Rail train obstacle detection warning signal transmission system and method

Abstract

The present invention discloses a rail train obstacle detection warning signal transmission system and method, which relates to the field of obstacle detection communication technology, and comprises a rail obstacle detection test module, a test data analysis module, a curve operation control module and a detection data communication module. The rail obstacle detection test module is used to perform a rail obstacle detection simulation test and obtain curve image data; the test data analysis module is used to obtain target operation parameters when the train is traveling on a curve during the test, and a test data association model is established according to the target operation parameters and the curve data; the curve operation control module is used to plan the travel speed and deceleration time of the train according to the test data association model; the detection data communication module is used to perform obstacle detection by using multi-sensor fusion, and when an obstacle is detected, communication is performed with an operation control center, thereby improving the accuracy and effectiveness of obstacle detection and ensuring the operation safety of the train.

Description

Technical Field

[0001] The present invention relates to the field of obstacle detection communication technology, and in particular to a rail train obstacle detection warning signal transmission system and method. Background Art

[0002] Urban rail transit is a high-capacity transportation system with a large number of passengers every day. It is one of the main means of transportation for urban travel. Therefore, it is very important to ensure the safe and reliable operation of rail transit. With the gradual popularization of fully automatic unmanned driving technology in rail transit, there is no driver on the train. Unexpected trains, abandoned equipment, dropped objects and other obstacles may appear in front of the train, which brings serious safety hazards to driving safety. Therefore, the train needs to have obstacle detection function;

[0003] However, in the prior art, random obstacles on the track can only be monitored by installing monitoring equipment on the trackside and using manual inspection methods such as rail vehicles and drivers to monitor the obstacles. There is a risk of failing to detect obstacles in time. By installing sensors on the train and using multi-sensor fusion to detect obstacles on the track and timely transmit abnormal warning signals to the train operation control center, the timeliness of obstacle detection can be effectively improved to maintain the safety of train operation. However, when the train passes through a curve, it is necessary to slow down to detect obstacles. Since it takes a certain amount of time for the sensor to acquire the track image, improper deceleration may result in the failure to collect a complete curve image during obstacle detection. Too sudden deceleration may also cause discomfort to passengers on the train, which can easily affect the effectiveness of obstacle detection on the track and the normal operation of the train. Summary of the invention

[0004] The object of the present invention is to provide a railway train obstacle detection warning signal transmission system and method to solve the problems raised in the prior art.

[0005] To achieve the above-mentioned object, the present invention provides the following technical solutions: A rail train obstacle detection warning signal transmission system, comprising a rail obstacle detection test module, a test data analysis module, a curve operation control module and a detection data communication module;

[0006] Performing a track obstacle detection simulation test through the track obstacle detection test module and acquiring the curve image data collected during the test;

[0007] The test data analysis module analyzes the curve image data to obtain the target operating parameters of the train when it is traveling on the curve during the test, and establishes a test data association model based on the target operating parameters and the curve data;

[0008] The curve operation control module is used to plan the speed of the train when it is actually traveling on the curve according to the test data association model, and to plan the deceleration time, and the train operation is controlled according to the planning results;

[0009] The detection data communication module is used to detect obstacles when the train is running on the track, and communicates with the operation control center when an abnormality is detected on the track.

[0010] Preferably, the track obstacle detection test module includes an obstacle detection test unit, a sensor fusion unit and a curve image data acquisition unit;

[0011] The obstacle detection test unit simulates a scenario in which a train is running on a track, and performs a simulation test on track obstacle detection during the train running process;

[0012] The sensor fusion unit uses sensor fusion technology to fuse visual sensors and radar sensors to detect track obstacles;

[0013] When performing simulation tests, the curve image data acquisition unit collects track image data acquired by sensors when a train travels at different speeds on curves with different radii.

[0014] Preferably, the test data analysis module includes an image splicing complete detection unit, a curve data acquisition unit, a train operation parameter acquisition unit and a test data association model establishment unit;

[0015] The track images acquired by the sensor when the train is running on the curve are stitched by the image stitching integrity detection unit, and the OpenCV technology is used to detect whether the stitched image can show the complete curve, that is, the image integrity is detected by the OpenCV technology;

[0016] The curve radius corresponding to the complete curve image that can be captured is collected by the curve data collection unit;

[0017] The train running parameter acquisition unit acquires the running speed parameter of the train when the train is running on the curve in the complete curve image;

[0018] The test data association model building unit performs fitting training on the collected curve radius and driving speed parameters to build a test data association model.

