A mine transport track anti-intrusion alarm method and alarm system

Abstract

The present application provides an anti-intrusion alarm method and alarm system for mine transportation tracks, which relate to the technical field of intrusion monitoring. The method comprises: first deployment, first collection, intrusion judgment and alarm steps; the system comprises: a first deployment module, a first collection module, an intrusion judgment module and an alarm module. The present application monitors the transportation tracks through different sensors, and can timely detect and prevent track intrusion incidents, thereby improving the safety of mine transportation.

Description

Technical Field

[0001] The present application relates to the technical field of intrusion monitoring, and in particular to an anti-intrusion alarm method and alarm system for mine transportation tracks. Background Art

[0002] In the mining environment, the safety monitoring of the transport track is a vital task. Due to the complex and changeable mining environment, there are many potential safety hazards, such as personnel entering the track area by mistake, objects sliding, etc., which may threaten the trains or mine cars on the transport track, thus causing safety accidents.

[0003] Traditional methods of safety monitoring of mine transport tracks mainly rely on manual inspections and single sensor monitoring. However, manual inspections have problems such as long inspection cycles, limited coverage, and slow response speed, making it difficult to detect and deal with potential safety hazards in a timely manner. Single sensor monitoring, although it can detect target objects to a certain extent, is often prone to false alarms or missed alarms due to the limitations of the sensor itself, and cannot meet the accuracy and reliability requirements of mine transport track anti-intrusion alarms. Summary of the invention

[0004] In order to timely detect and prevent track intrusion incidents and improve the safety of mine transportation, the present application provides a mine transportation track anti-intrusion alarm method and alarm system.

[0005] In the first aspect, the present application provides a mine transport track anti-intrusion alarm method, which adopts the following technical solution:

[0006] A mine transport track anti-intrusion alarm method, comprising the following steps:

[0007] First deployment: infrared sensors, microwave sensors and laser sensors are installed in the monitoring area. The laser beams sent by all laser sensors form a three-dimensional channel surrounding the transport track and the transport vehicle.

[0008] First acquisition: collecting the output data of the infrared sensor, recorded as the first data; collecting the output data of the microwave sensor, recorded as the second data; collecting the output data of the laser sensor, recorded as the third data;

[0009] Intrusion judgment: including first judgment, first calculation, second judgment and third judgment;

[0010] First judgment: judging whether the first data belongs to a preset range, if so, executing the first calculation step; if not, executing the first collection step;

[0011] First calculation: calculating the moving direction of the first target and the shortest distance between the first target and the transport track based on the second data;

[0012] Second judgment: judging whether the shortest distance is less than a preset safety threshold and the moving direction is toward the track, if so, executing the first alarm step; if not, executing the first collection step;

[0013] Third judgment: judging whether the second target has an intrusion trend based on the third data, if so, executing the second alarm step; if not, executing the first collection step;

[0014] Alarm: including the first alarm and the second alarm;

[0015] First alarm: output sound and light alarm signal;

[0016] Second alarm: Send an alarm signal to operation and maintenance personnel.

[0017] This application first uses an infrared sensor to detect whether the first target (a living intrusion target, i.e., a person or an animal) has entered the monitoring area, and then uses the second data collected by the microwave sensor to measure whether the first target has an intrusion trend (i.e., the direction of movement is toward the transportation track and the shortest distance is less than the preset safety threshold). If so, an audible and visual alarm signal is issued to drive the first target away; otherwise, the first acquisition step is re-executed to continue monitoring the monitoring area; at the same time, this application also uses the third data to determine whether the second target (i.e., inanimate objects such as stones) has an intrusion trend. If so, an alarm signal is issued to the operation and maintenance personnel, otherwise, the first acquisition step is re-executed to continue monitoring the monitoring area. This application achieves multi-dimensional and multi-level coverage of the monitoring area by integrating multiple sensors such as infrared sensors, microwave sensors, and laser sensors, and can prevent and promptly detect intrusion incidents, thereby improving the safety of the mine transportation process.

[0018] Optionally, before performing the first deployment step, the method further includes:

[0019] First modeling: acquiring point cloud data of the transport track and the surrounding environment of the transport track, and building a three-dimensional model through a surface reconstruction algorithm based on the point cloud data;

[0020] First planning: planning the positions of infrared sensors, microwave sensors and laser sensors in the three-dimensional model;

[0021] First simulation: Introduce different intrusion events into the 3D model to conduct intrusion simulation tests, obtain simulation results, and determine the orientation of the infrared sensor, microwave sensor, and laser sensor based on the simulation results;

[0022] In the first deployment step, infrared sensors, microwave sensors and laser sensors are installed in the monitoring area according to the positions in the first planning step and the orientations in the first simulation step.

[0023] This application obtains point cloud data of the transport track and its surrounding environment, and applies a surface reconstruction algorithm to establish a three-dimensional model. Then, the positions of infrared sensors, microwave sensors, and laser sensors are planned in the three-dimensional model, so that various sensors can cover the areas that need to be monitored. Then, different intrusion events are introduced to conduct intrusion simulation tests to simulate various situations in actual operation, so as to evaluate the rationality of the layout and orientation of the sensors. Then, the orientation of the sensor is determined according to the simulation results, so that intrusion events can be accurately identified and responded to in actual applications. Then, the sensor is installed in the actual monitoring area according to the position in the first planning step and the orientation in the first simulation step, which helps to reduce errors and omissions in the installation process and improve stability and reliability. This application realizes comprehensive monitoring of the transport track and its surrounding environment, which not only improves accuracy and reliability, but also reduces the cost of installation and maintenance, and provides a strong guarantee for the safe operation of the transport track.

[0024] Optionally, after the step of executing the alarm, the method further includes:

[0025] Second calculation: counting the number of executions of the first alarm step in each monitoring area, recorded as fourth data; counting the number of executions of the second alarm step in each monitoring area, recorded as fifth data;

[0026] First acquisition: acquiring fourth data greater than a preset number of times threshold, and recording it as new fourth data; acquiring fifth data greater than a preset number of times threshold, and recording it as new fifth data;

[0027] Fourth judgment: judging whether the new fourth data and the new fifth data correspond to the same monitoring area, if so, executing the second deployment step; if not, executing the first collection step;

[0028] Second deployment: Increase the preset safety threshold and execute the first acquisition steps.

[0029] This application uses the second calculation step to count the number of executions of the first alarm and the second alarm in each monitoring area, which helps to understand the alarm frequency of each monitoring area. In the first acquisition step, this application screens the fourth data and the fifth data that are greater than the preset threshold, so as to focus on those monitoring areas with frequent alarms and potential safety hazards. Then in the fourth judgment step, by judging whether the new fourth data and the new fifth data correspond to the same monitoring area, it is decided whether to execute the second deployment step or the first acquisition step. If the new fourth data and the new fifth data correspond to the same monitoring area, it means that the invasion trend of the second target in the area may be a potential safety hazard caused by the first target. Therefore, it is necessary to execute the second deployment step (increase the preset safety threshold) to achieve early warning, so as to drive away people or animals and reduce invasion events caused by the second target.

