction protection alarm solution that can deeply integrate front-end sensing with digital intercom communication to achieve full automation from train approach detection to alarm for all personnel, eliminate the delay of manual response, and be compatible with existing digital intercom equipment. Summary of the Invention

[0007] In view of the technical defects and drawbacks existing in the prior art, the present invention provides a railway construction protection alarm method and system based on wireless intercom relay to overcome the above problems or at least partially solve the above problems. The specific solution is as follows;

[0008] As a first aspect of the present invention, a railway construction protection alarm method based on wireless intercom relay includes the following steps:

[0009] Step S1: Monitor the train approach status of the railway construction section in real time through front-end sensing devices, and generate a trigger signal when the preset trigger conditions are met;

[0010] Step S2: The automatic trigger controller receives the trigger signal and simultaneously outputs a PTT control signal and an alarm audio signal to the digital intercom repeater host via a hardwired connection; wherein, the PTT control signal is used to forcibly preempt and start the repeater host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the repeater host;

[0011] Step S3: After the transmission channel is activated, the relay host modulates the alarm audio signal into a digital radio frequency signal and broadcasts it to the construction area through the antenna;

[0012] Step S4: The digital handheld walkie-talkie in the construction area receives the digital radio frequency signal on the preset channel and outputs an alarm voice.

[0013] In some embodiments, the front-end sensing device includes at least one of a train proximity warning radar, an infrared intrusion detector, a lidar, or a track circuit sensor; each type of front-end sensing device corresponds to different preset triggering conditions:

[0014] When the front-end sensing device is a train approach warning radar, the preset triggering condition is that the distance at which the radar detects the target object is less than or equal to a preset warning distance;

[0015] When the front-end sensing device is an infrared intrusion detector, the preset triggering condition is that the detector detects a target object entering a preset detection area;

[0016] When the front-end sensing device is a lidar, the preset trigger condition is that the lidar detects that the contour features of the target object match a preset train contour template;

[0017] When the front-end sensing device is a track circuit sensor, the preset trigger condition is that the sensor detects a change in the short-circuit state of the track circuit.

[0018] In some embodiments, when the front-end sensing device is a train approach warning radar, step S1 further includes a step of performing Kalman filtering on the raw observation data output by the radar to suppress measurement noise caused by electromagnetic interference and clutter reflection in the construction environment, and output the filtered stable distance and speed values.

[0019] The Kalman filtering process is based on the train distance-speed joint Kalman filtering algorithm, specifically including:

[0020] Establish a state vector X(k), with the train's distance d(k) and speed v(k) as state variables:

[0021] X(k) = ;

[0022] Based on the assumption of uniform linear motion, the discrete-time state transition equation is established:

[0023] X(k)=A X(k-1)+W(k)

[0024] The state transition matrix is ​​as follows:

[0025] A= ;

[0026] T is the radar sampling period, and W(k) is the process noise vector;

[0027] Establish a linear observation equation, using the range observation value z directly output by the radar. d (k) and velocity observation z v (k) constitutes the observation vector:

[0028] Z(k) = ;

[0029] The relationship between the observation vector Z(k) and the state vector X(k) is expressed as follows:

[0030] Z(k) = H X(k)+V(k);

[0031] The observation matrix is ​​as follows:

[0032] H= ;

[0033] V(k) is the observation noise vector;

[0034] Through iterative calculations of the prediction and update steps, the state estimate is corrected one by one, and the filtered stable distance and velocity values ​​are output. The stable distance value is then compared with the preset warning distance to determine whether the preset triggering condition is met.

[0035] In some embodiments, the Kalman filtering process sequentially performs a prediction step and an update step within each radar sampling period:

[0036] The prediction step includes:

[0037] Based on the optimal state estimate X(k-1|k-1) and its covariance matrix P(k-1|k-1) from the previous time step, predict the state and covariance at the current time step:

[0038] The state prediction equation is as follows:

[0039] X(k|k-1)=A X(k-1|k-1);

[0040] The covariance prediction equation is:

[0041] P(k|k-1)=A P(k-1|k-1) A T +Q;

[0042] Where Q is the process noise covariance matrix, based on the train's maximum acceleration disturbance a. max The results are calculated from the radar sampling period T:

[0043] Q= ;

[0044] The update steps include:

[0045] Obtain the observation vector Z(k) = [z d (k),z v (k)] T The observed value is used to correct the state prediction value obtained in the prediction step, so as to obtain the optimal state estimate at the current time:

[0046] Calculate the Kalman gain matrix:

[0047] K(k)=P(k∣k-1) H T (H P(k∣k-1) H T +R) -1 ;

[0048] State update equation (optimal estimate):

[0049] X(k∣k)=X(k∣k-1)+K(k) (Z(k)-H X(k∣k-1));

[0050] Covariance update equation:

[0051] P(k∣k)=(IK(k) H) P(k∣k-1);

[0052] Where I is the second-order identity matrix; R is the observation noise covariance matrix, based on the ranging accuracy σ of the radar. d and speed measurement accuracy σ v The calculation yielded:

[0053] R= .

[0054] In some embodiments, the Kalman filtering process further includes an outlier removal mechanism, which is executed before the update step, specifically including:

[0055] Calculate the observation residual vector at the current time:

[0056] ε(k)=Z(k)-H X(k|k-1);

[0057] The observation residual vector includes a distance residual component ε. d (k) and velocity residual component ε v (k), that is:

[0058] ε(k)= ;

[0059] Obtain the ranging accuracy σ of the radar. d and speed measurement accuracy σ v And calculate the distance from the residual threshold N according to the preset residual threshold multiple N. σ d and velocity residual threshold N σ v ;

[0060] When the absolute value of the distance residual is |ε d (k) | greater than the distance residual threshold, or the absolute value of the velocity residual | ε v When (k)| is greater than the velocity residual threshold, the current observation vector Z(k) is determined to be an outlier, the update step is skipped, and the result of the prediction step is directly used as the optimal state estimate for the current moment.

[0061] X(k|k)=X(k|k-1);

[0062] P(k∣k)=P(k∣k-1);

[0063] Conversely, when the absolute values ​​of the distance residual and the velocity residual are both not greater than the corresponding residual threshold, the update step is executed normally.

[0064] In some embodiments, the Kalman filtering process further includes a motion state constraint step, which is performed after the update step to ensure that the state estimate conforms to the physical motion laws of the railway train, specifically including:

[0065] Obtain the current optimal state estimate X(k|k)=[d(k),v(k)] output by the update step. T , where d(k) is the distance estimate and v(k) is the velocity estimate;

[0066] Apply a one-way motion constraint to the speed estimate: when the speed estimate v(k) < 0, it indicates that the train's reverse motion does not match the actual situation, and the speed estimate is forcibly corrected to v(k) = 0;

[0067] A non-negative distance constraint is imposed on the distance estimate: when the distance estimate d(k) < 0, it indicates that the train has passed the radar installation position, which contradicts the physical fact, and the distance estimate is forcibly corrected to d(k) = 0;

[0068] The state estimate after the motion state constraint step is used as the final state estimate result at the current moment, and is used in the prediction step of the next sampling period.