[0019] Preferably, the curve operation control module includes a target curve data substitution unit, a required driving speed prediction unit and a deceleration planning unit;

[0020] When the train is actually running on the track, the target curve data substitution unit obtains the curve radius on the route the train is currently running on as target curve data, substitutes the target curve data into the test data association model, and obtains a data output result;

[0021] The required speed prediction unit predicts the required speed of the train when it is traveling on a curve on the current route according to the data output result;

[0022] The deceleration planning unit obtains the current running speed of the train on the straight track connected to the curve as the initial running speed, and the default running speed of the train on the straight track remains unchanged. If the initial running speed is greater than the required running speed on the corresponding curve, the speed difference between the initial running speed and the required running speed is obtained, and the difference between the historical speed of the train when it was running on the straight track connected to different curves in the past under normal operation and the historical speed of the train when it was running on the corresponding curve is collected as the historical speed difference. The distance required from the position where the train started to decelerate when it was running on the straight track connected to the curve to the starting position of the corresponding curve is collected as the distance required for normal deceleration. The normal operation refers to the situation that the train does not suddenly decelerate during operation, and the starting point of the curve refers to the connection point between the straight track and the corresponding curve. The historical speed difference and the distance required for normal deceleration are fitted and trained to establish a curve deceleration planning model, and the speed difference between the initial running speed and the required running speed is input into the curve deceleration planning model. The time when the train needs to start decelerating on the straight track is analyzed according to the output result, and the train is controlled to start decelerating at the time when deceleration is required until it decelerates to the required running speed.

[0023] Preferably, the detection data communication module includes a track obstacle detection unit, an abnormality warning unit and a warning signal transmission unit;

[0024] The track obstacle detection unit detects obstacles on the curve using visual sensors and radar sensors when the train is traveling on the curve at the required speed;

[0025] When an obstacle is detected on a curve, an abnormal warning is issued by the abnormal warning unit;

[0026] The abnormal warning signal is transmitted to the operation control center, i.e., 0CC, through the warning signal transmission unit.

[0027] A method for transmitting a warning signal for detecting an obstacle on a railway train, comprising:

[0028] S01: Perform a track obstacle detection simulation test and obtain curve image data collected during the test;

[0029] S02: Analyze the curve image data to obtain the target operating parameters of the train when it is running on the curve during the test, and establish a test data association model based on the target operating parameters and the curve data;

[0030] S03: Perform speed planning when the train is actually traveling on a curve based on the test data association model, and perform deceleration time planning;

[0031] S04: Control the train operation according to the planning results, perform obstacle detection when the train is running on the track, and transmit abnormal warning signals to the operation control center when abnormalities are detected on the track.

[0032] Preferably, in step S01: a scenario of a train running on a track is simulated, and a simulation test of track obstacle detection is performed during the train running process. The sensor fusion technology is used to fuse the visual sensor and the radar sensor to detect track obstacles. When performing the simulation test, the track image data obtained by the sensor when the train runs on k curves with different radii at different speeds is collected, and the set of k curve radii is r={r1, r2, ..., r k}, during the test, the trains travel on the curve at the speeds in the set v, where v={v1, v2, ..., v n}, where n represents the number of velocity items in the set v.

[0033] Preferably, in step S02: the track images acquired by the sensor when the train is traveling on the curve are spliced, and the OpenCV technology is used to detect whether the spliced ​​images can show the complete curve. When the train is traveling on the curve with a radius in the set r, the sensor can acquire the complete curve image at the speed in the set V, and the set V is: V={V1, V2, ..., V k}, set V is included in set v, V1 means that when the train is traveling at speed V1 on a curve with radius r1, the sensor can obtain a complete image of the curve with radius r1, forming the training data set {(r1, V1), (r2, V2), ..., (r k , V k )}, after fitting the training data, establish the test data association model:

[0034] Y=a1*X+a2;

[0035] Wherein, a1 represents the bias of the test data association model, a2 represents the intercept of the test data association model, X represents the variable representing the curve radius in the test data association model, and Y represents the variable representing the train speed in the test data association model;