[0030] Optionally, after executing the fourth determination step and before executing the second deployment step, the method further includes:

[0031] Second acquisition: obtaining the timestamp of the sound and light alarm signal corresponding to the new fourth data, recorded as the first timestamp; obtaining the timestamp of the alarm signal corresponding to the new fifth data, recorded as the second timestamp;

[0032] Fifth judgment: judge in turn whether the difference between each first timestamp and each second timestamp is less than a preset time threshold, if so, execute the second deployment step; if not, execute the first collection step.

[0033] The present application obtains the timestamp of the sound and light alarm signal corresponding to the new fourth data (i.e., the fourth data greater than the preset threshold, corresponding to the first alarm number of a certain monitoring area) and the timestamp of each alarm signal corresponding to the new fifth data (i.e., the fifth data greater than the preset threshold, corresponding to the second alarm number of a certain monitoring area), and in the fifth judgment step, by comparing whether the first timestamp and the second timestamp are the same or similar, it is judged whether the sound and light alarm signal corresponding to the new fourth data and the alarm signal corresponding to the new fifth data are generated at the same time point or a similar time point, which helps to confirm whether the two data reflect the alarm situation in the same time period. If the timestamps are the same or similar, it means that the second alarm is caused by the first alarm. At this time, the second deployment step (increasing the preset safety threshold) is executed to expand the scope of the early warning, thereby reducing the possibility of the first target causing the invasion of the second target. If the timestamps are different, it means that the two data reflect the situation of different time periods or different events, and the first acquisition step is re-executed at this time. After adding the second acquisition and the fifth judgment steps, the present application can more accurately judge the relevance and time characteristics of the alarm data, so as to make more reasonable decisions, which helps to improve the accuracy and reliability of the alarm and reduce the possibility of false alarms and missed alarms.

[0034] Optionally, after performing the first simulation step and before performing the first deployment step, the method further includes:

[0035] Second collection: collect simulation data of laser sensors under each intrusion event;

[0036] Feature extraction: Use simulation data to extract intrusion features of different intrusion events;

[0037] Second modeling: construct an invasion model based on the invasion characteristics. The calculation model of the invasion model is as follows:

[0038] ;

[0039] in, is the output probability of the intrusion model; A is the intrusion event; is the i-th intrusion feature; n is the number of intrusion features; is the probability of the existence of the i-th intrusion feature; is the probability of an intrusion event occurring under the condition that the i-th intrusion feature exists;

[0040] Association: Establish the association between the intrusion features and the output probability of the intrusion model.

[0041] This application collects simulation data of laser sensors under each intrusion event, and then uses these simulation data to obtain detailed information about the intrusion event, such as the type, location, time, etc. of the intrusion. After collecting the simulation data, through data analysis and processing, the significant features that can distinguish different intrusion events, namely the intrusion features, are identified. These intrusion features include the time, type, location, etc. of the intrusion. By extracting the intrusion features, the key information required to classify the intrusion events can be retained while reducing the complexity of the data. After the intrusion features are extracted, the modeling step uses these intrusion features to build an intrusion model that can predict intrusion events. The intrusion model evaluates the possibility of different intrusion events by calculating the output probability. The association step links the extracted intrusion features with the output probability of the intrusion model, so as to facilitate the retrieval of the output probability of the intrusion model according to the intrusion features.

[0042] Optionally, after executing the third judgment step and before executing the second alarm step, the method further includes:

[0043] Third calculation: extracting the features contained in the third data, and calculating the similarity between the features and the i-th intrusion feature using a cosine similarity algorithm;

[0044] Sixth judgment: judging whether the similarity in the third calculation step is greater than a preset similarity threshold, if so, executing the calling step, if not, executing the iteration step;

[0045] Iteration: take the i+1th intrusion feature as the new ith intrusion feature until the preset stop condition is met and execute the first acquisition step;

[0046] Retrieve: Use the i-th intrusion feature to retrieve the output probability of the intrusion model;

[0047] Seventh judgment: judge whether the output probability described in the retrieved step is greater than the preset probability threshold, if so, execute the second alarm step; if not, execute the first collection step.

[0048] The present application extracts the features contained in the third data, and uses the cosine similarity algorithm to calculate the similarity between the features and the existing intrusion features, and quantifies the similarity between the two by the calculated similarity. Then, by comparing the calculated similarity with the preset similarity threshold, it is determined whether the third data is considered to be sufficiently similar to the i-th intrusion feature. If the similarity is greater than the preset similarity threshold, the two are considered to be similar, indicating that an intrusion event similar to the i-th intrusion feature has occurred; if the similarity is not greater than the preset similarity threshold, the two are considered to be dissimilar. If the result of the sixth judgment is yes (i.e., the similarity is greater than the threshold), the associated output probability is retrieved from the intrusion model using the currently matched i-th intrusion feature, i.e., the probability of an intrusion event occurring under the condition that the i-th intrusion feature exists, and then by comparing the output probability retrieved from the intrusion model with the preset probability threshold, it is determined whether to trigger an alarm. If the output probability is greater than the preset probability threshold, it is considered that the possibility of a subsequent intrusion event is relatively high, and the second alarm step needs to be executed; if the output probability is not greater than the probability threshold, it is considered that the current situation is insufficient to trigger an alarm. The present application improves the accuracy and flexibility of intrusion detection by calculating the similarity between the features of the third data and the known intrusion features and using the output probability of the intrusion model to determine whether an alarm is needed.

[0049] Optionally, after executing the sixth judgment step and before executing the calling step, the method further includes:

[0050] Fourth calculation: using the cosine similarity algorithm to calculate the similarity between the feature and the remaining intrusion features respectively, recorded as the sixth data;

[0051] Fifth calculation: obtaining sixth data greater than a preset similarity threshold and recording it as new sixth data;

[0052] Eighth judgment: judging whether the amount of the new sixth data is greater than a preset amount threshold, if so, executing the step of retrieving; if not, executing the step of updating the feature;

[0053] Feature update: All intrusion features are integrated into a feature set, and the features are added to the feature set, and the second modeling step is performed.