[0069] In some embodiments, the Kalman filtering process further includes a sampling period adaptive adjustment strategy, which is used to dynamically adjust the sampling period of the radar according to the distance between the train and the construction section, so as to reduce the computational load at long distances and improve the real-time performance of early warning at short distances.

[0070] The adaptive adjustment strategy for the sampling period is executed before the prediction step in each sampling period, and specifically includes:

[0071] Obtain the distance estimate d(k-1) in the optimal state estimate X(k-1|k-1) at the current moment, or obtain the final distance estimate after motion state constraints at the previous moment;

[0072] Based on the distance interval where the distance estimate is located, the corresponding sampling period value T is retrieved from a preset multi-level sampling period mapping table, and the state transition matrix A is updated with this sampling period value. and the process noise covariance matrix Q;

[0073] The multi-level sampling period mapping table defines the correspondence between multiple distance intervals and sampling period values. The shorter the distance interval, the smaller the corresponding sampling period value.

[0074] In some embodiments, in step S2, the forced preemption and activation of the relay host's transmission channel is achieved through the following sub-steps:

[0075] Step S2-1: After receiving the trigger signal, the automatic trigger controller converts the trigger signal into a high-level PTT control signal using its internally integrated level conversion module;

[0076] Step S2-2: The high-level PTT control signal is directly transmitted to the PTT control port of the relay host through a shielded hardwire. The shielded hardwire is used to suppress the degradation of the control signal by electromagnetic interference at the construction site.

[0077] Step S2-3: The PTT control port directly applies the high-level PTT control signal to the enable terminal of the relay host's transmit circuit. The priority of this enable terminal is higher than the processing priority of the relay host's receive demodulation logic and software protocol stack.

[0078] Step S2-4: In response to the high-level PTT control signal received by the enable terminal of the transmitting circuit, the transmitting channel of the relay host is immediately started. If the relay host is forwarding other voice calls at this time, the start operation of the transmitting channel will forcibly interrupt the current call, thereby realizing hardware-level preemption of the transmitting channel.

[0079] Step S2-5: At the same time as the transmission channel is activated, the automatic trigger controller synchronously transmits the alarm audio signal stored inside it to the audio input terminal of the relay host through the audio cable, so that the relay host can modulate it into a digital radio frequency signal.

[0080] In some embodiments, the method further includes a system reset step and a fault self-test step:

[0081] The system reset step is performed after step S4, and specifically includes:

[0082] After the train passes through the construction section, the trigger signal of the front-end sensing device disappears, and the automatic trigger controller stops outputting the PTT control signal after detecting that the trigger signal has disappeared.

[0083] In response to the cessation of the PTT control signal, the relay host shuts down its transmit channel and returns to the receive standby state, waiting for the next trigger signal;

[0084] The fault self-test step is executed independently of the system reset step, and specifically includes:

[0085] During system power-on initialization, the automatic trigger controller executes a power-on self-test program, sending test signals to the front-end sensing device and the relay host respectively, and checking whether the expected response signal is received, in order to verify whether the signal line connection with the front-end sensing device and the electrical connection of the PTT control line and audio line with the relay host are normal.

[0086] During normal system operation, the automatic trigger controller periodically generates heartbeat signals at preset time intervals and forwards the heartbeat signals to the digital handheld walkie-talkie through the relay host to indicate to the construction personnel that the system is in normal working condition.

[0087] When the power-on self-test program detects any electrical connection abnormality, or when the automatic trigger controller fails to receive the expected response signal multiple times in a row, the automatic trigger controller generates a fault alarm signal and broadcasts a fault alarm voice through the relay host.

[0088] As a second aspect of the present invention, a railway construction protection alarm system based on wireless intercom relay is also provided, which applies a railway construction protection alarm method based on wireless intercom relay as described above, the system comprising:

[0089] Front-end sensing devices are used to monitor the approach status of trains in railway construction sections in real time and generate trigger signals when preset trigger conditions are met.

[0090] An automatic trigger controller is electrically connected to the front-end sensing device via a signal line. It is used to receive the trigger signal and simultaneously output a PTT control signal and an alarm audio signal to the digital intercom relay host via a hardwired connection. The PTT control signal is used to forcibly preempt and activate the relay host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the relay host.

[0091] The digital intercom repeater host is connected to the automatic trigger controller via a hardwired connection. It is used to modulate the alarm audio signal into a digital radio frequency signal after the transmission channel is activated, and broadcast it to the construction area through an antenna.

[0092] At least one digital handheld walkie-talkie is placed at the construction site and operates on a preset channel to receive the digital radio frequency signal and output alarm voice.

[0093] The present invention has the following beneficial effects:

[0094] This invention provides a railway construction protection alarm method based on wireless intercom relay. It automatically monitors train approach status through front-end sensing devices, eliminating the need for manual monitoring and operation, thus removing the response delay of traditional manual protection methods that involve "alarm detection → manual judgment → manual button press." The automatic trigger controller simultaneously outputs PTT control signals and alarm audio signals via hardwired connection, ensuring that alarm commands reach the relay host with hardware-level priority, unaffected by the relay host's current operating status (such as forwarding other calls). The alarm audio signal is modulated into a digital radio frequency signal and broadcast via antenna, allowing all digital handheld intercoms within the construction area to receive and play the alarm voice, eliminating the need for construction personnel to wear dedicated receiving terminals. Attached Figure Description

[0095] Figure 1 A flowchart illustrating a railway construction protection alarm method based on wireless intercom relay provided in an embodiment of the present invention;

[0096] Figure 2 This is a schematic diagram of the framework of a railway construction protection alarm system based on wireless intercom relay, provided for an embodiment of the present invention. Detailed Implementation

[0097] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0098] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0099] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0100] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0101] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0102] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0103] To address at least one of the technical problems existing in the aforementioned related technologies, the present invention provides a railway construction protection alarm method based on wireless intercom relay. Figure 1 This is a flowchart illustrating a railway construction protection alarm method based on wireless intercom relay, provided by an embodiment of the present invention. The method includes:

[0104] Step S1: Monitor the train approach status of the railway construction section in real time through front-end sensing devices, and generate a trigger signal when the preset trigger conditions are met;

[0105] Step S2: The automatic trigger controller receives the trigger signal and simultaneously outputs a PTT control signal and an alarm audio signal to the digital intercom repeater host via a hardwired connection; wherein, the PTT control signal is used to forcibly preempt and start the repeater host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the repeater host;

[0106] Step S3: After the transmission channel is activated, the relay host modulates the alarm audio signal into a digital radio frequency signal and broadcasts it to the construction area through the antenna;

[0107] Step S4: The digital handheld walkie-talkie in the construction area receives the digital radio frequency signal on the preset channel and outputs an alarm voice.