[0036] Since the train needs to slow down when traveling on a curve, and considering that the train needs to use sensors to collect images on the track to detect and identify obstacles on the track while traveling, if the deceleration is improper, the complete curve image may not be collected. By simulating the train operation scene, the images collected by the sensor when the train passes through curves of different radii are tested at different driving speeds. After image stitching, it is tested whether the stitched image can display the complete curve. If the complete curve can be displayed, it is judged that when the train travels on the curve of the corresponding radius at the corresponding speed, the sensor can collect complete track image data, and there is no data missing, that is, the image of some areas cannot be collected, resulting in the problem that obstacles in the corresponding areas cannot be detected. By establishing a test data association model, it is used to judge the driving speed on curves of different radii during the subsequent actual operation of the train, ensuring that the sensor can obtain complete image data when the train travels on the curve, and improving the effectiveness of obstacle detection on the curve.

[0037] Preferably, in step S03: when the train is actually running on the track, the radius of a random curve on the current route of the train is obtained as R i , i represents the i-th curve on the current route of the train, and R i Substitute into the test data association model, let X=R i , the predicted speed of the train when traveling on a random curve on the current route is a1*R i +a2, the train's current speed on the straight track connected to a random curve is obtained as H. The straight track connected to a random curve means that the train will enter the curve connected to the corresponding straight track after passing the corresponding straight track. H is used as the initial speed. If H>a1*R i +a2, the speed difference between the initial speed and the required speed is H-(a1*R i +a2), when the train is running normally, the historical speed set of the train on different curves is collected as L={L1, L2, ..., L m}, L m represents the historical speed of the train when it travels on the mth curve in the past. The historical speed set of the train when it travels on the straight track connected to the corresponding curve is obtained as F={F1, F2, ..., F m}, and the historical speed difference set is {F1-L1, F2-L2, ..., F m -L m}, the distance required from the position where the train started to slow down when traveling on the straight track connected to the corresponding m curves to the starting position of the corresponding curve is collected as the distance required for normal deceleration. The set of distances required for normal deceleration is D={D1, D2, ..., Dm}, D m It means that under normal train operation, the train used to travel on the straight track connected to the mth curve at a speed of F m The distance required from the starting point of the deceleration to the starting point of the mth curve is calculated for the training data {(F1-L1, D1), (F2-L2, D2), ..., (F m -L m , D m )}After fitting, the curve deceleration planning model is established:

[0038] y=ρ1*x+ρ2;

[0039] Among them, ρ1 represents the bias of the curve deceleration planning model, ρ2 represents the intercept of the curve deceleration planning model, x represents the variable representing the speed difference in the curve deceleration planning model, and y represents the variable representing the distance in the curve deceleration planning model. Let x=H-(a1*R i +a2), the time when the train needs to start decelerating on the straight track is predicted to be: the distance required to reach the starting point of the curve connected to the corresponding straight track is ρ1*[H-(a1*R i +a2)]+ρ2, the deceleration begins;

[0040] After determining the speed that the train should travel on the curve, considering that the train needs to slow down when entering the curve after traveling on the straight track, but improper deceleration operation, that is, deceleration too suddenly will cause discomfort to passengers on the train, the train starts to decelerate when it is a certain distance away from the curve. The historical normal operation data of the train is collected through big data technology, and the curve deceleration planning model is established through data fitting. The timing of starting deceleration when the difference between the speed of the train on the straight track and the speed that should be traveled when entering the curve is different is planned. This can not only avoid the problem of deceleration too suddenly, but also ensure that the train runs at the speed that should be traveled when entering the curve.

[0041] Preferably, in step S04: the train is controlled to run normally on the current straight track, and the distance required for the train to reach the starting position of the curve connected to the current straight track is monitored. When the required distance is monitored to be ρ1*[H-(a1*R i +a2)]+ρ2, the train starts to decelerate until it decelerates to a1*R i +a2, control the train to move at a speed of a1*R on the curve connected to the current straight track i +a2: When the train is traveling on the current straight track and the curved track connected to the current straight track, the visual sensor and radar sensor are used to detect track obstacles;