[0054] The present application calculates the similarity between the feature extracted from the third data and all the remaining intrusion features. The calculation of the similarity helps to identify whether the feature is similar to a known intrusion feature, thereby determining whether it is a potential intrusion feature. The present application sets a similarity threshold to screen out intrusion features that are highly similar to the current feature, and counts the number of intrusion features that are similar to the current feature, which helps to quantify the similarity between the current feature and the known intrusion features. Then, based on the result of the fifth calculation, a threshold is set to determine whether the current feature is similar to enough known intrusion features. If the number of similar intrusion features exceeds the threshold, it is considered that the number of intrusion features similar to the current feature is sufficient, and then the retrieval step is performed; if not, it is considered that the number of intrusion features similar to the current feature is small, and then the feature update step is performed to add the current feature as an intrusion feature to the existing feature set to expand the feature set, and then the second modeling step is performed to improve the calculation accuracy of the intrusion model.

[0055] Optionally, after executing the first acquisition step and before executing the intrusion determination step, the method further includes:

[0056] The third collection: collect historical data of infrared sensors, which are recorded as the first historical data; collect historical data of microwave sensors, which are recorded as the second historical data; collect historical data of laser sensors, which are recorded as the third historical data;

[0057] First processing: performing standardization processing on the first historical data, the second historical data and the third historical data, and integrating the first historical data, the second historical data and the third historical data after the standardization processing into a historical data set;

[0058] Second processing: performing standardization processing on the first data, the second data and the third data, recording the standardized first data as the seventh data, recording the standardized second data as the eighth data, and recording the standardized third data as the ninth data;

[0059] Clustering: The k-Means algorithm is used to cluster the historical data in the historical data set to obtain six different clusters;

[0060] Mapping: Map the seventh data, the eighth data, and the ninth data to different clusters respectively to obtain three mapping results;

[0061] Ninth judgment: judge whether the three mapping results are all in line with expectations, if so, execute the first acquisition step; if not, execute the intrusion judgment step.

[0062] The present application collects historical data of infrared sensors, microwave sensors and laser sensors, and then standardizes the collected historical data and integrates them into historical data sets. Standardization helps to eliminate the dimensional differences between different types of sensor data, making the data more comparable. Then the first data, second data and third data currently collected are standardized and recorded as the seventh data, the eighth data and the ninth data respectively. Then the k-Means algorithm is used to cluster the data in the historical data set to obtain six different clusters. The current seventh data, the eighth data and the ninth data are respectively mapped to the six clusters obtained by clustering to obtain three mapping results. The mapping process helps to determine which categories in the current data and the historical data set are similar. Then, based on the mapping results, a judgment is made. If the three mapping results are all in line with expectations (that is, the seventh data, the eighth data and the ninth data are all mapped to the corresponding normal clusters), the first acquisition step is executed; if it is not in line with expectations, the intrusion judgment step is executed. The present application combines historical data to make a preliminary judgment on the data collected by each sensor before the intrusion judgment. Only when the current data is mapped to an abnormal cluster will the intrusion judgment be made, further reducing the probability of false alarms.

[0063] Optionally, after executing the mapping step and before executing the ninth determination step, the method further includes:

[0064] Determine the clusters: based on the mapping results, respectively obtain the clusters to which the seventh data, the eighth data, and the ninth data belong, record the cluster to which the seventh data belongs as the first cluster, record the cluster to which the eighth data belongs as the second cluster, and record the cluster to which the ninth data belongs as the third cluster;

[0065] Sixth calculation: calculate the distance between the seventh data and each original data in the first cluster, recorded as the fourth distance; calculate the distance between the eighth data and each original data in the second cluster, recorded as the fifth distance; calculate the distance between the ninth data and each original data in the third cluster, recorded as the sixth distance;

[0066] Seventh calculation: calculate the sum of all fourth distances, record it as the tenth data; calculate the sum of all fifth distances, record it as the eleventh data; calculate the sum of all sixth distances, record it as the twelfth data;

[0067] Tenth judgment: judging whether the tenth data and / or the eleventh data and / or the twelfth data are less than a preset threshold value, if so, executing the step of updating the center point; if not, executing the step of adding data;

[0068] Updating the center point: using the seventh data as the center point of the first cluster and / or using the eighth data as the center point of the second cluster and / or using the ninth data as the center point of the third cluster;

[0069] Data addition: adding the seventh data and / or the eighth data and / or the ninth data to the historical data set, and performing a clustering step.

[0070] Based on the mapping results, the present application assigns the seventh data, the eighth data and the ninth data to different clusters (the first cluster, the second cluster and the third cluster), and then calculates the distances (i.e., the fourth distance, the fifth distance and the sixth distance) between the seventh data, the eighth data and the ninth data and each original data in their respective clusters, thereby quantifying the similarity or difference between the data point and other data points in the cluster. Then, the sum of the distances of each data point to all the original data in the cluster to which it belongs (the tenth data, the eleventh data and the twelfth data) is calculated to reflect the overall similarity between the data point and other data points in the cluster. Then, it is determined whether the tenth data, the eleventh data and the twelfth data are less than the preset threshold. If it is less than the preset threshold, it means that the data point is more similar to other data points in the cluster and is more suitable as the new center point of the cluster. Therefore, the step of updating the center point is performed, and the data corresponding to the data less than the preset threshold is used as the new center point of the corresponding cluster. If the distance sum is not less than the preset threshold, it means that the data point is greatly different from other data points in the cluster and is not suitable as the center point of the cluster. Therefore, the step of adding data is performed, that is, the seventh data, the eighth data or the ninth data is added to the historical data set, and the step of clustering is performed. This application can more accurately evaluate whether a data point is suitable as a new center point of the cluster by calculating the distance and distance between the data point and other data points in the cluster and making a judgment, thereby improving the accuracy and robustness of clustering.

[0071] In the second aspect, the present application provides a mine transport track anti-intrusion alarm system, which adopts the following technical solution:

[0072] A mine transport track anti-intrusion alarm system, comprising:

[0073] The first deployment module is used to install infrared sensors, microwave sensors and laser sensors in the monitoring area, and the laser beams sent by all the laser sensors form a three-dimensional channel surrounding the transport track and the transport vehicle;

[0074] The first acquisition module is connected to the first deployment module for acquiring output data of the infrared sensor, which is recorded as the first data; the output data of the microwave sensor, which is recorded as the second data; and the output data of the laser sensor, which is recorded as the third data;

[0075] An intrusion judgment module, which is in communication with the first acquisition module, comprises a first judgment unit, a first calculation unit, a second judgment unit and a third judgment unit;

[0076] A first judging unit, used to judge whether the first data belongs to a preset range;

[0077] a first calculation unit, configured to calculate a moving direction of the first target and a shortest distance between the first target and the transport track based on the second data;

[0078] A second judgment unit is used to judge whether the shortest distance is less than a preset safety threshold and the movement direction is toward the track;

[0079] A third judgment unit, used for judging whether the second target has an invasion trend based on the third data;

[0080] The alarm module is connected to the intrusion judgment module for outputting sound and light alarm signals or sending alarm signals to operation and maintenance personnel.