[0108] It should be noted that the "simultaneous output" mentioned in step S2 does not mean that the two signals are output completely synchronously at an absolute time point. Rather, it means that in the same operation sequence of alarm triggering, after the automatic trigger controller outputs the PTT control signal to start the transmission channel, it immediately or within a very short time interval (usually within milliseconds) outputs the alarm audio signal to ensure that the transmission channel of the relay host has been established before the alarm audio signal arrives.

[0109] This invention provides a railway construction protection alarm method based on wireless intercom relay. It automatically monitors train approach status through front-end sensing devices, eliminating the need for manual monitoring and operation, thus removing the response delay of traditional manual protection methods that involve "alarm detection → manual judgment → manual button press." The automatic trigger controller simultaneously outputs PTT control signals and alarm audio signals via hardwired connection, ensuring that alarm commands reach the relay host with hardware-level priority, unaffected by the relay host's current operating status (such as forwarding other calls). The alarm audio signal is modulated into a digital radio frequency signal and broadcast via antenna, allowing all digital handheld intercoms within the construction area to receive and play the alarm voice, eliminating the need for construction personnel to wear dedicated receiving terminals.

[0110] In some embodiments, the front-end sensing device includes at least one of a train proximity warning radar, an infrared intrusion detector, a lidar, or a track circuit sensor; each type of front-end sensing device corresponds to different preset triggering conditions:

[0111] When the front-end sensing device is a train approach warning radar, the preset triggering condition is that the distance at which the radar detects the target object is less than or equal to a preset warning distance;

[0112] When the front-end sensing device is an infrared intrusion detector, the preset triggering condition is that the detector detects a target object entering a preset detection area;

[0113] When the front-end sensing device is a lidar, the preset trigger condition is that the lidar detects that the contour features of the target object match a preset train contour template;

[0114] When the front-end sensing device is a track circuit sensor, the preset trigger condition is that the sensor detects a change in the short-circuit state of the track circuit.

[0115] In one optional embodiment, the train approach warning radar is a millimeter-wave radar, and the preset warning distance is 2000 meters.

[0116] It should be noted that the "preset warning distance" is a trigger parameter defined specifically for train approach warning radar. For other types of front-end sensing devices, the spatial range parameters in their corresponding preset trigger conditions (such as the "preset detection area" of infrared intrusion detectors and the "train outline template" of lidar) may have different physical forms, but they are functionally equivalent to the "preset warning distance," used to define the spatial boundary of the target object entering the warning range of the construction section.

[0117] In the above embodiments, by supporting multiple types of front-end sensing devices, the system's environmental adaptability and deployment flexibility are improved. Specifically, different types of construction sites can select the most suitable front-end sensing devices according to their own characteristics. For example, in open track sections, train approach warning radar can be used to achieve long-distance (e.g., 2000 meters) train approach detection; in areas with obstructed visibility, such as tunnel entrances or curves, infrared intrusion detectors or lidar can be used to compensate for radar detection blind spots; in existing railway stations or switch areas, track circuit sensors can be used to achieve low-cost retrofitting using existing railway signaling facilities. Each device type corresponds to its own preset triggering conditions to ensure reliable alarm triggering in different application scenarios. For example, in scenarios where radar performance is degraded, such as nighttime construction or rainy / foggy weather, infrared intrusion detectors can be switched to as a supplementary sensing method to achieve reliable protection in all weather and all working conditions.

[0118] In some embodiments, when the front-end sensing device is a train approach warning radar, step S1 further includes a step of performing Kalman filtering on the raw observation data output by the radar to suppress measurement noise caused by electromagnetic interference and clutter reflection in the construction environment, and output the filtered stable distance and speed values.

[0119] The Kalman filtering process is based on the train distance-speed joint Kalman filtering algorithm, specifically including:

[0120] Establish a state vector X(k), with the train's distance d(k) and speed v(k) as state variables:

[0121] X(k) = ;

[0122] Based on the assumption of approximate uniform linear motion with random disturbances, a discrete-time state transition equation is established; wherein, the random disturbance is caused by unknown acceleration disturbances that the train may experience, and this disturbance is modeled using the process noise covariance matrix Q:

[0123] X(k)=A X(k-1)+W(k)

[0124] The state transition matrix is ​​as follows:

[0125] A= ;

[0126] T is the radar sampling period, and W(k) is the process noise vector;

[0127] Establish a linear observation equation, using the range observation value z directly output by the radar. d (k) and velocity observation z v (k) constitutes the observation vector:

[0128] Z(k) = ;

[0129] The relationship between the observation vector Z(k) and the state vector X(k) is expressed as follows:

[0130] Z(k) = H X(k)+V(k);

[0131] The observation matrix is ​​as follows:

[0132] H= ;

[0133] V(k) is the observation noise vector;

[0134] Through iterative calculations of the prediction and update steps, the state estimate is corrected one by one, and the filtered stable distance and velocity values ​​are output. The stable distance value is then compared with the preset warning distance to determine whether the preset triggering condition is met.

[0135] It should be noted that the "raw observation data" specifically refers to the distance and speed observation values ​​directly output by the train approach warning radar. In a preferred embodiment, the radar is a millimeter-wave radar, and its raw output data also includes information such as target azimuth and signal strength, but this information is not used as state variables in the Kalman filtering process of this invention. Furthermore, before Kalman filtering, the raw observation data can be converted to units and corrected to coordinates according to the radar's calibration parameters to ensure data consistency.

[0136] In the above embodiments, by introducing the Kalman filter algorithm into the radar signal processing stage, the pollution of radar measurement data by the complex electromagnetic environment and clutter interference at the construction site is effectively suppressed. Specifically, the algorithm uses the distance and speed of the train as state variables to establish a uniform linear motion model. Through iterative calculation of the state transition equation and observation equation, stable distance and speed estimates are extracted from the noisy raw observation data. Compared with the traditional method of directly using raw radar data, Kalman filtering can filter out measurement fluctuations caused by non-target factors such as reflections from construction machinery, birds flying overhead, and dust obstruction, so that the system only triggers an alarm when the train actually approaches.

[0137] In some embodiments, the Kalman filtering process sequentially performs a prediction step and an update step within each radar sampling period:

[0138] The prediction step includes:

[0139] Based on the optimal state estimate X(k-1|k-1) and its covariance matrix P(k-1|k-1) from the previous time step, predict the state and covariance at the current time step:

[0140] The state prediction equation is as follows:

[0141] X(k|k-1)=A X(k-1|k-1);

[0142] The covariance prediction equation is:

[0143] P(k|k-1)=A P(k-1|k-1) A T +Q;

[0144] Wherein, the process noise covariance matrix Q is based on the maximum acceleration amplitude a of the random disturbance. max The value is calculated from the radar sampling period T and used to characterize the non-modeled acceleration disturbances that the train may experience during approximately uniform motion.