[0042] Track obstacle detection is carried out through multi-sensor fusion. When an obstacle is detected, a warning signal is sent to the operation control center in a timely manner, and timely communication with the operation control center is carried out, which is conducive to improving the accuracy of obstacle detection and judgment, and helping to take preventive measures in time when encountering obstacles to ensure the safety of train operation.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention installs sensors on the train, detects track obstacles by means of multi-sensor fusion, sends warning signals to the operation control center in a timely manner when obstacles are detected, and communicates with the operation control center in a timely manner, which is conducive to improving the accuracy of obstacle detection and judgment, and helps to take preventive measures in time when encountering obstacles to ensure the safety of train operation;

[0045] By simulating the train running scene, the images collected by the sensor when the train passes through curves of different radii are tested at different driving speeds. After the images are stitched, it is tested whether the stitched images can display the complete curve. If the complete curve can be displayed, it is judged that when the train travels at the corresponding speed on the curve of the corresponding radius, the sensor can collect complete track image data, and there is no data missing, that is, the problem that the image of some areas cannot be collected, resulting in the inability to detect obstacles in the corresponding areas. By establishing a test data association model, it is used to judge the driving speed on curves of different radii during the subsequent actual operation of the train, ensuring that the sensor can obtain complete image data when the train travels on the curve, and improving the effectiveness of obstacle detection on the curve;

[0046] After determining the speed that the train should travel on the curve, considering that the train needs to slow down when entering the curve after traveling on the straight track, the historical normal operation data of the train is collected through big data technology, and a curve deceleration planning model is established through data fitting. The timing of starting deceleration when the difference between the train speed on the straight track and the speed that should be traveled when entering the curve is planned, which can not only avoid the problem of deceleration too suddenly, but also ensure that the train runs at the speed that should be traveled when entering the curve. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a structural schematic diagram of the railway train obstacle detection warning signal transmission system of the present invention;

[0048] Figure 2 The figure is a flow chart of the method for transmitting warning signals for obstacle detection of railway trains according to the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Example 1: Figure 1 As shown, this embodiment provides a rail train obstacle detection warning signal transmission system, the system includes: a track obstacle detection test module, a test data analysis module, a curve operation control module and a detection data communication module;

[0051] The output end of the track obstacle detection test module is connected to the input end of the test data analysis module, the output end of the test data analysis module is connected to the input end of the curve operation control module, and the output end of the curve operation control module is connected to the input end of the detection data communication module; the track obstacle detection simulation test is performed through the track obstacle detection test module and the curve image data collected during the test is obtained; the curve image data is analyzed through the test data analysis module to obtain the target operation parameters when the train is running on the curve during the test, and a test data association model is established based on the target operation parameters and the curve data; the speed of the train is actually planned when it is running on the curve according to the test data association model through the curve operation control module, and the deceleration time is planned, and the train operation is controlled according to the planning results; the obstacle detection is performed when the train is running on the track through the detection data communication module, and when an abnormality is detected on the track, it communicates with the operation control center.

[0052] The track obstacle detection test module includes an obstacle detection test unit, a sensor fusion unit and a curve image data acquisition unit;

[0053] The output end of the obstacle detection test unit is connected to the input end of the sensor fusion unit, and the output end of the sensor fusion unit is connected to the input end of the curve image data acquisition unit; the obstacle detection test unit simulates the scene of a train running on the track, and a simulation test is performed on the track obstacle detection during the train running process; the sensor fusion unit uses the sensor fusion technology to fuse the visual sensor and the radar sensor to detect track obstacles; when performing the simulation test, the curve image data acquisition unit collects the track image data obtained by the sensor when the train runs on curves with different radii at different speeds.

[0054] The test data analysis module includes an image splicing complete detection unit, a curve data acquisition unit, a train operation parameter acquisition unit, and a test data association model establishment unit;

[0055] The input end of the image stitching complete detection unit is connected to the output end of the curve image data acquisition unit, the output end of the image stitching complete detection unit is connected to the input end of the curve data acquisition unit, the output end of the curve data acquisition unit is connected to the input end of the train operation parameter acquisition unit, and the output end of the train operation parameter acquisition unit is connected to the input end of the test data association model establishment unit; the track image acquired by the sensor when the train is running on the curve is stitched by the image stitching complete detection unit, and the OpenCV technology is used to detect whether the stitched image can show the complete curve, that is, the OpenCV technology is used to detect the image integrity; the curve radius corresponding to the complete curve image that can be captured is collected by the curve data acquisition unit; the train operation parameter acquisition unit collects the driving speed parameters of the train when it is running on the curve in the complete curve image; the test data association model establishment unit fits the collected curve radius and driving speed parameters to establish a test data association model.