[0081] This application installs infrared, microwave and laser sensors in the monitoring area through the first deployment module, and the first acquisition module collects data in real time, and uses the intrusion judgment module to comprehensively judge the status, movement direction, distance from the transport track and intrusion trend of each target. Once an intrusion behavior or potential threat is detected, the alarm module will promptly output an audible and visual alarm signal or send an alarm signal to the operation and maintenance personnel to achieve efficient, accurate and reliable intrusion monitoring and protection functions.

[0082] In summary, the present application includes at least one of the following beneficial technical effects:

[0083] 1. This application first uses an infrared sensor to detect whether the first target (a living intrusion target, i.e., a person or an animal) has entered the monitoring area, and then uses the second data collected by the microwave sensor to measure whether the first target has an intrusion trend (i.e., the direction of movement is toward the transportation track and the shortest distance is less than the preset safety threshold). If so, an audible and visual alarm signal is issued to drive the first target away; otherwise, the first acquisition step is re-executed to continue monitoring the monitoring area; at the same time, this application also uses the third data to determine whether the second target (i.e., inanimate objects such as stones) has an intrusion trend. If so, an alarm signal is issued to the operation and maintenance personnel, otherwise, the first acquisition step is re-executed to continue monitoring the monitoring area. This application achieves multi-dimensional and multi-level coverage of the monitoring area by integrating multiple sensors such as infrared sensors, microwave sensors, and laser sensors, which can prevent and promptly detect intrusion incidents, thereby improving the safety of the mine transportation process.

[0084] 2. The present application obtains the timestamps (i.e., the first timestamp and the second timestamp) of the new fourth data (i.e., the fourth data greater than the preset threshold, corresponding to the first alarm number of a certain monitoring area) and the new fifth data (i.e., the fifth data greater than the preset threshold, corresponding to the second alarm number of a certain monitoring area), and in the fifth judgment step, by comparing whether the first timestamp is the same or similar to the second timestamp, it is determined whether the new fourth data and the new fifth data are generated at the same time point or at a similar time point, which helps to confirm whether the two data reflect the alarm situation in the same time period. If the timestamps are the same or similar, it means that the new fourth data and the new fifth data reflect that the second alarm is caused by the first alarm. At this time, the second deployment step (increasing the preset safety threshold) is performed to expand the scope of the early warning, thereby reducing the possibility that the first target causes the invasion of the second target. If the timestamps are different, it means that the two data reflect the situation of different time periods or different events, and the first acquisition step is re-executed at this time. After adding the second acquisition and the fifth judgment steps, the present application can more accurately judge the relevance and time characteristics of the alarm data, so as to make more reasonable decisions, which helps to improve the accuracy and reliability of the alarm and reduce the possibility of false alarms and missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 is a flow chart of Example 1 of the present application;

[0086] Figure 2 is a flow chart of Example 2 of the present application;

[0087] Figure 3 This is a flow chart from S04 second acquisition to S07 linking in Example 3 of the present application;

[0088] Figure 4 This is a flowchart of the third calculation in S61 to the seventh judgment in S69 of Embodiment 3 of the present application;

[0089] Figure 5 This is a flow chart of Example 4 of the present application. DETAILED DESCRIPTION

[0090] The following combination Figures 1 to 5 This application is described in further detail.

[0091] Embodiment 1: This embodiment discloses a mine transport track anti-intrusion alarm method, referring to Figure 1 The method includes: S1 first deployment, S2 first collection, S3 intrusion judgment and S4 alarm. First, various sensors are deployed in the monitoring area, then the data of these sensors are obtained, then intrusion judgment is performed based on these data to obtain judgment results, and then different types of alarms are performed according to the judgment results. This embodiment includes the following steps:

[0092] S1 The first deployment is to install infrared sensors, microwave sensors and laser sensors in the monitoring area.

[0093] Infrared sensors detect the presence of a first target (i.e., a person or animal) by detecting infrared radiation emitted by the first target. Microwave sensors emit microwave signals and receive reflected signals, and detect the movement speed and direction of the first target by measuring changes in the reflected signals. By emitting microwave signals multiple times and receiving reflected signals, the shortest distance from the target object to the transport track can be accurately calculated. Laser sensors detect small movements or position changes of the second target (such as inanimate objects such as stones and clods of earth) based on changes in received reflected signals.

[0094] There is no limitation on the positions of the infrared sensor and the microwave sensor, and the monitoring areas of the two sensors can cover the transport track and a range where the distance from the transport track is greater than a preset safety threshold.

[0095] There is no limitation on the position of the laser sensor either. The monitoring area formed by the laser beams of all laser sensors can surround the transport track and the transport vehicle. For example, the laser beams sent by the laser sensors form a three-dimensional safety island, which surrounds the transport track and the transport vehicle. When the second target passes through the three-dimensional safety island formed by the laser beams emitted by the laser sensors, it will affect the time difference and phase of the received signal, thereby realizing intrusion monitoring of the second target.

[0096] S2 is the first acquisition, which collects the output data of the infrared sensor, recorded as the first data; collects the output data of the microwave sensor, recorded as the second data; and collects the output data of the laser sensor, recorded as the third data.

[0097] S3 intrusion judgment includes S31 first judgment, S32 first calculation, S33 second judgment and S34 third judgment.

[0098] S31 is the first judgment, judging whether the first data belongs to the preset range. If so, it indicates that there is a first target in the current monitoring area, and it is necessary to execute S32 first calculation; if not, it indicates that there is no first target in the current monitoring area, and then execute S2 first collection.

[0099] S32 first calculation: calculating the moving direction of the first target and the shortest distance between the first target and the transportation track based on the second data.

[0100] Calculate the direction of movement: By analyzing the phase difference and time difference of the reflected signal, determine whether the first target is moving toward or away from the transport track.

[0101] Taking the Doppler effect method as an example, the process of calculating the motion direction of the first target is as follows:

[0102] S3211. Microwave sensors transmit a continuous wave (CW) or frequency modulated continuous wave (FMCW) signal.

[0103] S3212. Receive the microwave signal reflected from the first target and measure its frequency deviation.

[0104] S3213, calculating the movement speed v of the first target, the calculation model is as follows:

[0105] ;

[0106] in, is the speed of light; is the value of the frequency deviation of the reflected microwave signal; is the transmitting frequency of the microwave sensor; is the angle between the actual moving direction of the first target and the line connecting the microwave sensor and the first target.

[0107] S3214. By continuously measuring the frequency offset at multiple time points, the movement direction of the first target can be analyzed: if the frequency offset continues to increase, it means that the first target is approaching the sensor; if the frequency offset continues to decrease, it means that the first target is moving away from the sensor.

[0108] Taking the phase difference measurement method as an example, the process of calculating the moving direction of the first target is as follows:

[0109] S3221. The microwave sensor transmits a microwave signal and records the time of transmission.