[0145] Q= ;

[0146] In one alternative embodiment, a max =0.1m / s 2 When the sampling period T = 0.1s, the process noise covariance matrix is:

[0147] Q= ;

[0148] The update steps include:

[0149] Obtain the observation vector Z(k) = [z d (k),z v (k)] T The observed value is used to correct the state prediction value obtained in the prediction step, so as to obtain the optimal state estimate at the current time:

[0150] Calculate the Kalman gain matrix:

[0151] K(k)=P(k∣k-1) H T (H P(k∣k-1) H T +R) -1 ;

[0152] State update equation (optimal estimate):

[0153] X(k∣k)=X(k∣k-1)+K(k) (Z(k)-H X(k∣k-1));

[0154] Covariance update equation:

[0155] P(k∣k)=(IK(k) H) P(k∣k-1);

[0156] Where I is the second-order identity matrix; R is the observation noise covariance matrix, based on the ranging accuracy σ of the radar. d =2.0m and speed measurement accuracy σ v =0.3m / s Calculated as follows:

[0157] R= = .

[0158] By repeatedly executing the prediction and update steps in each sampling period, the filtered stable distance and velocity values ​​are continuously output.

[0159] In the above embodiments, the prediction step uses the state transition matrix A and the process noise covariance matrix Q to predict the current state and uncertainty based on the optimal estimate from the previous time step; the update step uses the Kalman gain K(k) to perform weighted fusion between the predicted and observed values ​​to obtain the optimal state estimate for the current time step. The process noise covariance matrix Q is calculated based on the train's maximum acceleration disturbance (e.g., 0.1 m / s²), reflecting the uncertainty of the train motion model; the observation noise covariance matrix R is determined based on the radar's ranging accuracy (e.g., ±2 m) and velocity accuracy (e.g., ±0.3 m / s), reflecting the sensor's measurement error.

[0160] In some embodiments, the Kalman filtering process further includes an initialization step, which is performed once after the system is powered on and before the prediction step is executed in the first sampling period, for determining the initial state estimate and its initial covariance matrix of the Kalman filter.

[0161] The initialization steps specifically include:

[0162] After the system is powered on and the radar output reaches a stable state, the first N valid observations of the radar output are continuously collected to obtain the range observations d1, d2, ..., dn for each sample.N and velocity observations v1, v2, ..., v N ;

[0163] Calculate the arithmetic mean of the N distance observations and the arithmetic mean of the N velocity observations to form the initial state vector:

[0164] X(0∣0)= ;

[0165] The initial covariance matrix P(0|0) is set as a diagonal matrix, and the values ​​of its diagonal elements are configured such that the diagonal elements of the Kalman gain matrix K(1) calculated in the first sampling period approach 1, so that the observations in the initial stage have a dominant weight in the state update.

[0166] The initial state vector X(0|0) and the initial covariance matrix P(0|0) are used as inputs to the prediction step in the first sampling period to start the iterative loop of the Kalman filtering process;

[0167] In an optional embodiment, N is 3, and the initial covariance matrix is:

[0168] P(0∣0)= .

[0169] It should be noted that the "stable state" refers to the state in which the radar, after power-on and completion of self-testing and warm-up processes, outputs distance and velocity observations with fluctuations less than a preset stability threshold. In a preferred embodiment, the criteria for determining the stable state are: the standard deviation of the distance observations from 10 consecutive samples is less than 0.5 meters, and the standard deviation of the velocity observations is less than 0.1 meters per second. When these conditions are met, the radar is considered to have entered a stable operating state and can begin collecting the necessary valid observations for initialization.

[0170] In the above embodiments, by explicitly defining the initialization steps of the Kalman filter, it is ensured that the filtering algorithm can quickly and accurately enter a stable working state after power-on. Specifically, after the system is powered on, it waits for the radar output to reach a stable state, and then continuously collects the first N valid observations (preferably 3), taking their arithmetic mean as the initial state vector. This avoids the influence of random errors that may be included in a single observation on the initial state. The initial covariance matrix is ​​set to a large value matrix that is much larger than the observation noise covariance matrix (e.g., 100 for the range direction and 10 for the velocity direction), making the Kalman gain of the first sampling period close to 1. The initial observations occupy the dominant weight in the state update, ensuring that the filtering algorithm can converge quickly. For example, on a cold winter morning, the radar output may be unstable due to the low temperature when it is first powered on. Waiting for the radar to stabilize before collecting data can avoid including abnormal data from the warm-up period in the initial state, allowing the filtering algorithm to output a state estimate close to the true value in the first sampling period.

[0171] In some embodiments, the Kalman filtering process is executed cyclically in each sampling period according to the following iterative process, and the iterative process is initiated after the initialization step is completed:

[0172] Step A: Obtain the distance estimate output after the motion state constraint step in the previous sampling period, execute the sampling period adaptive adjustment strategy, determine the sampling period value T of the current sampling period according to the distance interval in which the distance estimate is located, and update the state transition matrix A and the process noise covariance matrix Q based on the sampling period value T;

[0173] Step B: Perform the prediction step. Based on the optimal state estimate X(k-1|k-1) and its covariance matrix P(k-1|k-1) from the previous time step, predict the current state and covariance using the following formula:

[0174] X(k|k-1)=A X(k-1|k-1);

[0175] P(k|k-1)=A P(k-1|k-1) AT+Q;

[0176] Step C: Obtain the observation vector Z(k) at the current time and calculate the observation residual ε(k) = Z(k) - H X(k|k-1), execute the outlier removal mechanism to determine whether the observation vector is an outlier;

[0177] Step D: If the observed vector is not an outlier, then execute the update step, use the observed vector at the current time to correct the prediction result, and obtain the optimal state estimate X(k|k) and its covariance matrix P(k|k) at the current time; if the observed vector is an outlier, then skip the update step and directly use the output of the prediction step as the optimal state estimate at the current time, i.e., X(k|k)=X(k|k-1), P(k|k)=P(k|k-1);

[0178] Step E: Execute the motion state constraint step, apply physical constraints to the distance estimate d(k) and velocity estimate v(k) in the current optimal state estimate X(k|k)=[d(k),v(k)]T, and output the corrected final state estimate;

[0179] Step F: Use the final state estimate as input to the prediction step in the next sampling period, and return to step A to start the next iteration.

[0180] In some embodiments, the Kalman filtering process further includes an outlier removal mechanism, which is executed before the update step, specifically including:

[0181] Calculate the observation residual vector at the current time:

[0182] ε(k)=Z(k)-H X(k|k-1);

[0183] The observation residual vector includes a distance residual component ε. d (k) and velocity residual component ε v (k), that is:

[0184] ε(k)= ;

[0185] Obtain the ranging accuracy σ of the radar. d and speed measurement accuracy σ v And calculate the distance from the residual threshold N according to the preset residual threshold multiple N. σ d and velocity residual threshold N σ v ;

[0186] When the absolute value of the distance residual is |ε d (k) | greater than the distance residual threshold, or the absolute value of the velocity residual | ε v When (k)| is greater than the velocity residual threshold, the current observation vector Z(k) is determined to be an outlier, the update step is skipped, and the result of the prediction step is directly used as the optimal state estimate for the current moment.