[0056] The curve operation control module includes a target curve data input unit, a required driving speed prediction unit and a deceleration planning unit;

[0057] The input end of the target curve data substitution unit is connected to the output end of the test data association model establishment unit, the output end of the target curve data substitution unit is connected to the input end of the expected speed prediction unit, and the output end of the expected speed prediction unit is connected to the input end of the deceleration planning unit; when the train is actually traveling on the track, the curve radius on the current route of the train is obtained as the target curve data through the target curve data substitution unit, and the target curve data is substituted into the test data association model to obtain the data output result; the expected speed prediction unit predicts the expected speed of the train when it is traveling on the curve on the current route according to the data output result; the deceleration planning unit obtains the current speed of the train on the straight track connected to the curve as the initial speed, and the train speed on the straight track is assumed to remain unchanged. If the initial speed is greater than the expected speed on the corresponding curve, the initial speed is obtained. The speed difference between the initial speed and the speed that should be traveled is collected. The difference between the historical speed of the train when it was traveling on the straight track connected to different curves in normal operation and the historical speed on the corresponding curve is collected as the historical speed difference. The distance required from the position where the train started to decelerate when it was traveling on the straight track connected to the curve to the starting point of the corresponding curve is collected as the distance required for normal deceleration. Normal operation refers to the situation that the train does not suddenly decelerate during operation. The starting point of the curve refers to the connection point between the straight track and the corresponding curve. The historical speed difference and the distance required for normal deceleration are fitted and trained, and a curve deceleration planning model is established. The speed difference between the initial driving speed and the speed that should be traveled is input into the curve deceleration planning model. According to the output results, the time when the train needs to start decelerating on the straight track is analyzed. At the time when deceleration is required, the train is controlled to start decelerating until it decelerates to the speed that should be traveled.

[0058] The detection data communication module includes a track obstacle detection unit, an abnormality warning unit and a warning signal transmission unit;

[0059] The input end of the track obstacle detection unit is connected to the output end of the deceleration planning unit, the output end of the track obstacle detection unit is connected to the input end of the abnormal warning unit, and the output end of the abnormal warning unit is connected to the input end of the warning signal transmission unit; when the train is traveling on a curve at the required speed, the track obstacle detection unit uses visual sensors and radar sensors to detect obstacles on the curve; when an obstacle is detected on the curve, the abnormal warning unit issues an abnormal warning; and the warning signal transmission unit transmits the abnormal warning signal to the operation control center, i.e., OCC.

[0060] Example 2: Figure 2 As shown, this embodiment provides a method for transmitting a warning signal for detecting an obstacle on a railway train, which is implemented based on the warning signal transmission system in the embodiment and specifically includes the following steps:

[0061] S01: Conduct track obstacle detection simulation test and obtain curve image data collected during the test: simulate the scene of a train running on the track, conduct a simulation test on track obstacle detection during the train running process, use sensor fusion technology to fuse visual sensors and radar sensors to detect track obstacles, and collect track image data obtained by sensors when the train runs at different speeds on k curves with different radii during the simulation test. The set of k curve radii is r={r1, r2, ..., r k}, during the test, the trains travel on the curve at the speeds in the set v, where v={v1, v2, ..., v n}, where n represents the number of velocity items in the set v;

[0062] S02: Analyze the curve image data, obtain the target operating parameters of the train when traveling on the curve during the test, and establish a test data association model based on the target operating parameters and the curve data: splice the track images obtained by the sensor when the train is traveling on the curve, and use OpenCV technology to detect whether the spliced ​​image can show the complete curve. When the train is traveling on the curve with a radius in the set r, the sensor can obtain the complete curve image at the speed in the set V. The set V is: V={V1, V2, ..., V k}, set V is included in set v, V1 means that when the train is traveling at speed V1 on a curve with radius r1, the sensor can obtain a complete image of the curve with radius r1, forming the training data set {(r1, V1), (r2, V2), ..., (r k , V k )}, after fitting the training data, establish the test data association model:

[0063] Y=a1*X+a2;

[0064] * represents a multiplication sign, where a1 represents the bias of the test data association model, and a2 represents the intercept of the test data association model. A1 and a2 are solved according to the following formulas:

[0065] a1=[k*∑ k j=1 (r j *V j )-∑ k j=1 (r j )∑ k j=1 (V j )] / [k∑ k j=1 (r j ) 2 -(∑k j=1 (r j )) 2 ];

[0066] a2=[∑ k j=1 (V j )-a1*∑ k j=1 (r j )] / k;

[0067] X represents the variable referring to the curve radius in the test data association model, and Y represents the variable referring to the train speed in the test data association model;

[0068] S03: Based on the test data association model, the speed of the train is actually planned when it is traveling on the curve, and the deceleration time is planned: When the train is actually traveling on the track, the radius of a random curve on the current route of the train is obtained as R i , i represents the i-th curve on the current route of the train, and R i Substitute into the test data association model, let X=R i , the predicted speed of the train when traveling on a random curve on the current route is a1*R i +a2, the train's current speed on the straight track connected to a random curve is obtained as H. The straight track connected to a random curve means that the train will enter the curve connected to the corresponding straight track after passing the corresponding straight track. H is used as the initial speed. If H>a1*R i +a2, the speed difference between the initial speed and the required speed is H-(a1*R i +a2), when the train is running normally, the historical speed set of the train on different curves is collected as L={L1, L2, ..., L m}, L m represents the historical speed of the train when it travels on the mth curve in the past. The historical speed set of the train when it travels on the straight track connected to the corresponding curve is obtained as F={F1, F2, ..., F m}, and the historical speed difference set is {F1-L1, F2-L2, ..., F m -L m}, the distance required from the position where the train started to slow down when traveling on the straight track connected to the corresponding m curves to the starting position of the corresponding curve is collected as the distance required for normal deceleration. The set of distances required for normal deceleration is D={D1, D2, ..., D m}, D mIt means that under normal train operation, the train used to travel on the straight track connected to the mth curve at a speed of F m The distance required from the starting point of the deceleration to the starting point of the mth curve is calculated for the training data {(F1-L1, D1), (F2-L2, D2), ..., (F m -L m , D m )}After fitting, the curve deceleration planning model is established:

[0069] y=ρ1*x+ρ2;

[0070] Among them, ρ1 represents the bias of the curve deceleration planning model, ρ2 represents the intercept of the curve deceleration planning model, and the solution method of ρ1 and ρ2 is the same as a1 and a2;

[0071] x represents the variable representing the speed difference in the curve deceleration planning model, y represents the variable representing the distance in the curve deceleration planning model, let x=H-(a1*R i +a2), the time when the train needs to start decelerating on the straight track is predicted to be: the distance required to reach the starting point of the curve connected to the corresponding straight track is ρ1*[H-(a1*R i +a2)]+ρ2, the deceleration begins;

[0072] S04: Control the train operation according to the planning results, perform obstacle detection when the train is running on the track, and transmit abnormal warning signals to the operation control center when abnormalities are detected on the track:

[0073] Control the train to run normally on the current straight track, monitor the distance required for the train to reach the starting position of the curve connected to the current straight track, and when the required distance is ρ1*[H-(a1*R i +a2)]+ρ2, the train starts to decelerate until it decelerates to a1*R i +a2, control the train to move at a speed of a1*R on the curve connected to the current straight track i +a2 driving: When the train is driving on the current straight track and the curved track connected to the current straight track, the visual sensor and radar sensor are used to detect track obstacles. When an abnormality is detected on the track, an abnormal warning signal is transmitted to the operation control center;

[0074] For example, to establish a test data association model: Y=30X+51, when the train is actually running on the track, the radius of a random curve on the current route of the train is obtained as R i =0.3, unit: km, R i Substitute into the test data association model, let X=R i=0.3, and the predicted speed of the train on a random curve on the current route is a1*R i +a2=60, in km / h, the train's current speed on the straight track connected to a random curve is H=100, H>a1*R i +a2, the speed difference between the initial speed and the required speed is H-(a1*R i +a2)=40, establish the curve deceleration planning model: y=4x+20, let x=H-(a1*R i +a2)=40, and the predicted time when the train needs to start decelerating on the current straight track is: start decelerating when the distance required to reach the starting position of the curve connected to the corresponding straight track is 180 kilometers, control the train to run normally on the current straight track, monitor the distance required for the train to reach the starting position of the curve connected to the current straight track, and when it is monitored that the required distance is 180 kilometers, control the train to start decelerating until it decelerates to 60km / h, and control the train to run at a speed of 60km / h on the curve connected to the current straight track. When the train is running on the current straight track and on the curve connected to the current straight track, the visual sensor and radar sensor are used to detect track obstacles, and when an abnormality is detected on the track, an abnormal warning signal is transmitted to the operation control center.