[0110] S3222. Receive the microwave signal reflected by the first target and record the receiving time.

[0111] S3223. Calculate the phase difference between the transmitted signal and the reflected signal.

[0112] S3224. Calculate the position change of the first target relative to the sensor based on the phase difference and the propagation speed of the microwave signal.

[0113] S3225. By continuously measuring the position changes at multiple time points, the movement direction of the first target can be analyzed.

[0114] Calculate the shortest distance: According to the speed and position of the first target, calculate the shortest distance between it and the transport track. This distance is the straight-line distance from the current position of the first target to the outer edge of the track.

[0115] Taking the geometric method as an example, the process of calculating the shortest distance is as follows:

[0116] S3231. Determine the coordinates and direction of the microwave sensor according to the installation position and angle of the microwave sensor.

[0117] S3232. Obtain the relative position of the first target by measuring the phase difference or Doppler effect.

[0118] S3233. In the geometric model, the shortest distance between the first target and the transport track is calculated based on the coordinates and direction of the sensor and the relative position of the first target.

[0119] Based on the relationship between signal attenuation strength and distance, the process of calculating the shortest distance is as follows:

[0120] S3241. The microwave sensor transmits a microwave signal.

[0121] S3242. Receive the microwave signal reflected from the first target and measure its intensity.

[0122] S3243. Based on the known attenuation characteristics of microwave signals, a relationship model between signal strength and distance is established.

[0123] S3244: Estimate the distance between the first target and the sensor by substituting the measured signal strength into the relationship model.

[0124] S33 is the second judgment, judging whether the shortest distance is less than the preset safety threshold and the movement direction is toward the track. If so, it is considered that there is an intrusion, and then the first alarm S41 is executed; if not, it is considered that there is no intrusion, and then the first collection S2 is executed.

[0125] S34 is the third judgment, judging whether the second target has an invasion trend based on the third data, if so, executing S42 the second alarm; if not, executing S2 the first collection.

[0126] S4 alarm, including S41 first alarm and S42 second alarm.

[0127] S41 first alarm, outputting sound and light alarm signals for on-site warning to drive away the first target. The device for emitting sound and light alarm signals can be arranged on the sleepers of the transport track to drive away the first target, or can be arranged at a position close to the transport track.

[0128] S42 is the second alarm, which sends an alarm signal to the operation and maintenance personnel so that they can respond quickly and take necessary measures (such as sending an emergency brake signal to the transport vehicle).

[0129] This embodiment uses infrared sensors and microwave sensors to determine whether there is an intrusion behavior at the first target, and uses laser sensors to determine whether there is an intrusion trend at the second target, and formulates different alarm policies for living and non-living intrusion targets, which can achieve all-round monitoring and timely response to the area around the transportation track. The solution of this embodiment can not only improve the safety and efficiency of mine transportation, but also effectively prevent safety accidents caused by intrusion behavior.

[0130] Example 2: Reference Figure 2 The difference between this embodiment and embodiment 1 is that, before performing S1 first deployment, the method further includes:

[0131] S01 is a first modeling step, in which point cloud data of the transport track and the surrounding environment of the transport track are obtained, and a three-dimensional model is established based on the point cloud data by using a surface reconstruction algorithm.

[0132] Point cloud data is the basis for building 3D models and can be collected through a variety of methods, including but not limited to:

[0133] LiDAR: LiDAR measures the distance between an object and the sensor by emitting laser pulses and receiving reflected signals, thereby generating 3D point cloud data.

[0134] Structured light scanning: The structured light scanner projects a known pattern of light onto the surface of the transport track and its surrounding environment, captures the changes in the reflected light pattern through the camera, and calculates the point cloud data of the transport track and its surrounding environment based on the deformation of the pattern.

[0135] Stereo vision: Stereo vision uses two or more cameras to capture images of the same scene from different angles, calculates the depth information of objects by comparing the parallax in the images, and generates a three-dimensional point cloud.

[0136] After acquiring the point cloud data, the next step is to build a 3D model through a surface reconstruction algorithm. Surface reconstruction is the process of converting discrete point cloud data into a continuous surface model. Common surface reconstruction methods include mesh generation and voxelization.

[0137] Mesh generation: Connect point cloud data into triangular meshes to construct a surface model of the object.

[0138] Voxelization: Convert point cloud data into a 3D voxel grid, where each voxel represents a small cubic space. The surface is implicitly simulated by assigning effective distance field (SDF) values ​​to all voxels.

[0139] S02 first planning, planning the positions of infrared sensors, microwave sensors and laser sensors in the three-dimensional model.

[0140] In the three-dimensional model, all infrared sensors or microwave sensors can cover all transport tracks and the critical areas around them (the critical area refers to the area where the distance to the outer edge of the transport track is greater than a preset safety threshold).

[0141] In the 3D model, laser sensors are used to monitor whether secondary targets, such as rocks, have a tendency to invade the area where the transport track is located.

[0142] S03 is the first simulation, which simulates different intrusion events (such as people breaking in, objects falling, etc.), observes the response and alarm conditions of each sensor, and obtains simulation results (for example, whether each sensor can respond, whether the alarm can be correctly triggered, etc.).

[0143] The orientation of the infrared sensor, microwave sensor, and laser sensor is determined based on the simulation results so that each sensor can respond correctly and trigger the alarm correctly.

[0144] In the first deployment in S1, infrared sensors, microwave sensors and laser sensors are installed in the monitoring area according to the positions in the first planning in S02 and the directions in the first simulation in S03.

[0145] Then, the first collection of S2 to the alarm of S4 are executed in sequence, and after the alarm of S4 is executed, the following steps are also included:

[0146] S51 is the second calculation, which counts the number of times the sound and light alarm signals are issued in each monitoring area, recorded as the fourth data; and counts the number of times the alarm signals are issued to the operation and maintenance personnel in each monitoring area, recorded as the fifth data.

[0147] S52 first obtains, obtains fourth data greater than a preset threshold, and records it as new fourth data; obtains fifth data greater than a preset threshold, and records it as new fifth data.

[0148] S53 is the fourth judgment, judging whether there is new fourth data and new fifth data corresponding to the same monitoring area. If so, it indicates that there is an invasion of the first target and the second target in the area, and it is necessary to execute S54 second acquisition to further judge whether the invasion of the second target is caused by the first target; if not, execute S2 first acquisition.

[0149] S54: second acquisition, obtaining the timestamp of sending the sound and light alarm signal corresponding to the new fourth data, recorded as the first timestamp; obtaining the timestamp of sending the alarm signal corresponding to the new fifth data, recorded as the second timestamp.

[0150] S55 is the fifth judgment, judging in turn whether the difference between each first timestamp and each second timestamp is less than the preset time threshold. If so, it indicates that the invasion of the first target has led to the invasion of the second target, and then S56 is executed for the second deployment; if not, S2 is executed for the first collection.