[0187] X(k|k)=X(k|k-1);

[0188] P(k∣k)=P(k∣k-1);

[0189] Conversely, when the absolute values ​​of the distance residual and the velocity residual are both not greater than the corresponding residual threshold, the update step is executed normally.

[0190] In an optional embodiment, the residual threshold multiple N=3, and the ranging accuracy σ d =2.0m, the speed measurement accuracy σ v =0.3m / s.

[0191] It should be noted that the preset residual threshold multiple N is preferably between 2 and 5. When N=3, based on the 3σ principle of normal distribution, the probability of normal observations falling within the ±3σ range is approximately 99.73%. Therefore, only about 0.27% of normal observations will be misclassified as outliers, achieving a good balance between effectively removing outliers and retaining normal data. In practical applications, the value of N can be adjusted according to the electromagnetic interference intensity of the construction environment and the performance indicators of the radar. In environments with strong interference, the value of N can be appropriately reduced to improve the removal sensitivity.

[0192] In the above embodiments, by introducing an outlier removal mechanism based on the 3σ residual test, the contamination of the Kalman filter state by extreme outlier observations is effectively prevented. Specifically, this mechanism calculates the observation residual (the difference between the actual observation value and the predicted value) in each sampling period. When the absolute value of the distance residual or velocity residual exceeds three times the corresponding standard deviation, the current observation value is determined to be an outlier, and the update step is skipped directly. The predicted value is used as the state estimate for the current moment. This mechanism is particularly suitable for sudden interference scenarios common in railway construction sites: when a bird suddenly flies across the radar beam, it will generate a brief high-speed echo signal; when the construction robot arm swings, it will generate a sudden change in distance reading. If these abnormal data are used for state updates without discrimination, it will cause instantaneous distortion in the filtering results.

[0193] In some embodiments, the Kalman filtering process further includes a motion state constraint step, which is performed after the update step to ensure that the state estimate conforms to the physical motion laws of the railway train, specifically including:

[0194] Obtain the current optimal state estimate X(k|k)=[d(k),v(k)] output by the update step. T , where d(k) is the distance estimate and v(k) is the velocity estimate;

[0195] Apply motion direction constraints to the velocity estimate:

[0196] When the front-end sensing device is deployed in a unidirectional line section, if the speed estimate v(k) < 0, the speed estimate will be forcibly corrected to v(k) = 0.

[0197] When the front-end sensing device is deployed on a two-way track section, if the speed estimate v(k) < 0, its sign remains unchanged, but the reference frame of the distance estimate is flipped to reflect the trend of the train leaving the radar in the opposite direction.

[0198] Apply a non-negative distance constraint to the distance estimate:

[0199] When the distance estimate d(k) < 0, it indicates that the train has passed the radar installation position, and the distance estimate is forcibly corrected to d(k) = 0;

[0200] The state estimate after the motion state constraint step is used as the final state estimate result at the current moment, and is used in the prediction step of the next sampling period.

[0201] In the above embodiments, by applying motion state constraints, it is ensured that the state estimate output by the Kalman filter always conforms to the physical motion laws of the railway train. Specifically, the train can only travel in one direction along the track, so the speed estimate should not be negative; the straight-line distance between the train and the radar should not be less than zero. When the filtering calculation causes the state estimate to violate these physical constraints due to noise or model errors, the motion state constraint step forces it to be corrected to a reasonable boundary value. This mechanism is particularly important when the train has just passed the radar installation position: at this time, the actual distance value quickly approaches zero and then becomes negative (the train passes the radar position and continues to move forward), but the physical distance should be zero (the train is exactly below the radar) or increase again (the train moves away). For example, when the train passes the radar installation point at a speed of 80 km / h, the filtering algorithm may have a brief negative distance value within a few frames of data. The motion state constraint can immediately correct it to zero, avoiding logical errors caused by negative distances participating in subsequent calculations.

[0202] In some embodiments, the Kalman filtering process further includes a sampling period adaptive adjustment strategy, which is used to dynamically adjust the sampling period of the radar according to the distance between the train and the construction section, so as to reduce the computational load at long distances and improve the real-time performance of early warning at short distances.

[0203] The adaptive adjustment strategy for the sampling period is executed before the prediction step in each sampling period, and specifically includes:

[0204] Obtain the distance estimate d(k-1) in the optimal state estimate X(k-1|k-1) at the current moment, or obtain the final distance estimate after motion state constraints at the previous moment;

[0205] Based on the distance interval where the distance estimate is located, the corresponding sampling period value T is retrieved from a preset multi-level sampling period mapping table, and the state transition matrix A is updated with this sampling period value. and the process noise covariance matrix Q;

[0206] The multi-level sampling period mapping table defines the correspondence between multiple distance intervals and sampling period values. The shorter the distance interval, the smaller the corresponding sampling period value.

[0207] In one optional embodiment, the multi-level sampling period mapping table includes:

[0208] When the estimated distance is greater than the first distance threshold, the sampling period T = 0.2s;

[0209] When the distance estimate is between the first distance threshold and the second distance threshold, the sampling period T = 0.1s;

[0210] When the estimated distance is less than the second distance threshold, the sampling period T = 0.05s;

[0211] Wherein, the first distance threshold is 2000 meters, and the second distance threshold is 1000 meters.

[0212] It should be noted that in the first sampling period, since there is no optimal state estimate from the previous moment, the distance estimate used by the sampling period adaptive adjustment strategy is taken from the distance component in the initial state vector X(0|0) output by the initialization step. Starting from the second sampling period, the final distance estimate after the motion state constraint step of the previous sampling period is used.

[0213] In the above embodiments, an optimal balance is achieved between limited computing resources and the real-time requirements of early warning through an adaptive sampling period adjustment strategy. Specifically, when the train is far away (e.g., more than 2000 meters), it will take a long time for the train to reach the construction point, so the system uses a longer sampling period (e.g., 0.2 seconds) to reduce the processor's computing load and power consumption; when the train is at a medium distance (1000 to 2000 meters), a moderate sampling period (e.g., 0.1 seconds) is used; when the train approaches within 1000 meters, the shortest sampling period (e.g., 0.05 seconds) is used to update the state estimate at the highest frequency, ensuring that the system can respond quickly when the train approaches at high speed.