[0075] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for transmitting a warning signal for detecting an obstacle on a railway train, characterized in that: The following steps are involved: S01: Perform a track obstacle detection simulation test and obtain curve image data collected during the test; S02: Analyze the curve image data to obtain the target operating parameters of the train when it is running on the curve during the test, and establish a test data association model based on the target operating parameters and the curve data; S03: Perform speed planning when the train is actually traveling on a curve based on the test data association model, and perform deceleration time planning; S04: Control the train operation according to the planning results, perform obstacle detection when the train is running on the track, and transmit abnormal warning signals to the operation control center when abnormalities are detected on the track; In step S01: a scenario of a train running on a track is simulated, and a simulation test of track obstacle detection is performed during the train running process. The sensor fusion technology is used to fuse the visual sensor and the radar sensor to detect track obstacles. During the simulation test, the track image data obtained by the sensor when the train runs on k curves with different radii at different speeds is collected. The set of k curve radii is r={r1, r2, ..., r k }, during the test, the trains travel on the curve at the speeds in the set v, where v={v1, v2, ..., v n }, where n represents the number of velocity items in the set v; In step S02: the track images acquired by the sensor when the train is running on the curve are stitched together, and the OpenCV technology is used to detect whether the stitched images can show the complete curve. When the train is running on the curve with a radius in the set r, the sensor can acquire the complete curve image at the speed in the set V. The set V is: V={V1, V2, ..., V k }, forming a training data set {(r1, V1), (r2, V2), ..., (r k , V k )}, after fitting the training data, establish the test data association model: Y=a1*X+a2; Wherein, a1 represents the bias of the test data association model, a2 represents the intercept of the test data association model, X represents the variable representing the curve radius in the test data association model, and Y represents the variable representing the train speed in the test data association model; In step S03: when the train is actually running on the track, the radius of a random curve on the current route of the train is obtained as R i , i represents the i-th curve on the current route of the train, and R i Substitute into the test data association model, let X=R i , the predicted speed of the train when traveling on a random curve on the current route is a1*R i +a2, get the train's current speed on the straight track connected to a random curve as H, take H as the initial speed, if H>a1*R i +a2, the speed difference between the initial speed and the required speed is H-(a1*R i +a2), when the train is running normally, the historical speed set of the train on different curves is collected as L={L1, L2, ..., L m }, L m represents the historical speed of the train when it travels on the mth curve in the past. The historical speed set of the train when it travels on the straight track connected to the corresponding curve is obtained as F={F1, F2, ..., F m }, and the historical speed difference set is {F1-L1, F2-L2, ..., F m -L m }, the distance required from the position where the train starts to slow down to the starting position of the corresponding curve when the train travels on the straight track connected to the corresponding m curves in the past is collected as the distance required for normal deceleration. The set of distances required for normal deceleration is D={D1, D2, ..., D m }, for the training data {(F1-L1, D1), (F2-L2, D2), ..., (F m -L m , D m )}After fitting, the curve deceleration planning model is established: y=ρ1*x+ρ2; Among them, ρ1 represents the bias of the curve deceleration planning model, ρ2 represents the intercept of the curve deceleration planning model, x represents the variable representing the speed difference in the curve deceleration planning model, and y represents the variable representing the distance in the curve deceleration planning model. Let x=H-(a1*R i +a2), the time when the train needs to start decelerating on the straight track is predicted to be: the distance required to reach the starting point of the curve connected to the corresponding straight track is ρ1*[H-(a1*R i +a2)]+ρ2, the deceleration begins; In step S04: the train is controlled to run normally on the current straight track, and the distance required for the train to reach the starting position of the curve connected to the current straight track is monitored. When the required distance is detected to be ρ1*[H-(a1*R i +a2)]+ρ2, the train starts to decelerate until it decelerates to a1*R i +a2, control the train to move at a speed of a1*R on the curve connected to the current straight track i +a2, when the train is traveling on the current straight track and on the curve connected to the current straight track, the visual sensor and radar sensor are used to detect track obstacles.