[0151] S56 second deployment, gradually increasing the preset safety threshold to send out sound and light alarm signals in advance, warning the first target to stay away from the transportation track, reducing the intrusion of the second target caused by the first target, and executing S2 first collection.

[0152] This embodiment identifies whether the intrusion of the first target will lead to the intrusion of the second target by counting the number of executions of the alarm steps, comparing the timestamp difference, etc. If so, the preset security threshold is increased to reduce the threat of the first target to safe transportation. By adjusting the preset threshold and re-collecting data, it can dynamically adapt to changes in the security environment and improve security.

[0153] Example 3: Reference Figure 3 The difference between this embodiment and the second embodiment is that after executing the first simulation in S03 and before executing the first layout in S1, the following is further included:

[0154] S04 is the second collection, collecting simulation data of the laser sensor under each intrusion event.

[0155] When an intrusion event occurs, the laser sensor will emit a laser beam and receive the reflected signal. By analyzing the intensity, frequency, phase and other parameters of these signals, various simulation data about the intrusion event can be obtained. These simulation data include but are not limited to key information such as the location, time and type of the intrusion.

[0156] S05 Feature extraction: using simulation data to extract intrusion features of different intrusion events, the intrusion features include: intrusion time, location and type.

[0157] S06 second modeling, building an intrusion model based on the intrusion features, the calculation model of the intrusion model is as follows:

[0158] ;

[0159] in, is the output probability of the intrusion model; A is the intrusion event; is the i-th intrusion feature; n is the number of intrusion features; is the probability of the existence of the i-th intrusion feature; is the probability of an intrusion event occurring under the condition that the i-th intrusion feature exists.

[0160] S07 establishes an association, establishing an association between the intrusion feature and the output probability of the intrusion model, so as to retrieve the output probability of the intrusion model through the intrusion feature.

[0161] The association process involves corresponding and matching the extracted intrusion features with the output probability of the intrusion model, thereby obtaining a mapping relationship between each intrusion feature and the output probability of the intrusion model, so that the probability of the unknown event belonging to an intrusion event can be quickly retrieved based on its characteristic information (i.e., the output probability of the intrusion model mentioned in the subsequent steps).

[0162] Reference Figure 4 In other embodiments, after executing the third judgment in S34 and before executing the second alarm in S42, the method further includes:

[0163] S61: third calculation, extracting features contained in the third data, and calculating the similarity between the features and the i-th intrusion feature using a cosine similarity algorithm.

[0164] S62 is the sixth judgment, judging whether the similarity in the third calculation step is greater than the preset similarity threshold. If so, it means that the current feature has a high similarity with the i-th intrusion feature, and then S64 is executed for the fourth calculation; if not, S63 is executed for iteration.

[0165] S63 iterates, taking the i+1th intrusion feature as the new ith intrusion feature, until a preset stop condition is met, and then performing S2 first acquisition. The preset stop condition includes: reaching a preset number of iterations, or finding an intrusion feature with a similarity greater than a preset similarity threshold.

[0166] S64: fourth calculation, using a cosine similarity algorithm to respectively calculate the similarity between the feature and the remaining intrusion features, recorded as sixth data.

[0167] S65: fifth calculation, obtaining sixth data greater than a preset similarity threshold, and using the sixth data as new sixth data.

[0168] S66 is an eighth judgment, judging whether the quantity of the new sixth data is greater than a preset quantity threshold, if so, executing S68 to retrieve; if not, executing S67 to update the characteristics.

[0169] S67 feature update, integrate all intrusion features into a feature set, add the features to the feature set, and execute S06 second modeling, use the updated feature set to readjust the intrusion model to improve the accuracy and adaptability of the intrusion model, and then execute S68 calling.

[0170] S68 retrieves, based on the association relationship established in the association step S07, the output probability of the intrusion model is retrieved using the i-th intrusion feature.

[0171] S69 is the seventh judgment, judging whether the output probability is greater than a preset probability threshold. If so, it indicates that the second object has an intrusion trend, and then S41 is executed as the second alarm; if not, S2 is executed as the first collection.

[0172] This embodiment first extracts the features of the third data, and compares them with the existing intrusion features one by one using the cosine similarity algorithm. If the similarity exceeds a preset threshold, the similarity between the feature and the remaining features is further calculated and the number of features that meet the conditions is counted; if the number exceeds the preset threshold, the output probability of the intrusion model is retrieved using the matching intrusion feature and it is determined whether it indicates an intrusion trend; if any condition is not met, other features are iteratively searched, the feature set is updated and the model is rebuilt, or the data collection stage is returned to improve the accuracy and efficiency of rail transportation anti-intrusion.

[0173] Example 4: Reference Figure 5 The difference between this embodiment and embodiment 1 is that after executing S2 first acquisition and before executing S3 intrusion determination, it also includes:

[0174] S71 is the third collection, collecting historical data of infrared sensors, which is recorded as first historical data; collecting historical data of microwave sensors, which is recorded as second historical data; collecting historical data of laser sensors, which is recorded as third historical data.

[0175] The historical data of the infrared sensor are collected. These historical data reflect the changes in the infrared radiation in the monitoring area of ​​the infrared sensor and are used to detect moving people and animals (ie, the first target). These historical data are marked as the first historical data.

[0176] The historical data of the microwave sensor is collected. The microwave sensor detects the moving direction of the first target or the shortest distance between the first target and the transportation track by emitting microwaves and receiving reflected waves.

[0177] Collect historical data from laser sensors, which accurately measure the presence of inanimate intrusion by emitting a laser beam and measuring the reflection time.

[0178] S72: first processing, performing standardization processing on the first historical data, the second historical data and the third historical data, and integrating the standardized first historical data, the second historical data and the third historical data into a historical data set.

[0179] The second processing S73 uses the same standardization method as in the first processing S72 to standardize the first data, the second data, and the third data, and records the standardized first data as the seventh data, the standardized second data as the eighth data, and the standardized third data as the ninth data.

[0180] S74 clustering uses the k-Means algorithm to cluster the historical data in the historical data set and obtains six different clusters. The six clusters are: the cluster composed of the normal first historical data, the cluster composed of the first historical data during invasion, the cluster composed of the normal second historical data, the cluster composed of the second historical data during invasion, the cluster composed of the normal third historical data, and the cluster composed of the third historical data during invasion.

[0181] S75 mapping, mapping the seventh data, the eighth data and the ninth data to different clusters respectively, to obtain three mapping results.

[0182] S76 determines the clusters. Based on the mapping results, the clusters to which the seventh data, the eighth data, and the ninth data belong are respectively obtained. The cluster to which the seventh data belongs is recorded as the first cluster, the cluster to which the eighth data belongs is recorded as the second cluster, and the cluster to which the ninth data belongs is recorded as the third cluster.