[0214] In some embodiments, in step S2, the forced preemption and activation of the relay host's transmission channel is achieved through the following sub-steps:

[0215] Step S2-1: After receiving the trigger signal, the automatic trigger controller converts the trigger signal into a high-level PTT control signal using its internally integrated level conversion module;

[0216] Step S2-2: The high-level PTT control signal is directly transmitted to the PTT control port of the relay host through a shielded hardwire. The shielded hardwire is used to suppress the degradation of the control signal by electromagnetic interference at the construction site.

[0217] Step S2-3: The PTT control port directly applies the high-level PTT control signal to the enable terminal of the relay host's transmit circuit. The priority of this enable terminal is higher than the processing priority of the relay host's receive demodulation logic and software protocol stack.

[0218] Step S2-4: In response to the high-level PTT control signal received by the enable terminal of the transmitting circuit, the transmitting channel of the relay host is immediately started. If the relay host is forwarding other voice calls at this time, the start operation of the transmitting channel will forcibly interrupt the current call, thereby realizing hardware-level preemption of the transmitting channel.

[0219] Step S2-5: At the same time as the transmission channel is activated, the automatic trigger controller synchronously transmits the alarm audio signal stored inside it to the audio input terminal of the relay host through the audio cable, so that the relay host can modulate it into a digital radio frequency signal.

[0220] In the above embodiments, the problem of alarm signals being blocked by ordinary conversations is fundamentally solved through a hardware-level channel preemption mechanism. Specifically, the automatic trigger controller directly applies a high-level PTT control signal to the enable terminal of the relay host's transmitting circuit through a shielded hardwired connection. The hardware priority of this enable terminal is higher than the relay host's receiving demodulation logic and software protocol stack. Even if the relay host is forwarding routine conversations between construction workers, the PTT control signal can forcibly interrupt the current conversation and immediately start the transmission channel, ensuring that alarm commands are transmitted with the highest priority. Compared with traditional software priority schemes, hardware-level preemption does not require waiting for the software protocol stack to finish processing the current task, reducing the response time from tens of milliseconds at the software level to microseconds at the hardware level. For example, during busy construction periods, multiple safety personnel may be communicating intensively via walkie-talkies. If a train approaches, hardware-level preemption can forcibly insert alarm voice without affecting alarm broadcasts, and then resume normal conversation forwarding after the alarm ends.

[0221] In some embodiments, the generation and output of the alarm audio signal in step S2 are achieved through the following sub-steps:

[0222] Step S2-5-1: Pre-store at least one digital alarm voice file in the internal storage medium of the automatic trigger controller, the alarm voice file containing voice content for alerting construction personnel;

[0223] Step S2-5-2: At the same time or after outputting the PTT control signal, the automatic trigger controller reads the alarm voice file from the storage medium and sends it to the internally integrated digital-to-analog converter module;

[0224] Step S2-5-3: The digital-to-analog conversion module converts the digital alarm voice file into an analog audio signal;

[0225] Step S2-5-4: After the analog audio signal is amplified and level matched, it is transmitted to the audio input terminal of the repeater host through a shielded audio cable to ensure that the voltage amplitude of the analog audio signal is within the rated input range of the audio input terminal of the repeater host;

[0226] The output paths of the PTT control signal and the alarm audio signal are wired independently and in parallel within the automatic trigger controller. The two signals are transmitted to the PTT control port and audio input terminal of the relay host respectively via their respective shielded hardwires.

[0227] It should be noted that the "at least one digital alarm voice file" allows multiple voice files to coexist in the storage medium. In a preferred embodiment, different voice files correspond to different alarm scenarios or warning levels. For example, "An upward train is approaching, please immediately get off the track to avoid it" corresponds to an upward train trigger scenario, and "A downward train is approaching, please immediately get off the track to avoid it" corresponds to a downward train trigger scenario. The automatic trigger controller selects the voice file that matches the current alarm scenario from among the multiple voice files for playback based on the source or type of the trigger signal.

[0228] In the above embodiments, the clarity and reliability of the alarm audio signal are ensured by clearly defining the generation and output path of the alarm audio signal. Specifically, the alarm audio file is pre-stored in digital format in the storage medium of the automatic trigger controller, avoiding the delay and instability that may be caused by real-time speech synthesis; after digital-to-analog conversion, the analog audio signal is amplified and level matched before being transmitted to the audio input terminal of the relay host, ensuring that the voltage amplitude of the audio signal matches the rated input range of the relay host, avoiding modulation distortion due to insufficient signal or clipping distortion due to excessive signal; the PTT control signal and the alarm audio signal are wired independently in parallel within the automatic trigger controller, and the two signals are transmitted separately through their respective shielded hardwired connections, avoiding signal crosstalk. For example, in a noisy construction site, if the volume of the alarm audio is insufficient or the sound quality is unclear, construction workers may not be able to hear the alarm content clearly. Through level matching and power amplification, the output sound pressure level of the alarm audio can reach more than 100 decibels, making it clearly audible even in environments with loud machinery.

[0229] In some embodiments, the method further includes a system reset step and a fault self-test step:

[0230] The system reset step is performed after step S4, and specifically includes:

[0231] After the train passes through the construction section, the trigger signal of the front-end sensing device disappears, and the automatic trigger controller stops outputting the PTT control signal after detecting that the trigger signal has disappeared.

[0232] In response to the cessation of the PTT control signal, the relay host shuts down its transmit channel and returns to the receive standby state, waiting for the next trigger signal;

[0233] The fault self-test step is executed independently of the system reset step, and specifically includes:

[0234] During system power-on initialization, the automatic trigger controller executes a power-on self-test program, sending test signals to the front-end sensing device and the relay host respectively, and checking whether the expected response signal is received, in order to verify whether the signal line connection with the front-end sensing device and the electrical connection of the PTT control line and audio line with the relay host are normal.

[0235] During normal system operation, the automatic trigger controller periodically generates heartbeat signals at preset time intervals and forwards the heartbeat signals to the digital handheld walkie-talkie through the relay host to indicate to the construction personnel that the system is in normal working condition.

[0236] When the power-on self-test program detects any electrical connection abnormality, or when the automatic trigger controller fails to receive the expected response signal multiple times in a row, the automatic trigger controller generates a fault alarm signal and broadcasts a fault alarm voice through the relay host.

[0237] It should be noted that the "multiple consecutive times" is preferably three consecutive times. When the automatic trigger controller fails to receive the expected response signal from the front-end sensing device or the relay host for three consecutive times, it determines that the corresponding device has malfunctioned. This fault-tolerance mechanism can avoid false alarms caused by a single communication interruption, while ensuring that an alarm can be issued in a timely manner when the device actually malfunctions.