2. A railway train obstacle detection warning signal transmission system, applied to the railway train obstacle detection warning signal transmission method as claimed in claim 1, characterized in that: It includes a track obstacle detection test module, a test data analysis module, a curve operation control module and a detection data communication module; Performing a track obstacle detection simulation test through the track obstacle detection test module and acquiring the curve image data collected during the test; The test data analysis module analyzes the curve image data to obtain the target operating parameters of the train when it is traveling on the curve during the test, and establishes a test data association model based on the target operating parameters and the curve data; The curve operation control module is used to plan the speed of the train when it is actually traveling on the curve according to the test data association model, and to plan the deceleration time, and the train operation is controlled according to the planning results; The detection data communication module is used to detect obstacles when the train is running on the track, and communicates with the operation control center when an abnormality is detected on the track.

3. The rail train obstacle detection warning signal transmission system according to claim 2, characterized in that: The track obstacle detection test module includes an obstacle detection test unit, a sensor fusion unit and a curve image data acquisition unit; The obstacle detection test unit simulates a scenario in which a train is running on a track, and performs a simulation test on track obstacle detection during the train running process; The sensor fusion unit uses sensor fusion technology to fuse visual sensors and radar sensors to detect track obstacles; When performing simulation tests, the curve image data acquisition unit collects track image data acquired by sensors when a train travels at different speeds on curves with different radii.

4. The rail train obstacle detection warning signal transmission system according to claim 3, characterized in that: The test data analysis module includes an image splicing complete detection unit, a curve data acquisition unit, a train operation parameter acquisition unit and a test data association model establishment unit; The track images acquired by the sensor when the train is running on the curve are stitched by the image stitching complete detection unit, and the OpenCV technology is used to detect whether the stitched images can show the complete curve; The curve radius corresponding to the complete curve image that can be captured is collected by the curve data collection unit; The train running parameter acquisition unit acquires the running speed parameter of the train when the train is running on the curve in the complete curve image; The test data association model building unit performs fitting training on the collected curve radius and driving speed parameters to build a test data association model.

5. The rail train obstacle detection warning signal transmission system according to claim 4, characterized in that: The curve operation control module includes a target curve data substitution unit, a required driving speed prediction unit and a deceleration planning unit; When the train is actually running on the track, the target curve data substitution unit obtains the curve radius on the route the train is currently running on as target curve data, substitutes the target curve data into the test data association model, and obtains a data output result; The required speed prediction unit predicts the required speed of the train when it is traveling on a curve on the current route according to the data output result; The deceleration planning unit obtains the current speed of the train on the straight track connected to the curve as the initial speed. If the initial speed is greater than the required speed on the corresponding curve, the speed difference between the initial speed and the required speed is obtained. The difference between the historical speed of the train when it was traveling on the straight track connected to different curves in normal operation and the historical speed on the corresponding curve is collected as the historical speed difference. The distance required from the position where the train started to decelerate when it was traveling on the straight track connected to the curve to the starting point of the corresponding curve is collected as the distance required for normal deceleration. The starting point of the curve refers to the connection point between the straight track and the corresponding curve. The historical speed difference and the distance required for normal deceleration are fitted and trained to establish a curve deceleration planning model. The speed difference between the initial speed and the required speed is input into the curve deceleration planning model. The time when the train needs to start decelerating on the straight track is analyzed based on the output result. At the time when the train needs to start decelerating, the train is controlled to start decelerating until it decelerates to the required speed.

6. The railway train obstacle detection warning signal transmission system according to claim 5, characterized in that: The detection data communication module includes a track obstacle detection unit, an abnormality warning unit and a warning signal transmission unit; The track obstacle detection unit detects obstacles on the curve using visual sensors and radar sensors when the train is traveling on the curve at the required speed; When an obstacle is detected on a curve, an abnormal warning is issued by the abnormal warning unit; The abnormal warning signal is transmitted to the operation control center through the warning signal transmission unit.

Images