[0183] S77 is a sixth calculation, which is to calculate the distance between the seventh data and each original data in the first cluster, recorded as the fourth distance; calculate the distance between the eighth data and each original data in the second cluster, recorded as the fifth distance; calculate the distance between the ninth data and each original data in the third cluster, recorded as the sixth distance.

[0184] S78 is the seventh calculation, calculating the sum of all fourth distances, recording it as the tenth data; calculating the sum of all fifth distances, recording it as the eleventh data; calculating the sum of all sixth distances, recording it as the twelfth data.

[0185] S79 is the tenth judgment, judging whether the tenth data and / or the eleventh data and / or the twelfth data are less than a preset threshold. If so, it is considered that the new data point is close enough to the cluster to which it belongs and can represent the characteristics of the cluster, and then S80 is executed to update the center point; if not, S81 is executed to add data.

[0186] S80 updates the center point, using the seventh data as the center point of the first new cluster and / or the eighth data as the center point of the second new cluster and / or the ninth data as the center point of the third new cluster.

[0187] S81 data addition, adding the seventh data and / or the eighth data and / or the ninth data to the historical data set, and performing S74 clustering.

[0188] S82 ninth judgment, respectively judge whether the three mapping results meet expectations, if so, execute S2 first acquisition; if not, execute S3 intrusion judgment.

[0189] The three mapping results are in line with expectations, which means that the three data are mapped into the normal clusters formed by the historical data.

[0190] This embodiment first collects historical data from infrared, microwave and laser sensors and performs standardization processing, and then uses the k-Means algorithm to cluster the historical data to form six clusters. The newly collected data is also standardized and mapped to the corresponding cluster, and the distance sum between it and the original data in the cluster is calculated. Based on the comparison of these distance sums with the preset threshold, it is decided whether to update the center point of the cluster or add the data to the historical data set and re-cluster. Finally, depending on whether the mapping result meets expectations, it is decided whether to continue to collect new data or perform intrusion judgment. This embodiment clusters the historical data and maps the real-time data to the clustered clusters to determine whether there is an intrusion event. This method not only improves the accuracy and efficiency of intrusion detection, but also enhances adaptability. With the continuous addition of new data, this embodiment can dynamically adjust the distribution of clusters. By calculating the distance sum between the new data and the original data in the cluster and comparing it with the preset threshold, it can more sensitively capture abnormal data, thereby issuing an intrusion alarm in time.

[0191] Embodiment 5: This embodiment discloses a mine transport track anti-intrusion alarm system, the system comprising:

[0192] The first deployment module is responsible for installing infrared sensors, microwave sensors and laser sensors in the monitoring area. The laser beams sent by all laser sensors form a three-dimensional channel surrounding the transport track and the transport vehicle.

[0193] The first acquisition module is communicatively connected with the first deployment module, and is used for acquiring output data of the infrared sensor, recorded as the first data; acquiring output data of the microwave sensor, recorded as the second data; and acquiring output data of the laser sensor, recorded as the third data.

[0194] The intrusion judgment module is connected to the first acquisition module in communication, and includes a first judgment unit, a first calculation unit, a second judgment unit and a third judgment unit.

[0195] The first judgment unit is responsible for preliminarily screening the output data (i.e., the first data) of the infrared sensor to determine whether it belongs to the preset range. If so, it means that a person or an animal has entered the monitoring area, and the first calculation unit is triggered.

[0196] The first calculation unit is used to calculate the moving direction of the target object and the shortest distance between the target object and the transport track by using the output data (ie, the second data) of the microwave sensor.

[0197] The second judgment unit is used to judge whether the shortest distance is less than a preset safety threshold based on the result of the first calculation unit. If it is less than the threshold and the movement direction of the first target is toward the track, it indicates that there is an invasion risk for the first target.

[0198] The third judgment unit is used to judge whether the second target has an invasion trend based on the third data. The third judgment unit will judge whether the second target has a clear invasion trend based on the output data of the laser sensor (ie, the third data).

[0199] The alarm module is connected to the intrusion judgment module for outputting sound and light alarm signals or sending alarm signals to operation and maintenance personnel.

[0200] In this embodiment, infrared, microwave and laser sensors are scientifically installed in the monitoring area through the first deployment module, and the data of these sensors are obtained in real time by the first acquisition module (recorded as the first, second and third data respectively). Subsequently, the four units in the intrusion judgment module - the first judgment unit evaluates whether the first data is within the preset range, the first calculation unit calculates the movement direction of the target object and the shortest distance to the key area based on the second data, the second judgment unit determines whether the distance is lower than the safety threshold, and the third judgment unit evaluates the intrusion trend of the target object based on the third data. If it is judged to be an intrusion or there is a risk, the alarm module will immediately issue an audible and visual alarm or send an alarm signal to the operation and maintenance personnel. This embodiment uses multiple sensors to monitor intrusion events, reducing the probability of false alarms or missed alarms.

[0201] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A mine transport track anti-intrusion alarm method, characterized in that: include: First deployment: infrared sensors, microwave sensors and laser sensors are installed in the monitoring area. The laser beams sent by all laser sensors form a three-dimensional channel surrounding the transport track and the transport vehicle. First acquisition: collecting the output data of the infrared sensor, recorded as the first data; collecting the output data of the microwave sensor, recorded as the second data; collecting the output data of the laser sensor, recorded as the third data; Intrusion judgment: including first judgment, first calculation, second judgment and third judgment; First judgment: judging whether the first data belongs to a preset range, and if so, executing a first calculation step; If not, then execute the first acquisition step; First calculation: calculating the moving direction of the first target and the shortest distance between the first target and the transport track based on the second data; Second judgment: judging whether the shortest distance is less than a preset safety threshold and the moving direction is toward the track, if so, executing the first alarm step; If not, then execute the first acquisition step; Third judgment: judging whether the second target has an intrusion trend based on the third data, if so, executing the second alarm step; if not, executing the first collection step; Alarm: including the first alarm and the second alarm; First alarm: output sound and light alarm signal; Second alarm: Send an alarm signal to the operation and maintenance personnel; Second calculation: counting the number of executions of the first alarm step in each monitoring area, recorded as fourth data; counting the number of executions of the second alarm step in each monitoring area, recorded as fifth data; First acquisition: acquiring fourth data greater than a preset number of times threshold, and recording it as new fourth data; acquiring fifth data greater than a preset number of times threshold, and recording it as new fifth data; Fourth judgment: judging whether the new fourth data and the new fifth data correspond to the same monitoring area, if so, executing the second acquisition step; if not, executing the first acquisition step; Second acquisition: obtaining the timestamp of the sound and light alarm signal corresponding to the new fourth data, recorded as the first timestamp; obtaining the timestamp of the alarm signal corresponding to the new fifth data, recorded as the second timestamp; Fifth judgment: sequentially judging whether the difference between each first timestamp and each second timestamp is less than a preset time threshold, if so, executing the second deployment step; if not, executing the first collection step; Second deployment: Increase the preset safety threshold and execute the first acquisition steps.