[0238] In the above embodiments, the long-term reliable operation of the system is ensured through a dual guarantee mechanism of system reset and fault self-check steps. Specifically, the system reset step automatically restores the relay host to the receiving standby state after the train passes, enabling the system to continuously handle multiple train approach events without manual intervention. The fault self-check step includes two levels: power-on self-check and periodic heartbeat monitoring. The power-on self-check verifies the integrity of all electrical connections when the system starts up, preventing the system from operating with defects due to loose wiring or equipment failure. The periodic heartbeat signal is forwarded by the relay host to the construction personnel's handheld walkie-talkie, allowing the construction personnel to confirm that the system is in normal working condition at any time. When an abnormality is detected, the system automatically broadcasts a fault alarm voice to prompt the construction personnel to take emergency measures. For example, during nighttime track maintenance operations lasting several hours, the system sends a heartbeat signal every 30 seconds. The construction personnel can hear a regular "beep" sound through the handheld walkie-talkie to confirm that the system is normal. Once the heartbeat signal is interrupted more than 3 times, the system immediately broadcasts a voice alarm: "System failure, please pay attention to manual protection."

[0239] In some embodiments, after the relay host returns to the receiving standby state in the system reset step, a communication restoration step is further included:

[0240] The automatic trigger controller determines that the alarm event has ended after confirming that the trigger signal of the front-end sensing device has been continuously absent for more than the preset reset delay time.

[0241] In response to the end of the alarm event, the relay host exits the forced transmission state and re-enables its receiving demodulation logic and uplink signal forwarding function.

[0242] The relay host begins to receive uplink signals sent by construction workers through the digital handheld walkie-talkies again, and modulates the uplink signals to forward them to other digital handheld walkie-talkies in the same channel group, so that normal voice communication between construction workers is restored.

[0243] Based on the same inventive concept, this invention also provides a railway construction protection alarm system based on wireless intercom relay, see [link to relevant documentation]. Figure 2 As shown, the system includes:

[0244] Front-end sensing devices are used to monitor the approach status of trains in railway construction sections in real time and generate trigger signals when preset trigger conditions are met.

[0245] An automatic trigger controller is electrically connected to the front-end sensing device via a signal line. It is used to receive the trigger signal and simultaneously output a PTT control signal and an alarm audio signal to the digital intercom relay host via a hardwired connection. The PTT control signal is used to forcibly preempt and activate the relay host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the relay host.

[0246] The digital intercom repeater host is connected to the automatic trigger controller via a hardwired connection. It is used to modulate the alarm audio signal into a digital radio frequency signal after the transmission channel is activated, and broadcast it to the construction area through an antenna.

[0247] At least one digital handheld walkie-talkie is placed at the construction site and operates on a preset channel to receive the digital radio frequency signal and output alarm voice.

[0248] The digital intercom relay host is preferably a 400MHz digital intercom relay host, which is used to modulate the alarm audio signal into a 400MHz digital radio frequency signal and broadcast it to the construction area through an elevated antenna. Correspondingly, the digital handheld intercom is also preferably a 400MHz digital radio frequency signal.

[0249] It should be noted that the upper limit of the number of "at least one digital handheld walkie-talkie" depends on the channel capacity and transmission power of the digital walkie-talkie relay host. In a preferred embodiment, the digital walkie-talkie relay host adopts the DMR or PDT digital trunking communication protocol, supporting dozens to hundreds of digital handheld walkie-talkies to access within the same channel group. All connected handheld walkie-talkies can simultaneously receive the alarm broadcast as long as they operate on a preset channel with the same downlink frequency as the relay host, without the need for additional pairing or registration operations.

[0250] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A railway construction protection alarm method based on wireless intercom relay, characterized in that, Includes the following steps: Step S1: Monitor the train approach status of the railway construction section in real time through front-end sensing devices, and generate a trigger signal when the preset trigger conditions are met; Step S2: The automatic trigger controller receives the trigger signal and simultaneously outputs a PTT control signal and an alarm audio signal to the digital intercom repeater host via a hardwired connection; wherein, the PTT control signal is used to forcibly preempt and start the repeater host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the repeater host; Step S3: After the transmission channel is activated, the relay host modulates the alarm audio signal into a digital radio frequency signal and broadcasts it to the construction area through the antenna; Step S4: The digital handheld walkie-talkie in the construction area receives the digital radio frequency signal on the preset channel and outputs an alarm voice.

2. The method according to claim 1, characterized in that, The front-end sensing device includes at least one of train proximity warning radar, infrared intrusion detector, lidar, or track circuit sensor; each type of front-end sensing device corresponds to different preset triggering conditions: When the front-end sensing device is a train approach warning radar, the preset triggering condition is that the distance at which the radar detects the target object is less than or equal to a preset warning distance; When the front-end sensing device is an infrared intrusion detector, the preset triggering condition is that the detector detects a target object entering a preset detection area; When the front-end sensing device is a lidar, the preset trigger condition is that the lidar detects that the contour features of the target object match a preset train contour template; When the front-end sensing device is a track circuit sensor, the preset trigger condition is that the sensor detects a change in the short-circuit state of the track circuit.

3. The method according to claim 2, characterized in that, When the front-end sensing device is a train approach warning radar, step S1 also includes a step of performing Kalman filtering on the raw observation data output by the radar to suppress measurement noise caused by electromagnetic interference and clutter reflection in the construction environment, and output the filtered stable distance and speed values. The Kalman filtering process is based on the train distance-speed joint Kalman filtering algorithm, specifically including: Establish a state vector X(k), with the train's distance d(k) and speed v(k) as state variables: X(k)= ; Based on the assumption of uniform linear motion, the discrete-time state transition equation is established: X(k)=A X(k-1)+W(k) The state transition matrix is ​​as follows: A= ； T is the radar sampling period, and W(k) is the process noise vector; Establish a linear observation equation, using the range observation value z directly output by the radar. d (k) and velocity observation z v (k) constitutes the observation vector: Z(k)= ; The relationship between the observation vector Z(k) and the state vector X(k) is expressed as follows: Z(k)=H X(k)+V(k); The observation matrix is ​​as follows: H= ； V(k) is the observation noise vector; Through iterative calculations of the prediction and update steps, the state estimate is corrected one by one, and the filtered stable distance and velocity values ​​are output. The stable distance value is then compared with the preset warning distance to determine whether the preset triggering condition is met.