2. The mine transport track anti-intrusion alarm method according to claim 1 is characterized in that: Before performing the first deployment step, the method further includes: First modeling: acquiring point cloud data of the transport track and the surrounding environment of the transport track, and building a three-dimensional model through a surface reconstruction algorithm based on the point cloud data; First planning: planning the positions of infrared sensors, microwave sensors and laser sensors in the three-dimensional model; First simulation: Introduce different intrusion events into the 3D model to conduct intrusion simulation tests, obtain simulation results, and determine the orientation of the infrared sensor, microwave sensor, and laser sensor based on the simulation results; In the first deployment step, infrared sensors, microwave sensors and laser sensors are installed in the monitoring area according to the positions in the first planning step and the orientations in the first simulation step.

3. The mine transport track anti-intrusion alarm method according to claim 2 is characterized in that: After performing the first simulation step and before performing the first layout step, the method further includes: Second collection: collect simulation data of laser sensors under each intrusion event; Feature extraction: Use simulation data to extract intrusion features of different intrusion events; Second modeling: construct an invasion model based on the invasion characteristics. The calculation model of the invasion model is as follows: ； in, is the output probability of the intrusion model; A is the intrusion event; is the i-th intrusion feature; n is the number of intrusion features; is the probability of the existence of the i-th intrusion feature; is the probability of an intrusion event occurring under the condition that the i-th intrusion feature exists; Association: Establish the association between the intrusion features and the output probability of the intrusion model.

4. The mine transport track anti-intrusion alarm method according to claim 3 is characterized in that: After executing the third judgment step and before executing the second alarm step, the method further includes: Third calculation: extracting the features contained in the third data, and calculating the similarity between the features and the i-th intrusion feature using a cosine similarity algorithm; Sixth judgment: judging whether the similarity in the third calculation step is greater than a preset similarity threshold, if so, executing the calling step; if not, executing the iteration step; Iteration: take the i+1th intrusion feature as the new ith intrusion feature until the preset stop condition is met and execute the first acquisition step; Retrieve: Use the i-th intrusion feature to retrieve the output probability of the intrusion model; Seventh judgment: judge whether the output probability described in the retrieved step is greater than the preset probability threshold, if so, execute the second alarm step; if not, execute the first collection step.

5. The mine transport track anti-intrusion alarm method according to claim 4 is characterized in that: After executing the sixth determination step and before executing the calling step, the method further includes: Fourth calculation: using the cosine similarity algorithm to calculate the similarity between the feature and the remaining intrusion features respectively, recorded as the sixth data; Fifth calculation: obtaining sixth data greater than a preset similarity threshold and recording it as new sixth data; Eighth judgment: judging whether the amount of the new sixth data is greater than a preset amount threshold, if so, executing the step of retrieving; if not, executing the step of updating the feature; Feature update: All intrusion features are integrated into a feature set, and the features are added to the feature set, and the second modeling step is performed.

6. The mine transport track anti-intrusion alarm method according to claim 1 or 2, characterized in that: After executing the first acquisition step and before executing the intrusion judgment step, the method further includes: The third collection: collect historical data of infrared sensors, which are recorded as the first historical data; collect historical data of microwave sensors, which are recorded as the second historical data; collect historical data of laser sensors, which are recorded as the third historical data; First processing: performing standardization processing on the first historical data, the second historical data and the third historical data, and integrating the first historical data, the second historical data and the third historical data after the standardization processing into a historical data set; Second processing: performing standardization processing on the first data, the second data and the third data, recording the standardized first data as the seventh data, recording the standardized second data as the eighth data, and recording the standardized third data as the ninth data; Clustering: The k-Means algorithm is used to cluster the historical data in the historical data set to obtain six different clusters; Mapping: Map the seventh data, the eighth data, and the ninth data to different clusters respectively to obtain three mapping results; Ninth judgment: judge whether the three mapping results are all in line with expectations, if so, execute the first acquisition step; if not, execute the intrusion judgment step.

7. The mine transport track anti-intrusion alarm method according to claim 6, characterized in that: After executing the mapping step and before executing the ninth determination step, the method further includes: Determine the clusters: based on the mapping results, respectively obtain the clusters to which the seventh data, the eighth data, and the ninth data belong, record the cluster to which the seventh data belongs as the first cluster, record the cluster to which the eighth data belongs as the second cluster, and record the cluster to which the ninth data belongs as the third cluster; Sixth calculation: calculate the distance between the seventh data and each historical data in the first cluster, recorded as the fourth distance; calculate the distance between the eighth data and each historical data in the second cluster, recorded as the fifth distance; calculate the distance between the ninth data and each historical data in the third cluster, recorded as the sixth distance; Seventh calculation: calculate the sum of all fourth distances, record it as the tenth data; calculate the sum of all fifth distances, record it as the eleventh data; calculate the sum of all sixth distances, record it as the twelfth data; Tenth judgment: judging whether the tenth data and / or the eleventh data and / or the twelfth data are less than a preset threshold value, if so, executing the step of updating the center point; if not, executing the step of adding data; Updating the center point: using the seventh data as the center point of the first cluster and / or using the eighth data as the center point of the second cluster and / or using the ninth data as the center point of the third cluster; Data addition: adding the seventh data and / or the eighth data and / or the ninth data to the historical data set, and performing a clustering step.

8. A mine transport track anti-intrusion alarm system, the system is applicable to the method according to any one of claims 1 to 7, characterized in that: include: The first deployment module is used to install infrared sensors, microwave sensors and laser sensors in the monitoring area, and the laser beams sent by all the laser sensors form a three-dimensional channel surrounding the transport track and the transport vehicle; The first acquisition module is connected to the first deployment module for acquiring output data of the infrared sensor, which is recorded as the first data; the output data of the microwave sensor, which is recorded as the second data; and the output data of the laser sensor, which is recorded as the third data; An intrusion judgment module, which is in communication with the first acquisition module, comprises a first judgment unit, a first calculation unit, a second judgment unit and a third judgment unit; A first judging unit, used to judge whether the first data belongs to a preset range; a first calculation unit, configured to calculate a moving direction of the first target and a shortest distance between the first target and the transport track based on the second data; A second judgment unit is used to judge whether the shortest distance is less than a preset safety threshold and the movement direction is toward the track; A third judgment unit, used for judging whether the second target has an invasion trend based on the third data; The alarm module is connected to the intrusion judgment module for outputting sound and light alarm signals or sending alarm signals to operation and maintenance personnel.

Images