4. The method according to claim 3, characterized in that, The Kalman filtering process sequentially performs a prediction step and an update step within each radar sampling period: The prediction step includes: Based on the optimal state estimate X(k-1|k-1) and its covariance matrix P(k-1|k-1) from the previous time step, predict the current state and covariance: The state prediction equation is as follows: X(k∣k-1)=A X(k-1∣k-1); The covariance prediction equation is: P(k∣k-1)=A P(k-1∣k-1) A T +Q; Where Q is the process noise covariance matrix, based on the train's maximum acceleration disturbance a. max The results are calculated from the radar sampling period T: Q= ； The update steps include: Obtain the observation vector Z(k) = [z d (k),z v (k)] T The observed value is used to correct the state prediction value obtained in the prediction step, so as to obtain the optimal state estimate at the current time: Calculate the Kalman gain matrix: K(k)=P(k∣k-1) H T (H P(k∣k-1) H T +R) -1 ; State update equation (optimal estimate): X(k∣k)=X(k∣k-1)+K(k) (Z(k)-H X(k∣k-1)); Covariance update equation: P(k∣k)=(I-K(k) H) P(k∣k-1); Where I is the second-order identity matrix; R is the observation noise covariance matrix, based on the ranging accuracy σ of the radar. d and speed measurement accuracy σ v The calculation yielded: R= 。 5. The method according to claim 4, characterized in that, The Kalman filtering process also includes an outlier removal mechanism, which is executed before the update step, specifically including: Calculate the observation residual vector at the current time: ε(k)=Z(k)-H X(k∣k-1); The observation residual vector includes a distance residual component ε. d (k) and velocity residual component ε v (k), that is: e(k)= ; Obtain the ranging accuracy σ of the radar. d and speed measurement accuracy σ v And calculate the distance to the residual threshold N according to the preset residual threshold multiple N. σ d and velocity residual threshold N σ v ; When the absolute value of the distance residual is |ε d (k) | greater than the distance residual threshold, or the absolute value of the velocity residual | ε v When (k)| is greater than the velocity residual threshold, the current observation vector Z(k) is determined to be an outlier, the update step is skipped, and the result of the prediction step is directly used as the optimal state estimate for the current moment. X(k|k)=X(k|k-1); P(k∣k)=P(k∣k-1); Conversely, when the absolute values ​​of the distance residual and the velocity residual are both not greater than the corresponding residual threshold, the update step is executed normally.

6. The method according to claim 4, characterized in that, The Kalman filtering process further includes a motion state constraint step, which is executed after the update step. This motion state constraint step ensures that the state estimate conforms to the physical motion laws of the railway train, specifically including: Obtain the current optimal state estimate X(k|k)=[d(k),v(k)] output by the update step. T , where d(k) is the distance estimate and v(k) is the velocity estimate; Apply a one-way motion constraint to the speed estimate: when the speed estimate v(k) < 0, it indicates that the train's reverse motion does not match the actual situation, and the speed estimate is forcibly corrected to v(k) = 0; A non-negative distance constraint is imposed on the distance estimate: when the distance estimate d(k) < 0, it indicates that the train has passed the radar installation position, which contradicts the physical fact, and the distance estimate is forcibly corrected to d(k) = 0; The state estimate after the motion state constraint step is used as the final state estimate result at the current moment, and is used in the prediction step of the next sampling period.

7. The method according to claim 4, characterized in that, The Kalman filtering process also includes a sampling period adaptive adjustment strategy, which is used to dynamically adjust the sampling period of the radar according to the distance between the train and the construction section, so as to reduce the computational load at long distances and improve the real-time performance of early warning at short distances. The adaptive adjustment strategy for the sampling period is executed before the prediction step in each sampling period, and specifically includes: Obtain the distance estimate d(k-1) in the optimal state estimate X(k-1|k-1) at the current moment, or obtain the final distance estimate after motion state constraints at the previous moment; Based on the distance interval where the distance estimate is located, the corresponding sampling period value T is retrieved from a preset multi-level sampling period mapping table, and the state transition matrix A is updated with this sampling period value. and the process noise covariance matrix Q; The multi-level sampling period mapping table defines the correspondence between multiple distance intervals and sampling period values. The shorter the distance interval, the smaller the corresponding sampling period value.

8. The method according to claim 1, characterized in that, In step S2, the forced preemption and activation of the relay host's transmission channel is achieved through the following sub-steps: Step S2-1: After receiving the trigger signal, the automatic trigger controller converts the trigger signal into a high-level PTT control signal using its internally integrated level conversion module; Step S2-2: The high-level PTT control signal is directly transmitted to the PTT control port of the relay host through a shielded hardwire. The shielded hardwire is used to suppress the degradation of the control signal by electromagnetic interference at the construction site. Step S2-3: The PTT control port directly applies the high-level PTT control signal to the enable terminal of the relay host's transmit circuit. The priority of this enable terminal is higher than the processing priority of the relay host's receive demodulation logic and software protocol stack. Step S2-4: In response to the high-level PTT control signal received by the enable terminal of the transmitting circuit, the transmitting channel of the relay host is immediately started. If the relay host is forwarding other voice calls at this time, the start operation of the transmitting channel will forcibly interrupt the current call, thereby realizing hardware-level preemption of the transmitting channel. Step S2-5: At the same time as the transmission channel is activated, the automatic trigger controller synchronously transmits the alarm audio signal stored inside it to the audio input terminal of the relay host through the audio cable, so that the relay host can modulate it into a digital radio frequency signal.

9. The method according to claim 1, characterized in that, The method also includes a system reset step and a fault self-test step: The system reset step is performed after step S4, and specifically includes: After the train passes through the construction section, the trigger signal of the front-end sensing device disappears, and the automatic trigger controller stops outputting the PTT control signal after detecting that the trigger signal has disappeared; In response to the cessation of the PTT control signal, the relay host shuts down its transmit channel and returns to the receive standby state, waiting for the next trigger signal; The fault self-test step is executed independently of the system reset step, and specifically includes: During system power-on initialization, the automatic trigger controller executes a power-on self-test program, sending test signals to the front-end sensing device and the relay host respectively, and checking whether the expected response signal is received, in order to verify whether the signal line connection with the front-end sensing device and the electrical connection of the PTT control line and audio line with the relay host are normal. During normal system operation, the automatic trigger controller periodically generates heartbeat signals at preset time intervals and forwards the heartbeat signals to the digital handheld walkie-talkie through the relay host to indicate to the construction personnel that the system is in normal working condition. When the power-on self-test program detects any electrical connection abnormality, or when the automatic trigger controller fails to receive the expected response signal multiple times in a row, the automatic trigger controller generates a fault alarm signal and broadcasts a fault alarm voice through the relay host.

10. A railway construction protection alarm system based on wireless intercom relay, employing the railway construction protection alarm method based on wireless intercom relay as described in any one of claims 1-9, characterized in that, The system includes: Front-end sensing devices are used to monitor the approach status of trains in railway construction sections in real time and generate trigger signals when preset trigger conditions are met. An automatic trigger controller is electrically connected to the front-end sensing device via a signal line. It is used to receive the trigger signal and simultaneously output a PTT control signal and an alarm audio signal to the digital intercom relay host via a hardwired connection. The PTT control signal is used to forcibly preempt and activate the relay host's transmission channel, and the alarm audio signal is used to be modulated into a digital radio frequency signal by the relay host. The digital intercom repeater host is connected to the automatic trigger controller via a hardwired connection. It is used to modulate the alarm audio signal into a digital radio frequency signal after the transmission channel is activated, and broadcast it to the construction area through an antenna. At least one digital handheld walkie-talkie is placed at the construction site and operates on a preset channel to receive the digital radio frequency signal and output alarm voice.

Images